Matrix Isolation and Characterization of 1,1-Diazenes Citation Sylwester, Alan Paul (1986) Matrix Isolation and Characterization of 1,1-Diazenes. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/52cm-3b89. https://resolver.caltech.edu/CaltechTHESIS:10252019-122237282 Abstract CHAPTER 1 The photochemical generation, matrix isolation and direct spectroscopic characterization of H 2 NN 3 is reported. UV (VIS filtered) photolysis of carbamoyl azide 15 in a rigid glass (2-MTHF, 80°K) generates blue-violet 3. The electronic absorption spectrum of 3 reveals a structured absorption curve λ max = 636 nm, λ 0,0 = 695 nm for the n-π* transition of 3. This transition is blue shifted in a more polar glass (2-MTHF:nPrCN, 1:1, 80°K) to λ max = 624 nm, λ 0,0 = 681 nm. The argon matrix FT-IR spectrum of 3 shows bands at 2865.55, 2807.20, 1863.20, 1574.16, and 1003.07 cm -1 (1:2000, Ar, 10°K). The characteristic N=N stretch for 3 at 1574.16 demonstrates the considerable double bond character in the 1,1-diazene. Incorporation of a terminal 15 N label into H 2 N- 15 N 3- 15 N shifts the N=N stretch to 1547.64 cm -1 . The argon matrix infrared spectra of 3, 3- 15 N, 3-d 2 , and 3-d 2 - 15 N are reported. Thermal decomposition of 3 (2-MTHF, 90°K) affords 2-tetrazene (λ max = 260 nm), trans HNNH 1 (λ max = 386 nm) and an unidentified species (λ max = 480 nm). Subsequent thermolysis (>100 K) affords NH 3 , N 2 H 4 , N + H 4 N - 3 , H 2 and N 2 products. Oirect irradiation of 3 in a glass (2-MTHF = 77°K) with visible light affords H 2 , N 2 , and trans-HNNH 1. Photodecomposition of matrix isolated 3 (Ar, 10°K) with visible light in the presence of CO affords formaldehyde (H 2 CO), trans-HNNH 1, H 2 , and N 2 . This represents the first direct observation of thermal and photochemical interconversion of H 2 N 2 isomers. CHAPTER 2 Preliminary studies of the low temperature matrix isolation and spectroscopic characterization of 1,1-dimethyldiazene 7 and l, 1-di-isopropyldiazene 18 are reported. The UV (VIS filtered) photolysis of carbamoyl azides 13 and 17 in a rigid medium (organic glass, 80°K or Ar matrix, 10°K) provides a new general method for the photochemical generation of reactive 1,1-diazenes. This photochemical route is considered to proceed via the photo-Curtius rearrangement of a carbamoyl azide to an aminoisocyanate followed by photodecarbonylation to a 1,1-diazene and carbon monoxide. Electronic absorption spectroscopy (2-MTHF, 80°K) reveals structured absorption curves (n-π*) λ max = 556 nm, λ 0,0 = 643 nm for 7 and λ max = 504 nm, λ 0,0 = 620 nm for 18. 1,1-Dimethyldiazene 7 was independently generated by UV (VIS filtered) photolysis of (Z)-3,3-dimethyl-1-phenyltriazene-1-oxide 16 to afford 7 and nitrosobenzene. Matrix isolation FT-IR spectroscopy (Ar, 10°K) reveals the characteristic N=N stretch for 7 at 1600.96 cm -1 . Incorporation of a terminal 15 N label shifts this stretch to 1581.83 cm -1 . The N:N stretch for 18 at 1600.92 cm -1 is 15 N shifted to 1579.46 cm -1 . Photochemical decomposition of 7 (Ar, 10°K) yields the infrared bands of ethane and an unidentified species (U) which is subsequently photolyzed to ethane. The effects of substitution on the electronic transitions and R 2 N=N stretches of 1,1-diazenes correlates with the trends of the isoelectronic carbonyl compounds. Thermolysis of 7 and 18 (2-MTHF, 90°K) yields red-orange 9 (λ max = 464 n, ε ≃ 3000 M -1 cm -1 ) and 30 (λ max = 474 nm, ε ≃ 3000 M -1 cm -1 ), respectively. These species are tentatively identified as the azomethinimine tautomers of the 1,1-diazenes with α-hydrogens. Irradiation of 7 and 18 at their n-π* transitions in the visible (2-MTHF, 80°K) also initially yields 9 and 30, respectively, in addition to the hydrocarbon products expected from nitrogen extrusion. Subsequent bimolecular decomposition of 9 (E a = 8.2 ± 0.5 kcal/mol, log 10 A = 1.8 ± 0.6) yields tetramethyl-2-tetrazene 19. Bimolecular decomposition of 9-d 6 (E a = 8.6 ± 0.5 kcal/mol, log 10 A = 1.4 ± 0.6) reveals a deuterium isotope effect k H /k D = 6.7 at 190°K for loss of 9. Thermal decomposition of 30 affords hydrocarbon products 32, 33, and 34 expected for nitrogen extrusion from 1,1-diazene 18. The activation parameters for unimolecular decomposition of 30 are E a = 16.8 ± 0.5 kcal/mol, log 10 A = 11.8 ± 0.3. CHAPTER 3 The low-temperature 15 N NMR spectrum of the 1,1-diazene, N-(2,2,6,6-tetramethylpiperidyl)nitrene (1) is reported. The 15 N double- and mono-labeled 1,1-diazenes 1a and 1b were synthesized. The nitrene and amino nitrogens of 1 have resonances in dimethyl ether at -90°C at 917.0 and 321.4 ppm, respectively, downfield from anhydrous 15 NH 3 , affording a chemical-shift difference of 595 ppm for the directly bonded nitrogen nuclei. The chemical shift of the ring nitrogen is consistent with an amino nitrogen whose lone pair is largely delocalized. The large downfield shift of the nitrene nitrogen is consistent with a large paramagnetic term due to a low-lying n-π* transition. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: Chemistry Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Dougherty, Dennis A. Thesis Committee: Dougherty, Dennis A. (chair) Dervan, Peter B. Grubbs, Robert H. Marcus, Rudolph A. Defense Date: 22 July 1985 Funders: Funding Agency Grant Number NSF UNSPECIFIED W. R. Grace Fellowship UNSPECIFIED Record Number: CaltechTHESIS:10252019-122237282 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10252019-122237282 DOI: 10.7907/52cm-3b89 Related URLs: URL URL Type Description https://resolver.caltech.edu/CaltechAUTHORS:20160511-152019347 Related Item Article Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 11863 Collection: CaltechTHESIS Deposited By: Mel Ray Deposited On: 25 Oct 2019 22:47 Last Modified: 16 Apr 2021 23:16 Thesis Files Preview PDF - Final Version See Usage Policy. 62MB Repository Staff Only: item control page

MATR.IX ISOLATION AND CHARACTERIZATION OF 1, 1-:0IA'ZENES Thesis by Alan Paul Sylwester In Partial Fulfillment of the Requirements for the Degree of noctor of Philosophy California Institute of Technology Pasadena, California 1986 (Submitted July 22, t 985) ii To "Av Parents and to laura iii ACKNOWLEDGMENTS I would like to thank my research advisor, Peter nervan, for his advice, encouragement, and enthusiastic support of this work. I also want to thank the me mbe rs of the nervan ~ roup, past a nd present, for providing a diverse and stimulating environment in which to grow and learn. Special thanks go to Dan nuan, Pete Schultz, and Dan Mcintyre for help in the early years, !3urt Leland for his help and friendship and Jim Hanson for picking up the torch. Helpful discussions of matrix isolation with l'vbry Mandrich and members of the Chapman group at UCLA are much appreciated. I would like to thank ~ark Thompson for the Toepler pump analyses and \~attson Instruments for a great FT -IR. Thanks go to Sig and Eric of the glass shop, Tony and f1ill of the mechanical shop, and Tom in the electronics shop, for their frequent assistance in turning ideas into working reality. I am indebted to T)ot Lloyd for typing this thesis during a busy time. Financial support of the National "cience Foundation and a W. 'R.. Grace graduate fellowship are gratefully acknowledged. Final thanks go to Laura for her love and support and for "talking science" at ?. a.m. lV ABSTRACT CHAPTER 1 The photochemical generation, matrix isolation spectroscopic characterization of H2NN 3 is reported. and direct UV (VIS filtered) photolysis of carbamoyl azide 15 in a rigid glass (2-MTHF, 800K) generates blue-violet 3. The electronic absorption spectrum of 3 reveals a structured absorption curve A.max = 636 nm, A.o o = 695 nm for the n-TI* transition of 3. ' This transition is blue shifted in a more polar glass (2-MTHF:nPrCN, 1:1, 80°K) to Amax = ()24 nm, A. 0,0 = 681 nm. The argon matrix FT -IR spectrum of 3 shows bands at 2865.55, 2807.20, 1863.20, 1574.16, and 1003.07 cm-1 (1:-:?000, Ar, lOOK). The characteristic N=N stretch for 3 at 1574.16 demonstrates the considerable double bond character in the 1, 1-diazene. Incorporation of a terminal 15N la':lel into H2N_15N 3-15N shifts the N=N stretch to 1547.64 cm-1. The argon matrix infrared spectra of 3, 3-15N, 3d2, and 3-d2_15N are reported. Thermal decomposition of 3 (2-MTHF, 90°K) affords 2-tetrazene ( Amax = 260 nm), and an unidentified species trans HNNH 1 (A.max = 386 nm) fA max= 480 nm). Subsequent thermolysis + ( >100 K) affords NH3, N2H4, NH4N3, H2 and N2 products. Oirect irradiation of 3 in a glass (2-MTH'~=, 770K) with visible light affords H2, N2, and trans'!-JNI\1H 1. Photodecomposition of matrix isolated 3 (Ar, lOOK) with visible light in the presence of CO affords formaldehyde (H2CO), trans-HNNH 1, '!-J2, and N2. This represents the first direct observation of thermal and photochemical interconversion of H2''J? isomers. v CHAPTER 2 Preliminary studies of the low temperature matrix isolation and spectroscopic characterization of isooropyldiazene 18 are reported. 7 l, 1-dimethyldiazene and l, 1-di- The UV (VIS filtered) photolysis of carbamoyl azides 13 and 17 in a rigid medium (organic glass, 800K or Ar matrix, lOOK) orovides a new general method for the photochemical generation of reactive 1, 1-diazenes. This photochemical route is considered to proceed via the photo-Curtius rearrangement of a carbamoyl azide to an aminoisocyanate followed by photodecarbonylation to a 1,1-diazene and carbon monoxide. Electronic absorption spectroscopy (2-MTHF, 8QOK) reveals structured absorption curves fn-TI*) "-max= 556 nm, "-0 o = 643 nm for 7 and "-max= 504 nm, "-o 0 = 620 nm for 18. ' ' 1,1-Dimethyldiazene 7 was independently generated by UV (VIS filtered) photolysis of (Z)-3,3-dimethyl-1phenyltriazene-1-oxide 16 to afford 7 and nitrosobenzene. Matrix isolation FT -IR spectroscopy (Ar, lOOK) reveals the characteristic N=N stretch for 7 at 1600.96 cm-1. Incorporation of a terminal 15N label shifts this stretch to 1581.83 cm-1. The N:N stretch for 18 at 1600.92 cm-1 is 15N shifted to 1579.46 cm-1. Photochemical decomposition of 7 (Ar, lOOK) yields the infrared bands of ethane and an unidentified species (U) which is subsequently photolyzed to ethane. The effects of substitution on the electronic transitions and 'R.2N=N stretches of 1, l-diazenes correlates with the trends of the isoelectronic carbonyl compounds. Thermolysis of 7 and 18 (2-MTHF, 9QOK) yields red-orange 9 (/..max = 464 n, nm, £ £ ~ 3000 M-1 cm-1) and 30 <"-max= ~ ~000 ~.,-1 cm-1 ), respectively. 474 These species are tentatively vi identified as the azomethinimine tautomers of the 1, 1-diazenes with ahydrogens. Irradiation of 7 and 18 at their n-n* transitions in the visible (2MT'!-IF, 800K) also initially yields 9 and 30, respectively, in addition to the hydrocarbon · products expected from nitrogen extrusion. Subsequent bimolecular decomposition of 9 (Ea = 8.2 .±. 0.5 kcal/mol, log;10 A= 1.8 .±. 0.6) yields tetramethyl-2-tetrazene 19. Bimolecular decomposition of 9-d(; (Ea = 8.6 .±. 0.5 kcallmol, log10 A= 1.4 .±. 0.6) reveals a deuterium isotope effect kHikn = 6.7 at 190°K for loss of 9. Thermal decomposition of 30 affords hydrocarbon products 32, 33, and 34 expected for nitrogen extrusion from 1,1diazene 18. The activation parameters for unimolecular decomposition of 30 are E.a = 16.8 .±. 0.5 kcal/mol, log10 A= 11.8 .±. 0.3. CHAPTER 3 The low-temperature 15N NMR spectrum of the 1,1-diazene, N(?,2,6,6-tetramethyloiperidyHnitrene (1) is reported. The 15N double- and mono-labeled 1,1-diazenes 1a and 1b were synthesized. The nitrene and amino nitrogens of 1 have resonances in dimethyl ether at -9QOC at 917.0 and 321.4 ppm, respectively, downfield from anhydrous 15NH3, affording a chemical-shift difference of .595 ppm for the directly bonded nitrogen nuclei. The chemical shift of the ring nitrogen is consistent with an amino nitrogen whose lone pair is largely delocalized. The large downfield shift of the nitrene nitrogen is consistent with a large paramagnetic term due to a lowlying n-1r* transition. vii TABLE OF CONTENTS INTROOUCTION A. B. C. n. Theoretical Studies H2N2 Isomers • • • • • • • • • • • • • • • • • • • • • Generation and Chemistry of J., t -Diazenes • • • • • • • • • • • • • Kinetically Persistent 1, 1-Diazenes • • • • • • • • • • • • • • • • • • • • H2N2 Spectroscopic Characterization • • • • • • • • • • • • • • • • • • 3 9 17 E. H2N2 Chemistry •· •· · · · · · · · · · •••· · · · · · · · · · · · · · · · · · · · 25 29 F. Experimental Strategy • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 33 RESULTS AND DISCUSSION CHAPTER 1 Low Temperature Matrix Isolation and Characterization of H2NN • • • • • • • • • • • • • • • • • • • • • • • • • 40 A. R. Synthesis of Carbamoyl Azides • • • • • • • • • • • • • • • • • • • • • • • • Electronic Absorption Spectra of H2NN 3 41 and D2 N'N 3-rl2 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 41 C. D. E. F. G. H. I. J. Thermal necomposition of H2NN 3 • • • • • • • • • • • • • • • • • • • • Photochemical necomposition of H2NN 3 • • • • • • • • • • • • • • • Emission Spectroscopy • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Matrix Isolation Infrared Spectroscopy of H2NN 3 • • • • • • • • H2NN Product A.nalysis • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Matrix FT -IR Studies of Carbon Monoxide • • • • • • • • • • • • • • l?.S R S tucfies • • • · · • • · • · · • · · · · • · · · • · • • · · · · · · · · · • · · · · · • Thermal f'ecomposi tion Kinetics • • • • • • • • • • • • • • • • • • • • • • 46 47 54 56 98 105 116 118 K. Summary · · • · · · · • · • · · · · · · • · · · · · · • · · · • · • · • · · · · · · · · · · · 123 RESULTS ANT) DISCUSSION CHA.PTER 2 A. B. C. D. E. 1,1-Diazenes with a-Hydrogens. Matrix Isolation and Characterization of 1,1-0imethyldiazene and 1, 1-0 iisooropyldiazene • • • • • • • • • • • • • • • • • • • • • • • • • • • • Synthesis of Carbamoyl Azides • • • • • • • • • • • • • • • • • • • • • • • • Electronic Absorption Spectra of 1,1-Dimethyldiazene 7 and 1,1-niisopropyldiazene 18 • • • • • • • • • • • • • · • • · • • • • Thermal necomposition of l, 1-Diazenes 7 and 18 · • • • • • • · Photochemical necomposition of 1,1-0iazenes 7 and 18 • • · Matrix Isolation Infrared Soectroscopy of 1,1T)imethyldiazene 7 · · · · · · · · · • · • · · · • · · · · · · · · · • • • · · · · · F. G. H. I. Matrix Isolation Infrared Spectroscopy of 1,1Diisooropyldiazene 18 • · • • • • • • • • • • • • · • • • • • • • • • • • • • · • 1,1-Dimethyldiazene Product Analysis • • • • • • • • • • • • • • • • • • 1, t-niisooropyldiazene Product Analysis • • • • • • • • • • • • • • • • 1,1-f'\iazene Tautomerization • • • • • • • • • • • • • • • • • • • • • • • • • 126 127 127 137 140 142 173 174 182 185 viii TABLE OF CONTENTS (continued) J. K. L. Thermal necomposition Kinetics • • • • • • • • • • • • • • • • • • • • • • Matrix FT-I'R. Studies of Carbon Monoxide • • • • • • • • • • • • • • .Summary • • • • · • • · • · • • • • • • · • • • • • • · · • • • • • • • • • • • • • • • • • • 188 201 206 EXPE'R.IMENTALSECTION ••••••••••••••••••••••••••••••••••••• 208 REFERENCES AND NOTES • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 239 APPENDIX 253 Cl-fAPTER 3 l5N NMR Spectrum of a 1,1-Diazene. N-(2,2,6,6tetramethyipiperidyl)nitrene • • • • • • • • • • • • • • • • • • • • • • • 257 INT'R..ODlJCTIOt\J • • • • • • • · • · • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 259 RESULTS AND DISCUSSION • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 260 CONCLUSION 264 EXPERIMENTAL SECTION • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 265 REFJ:RF.NCES ANO NOTES • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 268 -1- INTRODUCTION The isomers of H2N2 1, 2, and 3 have been implicated as reactive intermediates in a number of chemical systems such as diimide reduction, hydrazine oxidation and nitrogen fixation. Much of our understanding of the structures and energetics of the 'H'2N'? isomers comes from theoretical efforts! due to the limited experimental2 characterization of these reactive species. H I I N=N.. I N=N \ H H 1 H 2 \+N=N:.. I - H 3a H\ I .. N-N· H 3b "T)iimide" generated in situ has been widely used as a reagent for the stereospecific reduction of multiple bonds)a The accepted mechanism for this transformation requires the cis HNNH isomer 2. Prior to this work, only the trans HNNH isomer 1 had been spectroscopically characterized. 4 Hydrazine (Hi'JNH2) is used in hundred million pound quantities annually as a fuel and as a corrosion inhibitor. The H2N2 isomers represent the formal oxidative intermediates in these hydrazine oxidations. Transition metal dinitrogen complexes and metal complexes of each of the 'H'2''J2 isomers have been prepared. These complexes have received considerable attention as possible models for the discrete steps in the fixation and reduction of dinitrogen to hydrazine and ammonia in the active site of nitrogenase enzymes. '5 -2- Spectroscopic characterization of each of the isomers of H2N2 is essential for direct studies probing their role as reactive intermediates in these and other chemical systems. Until 1978, l,l-diazenes6 3a or aminonitrenes 3b unlike their more stable l,2-dlazene7 isomers (a zo compounds) 1 and 2, were not usually isolated or characterized, but were assumed intermediates based on a substantial body of chemical evidence. The recent preparation of kinetically persistent 1,1diazenes fl., 5, and 6 by Dervan and coworkers have made possible the direct experimental characterization of the l, 1-diazene functional group. 8 These .::.{.+N=N: -A . 4 5 6 studies have provided much information concerning the electronic structure, the kinetics of thermal reactions and the photochemistry of 1, l-diazenes with tertiary alkyl substituents. Aided by the studies of kinetically persistent 1,1-diazenes, the method of low temperature matrix isolation9 is utilized for spectroscopic characterization of the parent l, 1-diazene, H2NN 3 (providing entry to the H2N2 energy surface), and other simple dialkyl kinetic persistence. These studies are described herein. 1,1-diazenes lacking -3- Theoretical Studies Theorists have devoted considerable effort to the H2N2 ener~y surface with regard to the electronic structures, relative energies and mechanisms for interconversion of the H2N2 isorners.1 concernin~ The most important question the electronic structure of H? NN 3 is the nature of its ground electronic state. . Unlike simple nitrenes (R-N .) the chemical behavior of aminonitrenes or 1,1-diazenes su~~ests a singlet ~round electronic state. Earlv ab initio calculations found a triplet nitrene-like ground state 3b for H2N"!\I as found experimentally for simple alkyl nitrenes. However, more recent calculations emplovin~ larger basis sets and confi~uration interaction found the singlet form 3a to be the ground electronic state. Table I. Sin~let- Triplet Energy Gap in H2N-N. Calculational \~ethod Ground State S-T Gap~ STn-3G 4-31G Triplet Triplet Triplet Triplet Singlet 26.3 11.7 5.2 2.1 1.6 1'3. 8 HF SCF 4-3lG-CI GVP,-CI Sin~let Reference Baird, Pople, Ahlrichs, Wagniere, Baird, Goddard, 1973.!.E 1978.!£ 1976.!.b 1973.!.2 19771i 1977~ ~kcal/mot-l, both states at equilibrium geometries. From the recent calculations of Davis and Goddard (GVB-CI), the following picture for the three lowest electronic states of H2NN emerges. The ground sta te is calculated to be a sin~let (1 A t) of olanar C7v geometry - 4- with a low lying triplet excited state (3A2) at 0.6 eV (13.8 kcal). The ground state singlet structure (S 0 ) results from stabilization due to delocalization of the amino nitro~en lone pair to the terminal nitrene nitrogen giving rise to + - substantial NN double bond character (HiN=l\f:). This NN double bond 0 character is reflected in the calculated 'JN bond length Re(N N) = 1.25 A (for 0 comparison the experimental Re(NN) for trans HNNH 1 is 1.'25 A),?i the large dipole moment (ll = 4-.036 D), and the calculated planar C2v geometry. To H~ Hro 50.7 kcal HH. •• Hh_)2 t 13.8 kcal Hro + H Pigure 1. GVI3-CI Calculations for H2NN. the experimentalist, the calculations suggest that H2NN should show an NN double bond stretch near the 1 500 to 1600 cm-1 region of the infrared (vN=l\f for trans HNNH 1 is 1529 cm-l in the Raman))f The first excited -5- singlet state (S t) is calculated to lie 2.2 eV (50.7 kcal/mol) above the ground state. This gap corresponds to an n- 'IT* electronic transition. Important to the experimentalist, the calculations predict that H2NN should be a colored species detectable by electronic spectroscopy with an n- 'IT* transition in the visible region near 560 nm. The S 1 and T 1 excited states have electronic configurations similar to . the lowest singlet and triplet states of nitrene (H-N ·). Interaction of the amino lone pair with the nitrene 1T electrons is equivalent in both spin states so the 37 kcal/mol separation between T 1 and S 1 corresponds to the triplet-singlet gap in nitrene. In the T1 and S1 excited state there are two opposing forces: pyramidalization which reduces antibonding interactions and planarization which gives rise to a favorable two-center three-electron bond. These two effects are calculated to be comparable leading to an optimal pyramidal geometry (21 o angle between the H2N plane and the N-N bond axis) with a calculated barrier of less than 1 kcal/mol (for comparison, the experimental inversion barrier for NH3 is 6 kcal/mol))O The calculated NN 0 0 bond length (1.37 A) for S 1 and T 1 (Re(NN) for hydrazine is 1.4-5 A) and the calculated dipole moment of 2.35 D are indicative of some two-center threeelectron bonding. energy surface has also been investigated theoretically)q,r,s,t,u The recent ab initio calculations of Casewit and Goddard (GVB-CI)1 t are summarized in Figure 2. The trans HNNH isomer 1 is found to be lowest in energy with a calculated heat of formation l1Hf(298) = 56.9 kcal/mol in good agreement with the latest experimental -6- value for the heat of formation of 1, t.Hf(298) =50.7 .:t2kcal/moJ.2mThe cis isomer ?. is calculated to lie 4.7 kcallmol hi~her than 1 and the 1,1diazene isomer 3 is calculated to lie 29.4 kcal/mol higher than 1. The N -H bond strengths were also calculaterl affording N-H bond strengths of 42 ~=N + H· t \ \ I I \ \ \ \ _dl \ I I t 71.4 t 29.4 \ \ \ I H \ \ \ \ + \ N=N/ H/ H N=N \ H F Figure 2. GVF\-CI calculations for the l1?.N2 energy surface and the diazenyl radical. kcallmol for H7NN 3, 71.5 kcallmol for 1, and 46.7 kcal/mol for 2. These calculations suggest that unimolecular interconversion of the H7..N2 isomers through bond homolysis requires surmounting substantial barriers. Various other theoretical treatments for unimolecular cis to trans HNNH isomerization find inversion barriersl a,c, f, h,i,k, m,q,r of 4 5-51 kcal/mol and rotational barriers! a,c,h,m of 55-84 kcal/mol. None of these calculated barriers is consistent with experimental 4.?. kcal/mol activation energy for gas phase -7- Scheme I trans-HNNH + trans-HNNH - trans-HNNH + H• H· + Nz + HzNNH l'..H = -15 kcal (1) H2NNH l'..H = -62.3 kcal (2) cis-HNNH + H2NN H l'..H = 4.7 kcal (3) HNN + H2NNH2 l'..H = -4.6 kcal (4) N2 + H2NNH l'..H = -86 kcal (5) N2 + H· l'..H = -24.1 kcal (6) .,..A' H2NN H + trans-HNNH .........._ - HNN + trans-HNNH HNN decomposition of 1 found by Willis and coworkers.l2 Casewit and Goddard! t have proposed a radical-chain decomposition mechanism for 1 (Scheme I) which finds the experimental barrier to be consistent with the calculated energy gap of 4.7 kcal/mol between cis and trans HNNH. However, a radical chain mechanism, for the reduction of olefins by "diimide" in the gas phase has been ruled out.l2 Correcting Casewit and Goddard's calculated N -H bond strength of H2NN for the difference between an N-H and N-CH3 bond strength ( D(HMeN-H) - D(HMeN-CH3) = 18.5 kcal/mol) 13 provides an estimate for the N-CH3 bond energy in 1,1-dimethyl diazene 7 of 23.5 kcal/mol. Application of this correction to the calculated N -H bond strength for trans HNNH 1 results in an estimate for the N-C bond strength in trans-1,2-dimethyl -~- Me \+ Me I • N=N -••~ N=N + Me· ...,..__ N=N I I I Me Me Me 7 8 diazene 8 of 52.9 kcal/mol. This estimate is in excellent agreement with the experimental value of 52.5 kcal/moll4 for the thermal decomposition of 8 in the gas phase. 4artree-'Fock (l-fF) calculations of 'Pasto and Chipmanlr result in a 40 kcal/mol N-'1-l bond strength for Hi'.JN 3 and a 26 kcal/mol C-1\J bond strength for 1, t-dimethyldiazene 7 in good agreement with Casewit and Goddard. t t 0ne must be cautious in applying GV'B-CI results to 1., 1-diazene thermochemical estimates as the 18.5 kcal/mol correction may not l:>e H \+ I N=N H • ---•~ N=N 1 + H· ...~-- H 3 1 -9- directly applicable to J,1-ciialkyl diazenes. + Qualitatively, the N-H bond - energy is T-J2N=N is lower than the N -H bond energy in trans HN=NH because + et - cleavage in H2N=N results in the formation of a full N=N double bond (HN=N·) without the need for charge separation. broken, substantial 7T -bonding is gained. Although an N-H bond is However, in T-IN=NT-1, a full N=N double bond already exists so only two-center three-electron bonding is gained upon breaking an N-H bond. Generation and Chemistry of 1,1-Diazenes In the last several decades, the chemistry of presumed 1, 1-diazene intermediates has been extensively investigated.6 There are numerous methods reported for generating the presumed 1, 1-diazene intermediate (Figure 3). The most versatile method is oxidation of 1,1-disubstituted hydrazines with oxidants such as tert-butyl hypochloritel5 and nickel peroxide.l6 Other methods include the reduction of N-nitrosamines,17 base induced thermolysis of 1,1-disubstituted sulfonyl hydrazines18 and difluoramine addition to secondary amines.l9 Other attractive routes involve the thermal or photochemical cleavage of N-amino sulfoximines.20 The intermediacy of l,1-diazenes in these reactions is supported from the resulting products of their thermal decomposition. The thermal reactions of 1,1-diazenes can be divided into three general classes: (a) fragmentations, (b) isomerizations, and (c) bimolecular reactions (Figure f.!.). The actual reaction pathway is dependent upon the mode of generation, structure, degree of substitution, and reaction conditions. In general, the fragmentation of 1, l- Fi~ure 3. Ior A +base .... fase t O R2 N-H Methods for ~eneration of t, t-ciiazenes. R2N=N-SMe 2 ~or hv R2 N-N NaN203 R2 N-H ~NF2 x- R2N-NHS02R r.y + R2 N=N -4 R2 N-NO (R 2 i!.i=N~cu3CI3 21 base I) H+ ~ R2N-NH2 + R2N=NH I I 0 - -11- R' R 'N-N:f I .,, I I R' II N-N~R'" I ~ R R-N=N-R -•.-- \ R" R' \+ \ I I N=N ...~.,._.....,.~ R N-N _ _..,.~ Hydrocarbon R I \oMSO \R, \ 9 N=N-SMe 2 R-N=N-R' I R Figure 4. 'R.eactions of I ,1-diazenes. diazenes is in accord with a mechanism involving radical or biradical formation with loss of nitrogen. The presence of radical stabilizing sub- stituents :nay favor fragmentation. pyrrolidylhydrazinel8c at Pyrolysis of N-sulfonyl-2,5-dimethyl- 4000C yields 1,2-dimethyl cyclo- -12- -Q- c( 1 HNS~ butanes, propene and nitro~en. Oxidation of 1,1-dibenzy1 hydrazine with mercuric oxide at 25oc yields dibenzyl and nitrogen. 21 The question of whether symmetrical 1, 1-diazenes decompose by sequential one bond or simultaneous two bond cleava~e remains unanswered.2'2 R· NsN R• .. R \+ N=N: .. I R R• ... I N=N" R Concerted fragmentation of 1,1-diazenes may occur in a number of cases. Por example, treatment of cis and trans 2,3-dimethylaziridines with difluoramine results in apoarent stereospecific extrusion of nitrogen from the -13- ~+ ) yN=N - 5 proposed 1, 1-diazene intermediates. 23 Recently Duan and Dervan have demonstrated a stereospecific 2,1-elimination pathway for 1,1-diazenes with B-hydrogens.'-4 JQ H:x-N: :. - C CH3 - ...,~: : D H Isomerization of 1,1-diazenes consist mainly of two reaction types: (a) rearrangement to hydrazones and (b) isomerization to 1,2-diazenes. -14- Hydrazone formation is a general reaction for a-hydrogens generated by several different methods. 6 1,1-diazenes with A pro tic solvent is Labeling studies have generally necessary for hydrazone formation. demonstrated that the nitrene nitrogen ends up as the doubly bonded nitrogen in the prorluct hydrazone.25 Cyclic I, 1-diazenes have also been shown to ring expand to isomeric hydrazones. \+ I N=N .. Although this rearrangement has been the H II OH _......c......._ c CHR2 10 I / R ~CR 2 -NH w II _......N~ R • CHR2 ~ I H 9 0 R \ +N-N 7 II ... N-N I \ / ~· ~CR R 2 4 ... NH II _......N~ R -cR 2 . . 'NH I _......N........_ • CR2 R -15 - subject of T)iaziridines extensive and research, 1,2-diazenes the have mechanism been ruled remains out uncertain. as possible intermediates.26 P>oth of these species have been shown to survive the reaction conditions necessary to generate hydrazones. To date the reaction is considered to proceed through t he intermediacy of an azomethinimine6 9 (analogous to the enol form of a ketone). The function of the protic solvent may be to facilitate the tautomerization of the 1,1-diazene. "R.earran~ement of 1, l-diazenes to 1,2-diazenes has been shown to occur when the product 1,2-diazene N=N functionality is incorporated into an aromatic system.2 7 1', l-Oiazene precursors have also been proposed to give rise to 1,2-diazenes as the result of a concerted 2,3-sigmatropic rearrangement of the intermediate 1, l-diazenes.28 ( 0} . (tjJl] 1'1imerization of 1,1-diazenes results in the formation of 2-tetrazenes. The formation of tetrazenes is a general reaction for l, 1-diazenes. 6 However, the mechanism may vary depending on the reaction conditions and mode of generation of the 1,1-diazene. In only a few cases is the mechanism -16- known.l),8 Neutralization of solutions of 1, 1-dialkyl diazenium ions leads to the formation of a colored species with the ultimate reaction product being the corresponding 2-tetrazene.29 The tetrazene product probably results, at least in part, from the reaction of a 1, l-dialkyl diazene with a diazenium ion giving a protonated t etraze ne which t hen deprot onat e s. Rees and coworkers30 + R2N=NH ----•~ TetH+ Tet Tet = R2 N-N=N-NR 2 isolated a tetrazane ll from the lead tetraacetate oxidation of N-aminophthalimide, which upon further oxidation yielded the corresponding tetrazene 12. T)irect bimolecular dimerization of kinetically persistent l, 1-diazenes to form tetrazenes has been demonstrated. 8 [o] R2 N-NH2 ~ + R2 N=N -4 ... R2 N-N ~ RzN-NH2 [o] R2 N-N=N-NR 2 12 4 R2N-N-N-NRt H H ' ' 11 - t7- l,l-f)iazenes substituterl with electron withdrawing groups have been shown to ado to lT -bonds and to sulfoxides, 6 su~gesting that these additions are electrophilic in nature. Addition of diacvl 1, 1-diazenes to olefins is greater than 9 5% stereospecific even at low alkene concentrations. '31 By analogy to carhene chemistry, this has ryeen taken as evidence for a singlet 1, 1-diazene ground state. Simple dialkyl 1, l-diazenes have not been shown to be trapped by alkenes or sulfoxirles. Kinetically Persistent 1,1-Diazenes Recently, direct studies of the 1,1-diazene functional group became available with the preparation of kinetically persistent 1,1-diazenes 4 and 5 by Hinsberg, Schultz, and Oervan.8f Additionally, the acyclic 1,1-diazene 6 was prepared by Macintyre and Dervan. 8g All three persistent 1, 1-diazenes are orepared from the corresponding 1, t-disubstituted hydrazines by oxidation ~.N=N: _J .. f~ 4 5 6 -U~- with tert-butyl hypochlorite in low temperature solution (-78 to -1200C) in the presence of triethylamine. The kinetic oersistence of these diazenes is due to the tetramethyl steric blockade which slows dimerization to the corresponding tetrazenes, the absence of a hydrogens which precludes any tautomerization/hydrazone rearrangement chemistry and their preparation at low temperatures, where nitrogen extrusion is prevented. As predicted from theoretical studies, these persistent 1, 1-diazenes are colored species, giving purple (4) and red (5 and 6) solutions. Absorption and emission spectra are shown in Figures 5, 6, and 7. The absorptions are n-n * transitions, which show an expected blue shift in the more polar sol vent isopropanol. The extinction coefficients are 20 .:t. 3. structure the in absorption curves The vibrational correspond to the N-N vibrational /\ I \ I I I I I I I I \ \ \ ~ \ ·;; . c \ £ \ \ I \ \ I \' I 420 Figure 5. 4~ 480 510 540 570 ', 7~ 780 810 840 Absorption f - ) and fluorescence (---) spectra of 4 in CFCl3 at -780C and -l960C, respectively. -19- r\ II \ I \ " I\ I \ I I I \\ \ I \./ I I 44 0 470 500 ~30 \ ~ • c ·e.. 0 ·;; ' ~ 560 ·~ \ I 4 10 ... \ 680 \ ../ \ 710 740 "' \ 770 "' BOO Wawelenoth ( nm I Figure 6. Absorption ( - ) and fluorescence (---) spectrum of 5 in CFCI3 at -noc and -196°C, respectively. ,.,.--·-·........ ' ·'\ ,,·~· I ; .' I I I i i ; i i ,I i .i iI 1100 \ \'. '. '. '. \ \ ... \ \ i \ i '\ \ I"£ \\ "' '\ 100 ~l•ml Figure 7. Absorotion spectrum of 6 at -125°C in Me20 ( - ) and nPrOH/Me20 (--). Fluorescence spectrum of 6 at -196°C in Me~O/Cn 2 ci 2 (-·-·-). -20- frequencies of the respective excited S1 states. The average vibrational spacin~s from these absorption curves are 1040 to 1250 cm-1. Weak emission occurs from S1 with a fluorescence quantum vield of 7 x 10-3 for 5.~h The lifetime of the excited state CS1 is about 20 nanoseconds with internal conversion being the primary decay rnode.8h No other electronic transitions (E > 50 ,_, -1 cm-1) were detected for 5. Infrared absorptions are observed at 159 5 cm-1 for the 6-membered ring diazene and at 1638 cm-1 for the 5-membered ring diazene in dichloromethane solution)~£ In order to characterize these infrared bands as the respective N-N double bond stretches, the diazenes were prepared with a terminal 15N label. As expected, the infrared bands were found to shift to (a) I I I I ,, I~ ... I II I 11 f I I1 I I I I I I II ,.., t'• : '\ ~ 1 ·, I 'I I1 \"' 1 ~ ..., (b) Itl,,. II ~ •I I ,, I II I / \ ,' I 1 M' 11 1~ 1 ~ I I I I I I I ~ l I I I I I I I II ,, 1638_/ 1800 1600 Frequency ( cm-1) Fi~ure 8. Infrared spectra of 5 (a) R214N=14N and (b) R 2 14N=15N at -780C (-and at -780C after warming to 250C ( ---). -?.1- lower energy as calculated from Hooke's law (Figure 8). The similarity of these infrared transitions to the 1576 cm-1 Raman active N=N double bond stretch for trans-l, ?-diazenes is indicative of substantial N=N double bond character in the 1, 1-diazenes. infrored (cm- 1) (nm) electronic 14 Q·- N=~ : cJ·- N=N : 14 14 15 >..ma• ( i PrOH) N=N 543 526 1595 1568 497 487 1638 1612 >-max (CH2CI2l N=N Warming the purple-red 1, l-diazene solutions results in decolorization and formation of rfecomposition products shown in Figures 9 and 10. ~imolecular decomposition (k2) is a low activation energy process (Ea 0. 9 kcal/ mol, log A = 3.8 .±. 0. 7 temperatures (< - 300C for 4 for 4 ), which and .5). At = 6.4 .±. predominates at low warmer temperatures the unimolecular nitrogen extrusion pathway (kl) dominates. The Arrhenius parameters determined for k1 in ether solution are Ea = 20.0 .±. 0.4 kcallmol and log A = 1 '3. 7 .±. 0.3 for diazene 4, 8f I:: a = 19.0 .±. 0.6 ancf log A = 12.4 .±. 0.4 for diazene .5, 8f and F. a= 13 kcal/mol for an assumed log A = 13 for diazene -22- Figure 9. TJnimolecular anrl bimolecular decompositions for 1,1-diazenes4 and 5. -23- .::{.N=N:- )= 7\6 Figure 10. Unimolecular decomoosition products for 1,1-diazene 6. The direct and triolet sensitizer:! photolyses of 1,1-diazene 5 with visible light yield the tetrazene and hydrocarbon products in 4:1 and 9:1 ratios, respectively.8h The high yields of tetrazene from photolysis have been shown to be the result of bimolecular reaction of the nitrene-like excited state S1,T1 with the ground state 1,1-diazene S 0 • A spin correlation effect has been demonstrated in the products resulting from direct and triplet sensitized photolysis of d,~-14.8h To date attempts to locate or directlv observe T1 of a 1,1-diazene have been unsuccessful. summarizes the excited state parameters for 1,1-diazene 5.8h Figure 11 -24- Products E fSJ) E (T1) 47-50 kcal (C:l-i?CI2) < 31 kcal k1= 3 x to5 sec-1 kyc 2 x 10~ sec-! (-780C) ko kDJM + kN2 '<N2 '3 x to'> sec-1 (-78C) ki)J\1 8 x 107 mot-! sec-1 (-78C) kt<;c << kN or J'1n4 sec-1 ' 2 Figure 11. Excited state parameters for kinetically persistent I, 1-diazene 5. ,,~ + N=N 14 -25- t I N=N.. .. I N=N \ H H 1 H 2 3 H2N2 Spectroscopic Characterization Numerous methods have been employed to generate H2N2 as a reducing agent for organic synthesis or for spectroscopic characterization and study) Among the more useful modes of generation for spectroscopic study are (a) microwave (or Tesla coil) discharge of hydrazine vapor,2f (b) (a ) H2N-NH 2 dischcrge ~ H2 ,N 2 ,NH 3 + M (b) H N-N/ 2 \ H2N2 Tos ( c ) H-N3 hv + ./ H-N· + N2H4 , NH 4 N3 -?6- oyrolysis of metal tosyl hydrazide salts?.g and (c) dimerization of nitrene . ('f-L1\.!·) oroducerl from photolysis of hydrazoic acid)c l-l;;>N?. has been demonstrated to be extremely reactive with a lifetime of several minutes at room temoerature (gas phase).?.e Thermal rlecomoosition of H2"J2 occurs raoirlly near liquid nitrogen temoeratures fneat).?g l-1i'17 is also efficiently decomposed ohotoche'llicallv either in condensed phase or as a vapor:Zd,1'2 All of the ab0ve methods for generation of Hi'J?. might be expected to initially produce a mixture of isomers 1, 2, anrl 3. Yowever, rlue to the high reactivitv of these soecies and the presence of su~stantial amounts of l!vrlrogen bonding sirle products (e.f:',., NH3, N?l-14) in these modes of generation, fir'-n exoerimental characterization is available only for the trans Table fl. ~nectral Characteristics H2N2 Isomers. .. H I I N=N.. I H H electronic (T IV /VI'S) Xmax fnm) 3% inf rarerl fiR) fcrn-l) 31'31 Raman fcm- t) 3tn 15~3 15?9 ~Tentative assignments. N=N \ H Reference ..,~ J (3074)~ ?.c, 7.f 2f -27- HNNH isomer 1 (Table II)),33 Various other ambiguous assignments have been claimed for the other H2N2 isomers under non-matrix isolation conditions only to l)e later reinterpreted as hydrogen bonded complexes of 1 and side products.2 Conclusive assignments have resulted from careful matrix isolation studies, the use of heavy isotopes (D, 15N)2c,2f to aid in the assignment of vibrational transitions, and observations of thermal and photochemical reactivity. Trans l-fNNH 1 has been characterized as a yellow species showing a non-structured absorption, A.max = 386 nm (neat, 770K).2j structured absorption, A. 320-440 nm (e: ~ In contrast, a 4 at 365 nm), is observed for a mixture of 1 (15%) in a large excess of ammonia in the gas phase.34 In 1964, Rosengren and Pimentel2c identified two infrared transitions for trans HNNH at 3131 cm-1 (weak, NH asymmetric stretch) and 1286 cm-l (very strong, NNH asymmetric bend), produced from photolysis of hydrazoic acid in a nitrogen matrix at 20 K. In addition, heavy isotopes (D, 15N) were incorporated to verify these assi?;nments. Tentative assi~nments were made for two cis HNNH bands at 3074 cm-1 (weak, NT-i stretch) and 1279 cm-1 (strong). In 1973, Bondybey and Nibler2f reported the infrared and Raman spectra of isotopically labeled 1 produced from the microwave or T esla coil discharge of hydrazine, followed by deposition in a nitrogen matrix at 12 K. Their infrared assignments were consistent with those of 'R.osengren and Pimentel, resulting in a conclusive vibrational characterization of 1. None of the tentative cis HNNH infrared bands were located. The characteristic N=N double bond stretch for 1 was located in the 'Raman at 1. 529 cm-1. However, -28- the 15'\j shifted multiple for this mode was not located. Additional Raman bands for 1 were found at 3128 cm-1 (weak, NH symmetric stretch) and 1583 cm-1 (very weak, NNH symmetric bend). Analysis of the gas phase rotational structure of the high resolution infrared spectrum2i of 1 has provided the following geometrical parameters: 0 0 Re(NN) = 1.25 A, Re(NH) = 1.03 A, < HNN = 106.90. These parameters are consistent with the established singlet ground state of planar geometry for trans HNNH. '2i Trans HNNH 1 has also been characterized by photoelectron spectroscopy,?!< vacuum UV soectroscopy,2o and by mass spectroscopy.2n Foner and Hudson?.m have recently reported an experimental value for the heat of formation of 1 ( l1Hf(298) = 50.7.:!:. 2 kcal/moD from a combination of ionization potential and appearance potential measurements. All of these studies indicate that H2N2 exists predominantly as the trans isomer 1. Wiberg has reported2g that pyrolysis of solid metal tosyl hydrazide salts followed by traoping of the less volatile side products ('\IH3, N2H4, -noc) as means of preparing clean 1 by condensation onto a cold finger at 77° K. According to Wiberg, a mixture of 1, 2, and possibly 3, is obtained from pyrolysis of metal tosyl hydrazides (M = Li, Na, T<, Rb) followed by condensation of the products. When M = Cs, however, Wiberg reports2i that an isomer "isodiazene" (3) is ol:>tained. This "isomer" is reported to be 13 kcal/mol higher in energy than 1 by appearance potential measurement in the gas phase. In addition, this "isomer" is reported to l:>e a colorless species, A. max = '260 nm (770K). A yellow addition product (!.max = 398 nm) forms -29- /M H2N-N"'-. Tos Amox = 386 nm M= Li, No, K Amox = 260 nm quickly with ammonia (77°K). slowly with 1 at 110°K. The same yellow addition product is formed No further characterization for this "isomer" has been reported. Wiberg's assignment for "isodiazene" seems suspect and not in accord with ab initio (GVB-CI) calculations! t for the H2N2 energy surface which predict 1,1-diazene 3 to be 29 kcal/mol higher in energy than 1. The UV transition P.·max = 260 nm) observed by Wiberg, is also inconsistent with the theoretical (GVB-CI) predictionln of an n-n* transition for 3 near 560 nm in the visible region and the observed n-n* (A.max = 500-5/tO nm) for kinetically persistent 1,1-dialkyl diazenes.8 The parent 2-tetrazene 25 HttN4, which would be expected to result from dimerization of 1,1-diazene 3 has been reported as a meta-stable species (A. max = 263 nm))5 The gas phase thermal and photochemical decompositions of HNNH 1 in the presence and absence of unsaturated compounds (e.g., alkenes, dienes) have been studied by Willis et al.2e,h,32,31t,36 Thermal and photochemical decompositions of neat and matrix isolated 1 have been noted by others.2 -30- The result of gas phase kinetic studies of the thermal decomposition of 1 (generated by microwave decomposition of hydrazine) as a dilute gas mixture (approximately 10-15% l-I2N2 in excess ammonia) with alkenes has been interpreted in terms of the following complex reaction (Scheme II))4,36 The thermal decomposition of 1 (monitored spectrophotometrically), for the most part, follows first order decay kinetics, excluding a fast initial nonexponential decay which is presumably due to self heating and surface effects. The rate determining step is assumed to be the unimolecular isomerization of trans 1 to cis 2 (eq. l), although theory predicts the barrier for this unimolecular transformation to be quite high.l Cis 2 is then assumed to be the species responsible for further reduction by the accepted concerted SchemeD (1) (2a) (2b) (3) (4) (5) adduct + olefin ---- product + olefin + N2 (6) (7) -31- Woodward-Hoffman allowed cyclic transition state to give hydrazine (disproportionation) (eq. 2a) or reduction of added alkene (eq. 3). Pasto and Chipmanlr have calculated (4-31G) 6Hfforthisconcertedreaction(eq.3)to be 26.7 kcal/mol. In addition, Pastols calculates (4-31G) 6 H :f for the disproportionation step (eq. 2a) to be 23.8 kcal/mol. Thus, both the unimolecular isomerization of 1 to 2 and reduction of alkenes by 2 are predicted by theory to have rather substantial barriers which are apparently not found experimentally. However, the measured first order decay of 1 is observed to be faster in the presence of added hydrazine or water and is dependent upon the structure of added alkene)4,36 The self reactions of H2N2 (disproportionation) are some 4-8 times faster than alkene addition (reduction) at 1oooc in the gas phase. The first order decay of 1 as a function of temperature in the presence of added 1,3-butadiene gives Arrhenius activation parameters (log A = 0.255, Ea = 5.70 kcal/mol)36 which are different from the activation parameters for the decay of 1 in the absence of added alkene (log A= 0.477, Ea = 4.20 kcal/mol))4,36 In addition, the measured first order decay shows a deuterium kinetic isotope effect. Thermal decomposition of D2N2 1-d2 in the absence of alkene gives Arrhenius activation parameters log A= 0.301 and Ea = 4.40 kcal/mol)4,38 The gas phase thermal decomposition of 1 is not affected by added 02 and no conclusive evidence for a radical chain mechanism for the thermal decomposition of 1 has been noted. The stereospecificity (syn reduction) of alkenes by N2D2 was not demonstrated in this work)4,36 Willis2e,34,36 finds only H2, N2, and hydrazine as thermal products of -32- 1 with no evirlence for formation of ammonia or ammonium azide. Wiberg2g, 37 reoorts that in addition to H2, N2, and hydrazine, ammonia and ammonium azide are also produced from thermal decomposition of neat 1 prepared from the pyrolysis of metal tosyl hydrazides. Wiberg has also noted a ve ry rapid reaction of 1 with 02 yielding HOOH and N2.2g 25 The ohotochemical decomoosition of 1 in the presence and absence of added alkenes has also been reported by Willis.32 Photochemically, 1 yields H2 and N2 as products (>97%). The quantum yield for loss of 1 in the gas phase ranges from 13-20. Apoarent quantum yields for alkene reduction are approximately 7. The presence of 02 suppresses alkene reduction supporting the interpretation of a radical chain mechanism for the photochemical decomposition of 1 and resulting alkene reduction. radical f·I\!2H) is implicated as a chain carrier. The parent diazenyl The rate constant for hornolvsis of diazenyl radical is estimated to be 3 x 1o3 sec 1 at room temperature. Assuming a pre-exponential factor of 1013, the activation harrier for diazenyl radical homolysis is estimated to be approximately 13 kcallmol. The lifetime of the n-iT* excited state of 1 is estimated to be 5 x -33- 10-12 sec. No evidence for photochemical trans 1 to cis 2 isomerization has been reported.?,32 No evidence for collisional deactivation of the excited state of 1 is noted and dissociation of the excited state of 1 to ·N2H + H· or N2 + 2H· is concluded to have a quantum yield near unity, although dissociation directly to N2 + H2 cannot be excluded.32 Primary photo- dissociation of 1 may occurs from high vibrational levels of the ground state reached by rapid internal conversion or intersystem crossing from the excited state.32 N2H2:f ~ ·N2T-i + H• 6H = 71.5 kcallmol N2H2:f ~ 6H = 4-7 kcal/mol N2H2:f - N2 + 2H· N2 + H2 6H = -57 kcallmol Experimental Strategy From the direct studies of kinetically persistent 1, 1-dialkyl diazenes8 several factors are apparent concerning the generation and direct spectroscopic characterization of the parent 1,1-diazene H2NN and its simple alkyl derivatives (e.~., 'R. = CH3, Ci!-i6, etc.). nue to the lack of a steric blockade to dimerization, H2NN 3 and its simple dialkyl derivatives are likely to dimerize to their corresponding 2-tetrazenes at near diffusion controlled rates. Although 3 and its simple dialkyl derivatives should be more stable to unimolecular nitrogen extrusion than the tertiary substituted -3/t- persistent 1, l-diazenes, the use of very low temperatures may be necessary to prevent other possihle decomposition isomerization). f)irect spectroscopic pathways (e.g., tautomerization, characterization is likely to require the use of low temperature matrix isolation techniques. The generation of highly reactive species (carbenes, nitrenes, radicals, etc.) isolated in a non-reactive material (nohle gases, organic glasses, polymers, etc.) has been a successful method for the spectroscopic characterization fUV-VIS, IR, Raman, ESR) of reactive intermediates.9 Low temperature matrix isolation in a solid inert gas matrix is generally the method of choice for preventing diffusion and resulting intermolecular chemistry of reactive species. As a result of the low temperatures (1 0-300K) necessary to prepare a rigid inert gas matrix (Ar, Kr, Xe, N2) the activation energy for thermal reactions is not available to the isolated species. The spectroscopic transparency of these inert host gases facilitates extensive characterization of reactive intermediates using routine spectroscopic probes. Among the techniques commonly used for the generation and matrix isolation of reactive intermediates are (a) photolysis of a matrix isolated photochemical precursor, (b) pyrolysis or microwave discharge decomposition of a thermal precursor in the gas ohase followed by co-deposition with matrix host gas, and (c) gas phase chemical reaction (crossed beam) of two volatile species followed hy co-deposition with host matrix gas onto a cold spectrescopic window. Photolysis of a suitable matrix isolated photochemical precursor is probablv the most widely utilized technique. A common feature -35- of the photochemical routes is the loss of a small stable molecule (e.g., N2, CO, C02) which can diffuse away and become a part of the matrix, leaving a reactive nitrene, carbene, radical or highly strained molecule as an isolated species. The chemically inert noble gases (Ar, Kr, Xe) are considered to form rigid matrices at temperatures below 30% of their melting points (Tm). In the temperature regime of 100 to 250K, even small molecules (two to three atoms) can be effectively trapped in an argon matrix. Prior to this work no general method for the photochemical generation of 1, 1-diazenes had been reported. However, a number of potential photo- chemical precursors to 3 and its simple dialkyl derivatives suitable for matrix isolation studies were given consideration. These a·r e briefly described below. Photoextrusion of nitrogen from an aminoazide might provide a general means of generating 1,1-diazenes in a low temperature matrix. However, the only successful synthesis of an aminoazide reported is the preparation of dimethylaminoazide.39 Dimethylaminoazide has been found to be very unstable thermally. Oxidation of 1,1-dialkyl hydrazines with nitroso benzene yields (Z)-3,3- -36- dimethyl-1-phenyl triazene-1-oxide ad ducts. 40 1-Phenyl-2-phthali mido-diazene1-oxide products has been shown consistent with to photochemically initial formation cleave of in solution to N-phthalimidonitrene (phthalimido-1,1-diazene) and nitroso benzene.40 The N-phthalimido nitrene so generated has been trapped as an adduct with olefins, dimethylsulfoxide and as its dimer the 2-tetrazene. Ultraviolet photolysis of a matrix isolated (Z)-3,3-dimethyl-1-phenyltriazene-1-oxide 16 (yellow, A.max = 326 nm, e: = 8700, A. = 227 nm, e: = 8200) might be expected to produce a 1, 1-dialkyl diazene (purple, A.max = 500-600 nm, e: = 20) and nitroso benzene (blue, A.max = 780 nm, e: = 45). With the use of appropriate visible light filters to prevent the anticipated photochemical decomposition of the colored products, the extinction coefficient of a 1, 1-dialkyl diazene product could be determined by comparison with the visible absorption of the nitroso benzene co-product. hv > Ph NO purple R = CH3 16 blue -37- Thermolysis of some N-amino aziridines yields products consistent with the formation of an olefin and 1,1-diazene.41 Vacuum-UV irradiation (J?4 nm, 147 nm) of the parent aziridine, ethylenimine, in the gas phase produces ethylene and nitrene with a quantum yield of 0.2 to 0.3. 4? The reverse thermal reaction has been demonstraterl in an an~on matrix. 43 Aromatic suhstituents on the aziridine rin~ might provide a chromophore , hv suitable for effecting this transformation with ultraviolet light. In order to generate a 1, 1-diazene by this method, the rate of nitrene like S 1 relaxation to the grouncl state So ( 1, 1-diazene) via fluorescence or internal conversion must exceed the rate of S 1 addition to the olefinic fragment produced. Results from this laboratorv with N-(2,2, 5,5-tetramethylpyrolidyl) nitrene 5 in solution have shown that addition to activated olefins is not a significant pathway upon direct irradiation of l,l-diazenes.8g In solution, photolysis of carbamoyl azides yields products consistent witll the intermediacy of aminoisocyanates, 44 which readily dimerize or are trapped with alcohols. 4 5 The resulting dimers of aminoisocyanates can be -38- Jl R2 N-N H OR photolyzed in alcohol solution to products consistent with the intermediacy of the aminoisocyanate monomer. 46 The photo-Curtius rearrangement of carbamoyl azides has been shown to be sufficiently facile to occur under matrix isolation conditions at low temperatures (60K, neon).47 The intermediacy of aminoisocyanates is supported by matrix isolation infrared spectroscopy with a characteristic infrared absorption at '2230 cm-1 attributed to the aminoisocyanate.4 7 Ultraviolet (UV) photolysis of a matrix 13 I.R. 2230 cm" 1 -39- isolated aminoisocyanate, produced by the photo-Curtius rearrangement of a carbamoyl azide, mi~ht be expected to produce an aminonitrene (1, l-diazene) and carbon monoxide (CO) just as photolysis of aryl and alkyl isocyanates produces nitrenes and carbon monoxide.6a,48 ? • co Hi'JN and its simple dialkyl derivatives are expected to be colored species with n-7r * transitions in the visible region. Photochemical generation of t, 1-diazenes in a rigid matrix is likely to require the use of visible light filters to prevent position of the 1,1-diazene product. subsequent photochemical decom- The generation of H2NN and simple dialkyl derivatives by the photo-Curtius rearrangement/photodecarbonylation of carbamoyl azides is described. -40- RESULTS AND DISCUSSION CHAPTER 1 Low Temperature Matrix Isolation and Characterization of H2NN -4-1- Synthesis of Carbamoyl Azides 05) Carbamoyl azide 15 was prepared carbazide hydrochloride with sodium by treatment nitrite.4-9 The of semi- resulting stable crystalline solid was purified by recrystallization and sublimation. Deuterated NoN02 15 0 0 II + D2 N~N=N=N= .. .. II + H2 N~N=N=N: .. .. 15 15 05-d2) was prepared by exl-)austive fi,T) exchange in excess Electronic Absorption Spectrum of H2NN 3 and D2NN 3-d2 For electronic absorption spectroscopy of H2l\!"1 3, a solution of carbomoyl azide 15 in dry, degassed 2-methyltetrahydrofuran (2-MTHF) was loaded into a 1.0 em oath length low temperature spectroscopic cell50 attached to the low temperature matrix isolation apparatus. The solution was loaded under a positive pressure of argon through Teflon tubing with syringe suction. Cooling the solution to 800K results in the formation of a rigid optically transparent glass. Irrarliation of 15 in a 2-MTI-JF glass at 8001(. -42- with ultraviolet (l]V) liJ:?;ht from a 1000 watt xenon lamp 51 through two Corning (uv transmitting, visible (VIS) A= 400-680 nm cut-out filters CS-7-54 filters ~ results in the loss of the UV transition due to 15 and formation of a blue-violet glass. Electronic absorption spectroscopy of this blue violet glass (~OOK) reveals a structured absorbance curve (Figure 12) in the visible Amax = \ \ \ '' ' ' \ / ,, '/ ,_,.,I I /-, \ ,\ I '"" I \ \ \ \ \ \ ' ..... _.... I.&J u z '' \ \ <! \ \ aJ a: 0 (/) \ aJ \ <! \ \ 500 600 700 WAVELENGTH (nm) Fi~re 12. Electronic absorption spectrurn of Hi'-JN 3 in ?-MTHF glass 80°K (-)and 1:1 2-MTHF/nPrCN glass 8f)OK (---). 636 The nm. spectrum reveals 695, 1)36, '587, 545, and 509 nm. fairly well resolved bands at The maximum absorption at 6'36 nm (4 _5 kcal/mon corresponds to the vertical transition for H7.NN 3. This is close to the n-lT* electronic transition of 560 nm (50.7 kcal/mon calculated by f)avis and Goddardln for HzNN 3. The first absorbance band at 695 nm (41 kcal/mon is tentatively assigned to the adiabatic transition A. 0,0. '52a This structured transition (A. max= 636 nm) is comparable to the structured transitions (Xmax = 497 to 543 nm) observed8 for kinetically persistent 1, 1-diazenes 4, 5, and 6. The average vibrational spacing between absorbance maxima in Figure 12 of 1315 cm-1 corresponds to the average vibrational spacing of the N-N stretch in the excited St state of 3. The orominence of the N-N stretching mode is indicative of a substantial change in the Re ("1-N) on excitation from So to S1. Using the ratio of force constants predicted by Davis and Goddardl n for the N -N stretches of So and St of HzNN 3, v(N-N) in S1 should be 1357 cm-1)2b The smaller experimental average vibrational spacing of 1315 cm-1 possibly suggests that pyramidalization may be more important than 2-center three electron bonding in stabilizing S 1 of 3. J-iowever, the separation of the first two absorbance maxima in the spectrum of 3 at 69 5 and 636 nm corresponds to a vibrational frequency for v(Nl\1) of St of 1335 cm-1 closer to the calculated stretching frequency. r:or comparison, the average vibrational spacing observed for kinetic::~.ll v persistent 1, 1-diazenes 4, 5, and 6 are 1040 to 12 50 cm-1.8 The position of an n-lT* transition involving non-bonding electrons is -44- Table Ill. Comparison 1, 1-11iazene Electronic Transitions. f-0,0 v (nm) (cm-l) F.(So-S 1) (max) (kcal/mol) 616 1315 45.1 540 1357~ 50.7 l,t-niazene a (Theory, GVP,-CI)~ Q + - .. N=N: b .5 43 610 1040 52.7 b 506 620 1200 56.6 b 497 572 1238 57.6 ~2-MTYF ~lass, 800K • .Q~~e20 solution. Refs. 8f,~. 9~. ef. 1n. QR.ef. 52b. sensitive to solvent polarity and hydrog~n bonding effects.53 Using a more polar glass forming solvent would be expected to shift the structured absorbance curve of 3 to higher energy. Irradiation of 15 in a rigid optically transparent glass consisting of a 1:1 mixture of ?-MTHF and n-butyronitrile (nPrCI\!) shifts the structured absorbance curve to higher energy affording Amax = 624 nm and A 0,0 = 681 nm (Figure 12). This blue shift of A0,0 14 nm to higher energy suggests that thermally equilibrated ground and excited states must be differentially solvated, consistent with differing degrees of electron delocalization of So and <; l· n-1T* transition of persistent 1, l-diazene For comparison, the 5 shifts from A 0,0 = 565 nm in Me20 to A0,0 = 557. nm in isopropyl alcohol solution.8e Irradiation of 15-d2 in a 2-MTHF glass at 800K affords a structured absorbance curve for D2NN 3-d2 similar to that obtained for 3 (Figure 13). The average vibrational spacing for 3-d2 of 1305 cm-1 is slightly smaller than the average vibrational spacing observed for 3. Table IV. l,l-T1iazene 3 Electronic Transitions. 1, l-T)iazene Solven~ Amax (nm) A0,0 (nm) F.. (0,0) (kcal/mol) (S 1) 3 2-MTHF 6% 695 41.2 1315 3 2-MTHF /nPrCN 624 681 42.0 1315 3-d2 2-MTHF 636 695 41.2 1305 ~~woK, glass. -46- 1 w 0 z ~ m 0:: 0 C/) m ~ 0 .0 700 600 500 WAVELENGTH (nm) Figure 13. t:Jectronic soectru:n of n2'\J i\l 3-d2 P-i\HHF, ~OO!<). Thermal necomposition of H2"JN Warming the ?-\HHF glass of 3 from 800K to the ooint at wliich the ?-~-HHF glass softens f J'900K) results in loss of the structured absorbance curve and the blue-violet color due to 3. As the structured absorbance <iue to 3 rliminishes, three A.nax = ?6n nm, new absorl')ances are ohserved to grow 'max= 3~6 nm, and A.max ~~f) nm (r:'igure 14). softened gl ass appears yellow-orange. in at The The very intense hanrl at 260 nm appears to have a verv large extinction coefficient as it very rapidly exceeds - '-l-7- the measurable range. 'By spectral comparison the 260 nm transition is assigned to the known35 A max = ?63 nm transition for the parent trans-2-tetrazene 25 the expected dimerization product of H2NN 3. The extinction coefficient for the Amax= 263 nm transition of this unstable species is unknown. For comparison tetramethyl-2-tetrazene 19 has coefficient Amax= 280, e: ~ 14-,000.58 The Amax= 386 nm transition is a rather large extinct ion assigned by soectral comparison to the known2i Amax= 386 nm transition for trans HNN'fi 1. The identity of the Amax~ 480 nm transition is un- certain at this point. Warming the softened glass further to 9 5-1 OQOK results in loss of the Amax 386 and Amax 4~0 nm transitions. Warming the now colorless solution to approximately 2200K results in vigorous gas evolution and formation of ammonium azide as an insoluble white precipitate. The parent tetrazene 25 has been shown to decompose to give hydrazine, ammonia, and ammonium azide as products.35 These products together with N2, l-l2, and CO are obtained overall from photolysis of 15 to form 3 followed by thermolysis as described. Full product analysis decomposition of 3 will be reported in a later section. behavior is found for 3-<f2. from thermal Similar thermal 0ther glass forming solvents gave similar results in the same temperature regimes (see Experimental). Photochemical Decomposition of H2NN 1, 1-T)iazenes have been reported to photodecompose when irradiated at their n-iT* transition in the visible)~g Irradiation of a blue-violet glass of 3 in 2-MTHF at 800K with visible light (VIS) using Corning CS-1-75 and t w u z <! a) 0::: 0 (/) a) <! 300 400 500 600 700 800 WAVELENGTH (nm) Figure 14. '2-MTHF glass of 3 warmed to 900K. Spectra taken at 10 min intervals. -49- two CS-3-70 filters (A. > 500 nm) results in loss of the blue-violet color and the structured absorbance curve of 3, A. max = 386 nm transition due to 1. together with grow+.h cf the Likewise, irradiation of the A. max = 386 nm and A. max = 480 nm transitions obtained from thermolysis of 3 in softened ?-MTl-f!:' at 900K, after recooling to 800K with visible light from Corning filter CS-l-75 (A. > 340 nm) results in the loss of these two transitions and the associated yellow-oran~e color. The major products from t LLJ u z <I: CD ~ (I) ~ 800 WAVELENGTH (nm) Figure 1.5. Absorption curve after warming :?-MTHF glass of 3 to 900K for 2 h and recooling to 80°K showing Amax = 386 and Amax = 480 nm transitions. -50- photo-decomposition of 3 in a 2-MTHF glass at 800K are H2 and N2. ~imilarly photolysis of the N2 products. >.max = 386 and 480 nm transitions affords H2 ancl Full product analysis for photochemical and thermal decomposition of 3 will be reported in a later section. One possible assignment for the \nax = 480 nm transition may be the as yet uncharacterized cis HNNH isomer 2. In general cis 1,2-diazenes are found to have longer wavelength, lower energy n-TI * transitions than their trans 1,2-diazene interactions.? isomers due to destabilizing lone pair-lone pair Calculations (GTO-CJ) of the electronic states of 1 and 2 suggest an energ;y difference of 0.56 eV (12.8 kcal/mon for the n- 1T * transitions with the cis 2 n-1T* at lower energy.67 This energy difference woulrl place the n-TI* transition for 2 at 467 nm relative to 386 nm for trans Amqx fnm) liE (kcal/mon E: 1. 8t Be 352 81.4 25 368 77.8 250 The molar extinction coefficient fs) for a cis 1,2-diazene is typically an order of magnitude greater than in the trans isomer. Irradiation of 1,2- diazene n-TI* transitions results in photointerconversion of cis and trans with quantum yields of 0.3 to 0.5. Higher pressures and longer wavelengths -51- generally decrease competitive isomerization yields. decomposition pathways and increase Direct irradiation of either isomer is considered to produce a common singlet excited state (S 1) which populates vibrationally excited cis and trans ground states (So~) with nearly equal probabilities) The hot ground states may then decompose or are deactivated by collision. Consistent with this interpretation, the use of shorter wavelength excitation would produce hot ground states with greater excess vibrational energy which would be expected to result in greater decomposition yields. This has been observed previously in some cases.? Attempts to verify the assignment of the Amax = 4-80 nm transition to 2 by photointerconverison of the Amax = 386 nm transition of 1 and the Amax = 480 nm transition are shown in Figure 16. The solid curve is that obtained after thermally decomposing 3 in softened 2-MTHF at 9QOK followed by recooling to 8QOK where 2-MTHF is again rigid (Figure 15). Irradiation of the Amax = 4-80 nm (4-70 .±. 10 nm) transition with for 3 h or broad-band monochromatic light irradiation (A >4-70 nm) for 1 h results in decrease of the Amax= 480 nm transition and slight growth of the Amax = 386 nm transition of 1. Photolysis of the Amax = 386 nm transition with light (380 .±. 10 nm) results in loss of the transition and a very modest growth in the monochromatic Amax = 386 nm Amax = 480 nm transition. Reversal of the order of photolysis results in similar behavior although the effect is less pronounced. No previous attempts at photoisomerization of 1 and 2 have been reported.2 However, Craig has reported successful photoisomerization of neat HNNCH3 in a low -52- 1 w u z <! m 0::: ~ m <! 300 400 500 600 700 WAVELENGTH (nm) Figure 16. fa) Comolete thermal decomposition of 3 at 900K (2-\nHF) recooled to 8QOJ< (--). (b) Photolysis (> 470 nm) 1 h, 800K f---t (c) Photolysis (3~() .:t 10 nm)? h, 800K (-···-). 800 -53- temperature (770K) polycrystalline matrix monitored by IR.68 The very slight growth in the Amax = 386 nm transition of 1 upon photolysis of the Amax = 480 nm transition is consiste~t with photochemical conversion of cis 2 with a large molar extinction coefficient to trans 1 with a small molar extinction coefficient. the conversion of cis Assuming e: for cis 2 is ten times e: 2 to 1 is nearly quantitative. trans 1, However, photolysis of the "-max = 386 nm transition of 1 gives only a very slight increase in the "-max= 480 nm transition amounting to only a few percent conversion of 1 to 2 with the assumed extinction coefficient ratio. Hence, these experiments must be regraded as very tentative and inconclusive. These observations may be consistent with vibrationally excited ground state cis 2 efficiently decomposing or alternatively may suggest a barrier for the excited state (S 1) collapsing to cis 2 which results in very inefficient cis production and greater decomposition yield from photolysis of 1. Using the assumed extinction coefficient ratio, the initital ratio of 1 to 2 from thermolysis of 3 at 900K is approximately 30:1 in favor of 1. If this is representative of an equilibrium mixture, a free energy difference of 0.61 kcal/mol in favor of 1 is indicated. This is compared with a calculated difference of 4.7 kcal/mol in the heats of formation of 1 and 2.1 t All attempts to corroborate this tentative photochemical interconversion of 1 and 2 by Raman and Resonance Raman spectroscopy in a 2-MTHF glass at 77 K in the N=N double bond stretch region 1400-1700 cm-1 were unsuccessful. Any further attempts must overcome experimental difficulties such as efficient sample photolysis with laser excitation and excessive Rayleigh -5l+- scattering due to the samoling rnethod. Since 1 has been characterized by Raman (\J N:l\1 for 1 is 1.529 crn-1, N2, JOCK) its assi~ntnent should be facilitated. The N:N stretch for 2 would be expected to lie at lower frequency as ohserved for cis and trans I ,?-dimethyl diazene.89 Emission Spectroscopy For emission studies, concentrated (J'O.l-0.5 M) samples of 3 were oreparerl bv prolonged UV (VIS filtered) ohotolysis ( 14-2f+ h) of a 2-MTT-p::glass of 15 in a 5 mm OJ") quartz tube immersed in a liquirf nitrogen cooled (77 K) Suprasil finger dewar. Irradiation of the intense blue-violet glass of 3 at its n-iT* transition 0.. max= 636 nm) with either monochromatic light (at various intense l-ig emission lines, A = 436, 546, 579 nm) or broad-band (),_ = l+80-650 nm) from a 2 50 watt mercury-xenon source gave no detectable emission from 3 at wavelengths 600 to 1100 nm. From the excited state studies8g of kinetically oersistent 1, 1-diazene 5 it was determined that the short lifetime (73 nanoseconds) of the n-iT* excited state <; 1 was controlled bv very fast internal conversion (kyc ~ 1o8 sec 1 for 5) to ground state So. 1,1-niazenes 4~g and 68f were found to exhihit structureless fluorescence spectra fin contrast to a structured emission for 5) consistent with efficient vibronic coupling in the pyramidal S 1 state for 4 and 6. In the absence of vibrationally induced surface crossings and a low lying iT ,n * triolet state, the rate of internal conversion should be determined by Franck-Condon overlap !!:. F. (state 1-state 2). 54,5 5 and therefore should be proportional to In 1, l-diazene 5 !!:. 1:. (S 1-So) is 50 kcal/mol and -55- kJc is J'108 sec-1.8g is 1 x 1o-3. In The quantum yield for fluorescence (¢F) 1,1-diazene 6 t.E (51-So) 109 sec-1 and ¢F is reduced to 1 x 1o-4.8g is 47 for 5 kcal/mol, kiC ~ Energy gap considerations alone may account for the increased kiC in 6, which results in a diminished fluorescent lifetime (TF 2_ 10-8 sec) and reduced ¢F. For comparison, in azulene L'.E (S 1-So) is 4 2 kcal/ mol and ki c is greater A rough based estimate upon energy than 1012 sec- 1)6 gap considerations alone affords krc > 1012 sec-1 for H2NN 3 where L'.E (51-So) is 41.5 kcal/mol. less rigid geometries of 4 and 6 may increase the Alternatively, the contribution of vibronic coupling which could increase krc· The presence of high energy N-H stretches (2865, 2808 cm-1) (see later section on infrared characterization) reduced ¢F. in 3 may result in efficient vibronic coupling and a Substitution of D for H in 3-d2 could decrease vibronic 3-d2 coupling and slow kJC resulting in an increased ¢F for measurable demonstrated fluorescence. This effect has been the isoelectronic molecule formaldehyde (H2C0))7 and in In H2CO substitution of D for H results in an increase in ¢F by a factor of J'20 in going from H2CO to D2CO. However, no emission was detected for 3-d2 again indicative of a very low ¢F in 3. Assuming kiC > 1Ql2sec-1for3and kF J' 105 sec- 1, as found experimentally for 1,1- diazene 5, affords a rough estimate of ¢F < IQ-7 in 3. This is consistent with no detectable emission for 3. -.56- Matrix Isolation Infrared Spectroscopy of H2NN 3 For infrared studies of 3, freshly sublimed carbomoyl azide 15 was cryo-pumped from a -200C ice/salt bath and co-deposited with ultra high purity (TJHP) ar~on onto a cesium iodide (Csl) inner window of the matrix isolation apparatus held at 200K (see Experimental Section for details). The resultin~ an~ on matrix of 15 was slowly lowered to 1OOK and its infrared spectrum recorrled. Figure 17 shows the Fourier transform infrared (FT -IR) spectrum of 15 isolated in an argon matrix (.r 1:2000, Ar)59 at 1ooK. Table V lists the major FT -IR bands for 15 in Fi~ure 17 and their assignments. Impurity peaks due to co-condensed atmospheric H20 and C02 are noted. The stronger infrared bands are found to exhibit "extra" low intensity multiple bands possibly due to some site splitting, complexes atmosoheric impurities or the presence of multiple conformers. with In general these "extra" bands were found to be of no consequence. Matrix ratios were adjusted to prevent the formation of hydrogen bonded "dimer" bands which are characteristically evident as lower energy multiple N-H stretch bands in the re~ion 34-00 to 3200 cm-1. A..chieving only two sharp N-H stretches at 34-53.75 and 357'2.60 cm-1 was indicative of a successful matrix for the desired photochemistry. Photolysis (2 h) of 15 matrix isolated in argon with TJV light from two Corning CS-7 -54- (UV transmitting, VIS A = 4-00-680 nm cut-out) filters results in the loss of all bands due to 15 and formation of new product bands. 60 Using FT -IR subtraction programs61 the FT -IR spectrum of 15 is subtracted from the FT -I'R. spectrum after UV photolysis of 15. The resulting spectrum clearly shows the positive infrared l. 00 ~.90 ,, t0 Vl.00 ~. ~.20 ~-3~ ~.Ll0 c e ~.50 c. n vL 5~ ~-7~ r b ~. f30 0 C' b R l . 10 l. 20 1. 3~ 1. ll0 ~igure 17. llfH10 PT -TP 3 r; ~ 0 soec trurn of 3~~~ - - T- 2~~0 t 5, I ?f) mm rlenosition Wc.venumbPr-s 25~0 r.r I :701)0, .1\r, I noT<). 1500 r---------T-- -- ------ T------- . 1000 - - ~----- ---- - - I I VI --.J -58- Table V. FT -IR Bands for 15 in an Argon \~atrix (1 :?000, lQOK) Threshold = 0.05 A. Peak Location f em- I) Intensity (A.) '3776.05 3756. ?9 3711.21 0.05 0.14 0.08 H?,O, V OH H?.O, V OH H7.0, V OH 357?.'>9 351)Q. 7() 0.89 0.12 V N-H 3453.75 3451. I 0 0.81 0.15 v N-f-1' 3376.37 0.06 v NH (dimer) 2448 •.50 0.1 (:) overtone 2'345.09 2339.06 0. 11 0.04 C02, V CO ?175. '38 ?164.53 ?160.91 2157.54 2152.96 0.06 1.44 0.66 0.?3 0.07 2135.36 0.0~ 2092.21 0.07 overtone t791. ~4 0.09 overtone 1764.60 0.15 1761.95 0.28 1757.37 1754.48 1746.76 1738.81 1733 • .50 0.91 0.63 0.30 0.26 0.15 1706.26 0.06 1623.58 0.15 Assignment C02, v CO V N3 (asym) \)co -59- Table V o Continued Peak Location km-1) Intensity (.1\) Jf)07o91 160?o36 Oo15 0 o14 15~6o?.1 1578o98 1576o09 Oo20 Oo10 Oo92 Oo35 1385o65 1382o99 1'370 22 Oo06 Oo06 Oo08 1339o36 1335o02 Oo09 Oo06 1323o21 1320o80 1315o25 1302o96 1.08 1. 22 1.09 1.04 P50o89 1248o24 Oo56 Oo58 124-3o18 Oo06 1230o88 t227o26 12? 1. 72 1197 13 Oo97 Oo15 Oo16 Oo26 v N3 (sym) 1098o05 Oo30 P NH 2 881.34 878o93 Oo10 0 o10 v C-N3 866o39 Oo10 763o94 761o 04 Oo ·:q Oo07 737 18 695o47 642o20 Oo32 Oo73 Oo60 l583o~O 0 0 0 Assignment Hi"\ o 0-H o -T\1Hz v NCO (sym) 0 N3 N3 NCN NCN tll.00 c e -0. Wj tll.05 til. 10 til. 1 s tll.20 n n b r 0 s b R tll.25 tll.30 til •.~S Figure l~. £1000 3000 2000 w,_venumber s 2S00 I F F 1500 I 7o H20 1000 nifference r:'T -TT.? soectrum of ororiucts from f JV (VTS filtererl) photolysis (?h) of 15 (J:?f)()f), A.r, Jr)O~) minus FT-fP SDectrum of 15. Suhtraction factor a = OJ,, (T)) = l--f?l\JI\J 1, (F)= 4 7 C0, (f)= H 7 N-1\J=~=0. 3[)00 ' C0 2 I co D I 0'\ I ·::> -0. 02 -121. 04 ~.00 Figure 19. Ll000 I Pl. 06 - , V'l.04 - c e ~.02 ~.08 n d. b r 0 s b R Pl.32 PI.3V'l PI.ZB Pl.26 . Pl. 2 4 Pl.22 Pl.20 V'l. 18 Pl. 1 6 V'l. 14 ~. 1 2 0. 10 I I 31100 2000 W<"\venumbers 2'J00 D 1500 D 1000 T)ifference PT -IR spectrum of H,"lN 3 (I)) and CO. UV (VIS filtereci) ohotolysis (2h) of 15 (1:?.000, Ar,-· l0°K) minus spectrum after VIS (>500 nm) l)hotolysis (? h) of 3. Subtraction factor a = t .0 5. 3r:i00 D 1\ co D I - I <3' -62- absorption bands . due to products from photolysis of 15 together with atmospheric impurities Y20 and C02 (Figure 1~). These bands are divided into four groups and are assigned based on their photochemical and thermal behavior together with isotopic shifts. These assignments are summarized in Table VI. The first group of bands at 3277.53, 2989.46, 2259.03, 2?12.0?., 2207.()~, 10?1.15, 95?.49, 868.08, 864.70 cm-1 grow in quickly with UV photolvsis of 15 and decrease with prolonged UV irradiation. assigned to the ( lal)eled fJ) in Table photochemical intermediate These are aminoisocyanate vn expected to result from the photo-Curtius rearrangement of 15. The most intense band at 2212.02 cm-1 is very likely the characteristic l'IJ=C=O asymmetric stretch for aminoisocyanate. A second group of procfuct bands at 2865.55, 2807.20, 2140.90, 1863.'20, 1574.11), and 1003.07 cm-1 also grow in with prolonged UV (VIS filtered) photolysis of 15 and are efficiently photolyzed away with visible irradiation at lOOK (> 500 nm). R.ecall that this filter combination resulted in decolorization of the blue-violet 1T using Corning 2-MTHF CS-1-75 glass of and 3 two and its CS-3-70 filters structured n- * absorption curve. Subtraction of the spectrum resulting from visible light (VIS) photolysis fJ. > 500 nm) from the spectrum before VIS photolysis results in positive absorption bands assigned to H2NN 3 ( labeled (0) in Table VI) (Figure l 9). The bands assigned to aminoisocyanate are not affected by VIS ohotolvsis of 3. The band at 2140.90 is assigned to carbon monoxide (CO) resulting from photolysis of the intermediate aminoisocyanate. Curiously, -63- H \+ - I .. N=N: co H 15 3 this 1.() banrf at 2140.90 cm-1 shows the same photochemical behavior as the group of bands assigned to H2~N 3. This will be addressed more fully in a later section. Two other groups of bands resulting form UV photolysis of 15 are observed to increase with the VIS photolysis of 3. One group of bands at ?1'39.94, ?137.22, and 2137.53 are also assigned to CO. The unique behavior of these CO bands suggest that they may be somehow related to the products from VIS photolysis of 3. This will be addressed in a later section. A final set of bands at 2865.55,2800.22, 1741.70, 1498.71, 1227.03, and 116g.45 cm-1 also increase in intensity with UV photolysis of 15. Surprisingly, these bands also increase with VIS photolysis of l-l'z~~ 3. Based on their photochemical behavior, their isotopic shifts and spectral comparison, these bands are assigned to forma1clehyde fH2CO) (labeled fF) in Table VI) an unexpected -64- photo! ysis product. later section. This will be addressed more fully in a For comparison, literature62 values for the infrared spectral assignment of H2CO ( l:~OO) in an argon matrix at lOOK are 2863, 2797, 1742, 1498, 1245, and 1168 (:tl) cm-1. All bands except the 1227.03 cm-1 0245 cm-1) band are within experimental error. o CH2 rocking mode of This band, assigned to the H2CO, has been observed to shift to higher frequency in a hydrogen bonding environment. For example, forming the dimer of formaldehyde (H2C0)2 matrix isolated in argon shifts this mode from 1245 to 1251 cm-1.62b This mode is found at 1239 cm-1 for H2CO monomer in a nitrogen matrix.62b Of the infrared bands assigned to 3 (Figure 19) the band at 1574.16 cm-1 is in the region where one would expect to find the characteristic N=N double bond stretch of H2NN. For the kinetically persistent 1, 1-diazenes 4 and 5 in CH2Cl2 solution, this characteristic N= N stretch is found 159 5 and 1638 cm-1, respectively.8e In order to assign these bands as the N=N double bond stretch, the terminal l'>N labeled 1, 1-diazenes were prepared. Incorporation of the 15N label results in the predicted Hooke's law shift of the 14 N = l 5N stretch to lower energy. For kinetically persistent 1, 1-diazene 4 the experimental 14N=15N double bond stretch was found at 1568 cm-1.8e This ?.7 cm-l shift to lower energy was identical to the Hooke's law calculated shift. In order to verify the 1574.16 cm-1 band observed in the FT -IR spectrum of 3 as the characteristic N=N double bond stretch for this species, a l 5N terminal label was desired. The Hooke's law calculated ratio v (l4N=14N)fl4N=l5N) is 1.0171 or a predicted shift for the 14N=15N stretch -65- Stirring carbomoyl azide 15 with an from 157lt.16 to 15lt7.69 cm-1 in 3. excess of ( l-15N) labeled sodium azide63 in dry acetonitrile afforded 15- 15N. The combination of 15N NMR, high resolution mass spectroscopy and FT -IR spectroscopy demonstrates a statistical incorporation of 15N into 15- 15N by azide exchange in solution. Due to the symmetry of the azide anion 15 in solution6lt the 15N label should be equally distributed between the two terminal positions of the azide moiety. Four equivalents of ( 1-15N) sodium azide results in an approximately 80:20 mixture of 15-15N: 15 by mass spectroscopy. 15N-NMR reveals two singlets at -lltO.lO and -262.32 ppm from 15N-CH3N02. (.r 1:2000) at Matrix isolation of 15-15N in argon lOOK (Figure 20, Table VII) reveals that the unlabeled asymmetric azide stretch at 216lt.53 cm-1 has 15N shifted multiple bands at 2159.23 cm-1 and 21lt1.15 cm-1 in an approximate 1:1 ratio consistent with the 15N label equally distributed between the two terminal positions of the azide moiety. Photolysis of an 80:20 mixture of 15-15N:l5 would be expected to yield a 60:lt0 mixture of H2NN 3 to H2N 15N 3-15N by the photo-Curtius -66- Table VI. Infrared Assignments for Products from Photolysis of 15 and 1515N fJ"1:?000, Ar, lOOK). 120 min hV UV (VIS filtered) 120 min h v VIS (A > 500 nm) 100 min hV VIS (A= 360-420 nm) Assignment 3776.05 3756.53 3710.96 3669.74 h, VOH h, VOH h, vOH h, VOH 3?77. 53 ?989.46 (1), VNH (1), VNH n65.55 2865.55 ?807.?0 2~00.?? (-) (+) (_) (+) (+) (+) D, Vl\lH F, VC-H D, Vl\lH F, VC-H 2345.09 ?.33Cl.04 C02, vCO C02, vCO ??59.03 ?:?.38.78 2212.0'2 ?207.68 ? 194. p~ I, V-14N=C=0 I*, V-15N=C=0 I, V-l4N=C=0 (asym) I, V-14N=C=0 I*, V-15N=C=0 2149.8? 2143.32 2140.90 2139.94 2137.22 2137.53 (-) f-) (+) (+) (+) (+) U~63.'20 (-) 1741.70 (+) 1623.58 1610.08 1607.67 1592.96 (-) CO/T, vCO CO/T, vCO CO/D, vCO CO, vCO CO, vCO co, vco D, oN-H (+) F, vC=O h, 80-H h, 80-H h, 80-H h, 80-H Table VI. Continued. I?n min hv uv (VI~ f i1 terecf) 1 ';74. 16 154-7.{,4 120 min hv VIS (I. >500 nm) 100 min hv VIS ( '-= 360-420 nm) n, v 14N:14N l1*,v l4N:l51\l (_) (_) 149~.71 1277.03 F, o CH2 l?~~.f)J (_) I?R~ .~ 1 (-) T, o 1\J:N-H (asym) T*, 0 151\l:N-H (+) (+) F, c CH2 (?) 1218.83 1168.45 Assignment (+) (+) F,o CH2 1146.03 (I) 1030.32 (I) l 021 • t 5 (I) 1003.07 1002. '35 f-) (_) 95?.69 %8.08 ~64.70 h l T) F T r.o (_) (+) () = H:{) = A. minoisocvanate (H?N-N:C:O), I* = (1_15N) = niazene (H 2 NI\l), T)* -= (n-15N) = Formalclehyde (H?CO) = Trans-!, ?.-cfiazene ('f1NN1-J), T* = (T _151\l) = Carbon monoxicfe = Oecreased = Increased = Tentative assignment n,o l4N=l4"1-Y n*,o 151\l= 14N-H (I) (I) (I) 0 II • - H2 N~N=N=N: .. •• H \+ - 15 I N=N: .. H 0 II + 3 hv - 5 N: H2 N"/"-.N=N=' .. .. ft + - H2 ~ 15 N=N=N: •• .. rearrangement and subsequent photolysis of the intermediate amino- isocyanate. Photolysis of matrix isolated 15-15N (1:2000, Ar, lOOK) (Figure 20, Table VII) with two CS-7-.54 filters for 2 h results in decrease in the FT -IR bands of 15-15N and formation of product bands. Figure 21 shows the product FT -IR spectrum minus the FT -IR. spectrum of 15-15N. These bands are summarized in the first column of Table VI. In addition to the product bands reported from the UV photolysis of 15, additional 15N isotope shifted bands are found at 2238.78, 2194.18, 1547.64, and 1002.35 cm-1. These correspond in intensity (based on 1.5N incorporation into 15-15N) to l5N shifted bands of product infrared bands at 2259.03, 2212.02, 1547.16, and 100'3.07 cm-1, respectively, observed from UV photolysis of 15. band at 1547.64 cm-1 corresponds to the calculated The new 1547.69 cm-1 v 14N= 15N stretch for 3-15N (Figure 23). Comparison with the Raman active v 14-N= 14N stretch for HNNH 1 at 1529 cm-1 2f demonstrates the - 69- c C' c(" tsl 1- ~ ....... c c ._, ' I v; c c. ~~ G.; "t. E E c(" ' I ,.... ~ Q., ~~ :::. ~ >' (/) ~ Q) tc. (".. ....0 c E ~ :j -.. c :s:l tsl Q) :.."l > N c 3 0': If' 2' -"" If' I '+-" c tsl tsl I:S: tor: E ~ ...,i:. c <1-' • c,... v; Y' c cr c I tsl tsl !- ~-~ LL r:r· '-'"' tor: 0 (".• 4) !:::: _, tsl tsl tsl o::::s- tsl tsl ~ tsl tsl C'1 co ts:: ~ !"'- l.D t.n o::::t ~ N") f-,J ~ ts:: ts:: ts:: ts:: ts:: ts:: ts:: - ~ cr _o (/) o ~- ....o o c o w tsl ~ ts:: tsl tsl ts:: .!?! IJ. -70- Table VII. FT-IR Bands for 15-15N in an Argon Matrix (1:?000, lOOK) Threshold= 0.05 A. Peak Location (cm-l) Intensity (A) 371 I . ? t 0.02 '3572.59 3569.7() 0.09 34-53.75 3451.10 0.77 0. 11 v NH 24-4-8.0? 0.05 overtone 234-5.09 2339.06 0.09 0.05 C02, v CO C02, v CO 2164-.53 ?.159.?2 2155.1}1 ?.152.96 ?.151.03 214-6.93 214-1.14?.138.01 2136.56 ?135.% 1.02 1.040.20 0.09 0.07 0.06 1.01 0. 140.16 0.14- v _14N=14N=14N (asym) v _l5f\):14N=14-N (asym) 1761.4-7 1757.70 1755.?.0 Assignment 0.~8 174-4-.~'3 0.20 0.61 0.55 0.26 1739.77 1738.57 o. to 0.10 11}23.58 1607.91 1602.36 0.06 0.06 0.09 1586.21 15g'3.80 1578.98 1572.95 0.08 0.15 0.71 0.06 1359.37 0.08 H?O, o OH H:7.0, o OH H20, o OH -71- Table VII. Continued. Peak Location (cm-1) Intensity (A) t321.52 111.5.25 1311.40 1309.50 1302. cp 1?99. sn 0.8? 1.016 0.311 0. ?89 0.48 0.27 1248.00 1230. ~~ 0.35 0.60 v N3 (sym) 1221.96 1208.22 0.60 0.40 v 15N=N="J (sym) v N=N= 15N {sym) 1196. ~9 1193.03 1191.35 0.10 0. 11 0. 28 v N=N= 15N (sym) v N=N=15N (sym) 1098.05 0.20 p NH2 881.34 878.93 0.05 0.05 v C-N3 855.()6 763.96 736.94 735.97 695.23 693.55 641.96 0.08 0.18 0. 13 0.13 0.34 0.38 0.47 Assignment v NCO (sym) 8 N3 8 N3 8 NCN 8 NCT\115 8 N15cN -0. 02 Figure 21. 3Vl00 --T 1 \v (l v e I[ I I. I* I 2000 u mb p r s li 1r/F 2500 I \1 D,F \ C02 I co I D IP I -H 20 I 1500 I.ll.l(.tf F I F I 1000 ',I I D I I Difference FT -IR spectrum of products from UV (VIS filtered) photolysis (2 h) of 15-15N:l5 (1:2000, Ar, lQOK) minus FT -IR spectrum of 15-15N:l5. Subtraction factor = 0.6, (D)= H2NN 3, (D*) = H2N_l5N, (F)= H2CO, (I)= H2N-N=C=O, (I*) = H 2 N-15N=C=0. 3S00 Ll000 .lJ "'· . -- I I . __ I . 1&1 I -H 20 I 0.02 c e 0.00 Vl. 04 - Vl. 06 - l n a b i 0.08 Vl. 1 0 s 0 0. 1 2 Vl.1Ll b r=i 0. 1 6 0. 1 8 0. 20 I C0 2 I I ..... 1'.) 0.06 r Vl.00 e -0.04 -0.02 0.02 c n (l b 0.04 0.08 s 0 0. 10 R b 0. 12 0. 14 0. 16 I Fi~ure 1?. 4000 --1 0.18 - 3000 2000 Wc.venumbers 2500 I D 1500 I D* D 1000 I f"iffP.rence PT-TQ. snectrum of H71\1N 3 (r"'): H~N-15N 3-l5P1J (n*) mixture a nd 1:0, IJV (VI<:; filt e n~ r!) ol,otolvs is (?h) of 15-Ll\1:15 (1:7nnn, :'\r, ln°K ) Mi'lus soP.ctrwn after VIS (> 500 nrn) photolvs is (?h) of 3:J-15N. Subtractio11 fa c tor ('t = 1.!0. 3S00 D I.D co D I I --J v.J e Figure ?.3. Wo.venumbers Wo.venumbers (a) l-l2 14N= l'-1-'\J stretch of 3 from Figure 1q. (b) H2 I 4N= l'-1-N stretch of 3 and H 2 14"1= 1 1 N stretch of J-l5N from Fi~ure ??. 1S00 1'..)6QJ 1540 1520 15QJQJ 1580 1560 1540 1520 150QJ I .,....... I el.QJQJQJ - - c lli.QJQJQJ - ~.QJ5QJ n 0. b r 0 s b R lli.QJ/Ill -75- substantial double bond character in the 1, 1-diazene N=N bond in accord with the calculation of Davis and Goddard.! n Photolysis of the products from UV (VIS filtered) photolysis of 15-15N with visible light (> 500 nm) results only in the loss of all bands assigned to 3- 15N and 3 and the accompanying CO stretch at 2140.90. Figure 22 shows the difference FT -IR spectrum for before minus after VIS photolysis of the 3- 15N, 3 mixture. Table VIII summarizes the infrared bands assigned to 3 and 3-15N. The strongest FT -IR band of 3 at 1003.07 cm-1 is also affected by the incorporation of a 15N label. This band is likely assigned to the in plane asymmetric bending mode of 3. This band shows a small 15N isotope shift of 0.72 cm-1 to 1002.35 cm-1. For comparison, the asymmetric in plane bend is also the strongest infrared band observed for HNNH 1 at 1286 cm-1.2c This mode also shows the small 15N isotope shift expected for a bending mode from 1285.8 to 1284.3 cm-1 (a shift of 1.3 cm-1). The infrared band at 1863.20 cm-1 observed for 3-15N shows no resolvable (0.25 cm-1 resolution) 15N shift. This band is therefore likely assigned to the out of plane torsional mode for 3. The N-H stretching modes at 2865.55 and 2807.20 cm-1 are also unaffected by a terminal 15N label in 3-15N consistent with their assignments. Due to its C2v symmetry H2NN should exhibit six vibrational modes all of which should be both infrared and Raman active. Five of these six modes have been located in the argon matrix FT -IR spectrum described. The sixth mode is presumably obscured by other infrared bands. Conclusive assignment of the vibrational modes of 3 must await full Raman characterization and polarization studies. -76- Table VIII. Infrared Bands for H2NN 3 and H2N 15N 3-l5N (Ar, lOOK). Assignment 2g65.55 v N-H 2~07.20 v N-H 1863.20 o =N-Y (out of plane) t 574.16 v 14N:14N 15ft 7.64 v 141\J:l5N l003.n7 0 l4N= l4N-H (in plane) l 002. ~5 0 l51\J=l4N-H (in plane) ( ) = Tentative assignment The remaining 15N isotope shifted product bands observed from UV photolysis of 15-l5N at 2238.78 and 2194. t 8 cm-1 are assigned, due to the magnitude of their shift and v _l5N=C=0 the 15N isotope shifted (v their relative isocyanate intensities, stretching _l4N=C:0) of aminoisocyanate (labeled (I*) in Table and 2212.02 cm-1. All of these bands are to modes vn at 2259.03 observed to grow in quickly with UV photolysis and decrease in intensity with prolonged UV photolysis. Photolvsis of loss of all H~NN 3 bands with visible light (VI()) 0.> 500 nm) results in assigned to 3 and formation bands 214<U~2, 2143.32, 1288.01, and 1218.83 cm-1. of new In addition, VIS photolvsis of 3 results in the growt11 of bands assigned to formaldehyde -77- (H2C0) at 2865.5~, 2800.22, 1741.70, 1498.71, 1227.03, and 1168.45 cm-1 and growth of two CO bancis at "l3q.94 and 2137.22 cm-1. The CO bands will be discussed in a later section. Photolysis of 3-15N with visible light results in formation ancl growth of the same product bands observed in the photolysis of 3 together with formation of a single l5N isotope shifted band at 1286.81 cm-1 (see second column of Table VI). VIS irradiation b 500 nm) of 3 in a rigirl ~-MTHF glass at ~OOK resulted in the growth of an electronic absorption at 3~6 nm assigned by spectral comparison to the known electronic absorotion of trans '-lNNH 1 at 386 nm. The infrareci band at l ~~8.0 l cm-1 and its l 5l\] shifted multiple at 1286.~ 1 cm-1 (a shift of 1.20 cm-1) corresoond to the strongest infrared transition reported by 'Rosengren and Pimentef2c for 1 at 1285.8 .:!:. 0.3 cm-1 in a nitrogen matrix. The incorporation of a single 15N isotope into 1 was found by Pimentel to result in a 15N shift of 1.3 cm-1 to 17.~4.5.:!:. 0.3 cm-1. For direct comparison, 3 was matrix isolated in N? (1:2000, 100K).65 VIS photolysis b '500 nm) results in formation of the corresponding band for 1 at ~ hv (VIS) 3 +co 1 e c n 0. b r 0 s b R - Fi~ure 24. ~.~00 ~.~34 127~ - - 127~ Wo.venumbers 1300 1290 1280 (a) nifference r:'T -TP soectrum of 'HNNH l from vyc;; ohotolysis of 3 (?. h, Ar, to°K). J:\efore minus after photolysis of 1 (1MJ-4?.0 nrn), 100 min. (b) nifferPnce PT -IP soectrum of HNN'-l 1 and J-15N from VT') photolysis of 3:315N (?_ h, A..r, JOOJ<). P,efore minus after photolysis of 1; J-l.5N 060-4?0 nm), 1no min. Wo.venumbers 1300 1290 1280 ~.000 ~.034 I I )0 ........ -79- 12%.02 cm-1 in .agreement with the 1285.8.:, 0.3 cm-1 band reported by 'R.osengren and Pimentel.?c infrared band \1 N-H) for 1 at All attempts to locate the other reported 3131 cm-1 (N2 matrix, very weak, broad, have been unsuccessful. This may be attributed to a low yield of 1 from VI<; ol-lotolysis of 3 (estimated at < 10%). Photolysis of trans HNNH 1 could he effected with either monochromatic lie;ht (380 .:t 10 nm) or more efficiently witl-l Cornin$! C<;-5-58 and C<;-7-'H filters 2-3h) (see Figure 24). In addition to loss = 360-410 nm, (A_ of the 1288.0 l cm-1 band (Ar) (and the 1286.81 15N multiple with t-15N) two CO stretches at 2149.82 and 214'3.32 cm-1 are also lost with photolysis of 1. This will be discussed in more detail in a later section. Photolysis of 1 (X = 360-410 nm) results in growth of absorbances due to H2CO and two CO bands at 2139.94 and ?.137.53 cm-1. photolysis of 3. Identical behavior was found with VIS The third column of Table VI summarizes the effects of photolysis of 1 and t-15N. None of the tentative cis HNNH 2 bands reported by 'R.osene;ren an Pimentel2c at 1279 and 3074 cm-1 were observed in the photolysis of 1 with either broad-band (X = 360-410 nm) or monochromatic fA, = 380 .:t 10 nm) light (A r or N2, 1QOK). The identity of one last minor product band from ohotolysis of 3 and 3-15N at 1218.83 cm-1 remains uncertain. In order to further support the infrared assignments for 3, l, and their photochemical decomposition products, 3-<i2 and 3-d2_15N were also prepared. Exhaustive H-T) exchange of 15 and l5-15N afforded l5-<i2 and 15-d2- 15N. Photolysis of l5-d2 0:2000, Ar, lOOK) (Figure 25, Table IX) and 15-d215N (1:2000, Ar, lOOK) (Figure 26, Table X) with UV light from two CS-7-54 -80- filters (A. = 400-680 cut-out) results in loss of the infrared bands due to the starting carbomoyl azides and formation of product absorbances assigned to aminoisocyanate-d2 (labeled (I) in Table XI), D2NN (3-d2 ), formaldehyde-d2 rn2CO) (labeled (F) in Table XI) and carbon monoxide (CO) listed in the first column of Table XI. Subtraction of the spectrum of the respective starting carbomovl azide from the product mixture FT -I'R. spectrum (Figure 27 for 15d2, Pigure ?8 for l5-d2_15N) clearly reveals the product absorptions. The first group of bands at 2259.03 and 2:?'31.06 cm-1 from 15-d2 show 15N shifted multiole bands at 2237.~1 and 2212.26 cm-1 a shift of 21.22 and 18)W crn-1, respectively. These bands are observed to grow in quickly with UV photolysis and decrease with orolonged UV irradiation. These hands are assigned to the isocyanate stretches (v "'l=C=O) for aminoisocyanate-d2 and the corresponding 15N shifted modes (v L5N=C=0). No other bands for aminoisocvanate-d., were located suggesting that the v N=C=O is the most intense infrared band for the aminoisocyanate as anticioated. A second group of hands at 2140.91), 2138.73, 2109.08, 1599.23, 1571.27, 1195.69, 913.64, ~Q9.~9, and 793.83 cm-1 from photolysis of 15-d2 are assigned to n2NN 3-d2 and CO at ?.140.90 cm-1. The behavior of the CO banrls will be discussed in a later section. These bands all decrease with VI5 irrarliation (A. > 500 nm) as ohserved for H2NN 3 and its accompanying identical CO hand at 2140.90 cm-1. The difference FT -IR spectrum resulting from the subtraction of the spectrum after VIS photolysis of 3-d2 is shown in Figure 29. This spectrum clearly shows the absorbances assigned to 3-d2 and CO. The bands at 1599.23 and 1571.27 cm-1 are in the region where the e c 0.00 171. 10 0.20 0.30 0.ll0 0.. n 0.50 0.60 0.70 b r 0 s R b 0.80 0.90 1. 00 Figure 25. I ll000 l 3000 I 2000 Wo..venumbers I 2'J00 1 r.J00 FT-IR soectrum of 1.5~, 120 mm deposition (1:?.000, Ar, l0°K). Y500 I 1000 I - I 00 -82- Table IX. FT _y-q_ Bands for 1.5-d2 in an Ar~on Matrix (J :2000, JQOK) Threshold= ().()5 1\~ Peak Location (em- l) Intensity (A) 3756.26 0.07 2n2. 21 2682.68 ?664.98 /.447.% 0.06 0.'33 0.06 0.40 0.06 2345.09 0.08 2195.69 ?.164.54 2157.98 ? 154.50 2151.57 0.41 0.06 0.07 1.04 0.93 0.75 1753.24 1751.45 1747.45 1726.22 0.20 0.31 0.96 0.11 16?.3.58 1607.91 0.08 0.08 1466.47 1420.55 0.11 0.09 134~.?.4 1342.78 1341.90 1340.64 1338.07 0.22 0. t 8 0.19 0.29 1.% 1315.'27 t 311.80 0.31 0.19 1243.12 1236.49, 1?.34. 20 1224.00 0.41 0.12 0.99 0.16 2524.U~ ?.1~5.5'3 Assignment v N-D v N-D v N3 (asym) v C=O H20, o 0-H H20, 0-H o v NCO (sym) \J N3 (sym) -83- Table IX. Continued Peak Location (cm-l) Intensity (A) 1136.35 0.15 c ND2 821.72 0.10 p N-D 727.22 0.14- c N3 659.23 0.4-4- r NCN Assignment Vl. 7Vl s Vl.S0 b vL 30 c e Vl.00 Vl. l 0 Vl.20 Vl.Ll0 n f.\ Vl.6Vl r 0 Vl.80 R b vL 90 1 . 00 1 . 10 :) f) -r- 3Vl00 Wtivenumbers I 2~00 (1 :2000, Ar, lQOI<.). FT -IR spectrum of l5-d2_15N:l5-d2 ("-'80:20 mixture), 120 mm derosition - -- , 0 0 Figure 26. Ll000 I I .j::" 00 -85- Table X. FT -IR. Bands for 15-d2-15N in an Ar~on Matrix (1 :2000, lQOK) Threshold 0.05 A. Peak Location (cm-1) Intensity (A) 2782.22 2682.69 2524.19 0.05 0.32 0.05 0.37 2345.09 0.06 2195.69 2185.55 2170.98 2168.36 2164.55 2157.97 2151.40 2146.28 2141.20 2136.58 2133.55 2131.76 2129.48 0.21 0.11 266~.98 Assignment \) N-D \) N-D 0.35 0.11 0.09 0.63 0.39 0.45 0.08 0.10 0.12 0.51 0.41 1753.25 1751.46 1746.76 1726.23 0.14 0.20 0.99 0.06 1466.47 1420.55 0.09 0.07 1348.04 1346.41 1343.76 1341.10 1337.87 0.12 0. 14 0.26 0.23 1.18 1315.28 1311.80 0.18 o. t7 \)-N=N=N (asym) \)-15N=N=N (asym) \)-N=N=15N (asym) \)-N=N=15N (asym) \) C=O \)NCO (sym) Table X. Continued. Peak Location (cm-1) Intensity (A) 1243.12 1240.01 1234 . 24 1231.59 0.16 0.13 1225.55 Assignment 0.65 v N=''J=N (sym) 0.23 0.47 0.10 v 15N=N=N (sym) 1220.44 1207.2?. 0.65 v N=N=15N (sym) 11%.35 0.18 8 ND2 g21. 73 0.06 P N-D 7?.7.23 0.11 8 N3 659.00 0.29 0.18 r NCN r Nc15N 656.75 0. 12 ~. 10 0 r ~.04 c e -0. 02 ~.0fZJ ~.02 ~.06 n ~ b ~.08 0. 14 b s ~. 16 R ~. 18 ~.20 ~.22 ~.24 Figure 27. 4fZJ0fZJ T~ II H20 I 7)~00 r I 2~00 ...I.. 1\ .J..II I ,,D I D20 'I I 1~00 II ... 'fT"T ' J~ lL. I.J. F-h,d 2fZJ0fZJ w~venumber-s .J IF F F .--H 20 I ... .d 1.. \\ r D-h,d 1 11 (]fZJ l"' D ..1 I co 2 T D of prociucts from !IV (VJS filtereci) photolvsis (QO min) of l5-d2 (J:?()r)r), A.r, 10°'<) minus snr:>c trum of 15-d2· <:;uhtraction factor n = 0.1 n. fn) = n7''JN 3-d1 , (F) = n 7 r:o, m = n~l\l-N=f:=O. ..1 /I D 20 nifferenc~ J.T -TR so~ctrurn I 3C.j fZJfZJ ~I.... ~I ..,.,... / C02 D ,D co I "" I 00 ~.06 n c e Pl.02 0.0121 -0. 02 ~.04 ~.08 0. 0.20 0. 18 0. 16 r'J. 14 ~. 1 2 ~. 10 b r 0 s R b r'l.24 0.22 ~.26 ~.28 ~.30 I 3000 F-h,d\ 2000 Wc..venumbers I 2500 F D D I 1Cj00 I' 1000 o-h,d ~ D 020 I D* II''~ g,,o*j'' F If H20 / D nifference r.T -TR spectrum of ororfucts from TJV (VT'; filtereo) ohotolv5is (qO min) of 15-rl?-l5N:t5-rf2 mixture ft:?nno, .1\r, tn°J<) minus soectrum of 15-d?t 5N: t 5-d? rnlxt11re. <;uhtraction factor a= 0.10. (T")) = T)71\fi\J 1-d?, (n*) = ni'l1""' 3-d2...:t5"'f, (n) = n?r-:0, (J) = n?li.J-"T=C=0, (l*) = n 7 1\f_l5N=r:=n. I '3Si210 Figure 2~. ~ Lll2]00 I /H20 C02 I co o* I I :>0 00 -89- Table XI. InfrarP.d Assignments for Products from Photolysis of 1.5-d2 and 1.5-~-1.5N (J'J:2000, Ar, lOOK). 90 min hv (VIc; f i1 tered) 240 min uv (A hv VIS >500 nm) 150 min hv (A= 360-420 nm) Assignment 3776.05 3756.53 3710.96 370 l. ~ 1 h, v Ol-1 h, v OH h, v OH h, v OH /.78?.14 d, v 0-D d,v 0-D ?n3.~0 co,,\) co ?345.09 ?.339.06 C02, v CO n59.o3 ?237.81 2231.06 2212.26 I, v _14N=C=0 I*, v _15N=C=0 I, v _141\l:C:O I*, v _1 5N=C=0 2179.96 F, v C-D 2149.82 2143.32 (-) D, v N-D co,\) co ?137.53 2109.08 CO/T, v CO CO/T, v CO co,\) co co,\) co 2140.90 2139.94 2138.73 (-) (-) (-) D,v N-D 2071.48 F,v C-D 1720.00 1697.58 F (H,J)), v C=O F,v C=O 16?3. ')~ h, cS OH h, cS OH 1607.67 -90- Table XI. Continued 90 min hv UV (VIS filtered) 240 min hv VIS (A > 500 nm) 150 min hv (A = 360-420 nm) D, o N-D (-) h, o OH 1572.95 1571.27 (-) 1552.22 (-) D, v N14:14N D, v N14=15N d, o 0-D d, o 0-D 1427.35 1413.1;1 1195.69 1194.72 h D, o N=N-D D*, o 15N=N-D (-) (_) (+) (+) F, o C-D 965.95 (-) 964.50 (-) (?) (?*) 947.39 945.46 (-) (-) T, o N=N-D (asym) T*, o 15N=N-D 913.64 912.92 (-) (-) D (H,D), o N=N-D,H D* (H,D), o 15N=N-D,H 899.89 899.18 (-) (-) D (H,D), o N=N-D,H D* (H,D), o 15N=N-D,H 793.83 793.11 (-) (-) D, o N=N-D D*, o 15N=N-D = H20 = A minoisocyanate-d? (])2N-N=C=O), I* = (1_15N) l) = niazene (D2NN), J)* = (D_L5N) F = Formaldehyde-d2 (D2C=0) T = Trans-l,2-diazene-d2 (T)NND), T* = (T_15N) co = Carbon monoxide d = n2o (_) = necreased (+) = Increased ( ) = Tentative assignment I Assignment -91- characteristic N:N stretch for 3 is found. Photolvsis of l5-d2-1.5N yields 3~-1.5N. Subtraction of spectrum after VIS photolysis of 3-d2_1.5N results in the difference FT -1~ spectrum clearly showing; the absorbances due to 3-d21.5N and CO (Figure 30). The resulting spectrum reveals l5N shifted bands for 3-d2_15N at 1552.2?, 1194.7~, 912.92, 899.18, and 793.11 cm-1. 1557..?'2 cm-1 corresponds to the Hooke's law 15N The shifted vT)214N=l5N stretch of the characteristic v T)214N=l4N stretch at 1571.27 crn-1. The remainder of the 15N shifted bands of 3-d2_15N are assigned to 1 r;N shifts of the 1\J:N-n bending modes of 3-d2 by analogy to 3. The bands at 913.64 anci 899.89 cm-1 and their 15N multiples are assigned to the H,D counterparts of the strongest banrls observed for 3 at 1003.07 cm-1 and 3-d2 at 793X~ cm-1 due to some 1.5-h,d (and 1.5-h,d-1.5N) irnpurity in 15-d2 (and l5-~-l5N). The infrarerl transitions for 3-d2 and 3-<f2-l5N are summarized and assignerl in Table XII. All six infrared modes of n2NN 3-d2 have been located. Failure to observe a 1 5N multiole band (0.25 cm-1 resolution) for the absorotion for 3-d2 at 1599.23 crn-1 is consistent with its assignment as the out of plane torsion mode. A third group of bands from UV photolysis of 15-d2 and 15-d2_15N at 21 39.94 and 2137 •.53 are identical to the CO stretches observed in the UV photolysis of 15 and 15-l5N. These CO stretches are observed to increase with photolysis of J-d2 and 3-d2_1.5N with visible light consistent with their assignment as CO's somehow associated with the photochemical decornoosition oroducts of 3, 3-15N, 3-d2, and 3-d2_1.5N. The final group of product bands from l)hotolysis of 1.5-d2 and 15-d2- _, ~- ~4 ~ s ·-0. 12 --0. 1 0 ·-0. 08 ·- 0. 06 c e -0.04 n 0. -0.02 0.00 I b 0.02 0 I H 20 Figure 29. Ll000 I -1 0.06 b R ~.08 ~- \~ ~ ~. 1 2 ~.14 3000 2000 Wo.venumbers l 2500 I II 1500 'Jt I H20 O 0 20 O 1000 0-h,d 0 T)ifference FT -lq_ soectrum of T)?''JN 3-d:z (1""'1\ and CO. UV (VIS filtered) photolysis (90 min) of l5-d2 0:?.000, Ar, JOO'f<) minus spectrum after VIS f > 500 nm) photolysis (r.,. h) of J-d2. Subtraction factor a =1.0. 3S00 020 I .. 0 I co I ..0 I N Pl. 1 4 Pl. 1 2 111.10 Pl.08 111.06 111.04 s 0 b t1 n c e - 0.02 111.00 111.02 Pl. 16 b r Pl. 1 8 R 111.20 111.22 111.24 111.26 H 20 W::it?:ure 10. I 4000 I 3000 2000 Wt!venumbers 2S00 D D D 1500 D-h,d 1000 : / o* II 0 20 D \ II o* It H20 I D nifference ~T -T~ spectrum of n?!\11\l 1-d2 (~): n?"'-l ""' 1-<f2 _t 5N (n*) mixture anrl r.0. TJV (vyc;; filtererl) ohotolvsis (q() min) of 15-<f2_l5N:l5-d:? mixture 'I :?0fl(), A.r, ff)OJ<) minus snectrum after VIc;; (>'5()0 nm) photolvsis f 1.J. h) of J-<f2:J-<f2-' 5N. Suhtraction factor a= 1.1 0. 3 [_j 0 0 / co o• I ·.D I Vol -94- Table XII. Infrared Transitions of D2NN 3-<f2 and D2N-15N 3-<f2_15N (Ar, 1OOT<:). fcm-1) Assi!:~nment 2138.73 v N-D 2109.08 v N-D 1599.23 0 N:N-D (out of plane) 1571. ?.7 v 14N:14N 1552.22 v 14N=15N 1195.69 0 14N=N-D 1194. 7?. o 15N=N-D (913.64) o N:N-l-I (H,D) (912.9?) 0 15N:N-l-I (H,D) (899.89) o N:N-D (l-I,D) (899.18) 0 15N=N-D (H,D) 793.83 o N:N-D 793.11 0 15N:N-D ( ) =Tentative assignment l5N show no 1-'N multiples. These bands are assigned by spectral comparison to formaldehyde-d2 (T)2CO) am' formaldehyde-h,d (f-l,DCO). These bands are listed in the first column of Table XI fF-d2, F-hd). These bands are observed to increase with VIS photolysis of 3-d2 and 3-d2_l5N (see second column Table XT) dem~:mstrating, as found for 3, that formaldehyde is a photochemical decomposition product of 3 in the presence of CO. This will be discusser! -95- fully in a later section. The observed infrared bands for isotopically labeled formaldehydes are compared with their literature62,66 values in Table XIII. Table XIII. Formalrlehyde Infrared 'Bands from Photolysis of 3 and 3-d2(h,d) Literature in the Presence of CO (Ar, lOOK)~ (cm-1). Comparison. H2CO {lit.)2 D2CO (lit. )Q 2865.55 n63.o ?.179.% 2176.8 2873.8 v C-H ?.800.22 ?.797. 1 2071.48 2069.1 ?.764.3 v C-H 1741.70 1742.0 1697.58 1697.8 1720.00 1723.2 v C:O 1498.71 1498.8 1099.1 1399.1 o CH2 (scissor) 1227.03 1244.8 987.1 1028.0 o CH2 (rock) ll6g.45 1168.0 9~8 .3 1074.0 o CH2 (out of olane) 9g4.99 H,nco {lit.)~ Assignment ~This work .:t0.05 cm-1 • .!?Ref. 62 (Ar, lOOK). ~Ref. 66 (Ar, lOOK). Photolysis of matrix isolated J-d2 with visible light b 500 nm) also results in formation of new infrared bands at 2149.82, 2143.32, 965.95, and 947.39 cm-1. The bands at 2149.82 and 2143.32 cm-1 are identical to two of the new bands observed in VIS photolysis of 3 and 3- 1.5N. These bands are assi~ned to CO's and will be fully discussed in a later section. The bands at 965.95 and 947.39 show 15N shifted multiples with VIS photolysis of 3-d2_1.5N at 964.50 (a 15N shift of 1.45 cm-1) and 945.46 (a n c e n b r 0 s b R 980 960 940 920 Pl.el00 - Pl.02eJ 980 960 940 Wnvenumbers 1000 920 (a) f'ifference FT -l"R. soectrum of T)I\!Nf) l-d2 from VIS photolvsis (4 h) of 3ci2 (Ar, l0°K). P.efore minus after photolysis of l-d2 060-420 . nm), 150 min. (b) f'ifference FT -IR. spectrum of f"\!\!Nf' 1-<fl and l-d2-15N from VlS photolvsis (4 h) of 3-d2:3-<f2_15"l mixture fAr, l0°J<). Before minus after photolysis l-<f2:l-<f2-15N (360-420 nm), 150 min. Wo.venumbers 1000 Fi~re 31. PJ.000 - Pl.PJ20 I I (]'\ './) -97- shift of 1.93 cm-1 ), respectively. The band at 947.39 cm-1 corresponds to the known strong infrared band for trans nNND l-d2 found by Rosengren and Pimentel2C at 946.2.:!:. 0.3 cm-1 in a nitrogen matrix. The reported 15N isotope shift for tl)is band is 2.0 cm-1 to 944.2 cm-1. This is very close to the observed 1.9'3 cm-1 15N shift to 945.46 cm-1 in an argon matrix. These bands are efficiently photolyzed away with either monochromatic light (A_= 3go.:!:. 10 nm) or broad band irradiation (A.= 360-410 nm) (see third column Table XI) consistent with the assignment to trans nN .~D l-d2 and 1- d2_l5N (Figure 31). Table XIV summarizes the observed FT-IR. assignments for l-fNNl-f 1, f)NND 1-d2 and their literature assignments)c The band at 965.95 and its 151\t multiple at 964.50 cm-1 are also photolyzed away with 1- Table XIV. Infrared Rands for HNNH 1 and DNNn 1-d2. Comoarison.2c 1,2-T)iazene 1 Observec:@ (cm-1) LiteratureQ (cm-1) 1288.01 128.5 . 8 Difference (Obs.-Lit.) (cm-1) +2.2 (+0.2)~ (1286.02)~ 1-15N Literature 1286.S!1 1284.5 +2.3 1.20 1.3 -0.1 l-d2 947.39 946.2 +1.2 l-d2_15N 945.46 944.2 +1.3 l5N shift 1.93 2.0 -0.07 15N shift -98- d2. The assignment of these bands remains uncertain. Attempts to selectively photolyze or interconvert these bands with the bands due to 1-d2 (as a possible assignment to 2-d2) have been unsuccessful. Photolysis of 1-d2 as for 1 results in the increase of the bands due to formaldehyde (D2CO) and two CO bands at 2139.9lj. and 2137..53 cm-1. Two CO bands at 214-9.82 and ? 14-3.32 are lost concomitant with the bands due to 1-d2. Identical behavior was observed in the CO stretch region with 1. This will be discussed in a later section. The third column of Table XI summarizes the photochemical behavior of l-d2. H2NN Product Analysis For product analysis, samples of 3 were prepared by photo-Curtius rearrangement/photodecarbonylation of carbamoyl azide 15 isolated in a rigid 2-\HHF glass contained in a 5 mm O.D. quartz tube immersed in a liquid nitrogen filled finger dewar at 770K. Gaseous products Hz, N2, and CO were identified by VPC coinjection and mass spectral analysis and quantitated by Toepler pump analysis. A.mmonia (NH3), hydrazine (N2H4-), ammonium azide + (NH4-N3), and formaldehyde (H2CO) were quantitated by standard colorimetric analysis. Product ammonium azide was compared with independently + synthesized NH4-N3 by FT -IR. Warming a blue-violet 2-\IJTHF glass of 3 from 800K to 900K results in the loss of 3 and formation of HNNH 1 (yellow, A max = 386 tetrazene (Hlj.Nlj.) 25 (A. max= 260 4-80 nm (see Figure 14-). nm) Warming and an further nm), 2- orange species A. max= results in loss of the -99- yellow-orange color to~ether with vigorous gas evolution and formation of an insoluble precioitate of ammonium azide. thermal rlecomposition of 3 reveals Analysis of the products for the formation of NH3, N2H4-, + NH4-N3 together with l-f2, N2, and CO. Carbon monoxide (CO) and N2 are the expected side procfucts for photo-Curtius rearrangement of 15 to aminoisocyanate and N2 followed by photo-decarbonylation to form 3 and CO. These products are summarized in Ta~le XV. Ammonia, hydrazine, ammonium + ~ + H2N=N 3 !:::. 3 H 2 N-N=N-N~ 25 NH 4 N3 .. N2H4 N2 + ~ NH3 + N2 + NH3 azide, and N2 are known thermal decomposition products of 2-tetrazene 25. Observation of these products together with a thermally labile intermediate 1- max = 260 nm Oarge extinction-coefficient) suggests that 3 dimerizes to form the expected 2-tetrazene 25 0... max = 260 nm) which subsequently + thermally cfecomposes to Nl-!3, N2H4-, and N2 and tautomerizes to NH4N3. The mechanism for the thermal decomposition of tetrazene 25 is unknown. Attempts to directly monitor its disappearance by electronic spectroscopy + are severely impeded due to simultaneous formation of NH4-N3 precipitate -100- Table XV. Products from UV Photolysis of 1.5 Followed by Thermolysis of 3~ Product Ratio2 co Ratio.Q o;,., Yield~ 29.0 1 Ji5 50.0 21.7 49.3 0.40 12.0 13.3 30.2 0 12 4.0 4.0 9 1 1 2.3 4.0 9 1 0 !I.J?H4 % Yiel~ 0.03 0 + ~H4!1.J3 0 12 0 4.0 0 H 2 co.1 1.50 m2; 1.5 in 300 L 2-MTHF, ( 15~ = 58 mM~ t1v 2-CS-7-54 filters, 5 h, 770T<; warmed to R.l!f. EMolar ratio, normalized. Avera~e of tl,ree determinations. Individual yields .:!:.1 0% of each. Total mass 97 .:!:. 10°1-- typically. ~~ormalized from ratio in b • .Q(c) Corrected for CO and 1 equivalent of N2 as side products of photo-Curtius/decaroony1ation of 1.5 ..... 3 + N2 +CO. ~Normalized from ratio in _Q • .!Formaldehyde not detected. ~Typically togetl,er with vigorous gas evolution. Hence, kinetic studies were not attempted. The rapid and spontaneous nature of the thermal decomposition of 25 may suoport a radical chain decomposition mechanism (together with rapirl self-heating) or the presence of a catalyzed pathway (e.g., trace acid). Correcting the yields of products from thermal decomposition of 3 for the formation of CO anci one equivalent of '\!2 as side products of the photoCurtius rearrangementldecarbonylation route to 3 reveals an added comoonent of 'l-f2 and N?. as products (see Table XV). The ratios of Hz, N2, and CO products from thermal and photochemical decomposition of 3 in 2-MTHF -101- are summarizer! in Table XVI. Visible light ( >340 nm) ohotolysis of a blue-violet glass of 3 results in decolorization and formation of H2, N2, and CO in a 0.98:1.9 5:1.0 ratio. The nearly quantitative vield of gaseous products is consistent with a sequence consisting of UV pl,oto-Curtius/ decarbonylation of 1.5 to 3 + N2 + CO followed by direct ohotolvsis of 3 in the visible to produce l-f2 + N2 in an Table XVI. Thermal and Photochemical Decomposition of Hi\1~ 3. l-f2, N2, and CO 'R.atios~ Conditions N2/CO H2/CO h v,~ 6~ 1.65 0.40 h \ 1,2 h v (A >_500 nm )j,~ 1.8~ 0.82 h \1,~ h v r A >340 nm )~,.!. l.q5 0.9~ h v,~ h v fA >S()O nm ),Q h v fA >340 nm )~,_!. 1.94 0. 91) h v f!JV + VIS)~,_g 1.91) 1..04 ~\~olar ratios from Toepler oumo analysis. CO yield 95.:!:. 10% from warmed fR..T.) reaction mixture. .2 hv f2-CS-7-54 filters, 5 h, 770K). ~Warmed to 'R .T. Other products summarized in Table XV. Qh v (CS-1-75, 2-CS-3-70 filters, S h, 770K). 'Remaining products summarized in Table XVII. ~l-f2CO not detected, see 'Experimental. .!.hv (CS-1-75 filter, .S h, 770K). gh v (no filters, 5 h, 770K). overall expected ratio for l-f2:\J2:CO of 1.0:2.0: 1.0. UV photolysis of 1.5 with no visible absorbing filters results in no blue-violet color due to 3 and a nearlv quantitative yield of l-f2, N2, and CO in a ratio of 1.04:1.96:1.0. -102- The added component of H2 and N2 in the photochemical ~eneration followerl hv thermolysis of 3 may represent actual thermal decomposition products from 3. Alterna-tively 'Y7 and N2 may be the result of secondary photolysis (S 1 or T 1) during photochemical generation of 3. Competing relaxation/ decomposition pathways from photoexdted arninoisocyanate to excited 3 (~ 1, T 1' and C0 followed hy decomposition of excited 3 to H? and "J? Prociucts represents a reasonable alternative source of H2 and N2. Additionally, exdterl 3 mav isomerize to 1 which would be efficiently decomposerl at the wavelengths used to generate 3 from 15. Irradiation of 3 with visible light (). > 500 nm) affords yellow 1 ().max = 3U; nm). Subsequent warming to room temperature results in loss of + I anrl formation of NY4N3, N2H4, NH3 in addition to H2, N2, and CO in a 0.~2: t .88:1.0 ratio. These prorlucts are summarized in Table XVII. Photolysis of 3 appears to afford 1 in approximately 10 to 15% yield based on the rerluced yielrl of H? when compared to that obtained upon photolysis (> 340 nm) of 3. Thermolysis of 1 appears to result in formation of NH3, + N7H4, and ~TH41\13 as observerl from thermolysis of 3, which to some extent + also yields 1. Wiberg found2g 1\12H4, NH3, NH4N3, N2 and H2 as thermal products from decomposition of neat 1 near 770K. Willis found'37 only N2H4, '1-l?, anrl "J? as thermal prorlucts for gas phase decomposition of 1 (1 00°C). Willis proposed that hvdrazine results from isomerization of trans 1 to cis 2 + followerl by disproportionation. The mechanism for '\!H3, N2H4, anrl NH4N3 formation from 1 is less obvious. T)imerization of 1 to form tetrazene 25 has been reported. 3 5 Subsequent thermolysis of 25 would explain the -103- observed product~ The ~reater proportion of N2H4 obtained from thermolysis of l oroduced from VIS photolysis of 3 compared with that obtained from thermolysis of 3 could result from competin~ isomerization to 2 followed by disproportionation. The orange A. max = 480 nm species from thermolvsis of 3 (tentatively identified as cis 2) is also seen transiently in thermolysis of 1 or educed from photolysis of 3 J VIS Photolysis (> 500 nm) of concentrated samples (0. '> M) of 3 and 3-d2 occasionally resulted in formation of the orange >..max= 4~0 nm species.69] Attempts to trap H2N2 Table XVll. Product f)ecomposi tion Products from P hotoisomer iza tion of 3 to 1 Fallowed by Thermal Decomposition. Ratio~ co % Yield!2 Ratio~ % Yiel~ 26 1\T2 l.8~ 50 44.0 49 H2 0.~2 22 41.0 46 NH3 0.02 0.5 1 1.1 N2H4 0.05 1.3 2.5 2.8 0.3 0.8 1.5 1.7 + NH41\T3 ~~~olar ratios see exoerimental. Tyoically 1. 50 mg 15 in 300 1-1L 2-MTHF, hv ?-C<;-7-54 filters, 5 h, 77 K; hv CS-l-75, 2-CS-3-70 filters, 5 h, 77 K; warmeri to R. T. QT otal mass yield 90 .:t 10% tvpicall y. Normalized from ratio in a. Average of two determinations each. Formaldehyde not detected. Individual yields .:t l 0% of each. ~Corrected for CO and 1 equivalent of N2 as side products of photo-Curtiusldecarbonylation of 15 -+ 3 + N2 + CO . .£!"'1ormalized from ratio in c. -104-- with a large excess of reactive olefin (which could result in reduction if cis 2 is present) are summarized in Table XVIII. Norbornene has been shown to be among the most reactive olefins toward diimide reduction70 (presumably cis HNNH 2). Thermolysis of 3 produced from VIS filtered photolysis of 15, in a 2-MTl-ft=:- glass containing a 50 to 100 molar excess of norbornene does result in a low but measurable yield of norbornane. VIS photolysis of 3 (>500 nm or >340 nm) in the presence of the same excess of norbornene followed by thermolysis results in approximately the same degree of reduction (approximately 7% of 1 equivalent of 15). However, UV photolysis Table XVID. Reduction of Norbornene. % Yiel~ Condition~,!?_ 15 d:;; ~ hv,Q ~ 1 50 7.1 ..±..0.3 h v,Q,_!_ 11 1 100 6.~ . ±. 0.3 hvj hv (>34-0 nm),g !J. 1 50 6.6 . ±. 0.3 h v,Q h v ( >500 nm ),_!:! !J. 1 50 6.7 . ±. 0.3 h v (full lamp)J_ 11 1 50 7.2 . ±. 0.3 ~Photolysis in 5 mm quartz sample tul:>e, Suprasil finger dewar, 77 K. QMolar ratio 15:norbornene (15) = 4.5 x 10-2 M, in 2-MTHF, 300 11L sample. ~% Yield based on ( 151, n-octane as internal standard, 15 m SE54 capillary, R.T. ..2hv, ~-CS-7-.54 filters, H20 filter, quartz optics, 4 h. ~Warmed to R.T. ...!Porms opaque, fractured glass • .?hv, CS-1-75 filter, 4 h. .b.hv, CS-1-75, 2-CS-3-70 filters, 4 h• ..!h v, quartz optics, H~O filter, 4 h. -I f) 5- of 15 with no visible filters (which gives no !)lue-violet color of 3) results in the same de!:!;ree · of reduction within experimental error. These studies imPlicate a mechanism other than direct H7N2 reduction of norbornene. T)irect H abstraction by an excited state of norhornene7 2 and/or H· arldition (prorluced from photolysis of 3) are possible alternative reduction mechanisms. This was not explored in greater rletail. Matrix FT-J'R. Studies of Carbon ~onoxide <;everal grouos of banrls showing no isotone shifts in the photo-Curtius rearrangement/photodecarbonylation of 15, 15-15N, 15-d2, and 15-d2_15N have been assigned to r:o, a· co-product in the photochemical generation of isotonically labelerl 3. The appearance of several groups of CO bands each mirroring the hehavior of the photogenerated diazenes 3 and 1 and their nhotochemical rlecomnosition products (H?, 1\J'-) suggests an intimate association or matrix site arrangement of these products with a contiguous CO molecule in an argon matrix.9 l UV (VIS filtered) nhotolysis of 15 affords CO, l\J2,anrl 3 in arlrlition to formaldehvde and aminoisocyanate FT -117,. bands (see Figure 18). Figure 32b shows the CO stretch region of the infrared spectrum obtained from UV (VIS filtered) photolysis of 15. Four CO bands at 2140.90, 2139.94, 2137.??, and 2137.53 cm-1 are observed. Photolysis of 3 with visible light (). >500 nm) results in loss of the CO band at 2140.90 cm-J with loss of the hamis r!ue to 3 fFigure 37d, growth of two C0 bands at 2139.94 and ? 137.51 cm-1 together with formation of two new CO bands of equivalent intensitv at 2149.82 and 214 1.3? cm-1. As described earlier, VI<; photolysis -106- of 3 afforcls mainly Y2 and 1\12 (:2-MTHF, 770K), trans HNNH 1 (2-MTHF, 770J< or Ar, tOOK), and formaldehyde '""2CO ( A..r, 1QOK). Formaldehyde very likely results from successive hydro~en radical reduction of co62a conti~­ uous to 3 anc1 will be discussed in a later section. Photolysis of 1 (A..r, lOOK) results in loss of its infrared bands, loss of the two CO bands at 2145.82 and ?143.'32 cm-t together with ~rowth of the two CO bands at 2139.94 and ? t 1 7 .5'3 (Figure 3'2cn. The Photolysis products of 1 are mainly H2 and N2 (2MTl-p:::·, 77 I() (Table XVJ). \~atrix isolaterl complexes of C:O with various small molecules has .,een ciernonstrated to result in shifts of the CO infrared stretch due to hydrogen bonding fe.g., H20, NH3), dipole-dipole (e.g., CO dimers) and electrostatic (e.g., N?) interactions. 7 1 ~orne of these effects together with the observed anri assigned CO hands from matrix studies of isotopically labelerl 3, I, and their Photodecomposition proclucts are listed in Table XIX. Matrix isolation of CO (1 :?000, Ar, lOOK) affords a single infrared .,anrl at ?t3~.?'5 cm-1 in argon (Figure 3'3a). In a nitrogen matrix (1:2000, N2, Jr:)OK) the C::O stretch is observed at 2139.69 cm-1 (Figure 33b). Matrix isolation of CO in a mixture of nitrogen and argon, CO/N2/ Ar (1 :20:1000, JOOl() shifts the CO stretch to ? 136.37 cm-1. The two CO bands which are observer! to grow in with pl)otolvsis of 3 anrl 1 is an an!;on matrix at 2139.94 anrl ? ! 37.53 cm-1 are likely assigned to "complexes" or sites of CO contiguous to ~.r, (possibly two N?.'s) and 1-J2 (photolysis products of 3 and 1). The nearlv one-to-one ratio of intensities of these bands suggests two equally formed 1\1 7 + H,lr:O sites in an argon matrix from photolysis of 1.5, 3, and l. -107- 0.55 0.50 0.45 R 0.40 b s 0.35 0 r b 0. n c e b 0.30 0.25 c 0.20 0. 15 d 0. 10 0.05 I 2150 2140 2130 2120 Wo.venumbers Fi~ure 32. <;uccessive FT-IR soectra of CO region (Ar, lOOK). (a) f)efore ohotolvsis of 15 (l:~WOO, Ar, lOOK). (b) After UV (VIS filtered) ohotolysis of 15 (2 h) to form '1;21\1"-J 3 and CO. (c) After VIS bSOO nm) ohotolysis of 3 f? h) to form H"'lNH 1. (d) After ohotolvsis of l f%0-4'-0 nm) 100 min. -108- Table XIX. Matrix Isolated Carhon Monoxic:ie (CO) Infrared 5tretch Frequencies. CO/Matrix (R.atio) Ohs.~ v CO (cm-1) ('J)I !i..r ( 1:?000) CO/'N? Lit-2 v CO (cm-1) (~atio) 21'3~.25 2138.40 ( 1:1000) (1 :'?000) '2n9.69 2139.69 (1:1000) COI"J?_!'\r (1:9:1~00) ? 13h. '3? 2136.60 ( 1:20:2000) COIH?.nl Ar (l :?:'?000) 214g.60 214R.R (2: 5: t 000) 214'3.'3 (1:1:1000) ?l '39.9 c r_n/I\1 Y 3 1A.r r_n "c:iimers"IA.r c 2139.40 \.OIH?_"J 1'.T I Ar -rl 2140.90 CO/T)?_'r\.1 "-J I Ar -d CO/H?'r\.!'f\T 1"3? -rl C01t-4N'r\.JYI A.r -d CO/t-!)1\l'r\.Jnl Ar rl COI.!-Y'r\.11\Jf-f I"'J2 d 2149J~2 214'3.32 2149.82 ?.143.32 214'3.~0 21 V>.~4 C0 ll-J 7.IN? l"l? I A. r d 2139.94 ?13~.97 ?.137. 53 d 2139.94 2137.53 d 2139.1)9 ~This work (tOOK) • .!?R.ef. 71 (JOOK). ~C0/Ar warmerl to 300K and back to lOOT<. .£Matrix ratio unknown estimated at (1:1:>2000) from photolysis of 15 (rvi:?OOO, fi)OK). -109- 0.30 0.28 0.26 0.24 0.22 R 0.20 b 0. 18 s 0 0. 16 r 0. 14 b 0. 12 a 0. 10 n c e 0.08 0.06 0.04 b 0.02 0 . 00 -0.02 -0.04 2150 2140 2130 2120 Wo..venumbers Fi~ure33. fa) FT-T'Q snectrum of r:n (J:?OOO, .1\r, soectrum of l.n (1 :?000, 1\J,, l OO!<). lQOK). (b) FT-!1~ -110- ~.50 ~.55 3.50 0.45 R b 0.40 s 0. 35 0 r b a n c e ~ .30 3 . .25 0 . .20 b 0. 15 0. 10 c 0.05 2150 2140 2130 2120 Wo.venumbers Figure 34. (a) FT -TR soectrum of CO region after UV (VIS filtered) ohotolysis (?h) of 15 (1:2000, Ar, l()Ol() to form 3 and CO. (b) After warming (a) to 350'K 3 min back to li)O'f<:. (c) After warming (b) to ~50K 10 min hack to 1()O'f<:. - t 11- ~.50 0.Ll5 0.Ll0 R b s 0 0.35 0.30 r b 0.25 a. n c 0.20 e 0. 15 0. 10 0.05 2150 2140 2130 2120 Wo.venumbers Figure 3.5. Successive FT -I~ spectra of CO re~ion (N2, 10°K). (a) Before ohotolvsis of 15 (1:2000, N,, lOOK). (b) After UV (VII:) filtered) photolvsis of I 5 (? h) to. form 3 anrl CO. (c) After VIS (> 500 nm) ohotolvsis (2 h) of 3 to form l. (d) After photolysis of 1 (160-4?0 nm), ? h. -ll2- The growth of these r:n hands with seauential photolysis of 15 to 3 to 1 and l to oroducts fH7, 1\J?, 42~()) supports the assignment of these CO bands at ?.139.q4 anrl?! 37.53 em- I to two r.0/'12 sites. \'~:'arming an argon matrix of 3 to '350'< allows rliffusion of small molecules (e.e;., CO, N7.) and results in broadening and "merging" of the CO stretches (Figure 34 ). As an adrled test for these uniaue N?_ICO site assignments, UV (VI'i filtered) ohotolysis of 15 in a nitrogen matrix C.r 1:?000, N2, l I)Ol() affords only two CO bands at 2139.69 ---- - anrl 7135.~4 cm-1 (Figure 3'ib) together with hanrls due to 3.6 .5 VIS photolysis of 3 results in a diminisherll:>anrl at ?.135.~4 em-! (Figure 35c) together with loss of the infrared bands of 3 anrl growth of the CO banrl at 2139.69 cm-1. This is consistent with loss of the unique N2/CO sites found in an an~on matrix rlue to replacement of .A.r with N? molecules (degeneracy). VIS ohotolysis of 3 in a nitrogen matrix also results in formation of two r:o hands at ?.l'35.~lJ. and 2l43.SW cm-1 (Figure 35c) (analogous to the 2149.1P and ?.143.32 cm-1 banns in A.r, tOOK) together with an infrared band of l. Photolysis of l results in loss of these bands (Figure 35d). Hydrogen bonding to CO (e.g., H70, Nl-i3) shifts the CO stretching frequency to higher energy (Table XIX). The shift is '!-l-bond donor dependent (e.g., oK:a). 7 3 The stahilization results from H-bonding to the 5a lone nair of en at carbon which reduces anti!)oncling character in the CO bonrl (higher energy stretch). 7 3, 7 4 T)ipole-dipole stabilization of CO also results in a higher energy CO stretch. The small dipole moment of CO ( 11= 0.12 T)) is oriented toward carbon.75 Matrix isolated CO dimers (headto-tail) sl-tow a higher energy infrared stretch than free CO (Table XIX). The -113- H \+ - I N=N: . .. . H Ar Matrix N2 Matrix CO band at ?.140.90 cm-1 ( Ar, lOOK) which is lost with VIS photolysis of 3 is consistent with C::O stabilized by dipole-dioole interaction and/ or hydrogen bonding to 3. The large calculated dipole moment for 3 (4.036 T)) 1n should lead to favorable dioole-dipole stabilization of C::O in the proper orientation. The corresponding CO band in a nitrogen matrix is found at 213 .5.~4 cm-1 which is destabilized Oower energy stretch) compared to CO in N2 (2139.69 cm-1 ). The destabilization may be due to the inability for CO to rotate to a favorable dipole-dipole or hydrogen bond stabilized orientation to 3 in N2 at JOOT<. The two equivalent intensity CO bands at ?.149.82 and 21433?. cm-1 formed from VIc;; photolysis of 3 (Ar, lOOK), which are lost with subsequent o,_,otolysis of l suggests their assignment to two types of stabilized CO/l complexes. The magnitude of the higher energy shifts correlates well with the hvdrogen bond stabilization of CO encountered with r;2o or NH3 in an -114- I s- s· H '·I N=N H hv c-o H \~=N N=N CeO H Of s- s· H\ 3 N=N "linear l I s- s· H ' II c~o \H one allfl>lex ; 't/ c 8- N-N I hv I H H Of 0 8.. 1 C 8- H 3 0 8.. c'* 8- \-N II ' \ hLo CXJ1l)le xes non-linear " argon matrix (Table XIX). Photoisomerization of 3 to l could occur as a simple 1,2 H shift94 in electronically excited 3 (S1, T1) or alternatively, through bond cleavage to diazenyl radical (H"1N·) and H• followed by recombination.9'> If in the C0/3 complex CO and the N=N double bond of 3 lie coaxial (C'2V symmetry) then the complex of CO and 1 resulting from 1,2 H shift of excited 3 should result in a single complex of CO and 1 by symmetry (assuming no rotational diffusion). However, if the CO axis is nonlinear relative to the C'2 axis of 3, the photoisomerization of 3 to l should H \ ~c I N-N.. H ~0 H \+ - hv H I N=N 3 -11 5- result in two C0/1 complexes due to symmetry. Photolysis of aminoiso- cyanate to H21\1N 3 plus CO woulcl be exoected to yield a non-linear arrangement of CO and 3 due to the anticipated bent geometry of aminoisocyanate. 44 Observation of two CO stretches assigned to two complexes of l and CO formed with equal probability (1:1 intensities) at 2149.82 and 2143.32 cm-1 is consistent with a photochemical 1,2 H shift of 3 to 1 in a non-linear Scheme m hv (2139.5, 2137.5 cm-1) C:!!Q (2140.9 cm- 1) ,,- ... , C=O ~N2,' ... _, H \.N=N- + I H J (2149.8 cm- 1) (2143 3 cm- 1) C!!!O . C==O ® ,-~--, I ' I ,_, I + I H N=N I H ,-N-,' -- \'- 2'I 1 ,--, I \ N2' ,_, I -116- complex with CO. The lower energy CO stretches observed in a nitrogen matrix 2135.84 and 2143.~0 cm-1 su~gest a less stabilizing orientation of 1 and CO in 1\!?. produced from photolvsis of a destabilized 3/CO complex. Scheme III summarizes the CO matrix behavior. ES~ Studies The rlerivative electron soin resonance (ESR) spectrum of a 2-VITHF glass of 3 (I ()OK) reveals a structural signal in the ~ = 2 reg;ion, an apparent suoemosition of two si~nals (Figure 36a). The anisotropic nature of the glass orecludes a orecise determination of the J'20 gauss hyperfine splittings. The F.c;;-q_ soectrum of a 2-MT'!-ft::' glass of 3-d2 reveals an identical signal. VIS irradiation of 3 or 3-d2 with ().. > 500 nm, J' t OOJ<.) results in growth of the structural portion of the signal with the center doublet portion unchanged. The structured oortion of the signal corresponds to 2-MTHF radical(s) . . independently produced from photolvsis of trans-1,2-dimethvldiazene 8t or di-tert-butvl oeroxide in a 2-V!THF glass (770K) (Figure 36b). The identity of the center doublet portion of the soectrum is unknown.76 A similar spectrum to "~='igure 36a has been obtained from y-irradiation of a 2-MTHF glass.77 Prolonged T.JV (VIS filtered) photolysis of a 2-MTHF glass alone (770K) under conditions used to generate 3 does not yield the soectrum of "~='igure %a. The formation of ?-1\HHF radical(s) in ohotolysis of 15 or 3 is consistent with the generation of H• from 3 followed by abstraction from the matrix. VIS ohotolvsis of 3-~ prepared form 1.5-d2 (>97% d2) results in a mixture of n7:Hr:":H2 in a ratio of 75:23:2 by mass soectroscoov. This is consistent with -117- ------------ -, \.'- ,-, _,/ / _ ....... ,.,----- ------- ~--------------~ 3160 3220 3300 3380 Field (Gauss) Figure 36. (a) E'5'R spectrum of hlue-violet glass of H2NN 3 (2-\HHP, t0°l(). (b) t.'5'R spectrum (lOOK) of 2-"..nl-JF radical(s) produced from photolysis of either trans-1,2-dimethyldiazene 8t or ditert-butvl peroxide in 2-MTH~ glass, 770K. · -tl~- 4 abstraction from 7.-MTHF solvent bv T).. Failure to observe formaldehyde as a orocfuct from 3 in ?-11.iTl-IF mav reflect a preference for H abstraction from solvent to l-f· addition to CO. However, in the absence of an abstraction pathway f 1\r, l QO'I() nhotolysis of 3 and 1 in the presence of CO then affords formaldehyde. No signal attributable to diazenyl radical (HI\lT\1•) was observecf from VTS photolysis of 3 in ?-MTHF (1 OOK)78 or in an argon matrix I OOK.79 I\ concerted photochemical reduction of C0 hy 3 or 1 cannot be ruled out. Thermal Decomposition Kinetics The difficulties and perils associated with kinetic studies in rigid and softened organic glasses have been noted by others.81 Frequently, very raoid non-exponential decays are intermediates in these media. noted in the decomposition of reactive A rigorous interpretation of this behavior is lacking, but may be ascriber:l to matrix site effects or contributions of quantum mechanical tunnelling. The rapid thermal decomposition of H2NN 3 in a softened ?-~'HHF glass at 90.:!:. 1OK was monitored spectroscopically at f.36 nm. J:arlv kinetic points prior to temperature equilibration (.r30 were discarded. min) A linear least sauares analysis reveals an apparent bimolecular decomposition Pathway for 3 from the "best fit" !/absorbance versus time plot (Figure 37). Assuming a molar extinction coefficient ( e:) of ?.0 M-1 em- l (for kinetically persistent l, 1-diazenes e: = 20 .:!:. 3 M- I em -1) yields a corrected bimolecular rate constant of 7.2 x 1o-4 M- l sec-1 (r2 = 0. qq l) at 90 .:t 1OK. The estimated diffusional rate constant for 2-MTHF at -119- 90 .± 1OK is ~ x Io-6 102 M-1 sec-1.82 Hence, the observed bimolecular decay of 3 (k2) is apparently greater than or equal to the estimated diffusional rate constant. T)ue to the extreme variation of solvent viscosity with temperature82 the rate of decay of 3 was not determined as a function of temperature. - 120 - • E c ~ 1'0 ~ w u z <t CD a:: /. • 0(f) CD <t ......... o2NN • • 20 • • 40 • • 60 • 80 TIME (min) Figure 37. P,imolecular kinetic olots for decomposition of 3 anci 3-d2 at 90 .± l Ol( in ?- MTYF from same initial concentration at 80°K. -120- Thermal decomposition of Di'JN 3-d2, prepared indenticallv to Hi'!N 3, in softened 2-~-"THF at 90.:!:. l OK affords the "best fit" kinetic plot shown in Figure 37. The corrected bimolecular rate constant of 6.R x l0-5 M-1 sec-1 ( r? = 0. 998) is indicative of a substantial deuterium kinetic isotope effect kH/kn ~ 10.6 + 3 at 90.:!:. 1°K. 'Simple head to head dimerization of 3 to afford 2-tetrazene 25 (as is observed for kinetically persistent 1,1-diazenes in solution)~g would be expected to show only a small inverse secondary isotope effect. 25 3 For a hybridization change from sp3 to sp2, a normal secondary isotooe effect is typically kHikT) ~ 1.1 to 1.4 per deuterium at 298°K. This would correspond to kHiko = J .4 to 3. t per deuterium at 90°K. For comoarison, a normal primary isotope effect kH/ko ~ 5 at 298°K corresponds to a kH /kr, ~ ?0() at 9()0K. Clearly a simple head-to- head dimerization of 3 to 25 at 90°K fso2 to so\ kHII<o ~ 0.71 to 0.32 per deuterium) would not account for the ohserved isotope effct for bimolecular decomposition of 3. -!21 - The observed kinetic isotope effect mav be consistent with the mechanism proposed in Scheme IV. Scheme JV H \+ - N=N: !-" .. H~ )t - +' N=N H 3 N2 25 H \ H +; N=N .. \ H H \ l .. N=N .. \ H 1 2 -122- An initial concerted w2s + n2s + cr2s head-to-tail dimerization of 3 to afford intermediate 31 as the rate determining step would be expected to show the observed bimolecular decomposition kinetics and a deuterium isotope effect of approximately kr;/ko !::: 200/(3.2)2 !::: 20. comparable to the observed value kr;/ko = This value is 10.6 .:!:. 3 at 90 .:!:. 1OK. Tautomerization of 31 to afford '2-tetrazene 25 competitive with hydride elimination and subsequent disproportionation to afford trans HNNH 1, cis HNNli 2, H2 and N2 or addition to afford hydrazine would account for the products observed from thermal decomposition of 3. Unfortunately, kinetics for formation of l (A.max = 386 nm), 2 (A max= 480 nm) (tentative identification) and tetrazene 25 gave inconsistent results. Subsequent thermal decomposition of 1 and the Amax = 480 nm species at 90 to 1OOOK were also inconsistent and do not support a clear interpretation. The mechanism of Scheme IV may be consistent with the low activation energy (Ea < 5 kcal/mon necessary for the facile thermal decompositions observed at these low temperatures (assuming log A < 5). However, any mechanistic considerations derived from the limited kinetic data oresented must be tempered with the consideration of alternative mechanisms. Due to the presence of radicals (e.g., ·MTHF') in samples of 3 prior to thermolysis f see previous section) a facile radical chain decomoosi tion of 3, I, 2, and 25 cannot be excluded. This alternative mechanism could provide the necessary raoid, low Ea pathway for the observed decompositions. Unsuccessful attempts were made to prepare 3 isolated in a "rigid" high temperature matrix (e.g., adamantane, polymethylmethacrylate, -123- polyethylene) in order to investigate the "non-diffusive" thermochemistry of 3. unimolecular It was hoped that activation parameters could be derived for comparison with theoretical calculations for the N-H bond strength of 3 and the H2''12 energy surface. However, in all cases the blueviolet color of 3 was lost well below 2000'K, resulting in formation of the yellow color of trans 1. It is unknown whether this isomerization of 3 to 1 is the result of diffusion of 3 in these "rigid'' matrices, well below their softening points or whether a low activation energy pathway for unimolecular decomposition of 3 (e.g., tunnelling) exists. The yellow color of l persists for several days at ~9~0K in a polymethylmethacrylate matrix. This is suggestive of slow diffusion in this matrix. In contrast, the blue-violet color of 3 in this matrix is gone in less than 20 seconds upon warming from 770K to 120oK. Summary The low temperature matrix isolation and direct spectroscopic characterization of the parent 1, 1-diazene H2NN 3 has been ciescribed. The UV (VIS filtered) photolysis of carbamoyl azide 15 in a rigid medium (2-MTHF glass, 8QOK or A r matrix, 1QOK) provides a new method for the generation of reactive 1, l-diazenes. This photochemical rearrangement is considered to proceed by the photo-Curtius rearrangement of 15 to aminoisocyanate followed by photodecarbonylation of the aminoisocyanate to afford H2NN 3 is a blue-violet species A-max = 636 nm and carbon monoxide. H2NN displaving the characteristic now structurecf absorption curve -1?.4-- of a 1, 1-rliazene. The characteristic N=N double bond stretch for 3 is located at 1574.16 cm-1. This is consistent with the theoretical (GVB-CI) prediction for H2NN. Isotopic labeling (T), 15N) has provided identification of the infrared active modes for 3, 3-l5N, 3-d2, and 3-d2-l5N. No emission was detected of 3 at its n-n * transition consistent with very rapid from irradiation internal conversion (kyc ..r 1Ql2 sec-1) controlling the lifetime of the excited state (S 1). This in turn is consistent with the relatively small So-S 1 gap of 41.5 kcal/mol in 3. The products decomposition of 3 were determined. of thermal and photochemical Thermolysis of 3 (2-MTHF, 900K) affords tetrazene 25 (A. max = ?60 nm) at a nearly diffusion controlled rate as expected. Additionally, thermolysis of 3 affords yellow trans HNNH l (A max = '386 nm) and another species Am ax = 480 nm tentatively identified as cis HNNH 2. The mechanism for formation of these unexpected products is unknown but suggests that 3 is indeed a higher energy species than 1 and 2. Photolysis of 3 (2-MTH.F, 800K and Ar, lOOK) affords some trans HNNH l together with J-J2 and N2. Photolysis of 3 in the presence of CO (Ar, 10°K) also affords formaldehyde (H2CO). The photochemical isomerization of 3 to 1 can be compared to the proposed isoelectronic photochemical isomerization of H2CO to hydroxycarbene which has been invoked to explain the much disputed formaldehvde photochemistry.66,94 Attempts to detect diazenyl radical (HNN·) in the photodecomposition of 3 by ES'R. and matrix FT -IR were unsuccessful. A tt~rnpts at photochemical isomerization of trans 1 to cis 2 were inconclusive. The observation of several FT -I'R. stretchin~ frequencies for CO mirroring the photochemical behavior of diazenes 3 -125- anrl l and t'"teir ohotorlecomposition ororlucts in solid matrices (Ar, N2, 1()Ol() suggests an intimate association of these diazenes and the CO side product in a rigid low temoerature matrix. -126- RESlJL TS ANT) DISCUSSION CHAPTF.R 2 l, 1-niazenes With a -Hydro~ens. Matrix Isolation anci Characterization of l, l-'Dimethyldiazene ancf I, t -T)iisopropyldiazene -1?7- Synthesis of Carbamoyl Azides 'Reaction of the requisite dialkyl amine with phosgene at low temperatures affords dialkyl carbamoyl chlorides 214 7 and 22)D Carbamoyl azides 13 and 17 are easily prePared ~y the reaction of sodium azide with the corresnonding carbamoyl chlori<ie in dry acetonitrile.47 Preparative VPC affon·is analvticallv pure carbamoyl azi<ies 13 and 17. R2N-H 21 13 22 17 Electronic Absorption Spectra of 1,1-T)iazenes 7 and 18 P or dilute elP.ctronic absorption SPectroscopy of l, 1-<iiazenes 7 and 18, a (..r 5 x l o- ""3 ~~) 10? solution of the freshly purified carbamoyl azide in dry, degassed ?.-methyltetrahydrofuran (2-~~TJ-lF) was loaded into a l.O em path length low temperature spectroscopic cell '>IJ attached to the low temperature matrix isolation apparatus. The solution was loaded under a positive pressure of argon through Teflon tu!Jing with syringe suction. Cooling the solution to MO'!( results in the formation of a rigid oPtically transparent glass. -12 8- Irradiation of carbamoyl azide 13 in a 2-MTHF ~lass at 8001( with ultraviolet (UV) light from a 1000 watt xenon lamp51 through two Corning CS-7-54 ( UV transmittin~ , visible (VIS) A 400-680 nm cut-out) filters results in loss of the UV absorotion of the carbamoyl azide and formation of a redorang;e oran~e Electronic ~lass. glass absorption spectroscopy reveals a structured absorption of t his red- curve Amax = 556 nm (51.5 kcallmoi' and A 0,0 = 653 nm (43.9 kcallmol)52a assigned to 1,1-dimethyldiazene 7 together with a structureless absorption due to another species A max= 464 nm (Figure 38). Similarly, irradiation of carbamoyl azide 17 in a dilute glass (!- 5 x 1o-3 M, 2-MTHF, 800K)102 affords a purple glass. Electronic absorption spectroscopy reveals a structured absorption curve A max= 504 nm f52.7 kcal/mol) and A. 0,0 = 620 nm (46.2 kcal/mo1)52a of 1,1diisopropyldiazene 18 (Figure 39). These structured absorptions are close to 0 R2~N3 hv ~ 2 N-N=C=o] + N2 ! + R2N=N +CO R = CH 3 13 7 R = (CH 3)2CH 17 18 -129- lLJ (.) z <l: CD 0: 0 en m <l: 0.0 400 Fi!!Ure 38. 500 600 WAVELENGTH (nm) 700 Electronic absorotion spectrum of 1, 1-dimethyldiazene 7 and Amax =464 nm co-product (~-MT4'1:' glass, ~OOK). -130- w (.) z <! m 0::: 0 (/) CD <! 0.0 400 500 600 700 WAVELENGTH (nm) Fi~ure 39. Electronic absorption spectrum of 1, 1-diisopropyldiazene 18 (2MTHF glass, 8QO'f() -1'31- Table XX. l, t _T')ia?:ene T:lectronlc Transitions. v em- 1 1, 1-1""\iazene H \+ N=N I P.f';o-"1) (max) (I< can a (;% {,95 1'3 t 5 lj.), 1 '55{, 653 1115 51.5 '5/g (){lj. 1340 5LL.6 '50lj. '5~0 1306 '56.~ ')lL '3 610 1040 '5?.7 XN=N b 7\ '506 670 17.00 5{,,6 ~N=N lj.97 572 123g 57.6 H H3 C\+ I _ a N=N H3 C c CN=N ~+N=N a Q=N b -1 b ~This work. ?- ~ HHF= gl ass, 800'<. 2"-~e?O solution. -q ef. ~e,f. ~~ef. ~4. 7.~HHF glass, ~')O K . -132- the n-1T* transition calculated by navis and Goddardl n for H2NN 3 near 560 nm. The kinetically persistent t, 1-diall<yl diazenes 4, 5, and 6 prepared earlier in the T""'ervan group were found to have similar structured absorptions fl..max = 4'n to 543 nm) clue to this n-1T* transition in the visible.~e,f The parent t, 1-diazene 3 has also been shown to be a blue-violet species with a structured absorption Amax= 636 nm in a 2-\~THP glass at 8')0K. 4 The characteristic n-1T* transitions for t, t-diazenes are summarized in Table XX. The structure in the absorotion spectrum of l, 1-diazenes is assigned to the vibrational sryacing of the excited state ('i 1). The prominence of this 1\1-1\1 stretching mode is indicative of a substantial change in Re(NN) on excitation from ground state ('So) to 'it. The calculations of navis anrl Goddard In im!icate an optimal oyramidal 2;eometry for S 1 of 3. average vibrational spacing for 3 is 1315 cm-1. The experimental Comparison with the calculated "'1-N stretch for 'St of 3 (13'>7 cm-l)'>?b may indicate a preference for pyramidalization over two center three electron bonding in S 1 of 3. The average vibrational spacing in S t of 1, l-dimethy1diazene 7 is 1315 cm-1 comparable to that of 3. 'Ply comparison, the average spacing in S 1 of 1, tdiisoPropylriiazene is 1306 cm-1. The smaller vihrational spacings observed for the kinetically persistent l, 1-diazenes with tertiary alkyl substituents in solution (t ()1;0 to 1?38 em- l) may be reflective of the poorer resolution of their vibrational structure. In general, pyramidalization should stabilize I, 1diazene 'S 1 states bv relieving nitrogen lone P?ir-alkyl group repulsions. These interactions are certain! y more severe in tertiary substituted I, 1diazenes. The observed lower St N-!\1 vibrational frequencies in the I -( ~+N=N CN=N H3 C N=N H3 C\+ I H _ N=N \... H 1, 1-'"'iazene a d a a 504 5n ')')t; 4% fnm) A max 56.~ '>4J, ') 1 • ') l~ ') • 1 fkcan F.r"n-" 1) 99.4 95. ') Jn?.4 9?.4 rt<can '=:J"o-.;; 1) 7~8 1()0 279 110 A. max (nm) c -\C=O -! Cc=o ~c I H3 C\ C=O C=O H/ \ H Carl)onyl Tahle XXT. Comparison of 1, 1-'"'iazene anrl ~orresponrling r.arhonyl P.lectronic Transitions. I I \,>) \,>) .- b b '+97 ')(),; ')7 J) 'iiS.f.; "i? .7 Ff'in-" 1) <t<can 96.7 q,; .l.j. 91.9 <kcan FJSo-S l) ?.96 ?97 105 A max <nrn) MTHF glass, ~()OK. ~This work. ?-~AT4~ glass, ~noK. ~~,e?() solution. 'q~f. ~~,f. ~n-A.lkane solution. \I XN=N 7\ ')4 '3 Q=N b fnm) Amax 1, 1-niazene Table XXI. Continuer!. c q~f. ~'5. s!qef. 8'L 7\ ~C=O Q=o Carbonyl I ?.- I ·~ vJ - -135- kinetically persistent 1, 1-diazenes would be consistent wit!) this exolanation. However, in the absence of tertiary alkyl-lone pair repulsive interactions the average vibrational frequency in S 1 should follow a trend indicative of the necessity to relieve alkyl-lone pair repulsions. Small grouos (e.g., H, Me) sl)ould require less oyramidalization. Larger groups (e.g., i-Pr) should require more pyramidalization resulting in a lower S 1 vibrational frequency (as is observed'. In tl)e unsuhstituted five-membered ring 1, 1-diazene pyramidalization is restricted resulting in more two center three electron bonding and higher "1-'\1 vibrational frequency in S 1· TI The electronic n- * transitions of 1, 1-diazenes show a trend similar to that o'bserved in the isoelectronic Carbonyl COmPOUnds (Table XXI). l, 1-Di methyldiazene 7 was also prepared by UV (VIS filtered) ohotolysis of (~)-3, 1-cfimetl)vl-l-phenyltriazene-1-oxide40 16 in a rigid 2\~THF glass f~QOJ<) affording the structured absorption curve of 7 (purple, Amax = 5 56 nm) together with the structured absorption of nitroso'benzene (blue, Amax 780 nm, e: = 4 5) (Figure 40). Comparison of these absorhances (CH 3 )2 N-N~O '-'N7 I hv H3 0 c, ~ N=N + N I I ~c Ph Ph 16 7 -1%- t w u z <! CD 0::: 0 (f) CD <! 500 600 800 WAVELENGTH (nm) Figure 40. l:lectronic ahsorption spectrum of 1, l-dimethvldiazene 7 (A max = '5 56 nm' and co-Product nitrosobenzene ().max = 7~0 nm) from UV (VI~ filtered) photolysis of 16 (2-MTI-JF, 800J<). -1'37- orovides an estimate for E = 6 (l\~-1 cm-1). the molar extinction coefficient for 7 of Warming the glass gives a nearly quantitative (>90%) yielcl of tetramethvl-?-tetrazene 19 and nitrosobenzene. Thermal Decomposition of 1,1-Diazenes 7 and 18 Warming a 2-1\~Tl-p= glass of 7 to softening at 900!<: results in loss of the structured absorption curve clue to 7 (Figure 38) and formation of a redorange soecies (;\max = 464 nm) (Cigure 4 t) torether with growth of the UV absorotion due to tetrarnethvl-?-tetrazene 19 (Am ax= ?~0 nm). Warming this of softened glass to greater than PWOt< results in loss tl-)e Amax = 464 nm absorption anci growth of the UV absorption due to tetrazene 19. Tetrazene l9 is obtained in nearly quantitative yield from 13 overall. "imilarlv, warming a '-~~THF glass of 1, 1-diisooropyldiazene 18 to 900K results in loss of the structured absorption curve clue to 18 (Figure 39) anrl formation of a rerl-orange species (:\max= 474 nm) (Fi~ure 42) together with growth of the UV absorption due to tetraisoprooyl-2-tetrazene 20 ( Amax = ?90 nm). Warming this softened glass to greater than ?.00°K results in loss of the red-orange species (:\max= 474 nm) anrl formation of products exoected from thermal decomposition of 18. diazenes 7 Attempts to detect the 1,1- and 18 in these warmed red-orange solutions in possible equililxium with the Am ax = 464 or Amax = 474 nm species, respectively, by electronic spectroscopy were unsuccessful. -11~- LJ.J (.) z <( CD 0:: 0 (/) CD <( 0.0 500 400 700 600 WAVELENGTH (nm) Pigure 41. Thermolysis intervals. of 1,1-r!iazene 7 (2-MTT-fF, 9QOJ<), 10 min -139- w (.) z <l: CD 0:: 0 (/) CD <l: 0.0 500 400 700 600 WAVELENGTH (nm) Fi~re 42. Thermolysis intervals. of l, 1-c!iazene 18 (?.-".~THF, 90-°K', 10 min -140- Photochemical T)ecomposition of 1,1-l'iazenes 7 and 18 1,1-Diazenes have been shown to decompose when irradiated at their n-TI* transitions in the visible. Irradiation of a "2-MTl-JF f!;lass of 7 with visible liP.:ht (VI<;, 2 h) from Corning filters CS-1-75, CS-3-71), and CS-4-% ex. 470-410 nm) results in loss of the structured absorotion of 7 and growth of the absorotion Amax = 444 nm due to the red-orange species. F:.xtenrled VI<; ohotolvsis results in no rietectable necrease in the 464 nm ahsorption. A mixture of ethane and tetrazene 19 are obtained upon warming the glass to room temoerature. Ethane is the expected photochemical decomposition oroduct of 7. <;imilarly, VIS ohotolysis of a ?.-\.~Tl-JF glass of 18 with Corning filters CS-1-75 and CS-1-70 b 470 nm) results in loss of the structured absorotion of 1~ ann formation of the rf'ci-orange species Amax 474 nm. Extended VI<; irradiation flO h) results in no detectable loss of the Amax = 474 nm transition. A mixture of hvdrocarbon products expected from thermal and photochemical decomoosition of 18 together with tetrazene 20 are obtainen upon warming the red-orange glass. "~=ull pronuct analysis for thermal anrl photochemical necomoosition of 7, 18 and kinetics for thermal necomoosition are reoorten in a later section. Similar behavior was obtainerl in several other !!lass forming solvents incluning 3-methyl pentane, triacetin ann methvlmethacrvlate. One oossible explanation for the red-oram~e species obtained from 1, 1-diazenes 7 and 18 mif!;ht be the cis-?-tetrazene isomers of the corresponding trans-2-tetrazene oroducts ultimately observed. However, UV (VIS filtered) irradiation of trans-2-tetrazenes 19 and 20 in a rigid 2~~Tl-JF glass (~()OK) gave no red-orange coloration and no detectable visible -141- ! !:J., hll [ Amax =474 nm] l ,\+N=N- -< 18 hll ')lo ) ) >-< 32 33 34 AN-N=N-N>-- --< 20 y N2 -142- transitions between '350 and 750 nm. This is consistent with exclusion of the corresoondinf; cis-?-tetrazenes as the species responsible for the Amax 464 and 474 nm transitions from 1,1-diazenes 7 and 18, respectively. In andition, warming a rigirl glass of 7 or 18 in triacetin to 1500l<: (still rigid), results in loss of 7 ancl 18 and formation of the Amax = 464 and Amax = 474 nm SPecies res9ectivelv. Hence, diffusion of the 1, 1-diazenes is apparentlv unnecessary for the ohserved thermal decomposition. Matrix Isolation Infrareri Spectroscopy of I, 1-0imethyldiazene 7 For infrarerl studies a gas mixture of freshly purified dimethvl carbamoyl azide 13 in ultra high Purity an~ on (1: 1400, A.r) was deposited onto a cesium iodide (Csl) inner window of the matrix isolation apparatus at 20oK. The resulting argon matrix of 13 was slowly lowered to 1QOK. Figure 43 shows the FT-JR. spectrum of 13. Table XXII summarizes the assignments for the infrared bands of 13 and co-condenserl atmospheric impurities (H20, CO?.)· Irrarliation (I h) of matrix isolatec113 wit!, UV (VIS filtered) light from two r::orning CS-7-54 filters results in loss of the infrared banrls of 13 and formation of product bands. ~ubtraction of the FT -JR. spectrum of 13 from the FT -JR. spectrum resulting from UV photolysis of 13 clearly shows these prorluct bands (Figure 44). These bands are divided into five groups. A.ssignments are based on thermal and photochemical behavior and comparison to authentic ar~?on matrix infrared spectra. The first column of Table XXIII lists the prorlucts from UV photolysis of 13 and their assignments. The first group of bands labeled (I) are assigned to dimethylarninoiso- 0.90 0.80 b s 0.50 0.ll0 0.30 0. c e 0.00 0. 1 0 0.20 0.60 b n 0.70 r 0 1. 00 R 1. 10 1. 20 1. 30 Figure 4'3. 4000 3000 2500 2000 _T_____ __ 1500 I 170 mm rlenosition ( 1: l 400, A r, Jr') 0 !<:). Wc.venumbers FT -l"R so~ctrum of tl, 3() 0 0 ~- -1 1000 I I ""' I +=" -1 '~4- Table XXII. FT-IR Banns for 13 in Argon Matrix (120 mm) (1:1400, 100f() Threshold 0.0 5 A. Peak Location (cm-1) Intensity (A) ?.949.45 ?943.?.5 ?Q36.07 ?923.22 0.10 0.()9 0.05 0.05 v C-H ?345.03 2339.00 0.08 0.05 C02, v CO C02, v CO ?? "· 05 ??0~.53 0.06 0.06 ?156.75 ?15?..91 1.'36 1.20 0146. J?.) (2133.80) (21?R.99) 0. 14 0.18 0.12 1713.47 1705.59 1.36 0.51 l502.8g 0.56 l466.g3 1461.09 14 39.14 1434.59 1432.q2 1410.45 0.07 0.09 0.08 0.32 0. t 5 0. 11 1408.85 1407.49 0.34 0.?2 13n. 37 0.64 1.10 0.34 1396.?:7. 1386 .'~?. Assignment \1 N3 (asvm) \) f\13 \1 NCO (asym) v NCO (sym) -14'l- Tal:>le XXll. C:ontinued. Peak Location ( cm-1) Intensity (A.) 1?40 .M.l l. 39 0.79 1?%. 59 ( 1?1'2.4'2) 1?.28. 31 r 1n t. ?.~) 0 . _56 0.63 0.10 ( J?f)'3. 7'2) 0.17 0.09 0.09 li'3~L 'l3 0.50 11%.~9 11?4. 99 1 1?1 .41 0.89 0. 14 0.39 1071.61 0.06 ( 1?.15. 58) (121'2.51) tr.lf:l9. sn 1066.?1 0.12 0.()5 0.07 r) 729.76 7?6. 51 0.08 0. 1~ 648. p 644. ~ .5 0.14 0.06 = l '>N labelerl (lJ_l5N) as imouritv Assignment \J C-1\J ( asym) \J N-C (sym) 0.06 0.04 0.02 0.00 -0.02 -0. 04 n c e 0. 14 0. 12 0. 10 ~. 16 ~. 18 ~.20 ~.08 ~ b r 0 s R b ~.22 ~.24 ~.26 ~.28 ~.30 H20 Fi~ure 4lf.. Ll000 -1 - 0.' n. 3000 E ~ 2000 \v n. v e n u m b e r s 2500 \ C0 2 I II 1500 ,,, I~ A l HzO Ei' D II D D I 1000 H D T'ifference f.'T -IR spectrum of ororlucts from I.JV (\ff" filtered) photolysis (J h) of ll (1:1400, 1\r, lOOK) minus Sl)ectrum of 1'3. "ubtraction factor a= 3 [) 00 E co I I ~ ~ - - 0 . •1 0 ·- 0. 0 8 --0. 0 6 --0. 0 4 I ·-· Figure 45. Ll000 0. 00 -- el. 0 2 c e I =L ~- 02 n 0.04 0.06 b ~ 0.08 ~. 10 ~. 1 2 ~. 14 . eJ. 1 6 r 0 s R b Pl. 1 8 Pl.20 vL 22 el.24 3000 ---r 2000 Wc.venumbers 2500 1500 1000 I lL~ nifference r:'T -TR snectrwn of t, t -rlim ethvlr!iazene 7 anrl CO. I IV (VI~ filterer!) ohntolvsis (1 h) of l3 0:1400, .1\r, lnO'<) minus spP.ctnl'n aftP.r VTS (1-1-70-610 nm) photolvsis (t h) of 7. Subtraction factor a= J.r). 3 [J 0 0 '·····-~ ~L I co I I "" ~ - -148- cyanate, the expected product from the photo-Curtius rearrangement of 13. These hands grow in quicklv with UV photolvsis of 13 anrl decrease with extenrled UV photolysis. Photolysis with wavelengths 340-410 nm also results in decrease in their intensity (see third column of Tahle XXIIJ). These bands are unaffected by warming the matrix to 30-350'1(. The most intense band of (J) at ??. 19 cm-1 is likely assigned to the isocyanate stretching frequency. '!=:'or comparison, Lwowski47 reported this stretch at ??30 .:t ? cm-1 in a neon matrix at 60K. A second group of bands are observed to grow in with UV (VIS filtererl) irradiatio11 of t 3. These banrls assi!::>;ned to l, 1-dimethyldiazene 7 ~ecall that VtS irradiation of 7 in a rigid ?-TI.~TH'!=:' glass f~0°K) resulted in loss of 7. The infrared banrls assigned to 7 are are labeled fn) in Table XXJII. all efficiently photolvzed away with VI() irradiation at l QOl( using Corning filters CS-1-75, C:S-1-70, anrl CS-4-96 (). 470-610 nm). Subtraction of the FT -IR. spectrum after VI<:; photolysis from the spectrum before VIS photolysis results in positive absorption bands assie;ned to 7 (0) (Figure 45) together with a third group of bands at 2142.0?. and 2134.7 5 cm-1 assigned to carbon monoxirle (CO). All of the hands assigned to 7 are unaffected by warming the matrix to 30-3 ;ol(. The third group of hanrls, assigned to CO, show the same behavior as the bands assignerl to 7 fn). This type of CO behavior was demonstrated in the infrarerl studies of isotopically laheled l-I~l\lN 3 (see Chapter 1). The formation of two distinct CO bands mirroring the behavior of l, 1-diazene 7 will he discussed in a later section. - 14-9- A fourth group of product bands from UV photolysis of 13 also increase in intensity with VIS photolysis of 7 in argon at 1ooK. These bands labeled (E) in Table XXIII, are assigned to ethane. Ethane is a product of VIS photolysis of 7 in a 2-MTT-lF glass (800K). A fifth group of product bands from UV photolysis of 13 remain unaffected by VIS photolysis of 7. Irradiation with wavelengths 340-410 nm 26 also has no effect. These bands are tentatively assigned to the unknown Nmethyl-diazetedinone 26, labeled (A) in Table XXIII. The strongest band observed for 26 (Figure 44) at 1815.06 cm-1 is consistent with the stretching frequency of a carbonyl in a four-membered ring. For comparison, dimethyl- 0)-(0 N-N I' 27 28 r- -150- diazetedinedione 27 has a carbonyl stretch at 1780 cm-1 in CHCl3 solution.~6a 1\!,N'-diisopropyl-1,?-diazetedinedione 28 has a C=O stretch at t~l5 cm-l,81)b whereas B-lactams !1iazetedinone 26 have a stretch near 1800 cm-1.87 would result from intramolecular insertion of dimethyl carbamoyl nitrene into a methyl C-H bond. This intramolecular insertion is orecedented for diethyl carbamoyl carbene yielding the corresponding fourand five-membered intermolecular or ring comoounds.88 intramolecular dialkyl No previous carbamoyl evidence nitrene for insertion chemistry has been noted. 4 5b,6 Attempts to isolate and fully characterize 26 have not been successful to date. Of the infrared bands assigned to 1, 1-cHmethyl diazene 7 the major hand at 11)00.96 cm-1 and a minor band at 1590.83 cm-1 with a shoulder at 1591.53 cm-1 are all in the region where one would expect to find the characteristic N=N double bond stretch of the 1,1-diazene. This N=N stretch has been shown in the kinetically persistent 1, 1-dialkyldiazenes in solution and in the oarent 1,1-diazene l-J2NN in an an=!on matrix to be Hooke's law shifted to lower energy with the incorporation of a terminal 15N label at the nitrene nitrogen. In orc{er to verify the assignment of the characteristic N=N double bond stretch for 7 a 15N terminal label was desired. 'Reaction of dimethyl carbamoyl chloride 21 with ( J-15!\)' !labeled sodium azide63 in dry acetonitrile afforded 13-l5N. The combination of 15N NM'R., high resolution mass snectroscooy anrl FT -I'R. soectroscooy demonstrates an equal distribution of the 15N lahel between the two terminal positions of the azide moiety. I ')N N~~R. reveals two singlets at -141.40 and -?6,:;.02 porn from 15N-CH3N02. -151- Table XXIII. 60 min hv TJV (VI'S Filtered) Infrared 'Rands for Products from Photolysis of 13 and l3-15N (1:1400, Ar, 100J<) (cm-1). 60 min hv VIS (A 470-610 nm) 60 min hv (A 340-410 nm) ~776.01 Assignment H70, V 0-H H'-n, V 0-H H20, v 0-H ~756.53 ~711.10 ~184.?4 (_) n (overtone) ~037.18 (-) l) (?) 3009.44 (-) (U) T)\~, v C-H ~004.64 (_) ~000.50 ( +) (+) ( +) (+) <+) (+) (+) (+) E, v C-H E, v C-H E, v C-H E, v C-H 2%7.7.8 (_) (U) 7.9 58. 1 5 2955.91 (+) ( +) (+) (+) E, /.94?.97 ?941. 17 /.938.?1 2933.79 (_) (_) T), v C-H (-) (-) D, v C-H D v C-H ?9?~.06 (+) (+) E, v C-H ?909.4? (_) (U) ( +) ( +) ( +) <+) <+) (+) E v C-H E,' v C-H E, v C-H 299l.l'i 29~2.61 2980.53 ?.900.38 n95.21 n90.90 2345.00 2339.01 E v C H ' \) - C-H D, v C-H ' C02, v CO C02, v CO -152- Table XXIII. r:ontinued. 60 min hv l JV (VI<; Filtered) 60 min hv VIS (A 470.:.610 nrn) ??19.49 2207.()1) 60 min hv ( A 340-4 t 0 nm) Assignment (_) (-) I*' V-N 15:C:O I, V-N14:C:O co, \)co ?147.0? ?117 .~0 ?.135.93 2134.75 21'32.95 co,\) co (_) co,\) co co,\) co co,\) co 1~15.06 (A, v 14NCO) 1811.97 (A*, v 15NCO) 1623.60 1607.73 H20, H20, 1600.96 159 I. 53 15(}0.83 (-) (_) 15~1.~3 (-) (-) 157?.26 D11,~, v 14N= 14N Dm, v 14N:l4N Dm, v 14N=l4N n,~* v 14N= 15N . . ' Om*, v 14N:l5N (_) (- ) (-) 1473.12 1471.f;2 1467.~4 1463.03 143~.16 14?3.25 I, o Cl-13 E, oCH 3 I, o Cl-f3 E, o Cl-f3 1464.97 1448.50 001-f o OH (-) (_) (-) (U) 11M, o CH 3 DM, CH3 o 1430.01 (U) (-) D (?) E, w CH3 1380.12 1375.12 E, w CH3 1369.10 (U), w CH3 -153- Table XXIll. f:ontinued. 60 min h\) uv (VIS Filtered) 60 min h\J VII:) (A 470-610 nm) 60 min hv (A ::'-40-41 0 nm) 1358.17 (_) o~.1, w CH3 11~4.35 (_) I1M, P CH 3 1158.79 1154.74 * 1078.77 l ')77 .80 * 946. t~ q37.02 8 59.37 8 57.56 856.49 (-) (-) (-) (-) (-) (-) (_) ~49.74 (-) (_) (-) ~45. 90 (-) Assignment o o (U)., CN 14!\J (U*), CN15N (U) (U*) (U) (U) I1M, \J C-N I'M, \J C-N f) (?) I"m, \J C-N !)m 8??..27 (+) (+) <+) E E, \J C-C E 818.16 (_) I 606.67 605.56 (+) (+) (E) ~?9.?0 8?5.~5 T)M = 1,1-nimethyldiazene 7, T)~~* =7-l5N nm = ~.~in or species (7), nm * =nm_l5!1..J t:: = l:-:.thane I = Oimethylaminoisocvanate, I* =I-151\1 A = 1\1-Methyldiazetedinone (26), A* = A-15N ( ) = Tentative assignment (+) = Increased with ohotolysis (-) = necreased with photolysis (U) = Unidentified. (U*) = fU_15N). (E) -15ft- 21 l\~atrix isolation of 13-15N in argon 0:1'+00, Ar, lOOK) (Figure '+6, Table XXIV) reveals the two 15N labeled asymmetric azide stretches at ?.151J)ft and ?1'3'3. 7 '+ cm-t in an aooroximate 1:1 ratio. Photolysis of a 13-15N would be exoected to yield a 1:1 mixture of unlabeled 1,1-diazene 7 and terminally labeled 7-l5N bv the ohoto-f:urtius rearrangement and suhsequent ohotodecarbonylation of the intermediate dimethylaminoisocyanate. 0 II + 5 N: (CH !2 l W".)o4'-N=N::l .. .. hv 0.90 ~.80 0.70 s 0 c e 0.00 0. 10 0.20 0. 30 0.Ll0 0.50 c. n 0.60 b r b R 1. 00 1. 10 1. 20 l . 30 Fi~ure 46. 4000 3000 2000 Wc..venumbers 2500 r 1500 FT-TR soec trurn of tJ_15N, PO rnm denosition (l:l400, .A.r, tr)OI<.). 3S00 --- ------ - ·· 1 ---- - ~ -- --- - - --- T -- --- - -- --T - - -·-- --~ I VI I \ Jt - -1 '56- Table XXIV. FT-IR. Bands for tJ-l.5N in Argon ~~atrix (120 mm) (1:1400, lOOT<:) Thresl-tolrl 0.05 A. Peak Location (cm-1) Intensity (A) 3756.37 3711.02 0.07 0. 0 5 H20, v OH H20, v OH 2q49.45 0.09 0.06 v C-H 294~.~4 2935.50 0.05 ?.n3.n Assignment 0.05 '2 345.03 7.339.00 0.08 0.06 ?1~7.43 o. 14 2183.?4 ?174 . 7.7 0.09 0. 10 (?!57 .43) ?.15l. 04 ?148.07 ?146.00 2133.74 ?. 1?9. 05 0.15 1.0?. 0.60 0. 51 0.96 0.70 1724. l3 17?1 .45 1712.93 1704.0?. 1692.51 0.16 0. 17 1. 31 o. 54 0.07 v NCO fasym) 1623.65 1607.77 0.08 0.08 H7.0, 8 OH H20, o OH 1502.78 0.53 1466.42 146'2.11 1439.04 1434.45 1432.76 1410.45 l403.sn 1407.49 0.07 0.09 0.06 0.30 0. 14 0.11 0.34 o.n C02, v CO co 2 , v co v-1\1 15=N=N (asy m) V-N=N=l511,J (asym) -157- Table XXIV. Continued. Peak Location (cm-1) Intensity (.A.) 1'3<) 8 .13 1395.<)<) 1385.92 0.64 1.12 0.32 ( 1241). 39) 1233.60 1?.?9.73 1?~5. 75 P 20.q4 1? 1.5. 53 PP.52 t ?07 .SP 1203.75 0.13 0.92 0.67 0.34 0.38 0.98 0.60 0 . 33 0.54 11%.03 1114.78 1ll7.10 111?..15 0.8~ 0. 19 0.05 1071.6? 1069.!1.1 1066.23 0Jl6 0. 13 0.05 y CH3 973.41 970.53 0.07 0.05 v C-1\J (asvm) v C-l5N (asym) 7/.9.69 726.~2 0.05 0.15 cS 648 .I? 644.85 0.12 0 .l 0 () = Unla':>eled (13) as imourity Assignment v NCO (sym) v-1\l15="J=N (sym) v-"J="J= 1 '51\J (svm) p CH3 0.~9 N3 v "J-C (sym) -158- TJV (VIS filtereci) photolvsis J3-l5N (I h) (1 :1400, Ar, lOOK) results in loss of the infrared banrls of 13-15"l and formation of prociuct banrls. Figure 47 shows the product FT -IR. spectrum minus theFT -JR. spectrum of 13-15N. These prociuct bands are summarized in the first column of Table XXIII. In addition to the product bands from photolysis of 13, additional 1'>!\J shifted bands are found at ??07.00, 1~1 1.97, 15glX3, and 1572.26 crn-1. These hands correspo11d in intensity to 15N shifted multiples of unlal:>elerl product ba11ds at ~?1 9.49, 1815.06, 1600.96, and 1590JD cm-1, respectively. The major band at I 58 J.Sn cm-1 and a minor band at 1572.26 cm-1 are lost with VIS irradiation together with the banrls of unla~eled 7 and are assignee! to 7-15N. <;ubtraction of the FT -JR. soectrum after VI<; photolysis from the FT -IR spectrum before VI<; photolvsis results in the spectrum of 7, 7-15N, and bands due to carbon monoxirle fCO) (Figure 48). 1600.% and l5~l.~3 cm-1 are The more intense bands at assigned to the characteristic v 141\1= 14'-J and v l 4N= 151'1,J stretches of 7 and 7 _l5N, respectively. Figure 49 shows the '1=N stretch region. The observation of a minor hand at 1590.8 3 cm-1 which shows the same behavior as bands due to 7 and also shows a large 15N shift to 1572.26 cm-1 is surprising. One possible explaination for this woulrl he matrix site splitting of the bands of 7 yielding small multiole bands of 7 due to different sites. The minor band at 1590X3 em-! correspond to the 1600.96 cm-1 v environment. woulrl 14"'-J= 14'\1 stretch in a different site However, attempts to remove this possible site splitting by warming the matrix to 30-350K were unsuccessful. This observation is consistent with assigning the l59o.sn cm-1 bane! to another species with a l\!:'1\T stretch which is also efficiently photolyzed with VI<; irradiation - - · c e - I I 3000 1 E ~ [1 I* A*' 1S00 '· \ r~ E Ei ," D H20, /i o* / D \.1 A 2000 w~venumber s 2500 \ C0 2 I 1000 I h D nifference FT -IR. spectrum of products from UV (VIS filtered) photolysis (1 h) of 13-15N (t:Jl+OO, Ar, lOOK) minus spectrum of 13-15N. Subtraction factor a= 0. 16. 3[)00 H20 '4 Fi~ure 47. 4000 -0.05- ~.00 n 0.10 - 0. 15 ~. 20 0.05 c.. b i 0 s b R ~.25 ~. -~0 co I VI •..!) I ,_ e c c. n b r 0 s b R -0. 10 --0 . 05 0.00 0.05 Fi~ure 4~. 3000 2000 Wc.venumbers 2500 I I 1500 i/' liD* D 1 1000 (1 T1ifference FT -IR spectrum of 7, 7-l'>N and CO. UV (VI" filtered) photolysis h) of 13-1511\l fl:l400, A.r, tOOK) minus soectrum after VIS (470-610 nm) ohotolvsis (1 h) of 7:7-15N mixture. "ubtraction factor a = 1.0. 3500 - - ----·-- ·---- ~-- - - - ·----- -~ ---- - 4000 0.10 - 0. 15 0.20 0.25 co I I 0 ( ]\ - -1~1- Table XXV. Infrared ~ands for 1, 1-nimethylrliazene 7 and 7-15N (Ar, lOOK). Peak Location fcm-1) Assir.;nrnent ~1~4.24 (overtone) ~0'37. l3 (?) 3004.64 M, v C-11 ?.94?..97 2941.17 /.939.?7 ?93~.?3 ?.933.79 1600.96 11~, v 14N: 14N l5Q l. 53 1590.'39 m, \J 14N:l4N m, v ! 4"1= 14N l5R1.~3 1572.'26 1438.46 M, o Cl13 M, o CH3 1423.25 (?) 1358.17 M, w CH3 11~4.35 M, P Cl13 859. ~7 ~57 •.56 M, v C-N M, v C-1\J ~56.49 (?) 849.74 m, v C-N m, V C-N 144~.50 ~45.90 M = Major soecies (7, 7-15N). Tentative assignment. m = Minor species. (?) = Uncertain. () = n c e n b r 0 s b R liL 001?1 - ~. 140 156~ Wo.venumbers 1621?1 16liJI?I 1580 1540 q_? fa)~? 141'1J=141\J stretch of 7 (nM) and minor comoon~nt rnm) from Figure 45. (h) 14'\J= 111-!\J stretch of 7 and R? 14!\J= l5N stretch of 7-l 5N and minor comoonent stretches from J::'igure 4~ • . Wo.venumbers 1621?1 161?11?1 1580 1560 1541?1 Figure 49. 0.000 - 0. 140 I I J'\ N -- -16~- 0. 470-f1 10 nm) or its assignment to 7 in a different matrix site which is not altered f)y warming to 30-150K. The band at 1590.83 crn-l and other additional "minor" banrls which show matrix behavior similar to 7 and 7-15"1 are labeled (f)m) in Table XXIH. The major bands (greater intensity) assigned to 7 anrl 7-15N are labeled (T)~~). Table XXV summarizes the infrared bands of 7 anrl 7-15N ami their tentative assignments. Only the N:T\.1 double bond stretch mode shows a 1 5'] shift. This is consistent with the assignment of these bands to 1, 1-dimethyldiazene 7 and 7-151\1 with 1 '>1\J incorporation onlv into the terminal "nitrene" nitrogen. Two other orodtJct bands from UV ohotolysis of 13-15N also show a 151\l shift to lower energy. The asymmetric v 141\J=C=O of dimethylaminoisocyanate (labeled isocvanate stretch in Table (I) XXIIIl at 7.219.49 cm-1 shows a 15N shift to '2207.00 cm-t. For comparison, t'"le isocyanate stretch for the oarent aminoisocvanate ('1-l;?N-N:C:()) was founrl to t 5N shift from ?'21?..0?. cm-1 to ?194.18 cm-1 in an argon matrix at t QOT( (see Chaoter 1). In addition, the band at 181 5.06 cm-1 shows a small l 5N shift to l~ll.q7. This hand is tentatively assigned to the asvmmetric T\JC() stretch of N-methyl diazetedinone 26 (labeled (A) in Table XXIII). Photolysis of 7 with visihle lio.:ht (VIS) 0 . 470-610nmhesults in loss of all infrared hands of 7 and formation of new bands assigned to ethane (labeled (t:,) in Table XXIII) to~?;ether with a CO band at 2137.80 cm-1. Ethane and nitrogen are the exoected ohotolysis products of 1, 1-diazene 7. t:thane is a product of photolysis of 7 in a 2-MTl-JF glass (80°K). A I ·-0. 020 ·-0 . 0 3 0 0. 3S00 -0.010 b n c e -0.080 -0. 07~ -0.060 ·- 0.050 -0.040 Figure 50. I ll000 I I 2000 I I Wc.venumbers I 2500 I I I co I 1500 I II I ~I I u u 1000 Iu u u f 1 h) of 7. So~ctrurn spectrum of ({J) resulting from VIS (470-61 r) nrn) photolysis resulting frorn VTS ohotolvsis of 7 minus soectrum after ohotolvsis (J h) of (lJ) f140-'J.1 0 nm). Subtraction factor a= 1.0. r"'iff~rence FT _yq I 3000 I 0.000 r -J IU 0.010 ~.020 0 s b u 0.040 ~- ~3~ ~ R 0.050 ~.060 _J 0.070 ~.080 I I .._ I ~ 0\ ~.020 ~.010 ~.000 -0.010 -0.020 -0.030 ·-0. 040 ·-0. 050 r b ~ n -0.070 -0.06~ ~.030 0 c e ~.040 ~.050 ~.060 s b R ~.070 ~.080 ~.090 ~. l00 Figure 51. I Ll000 3~00 2000 \v c. v e n u m b e r s 2500 1500 u u u 1000 nifference r::"r -I~ soectrum of (lJ) and (TJ_l'>I\T) resulting from VIS (470-1; If) nm) o~otolvsis (t h) of 7:7-' '>N mixture. Soectrum resulting from VIS o~otolvsis f1 ~'of 7:7-15N minus snectrwn after ohotolysis (I h) of (T n:ru_l5'\J) mixtllre (-:l,ll-0- 1-1-1 'l nm) subtraction factor a= 1.0. 3S00 co I I (]'\ VI -166- Table XXVI. Comparison Ethane Infrared 13ands CAr, lOOK) (cm-1).~ hv VIS 1, l-'1iazene 7 hv uv _!-1,2-niazene 8t 3000.50 2991.15 3001.15 2991.87 2982.65 ?no.53 2'l58.15 29 55.91 79'>1.91 2982.05 2979. 50 ?9?~.00 2900.'3~ ?895.?.1 7893.02 ?.890.QO Ethane/Ar Assignment 3000.50 2991.97 2986.74 2Q84.16 ?Cl79.84 v C-H 2960.~1 2956.?3 295?.64 2926. 'l6 2901.74 ?.89.5.57 n9t.17 2951.26 2'l20.~3 2893.75 2891.52 1467.84 1464.97 1468.00 1380.12 1'375.12 1179.80 1378.17 1374.50 832.16 829.20 825.85 82?.?7 829.(.,4 82R. 30 R?.6.13 822.26 S1.20.34 818.09 ( 606 .67) (605.56) (610.03) (608.?8) ~1:1400, 1467.40 1464.21 Ar, 10°K. ( ) Tentative assignment. Close to strongest observed bands for methyl radical (CH3·) at 60?.-6?0 cm-1 (Ar, 10°K). 'Ref. 90. -167- comparison of the infrared spectrum of matrix isolation ethane (1:1400, Ar, lOOK) with that produced from 7 is made in Table XXVI. Some additional unidentified minor product bands (labeled (U) in the second column of Table XXIII 1 are also obtained from VIS photolysis of 7. VIS photolysis of 7-1.5N produces the same set of product bands to~ether with 15N shifted multiple bands at 1154.24 and 1077.80 cm-1. The 4 •.55 cm-1 15N shift of the band at 1158.79 cm-1 to 1154.24 cm-1 suggests its assi~nment to a bending mode involvin~ a 15N labeled nitrogen. The same is true for the band at 1078 .77 cm-l, which shows a 0.97 cm-1 15N shift to 1077.80 cm-1. Photolysis of (U) with wavelengths '340 to 410 nm results in loss of these unidentified product bands and ~rowth of hands due to ethane. Subtraction of the FT -IR spectrum after 340-410 nm photolysis from the FT -IR spectrum of (U) before photolysis results in the difference spectrum of Fi~ure 50. Positive absorbance bands of (U) and a CO band at 2l32.95 cm-1 are clearly displayed. Additionally, bands due to dimethylaminoisocyanate are also apparently photolyzed with light 340 to 410 nm. The negative absorhances correspond to ethane and a CO band at 2137.80 cm-1. The behavior of CO bands will be discussed in a later section. Similarly, Figure 51 shows the differenced spectrum resulting from 340-410 nm photolysis of 15N-labeled (U). The observation of ethane as a product from photolysis of (U) between 340 and 410 nm suggests the 1,2-dimethyldiazene isomers 8 as possible candidates for (U). The argon matrix infrared spectrum of trans-1,?.-dimethyldiazene 8t (1:1400, Ar, lOOK) (Fi~ure 52, Table XXVli) easily rules out 8t as the species responsible for the bands labeled (U). Photolysis of matrix isolated 8t (X 340-410 nm) results in loss e c n c.. b r 0 s R b ~.00 ~.05 ~. 10 ~. 15 ~.20 ~.25 ~.30 ~ • _3 C) ~.40 Figure 'P. 4000 ~-- ,- 8t CH3 --- N=N\ 3000 2500 -, I 2000 Wc.venumbers ---- -l 1500 FT -T-q_ soectrum of trans-l, ?-cii methvldia?.:ene 8t ( 1: 1400, A r, )r)Ol(). 3S00 H3 c\ 1000 I ()0 3\ I ,_ c e n -- 0. l 0 -0. 1 2 -0. l 4 ·- 0. 0 8 --0 . 0 6 ·- 0. 0 4 --0. 02 Pi~ure 5l. 4000 3S00 2500 200~ Wo.venumbcr s 1500 I 10~0 T)ifference FT _y-q soectrum of ethane/\.() from VJ<; ohotolvsis of I, 1-dia7.ene 7 tr.n. <;oectrum a.fter " " ohotolvsis of 7 anrl su~seauent ohotolvsis of (1 J) minus SDf>ctrum hefore VJ<; ohotolvsis of 7. <;uhtraction factor a = 1.0. 3000 0.00 0. I \{) I b J\ E 0.02 E r H20 0.04 0 I 0.06 s E V1.08 V1. 1 0 E ,_.,._ R b V1. l 2 0. 14 V1. l 6 vL 1s co -170- 1. 00 0.90 0.80 R b s 0. 70 c ~.50 0 Y' b a 0.50 n 0.40 c e 0.30 b 0.20 0. 10 0.00 3000 2950 2900 Wo..venumbers Figure 54. Comparison ethane FT -I'R. spectra of the C-H stretch re~ion. (a) Ethane (1:1400, Ar, lQOJ<:). fb) F.thane from VI() photolysis of I, 1-diazene 7 (Figure 53). (c) Ethane from photolysis <340410 nm) of trans-1,2-diazene 8t (1:1400, Ar, lOOK) 1 h. -171- Table XXVII. Infrared !='>an cis for 1,2-Nethyldiazenes 8t and 8c (cm-1 ). T rans-1, 2-fHazene 8t~ Cis- I ,2-J:":''iazene 8~ 299?.91 (m) 301~ (m) ?9~~.{;5 (m) 291-iO (s) ?.9~l.4f> (w) 2910 fm) ?9?h.l2 (s) 1 561 (w) 29?.3. /(., (w) 1477 n54.03 (w) 1440 fs) 1441.52 (s) 1430 (m) (m) 1436.76 (s) 1':379.46 (m) 1370 ( s) 1177.72 (w) 1159 (s) 11 og. n (w) 1110 (m) 1008.26 fm) 935 (m) ~This work (1:1400, A.r, l0°K). !?.R.ef. 89 (gas). w = weak, m = meciium, s = strong. of tt,e hands of 8t and formation of bands due to ethane. Subtraction of the FT -IR. spectrum after photolysis from the FT -IR. spectrum before photolysis of 8t clearly disPlays the infrared spectrum of ethane produced from photolysis of 8t (Fi~ure 53, Table XXVI). No bands of OJ) were obtained from photolysis of 8t. 1\! o infrared bands attributable to cis-1, 2-dimethyl diazene 8c were observed to ~row in the photolysis of trans-1 ,2-diazene 8t. Table -172- XXVI si-Jows a comparison of the ethane infrared hands produced from photolysis of l, 1-diazene 7, 1,2-cfiazene 8t, and argon matrix isolated ethane (1:1400, t\r, toOK). rlose insoection of the C-'1-J stretch region (Figure 54) reveals that infrared soectrum of matrix isolated ethane is noticeably differe11t from that of ethane prorluced ohotochemicallv from either 7 or 8t which give nearly identical ethane product spectra. molecule oro~uced The single nitrogen tor?.ether with ethane from photolysis of 7 and 8t aoparentlv oerturhs the matrix site around ethane in solid argon. Although 8t can be ruled out as a oossible assignment for (IJ), the cis-1, 2-diazene iso:ner 8c remains a possil:>le choice. Table XXVII shows a comparison of the infrared band for cis and trans 1, 2-diazenes 8t and 8c. The bands for 8c correlate fairly well with those for {U) in Table XXIII. A, nether (J.. max possible candidate for {U) would be '1-J, ethyl- I ,?-diazenel 03 = %7 nm) which would also vield ethane as a photodecomoosition product. However, no infrared band attributable to an N-H stretch is oJ-,serve~ for (TJ' in the region 2~00-3400 cm-1. Arlditionallv, comparison with ti-Je literature matrix infrared spectra for cis- and trans-f-I, methvl-1,?.diazene do not suooort 'Y, ethyl-! ,2-diazene as a candidate for (U))3 Tetrameti-Jyl-?-tetrazene 19 was also ruled out as a possihle candidate for (I.J) hv comoarison with its argon matrix infrared spectrum (see f:xoerimental). UV (VI5 filtered) photolysis of 19 (I :700, Ar, lOOK) results in formation of products dimeti-Jylamino rarlicals.93 from disproportionation/recombination The disproportionation of two products N-methvl- methvlenimine (\1 -'\T=Cl-lz 1643 cm-l) and rlimethyla.rnine (\1 NH = 3340 -173- cm-1) are also easily ruled out as candidates for (TJ). Matrix Isolation Infrared Spectroscopy of 1,1-Diisopropyldiazene 18 UV (VIS filtered) photolysis of matrix isolated diisopropyl carbamoyl azide 17 and 15N labeled 17-1.5N yielded infrared product bands for comparison with those obtained from photolysis of 13 and 13-1.5N. A band assiJ?;ned to v _l4N=C=0, the characteristic isocyanate stretch of diisopropylaminoisocyanate, was found at 2241.91 cm-1. Photolysis of 17-15N resulted in a 1 51\T shift to 2231.5 5 cm-1 for the 15N labeled isocyanate stretch v-15N=C=0 (Figure 55). UV photolysis of 17 also yields a minor product infrared band at 1815.95 cm-1 and a major product band at 1795.94 cm-1. These bands are tentatively .assigned to the carbonyl stretches of 29 and 30. The four and five-membered ring compounds would result from intramolecular C-H insertion of the diisopropyl carbamoyl nitrene into a methine or methyl C-H bond, respectively. Ao --{jl-N3 Photolysis of l7-1.5N results in a 15N shift of the hv 17 \N-f.o -(_A\ / [~)'N] AN-N=C=O -{ lh· 1 ~=( A+ - co -( N=N H 30 29 18 -174- minor hand at 1~1 5.95 cm-1 to 1~10.45 cm-1 in support of this asshz;nment. The corresooncling four-meml:>ered rl.iazetedinone 26 obtained from photolysis of 13 showeci a 15N shift from 1~15.n~ to 1811.q7 cm-1 (Fie;ure 56). The characteristic '1\1:1\! double bond stretch for J, t -diisooropyldiazene 18 is found at 1600.9? cm-1. Photolysis of 17-l5N results in a Hooke's law shift to 1579.46 cm-1 (Figure 57). The characteristic J., 1-diazene N=N double bond stretch frequencies are summarizer:! in Tal-,le XXVIII. A comparison of the 1, 1-c!iazene 'I\1:N double bond stretch frequencies with the isoelectronic carbonyl stretch frequencies is marle in Table XXIX. The N:l\1 stretch frequencies for 1, 1-diazenes parallel the trend in the isoelectronic carbonyl comoounds. One effect in this trend is indicative of increasing p-orbital character in the N-C bontis with decreasing CNC bond angle. This results in increased s-orbital character of the 1\11\! bond and higher energy N=N stretch frequencv. ()uoerimposed on this is an effect apoarent inductive/ hyperconiugative sta.,ilization of the positive charge on the amino nitrogen of the l, 1-diazene. Crreater electronic donation (alkyl vs. H) results in a + - stronger '1\l:N stretch frequency. double bond and a higher energy The onnosite electronic effect is realized in the iso- electronic carbonyl compounds l,t-nimethyldiazene 7 Product Analysis '~=or oroduct analysis, samples of 7 were orepared by UV (VIS filtered) photolysis of 13 in a rigid ?-MTHl= glass contained in either a 5 mm o.n. quartz tube immersed in a liquid nitrogen filleci finger dewar or contained in - 22AI'l 226V1 22~1'1 2221'1 V1.1'li'J0 22AI'l 2261'J 2240 2220 2201'1 Figure 55. w~venumbers TJV (vyc;; filtererl) nhotolvsisof t7. FT _yq snectra of diisonronvlaminoisocvanate. (a) - 1'~N=C:=0 stretch from (h) _t4l\T=C=0 anrl _l5N=C=0 stretches from TIV ("Ic;; filtered) nhotolvsis of t 7-l5N. w~venumbers I V1 c e ·""'I ~.1'101'1 1'1.1'155 n t1 b r 0 s b R 1'1 • 1 II'! c e n r1 b r 0 s b R Fieure '),;. ~.~~00 'LI:'l1Se! w~venumbers 1840 1820 1800 1780 1760 -~ ohotolvsis of 17. (b) r:rom T.TV (VI~ filtererl) ohotolvsis of l7-l5N. FT-TP. snectra of carhonyl stretches for 29 anrl30. fa) From lJV (VIS filtererf) Wr1venumbers 1841:'1 1820 1800 1780 1760 0.01:'100 0.0090 I ~ "" I ,__ c e n ~ b r 0 s R b Pigure 57. 0.0000 0.0170 Wo.venumbers W~ v e n u m.b e r s fa) "R7 14\l=liJ.!\1 stretcl-) of 1~ from UV (VJ~ filtererl' ohotolvsis of 17. (I)) l(2 l4!\t=f4N anrll(., 141\j=l';N stretcl-)es of 18 anrl p~_15N from TJV (Vl~ filtererl) ohotolvsis of l7~l5N. 1600 1580 1560 1540 1520 1600 1580 1560 1540 1520 0.0000 0.0170 I I ...,..., -1 73- Table XXVIIT. 1, 1-f' iaze ne In fra red Tran siti on s. 1, 1. -~""'~i azen e H \+ N=N I a l.574 J54 ~ 1'5 48 1571 155? 151J.5 N=N I ?,()J 15~2 1'5 74 ~+N=N a -( 160 1 1'57 9 1.574 15q 5 15{)8 ! 56~ 1f!l ~ ~ lf!l12 1610 H 0 \+ N=N I a D H3 C\+ I H3 C Q=N ~N=N a b b -179- Comparison of 1,1-Diazene anci Isoe1ectronic Carbonyl Infrared Transitions. Table XXIX. l, t-niazene 0 \.NaNI 0 Car~onyl 0 1571~ 169~~ 'c=o 0/ H H \ \+ N=N I 157ft~ Q=N 1595Q 1690~ -\.N=N -I 1601~ 1713~ --\C=O 1601.~ 1719~ ~c\ C=O )74:?~ H/ H H3 C\+ I _ N=N Q=o -I I ~c H3C ~N=N C=O 16382 17?.5~ ~C=O -180- the low temperature TJV /VI<; soectroscooy cell at ~f) OK. Warming a ?-~~THF glass of 7 results in loss of 7 and growth of transition due to tetramethyl-?-tetrazene 19 and a Arnax = 464 nrn. red-orange species Warming to greater than 1700K results in decolorization. Product analvsis hy analytical vapor phase chromatograohy (V'PC) reveals tetramethvl tetrazene 1'~ as the major oroduct together with ethane, N2 and c:n. Photochemical rlecomoosition of 7 with visible light (Vl') (470-61 0 nm) (~-~HHJ:;', gooK) also results in loss of 7 and formation of the red-orange soecies Amax = 464 nm. The Amax = 4f.4 soecies appears unaffected by futher visil)le ohotolvsis (>340 nm). Warming to decolorization and VPC analvsis reveals a greater yield of ethane than obtained from thermolysis of 7 Table XXX. 1, l-T)imethvlrliazene 7 necomposition 'Products~ Cone. 13 (~~) Conditions x to-1 ThermaJb,c X J()-1 hv(VI<;)I),rl 6.1 5 x 1o-' Thermal~ 9.3 5 x 1o-3 hv(VIS)e,g 19 1 1 9.7 6.0 ~Molar V'PC ratios, n-Octane internal standard. EOuartz finger dewar, 77 K, hv UV (2-CS-7-:54 filters), 3 h, 0.1 M, 13 2-MTHF • .fhv UV, warmed to 'R..T ., 95% conversion, 9 5% product yield. 1hv TJV, hv (VJS) (C:S-1 -75, C5-3-70, C<;-4-% filters), 3 h, warmed, 9'2 % conversion, Q{.o~ ororluct vielrl. ~UV !VIS soectroscopic cell, 82 K, hv lJ" (?-CS-7-54 filters),~ h, 5x to-3 M, 13 2-~~THF • .!.hv UV, warmerl, 92% conversion, 94% prorluct yield. g hv UV, hv (VIS), 3 h, warmerl, qoo.t, conversion, 92% product vield. -181- together with tetrazene 19 as the major product. The product ratios for thermal and photochemical decomposition of 7 are listed in Table XXX. r T etrazene 19 was synthesized independently by oxidation of 1,1dimethyldiazene. Tetrazene 19 isolated from decomposition of 7 by preparative VPC was compared spectroscopically (lH NMR, UV, IR) and by analytical VPC coinjection. Methane, 1,'2-dimethyldiazene 8 and formaldehyde dimethyl hydrazone 10 were not detected by l H NM'R or analytical VPC (SE-54 capillary) from decomposition of 7 in 2-MTHF. Ethane was the major photochemical decomposition product of 7 in an argon matrix at lOOK. It can be inferred that the red-orange A. max= 464 nm species which reuslts from thermal and photochemical decompositions of 7 in 2MTT;C ultimately yields tetrazene 19. The activation ener~y for nitrogen extrusion fro:n 7 is estimated to be 23-26 l<cal/mol. Hence, thermal decomposition of 7 to afford ethane and N2 is unlil<elv at the low temperatures where the structured absorption of 7 is lost ( < l OQOl(). The ethane detected in the products from UV photolysis of 13 -1~? - to produce 7 followed l->y thermal decomoosition of 7 very likely results, at least in part, from competitive decomposition/relaxation of excited 7 produced photochemically. 1,1-Diisopropyldiazene 18 Product Analysis Warming a 2-MTHF glass of 18 results in loss of 18 and formation of a red-orange species Amax = 474 nm. Warming to ~reater than 180°K results in decolorization. Product analysis by capillary VPC reveals tetrazene 20 and hydrocarbon products 32, :n, and 34. Photochemical decomposition of 18 with visible light (VIS) (>470 nm) also results in loss of 18 and formation of the red-orange species Amax = 474 nm. The products for thermal and photochemical decompositions of 18 are summarized in Table XXXI. Tetrazene 20 is the exoected dimerization product of 1, 1-diazene 18. Hydrocarbons 32, 33, and 34 are the expected products of thermal and photochemical decomposition of 18 to two isopropyl radicals and N2. unlikely that 1,1-diazene 18 extrudes N2 at the low It is temperatures ( <1 QQOK) where the structured absorption curve of 18 is lost and the red- I ~.N=N -< 18 [ Amax •474 nm] ! ~ ~ ) >-< 34 33 32 ~ >. )- N-N-N-N -< 20 '\ Nz -183- orange A max = 474 nm species grows in. Thus, it could be inferred that the Amax = 4-74 nm species which subsequently ultimately yields the observed hydrocarbon decomposes (>20QOK) products 32, 33, and 34 (at least in part). Higher initial concentrations of 18 afford a greater ratio of tetrazene 20 to hydrocarbons decomposition of 18. from both thermal and photochemical For comparison, I, 1-di-tert-butyldiazene 6, showed only unimolecular thermal decomposition to hydrocarbons 35, 36, and 37 at T)irect dimerization of 6 to tetra-tert-butyltetrazene was , :::{ + - .. N=N: 7\ •)-H ... 35 6 observed. 8 f not . H . >= 36 The ratio of disproportionation 37 products 32 and 33 to recombination product 34 from thermal decompositoin of 1,1-diazene 18 is comparable to that obtained from 1,1-di-tert-butyldiazene 6. Comparison with the thermal and photochemical decomposition of 1,'2-diisopropyl diazene is made in Table XXXII. The greater ratio of disproportionation to recombination products from 18 (or the red-orange A max = 474 species) may favor a comoetitive thermal decomposition of 18 via a B-hvdrogen elimination. Subsequent radical chain decomposition of the H, -~~~- Table XXXI. I ,1-f'liisoorooyldiazene f'lecomoosition Products.~ Conditions 32 33 34 20 ThermaJ!2,.E 1'> 13 1.0 4.8 11 12 3.8 1.0 hv(VIS)~ 2.8 3.3 1.0 1.0 Therma~,_g 16 15 1.0 160 ~A.~olar VPC: ratios n-dodecane and n-oentane internr.~.J standards. QUV /VI'S soectroscooic cell, 3-x 10-1M, 17 2-~;;;YHF, ~2 K, hv UV (2-CS-7-5~ filters), 3 h. .Ehv UV, warrnerl, 95% conversion, 9~% product yield. ~hv UV, hv (VIS) (C:S-1-75, CS-1-70), 10 h, warmed, 95% conversion, 9'3°,f, oroduct yield. ~Ouartz fin~er dewar, 77 K, 0.1 M 17 2-MTHF, hv TJV (2-CS-7-5~ filters), 3 h. .ftw (VIS) (CS-1-75, CS-3-70 filters), 10 h, warmed, q9o.?, conversion, 9~% product yield. _ghv UV, warmed, 96°-6 conversion, q?o~c, ororluct yield. Table XXXII. Molar Product Ratios for Some Related T)ecompositions. Conditions 32 33 34 1 ,?-T1iisopropyldiazen~ hv, R. T ., octane l.O 1.2 1.6 1 ,2-T)iisoorooyldiazen~ hv, R..T., gas 1.0 1.0 1.6 T"liisopropyl keton~ hv, toooc, gas 1.0 1.0 3.5 Comoound ~Cis or trans, qef. 104. QR. ef. 10 '5. -135- isopropyl-1,2-diazene would give the observed disproportionation products 32 Although a concerted B-elimination and 33. pathway cannot be ruled out, the formation of 2,3-dimethylbutane appears to indicate a nitrogen extrusion pathway for 18. A.. A+ - N=N N=N: ~Hj + \ ) H 32 18 1,1-Diazene Tautomerization The identity of red-orange species 1.J74 nm produced thermally or Amax = 4.64 nm photochemically from and Amax = 7 and n, respectively, which apparently yields tetrazene and hydrocarbon products (the expected thermolysis products from 7) of the 1,1-diazenes, is intriguing. The tautomer of a 1, l-diazene 7 (analogous to the enol form of a ketone) is azomethinimine 9. Stabilized azomethinimines with N-aryl substituents Ph H3 C\.+ - N=N ~c/ 7 \+ /J N-N \ 9 H 38 -1%- such as 38 have been prepared.96,97 These species are known to be deeply colored (red-orange, Amax = 4-60 nm, E ~ 10,600). Azomethinimines have also been postulated as intermediates in a number of chemical reactions. The reaction of alkyl and aryl lithium reagents with N-nitrosamines affords transient deeply colored species which can be trapped to give products consistent with the intermediacy of an azomethinimine.98 Neutralization I) Phli 2)~0 of rlimethyldiazenium salts with base in aqueous solution affords a transitory red species.99 Tetramethyl tetrazene 19 is the major product. This transient red color has been interpreted as evidence for 1,1-diazene 7. However, competitive deprotonation of dialkyl diazenium salts is also likely to occur at the a C-H resulting in formation of the tautomer of the 1,1-diazene, an azomethinimine.1 00 ! (Me)2 N-N=N-N (Me) 2 19 -187- Theoretical treatments for azomethinimines tend to indicate a singlet biradical grounrf state analogous to that found for the isoelectronic ozone. 9 7 The relative heats of formation of a 1,1-diazene and its azomethinimine tautorner have not been calculated. The electronic transitions for the deep red-orange species obtained from the corresponding l, l-dialkyldiazenes with a.-hydrogens are compared in Ta!)le X XXIII . These transitions follow the trend expected for substituent Table XXXHI. 'F.lectronic Transitions for Proposed 1, l-Diazene Tautomers. Tautomer A. max (nm\~ '· 464Q /J N-N \ --\.N-N-{ CN-N H 474Q \ ' H 491~ H <)N-N, 496~,~ H ~?-\Hl-rr:~· solution. 2This work. ER.ef. ~4. !i<:ihows thermochromism (A. max = 466 at 900K, A-max = lJ.Q6 at ?.OOOK). reversible -1~~- effects on a 'IT -'IT* transition (J" 5 nm red shift per alkyl suhstituent) as predicted by Woodwards rules for alkenes. 8 7 The observation of these deeply colored species k ~ 3000 ~-1 cm-1) in a number of solvents, the substituent effects on the electronic transitions, the observed thermal chemistry, and a comparison with the spectroscopic properties of stabilized azomethinimines suggests their tentative assignments as the tautomers of 1,1-dialkyl diazenes (azomethinimines). The apparent facile thermal rearrangement of 7 to 9 and R \+ I N=N R R =Me R = i- Pr \+N-N- R'~ \H R' 7 R' = H 9 ....... - R' = Me 30 ..... 18 18 to 30 at 90°'K suggests a verv small barrier for this transformation. Observation of this transformation in a triacetin glass well below the softening temperature may be indicative of a unimolecular pathway. Failure to detect 1, 1-diazenes 7 and 18 in equilibrium with 9 and 30 by electronic spectroscopy suggests that 9 and 30 are indeed lower in energy than the corresponding 1, l-diazenes. For comparison, a free energy difference tJ. G = I kcal/mol at 90°K corresponds to an equilibrium constant K = 268. Thermal Decomposition Kinetics The kinetics for tJ,ermal rlecomoosition of 1, 1-diazenes 7 and 18 in a - 1~9- softenerl. ?.-1\nHP glass (see Figures 4-1 and 4-2) do not bear interpretation due to the limited ciata attainable anci baseline effects due to overlap of the very strong transitions attributable to 9 and 30. In addition, the kinetics for growth of the A max= 4-64- nm and Amax = 4-74- nm transitions of 9 and 30 at 900K in a softened 2-MTHP glass are non-exponential and fit no simple kinetic treatment. The non-exponentiality may be attributable to the non- fluid nature of the softened glass (viscosity or site effects).81,82 Kinetic treatments for presumed site effects on unimolecular decompositions in rigid media (e.~., In A vs. t112, t113, etc.) did not yield interpretable data.lOl Although kinetic treatments at 900K proved inconclusive, thermal decomposition of 9 and 30 at warmer temperatures in 2-MTHF solution gave excellent kinetics. Thermal decomposition of 9 monitored at 4-64- nm gave bimolecular decomposition kinetics over a range of initial absorbances (Figure .58, Table XXXIV). Attempts to measure kinetics for formation of tetramethvl-?.-tetrazene 19 (Amax ~ no nm) which grows in with decay of 9 were unsuccessful presumably due to an overlapping UV absorption due to 9. 0 II + (D 3 C) 2 N~N=N=N: .. .. 13-d 6 -190- E - 4 .5f- c - - 1.5- . -···-··-·-·-·-· 20cPK ===·-·-·-·-·-·-·-·-·- -·- 0 .5 0 ·- ·-·-·- - - • - • - • - • - I 90° K I I 10 20 J 180°K 30 40 TIME (min) Fi~ure 58. Rimolecular decay l<inetics for Am ax = 464 nm species (9-h6) from thermolysis of 7 (?-MTHF). Table XXXIV. Second Order 'q_ate Constants for necomposi tion of 9-t\6. a ~?.- MTl-JF. Teml)erature (°K) k2 (A- 1 sec-1) r2 180 5.45 x to-5 0.999 190 1.40 x 1o-4 0.999 ?()I) 4.0' x 1o-4 0.999 ?JO 1.49 x to-3 0.999 -191- -6.5 -7.5 N ..X c __j -8.5 -9.5 -I 0 .5 L..,___ ____JL..__ _ ____J _ _ __ _ . L_ _ ____.__ _ _____. 4.65 4.85 5.05 5.25 5.45 5.65 1000/T(K) Figure 59. Arrhenius ptot for bimolecular decomposition kinetics of "-max= 464 nm species (9-il6) from 7 in 2-MTHF. For comparison, fullv deuterated (>99% d6) 13-ci6 was prepared. UV (VIS filtered) photolysis of 13-% in a 2-MTJ..fF glass at gooK afforded a structurec! absorption curve of 7-~ (A.max = 556 nm) identical to that obtained for 7. Warming the glass to 9QOK afforded 9-~ "-max= 464 nm. Prom the same initial absorbance the rate of formation of 9-h(; (tJ.A/6. t) is only a factor of 'V3 times the rate of formation of 9-cf6. This surprising result -19'2- woulrl be consistent with the lack of a primary deuterium kinetic isotope effect for tautomerization of 7 to 9 and the presence of only secondary isotone effects. For a change in hybridization from sn3 to so', a normal secondary rleuterium isotope effect ranges from kl--{ /kl") = t .4 to 3.1 per deuterium at 900T<. The thermal decomposition of 9-~ shows a strong deuterium isotope effect (J:'igure 60, Table XXXV). decomposition of 9. At 1900K kH/kD = 6.7 for bimolecular The magnitude of the deuterium isotope effect is consistent with a combination of primary and secondary isotone effects. A normal primary isotope effect of kH/kn = 5 at 2980T< would correspond to kJ--J /kn = 1?.. '5 at 1900T<. For a change in hybridization from sp2 to sp 3 a normal secondary isotope ranges from kH./ko = 0. 59 to OJD per deuterium at t 900K. A.. combination of two secondary isotone effects and a nrimary isotope effect would be exnected to range from kHikT) = 4.3 to 8.6. The observed k4 lkn = 6. 7 at l 900K is well within this range. 1\ssuming an extinction coefficient of allows calculation E: ~ 3000 ~-1 cm-1 for 9 of Arrhenius activation parameters for bimolecular decomposition of 9. For 9-h6 J:a = ~.2.:: 0.5 kcallrnol anri lo~10 t\ = 1.8.:: 0.6 (6Hf = 7.~ kcal/mol, t~~:f = -51.3 e.uJ. For 9-~ Ea =8.6 .±. 0.5 kcal/mol and log 10 A = 1.4 .=: O.t;. The apnarent low .1\ factors are consistent with severe stereochemical constraints on the transition state. For comparison, direct dimerization of kinetically persistent diazene 4 Ea = 6.4 .=: 0.9 kcallmol and logto t\ = 3.8.:: 0.7. ~echanistically, the presence of a deuterium isotone effect for -193- . . -·-·- -·-· .-·- -·-·- 2CXJ'K 0 .6-- • - • - • - • - • - • - • - • - • - • 190oK I I 10 20 - I 40 TIME (min) Figure 60. Bimolecular decay kinetics for Xmax= 464 nm species (9-~) from thermolysis of 7-<i6 f2-"'.nl-JF), Table XXXV. Second Order 'R.ate ~onstants for necomposition of 9-~~ ~2-MTI-JF. T emoerature (OK) k7. (A.- 1 sec-1) r2 190 2. to x Jo-5 0.997 7.00 5.83 x 1o-5 0.999 210 1. 52 x 1o-4 0. 99 .5 215 3.20 x to-4 0.999 -194- -7 . 7~------~--------~--------~--------~ -8.4 -9 . 1 N .X c: _j -9.8 -10.5 .""' -II .2 L--------'-----~------'--------' 4.55 4.75 4.95 5.15 5.35 1000/T(K) Figure 61. Arrhenius plot for bimolecular decomposition kinetics of "-max= 4{,4 nm soecies (9-~) from 7-~ in 2-MTHF. -1.95- bimolecular loss of 9 may be consistent with a mechanism involving an eauilibrium of 9 and 1, 1-diazene 7 followed by dimerization of 7 with 7 (unlikely, no isotope effect expected) or dimerization of 7 with 9 to form tetrazene 19 (Scheme V). .A. mechanism involving rate determining tautomerization of 9 to 7 followed by dimerization of two 7's should show unimolecular decay of 9 (see 1\ooendix Mechanism .A.) • .A. rapid eauilihration of 9 to 7, with equilibrium favoring 9, followed by rate rletermining dimerization of 9 and 7 to form tetrazene 19 should show the observeci bimolecular kinetics Scheme V Me\+ Me I kl N=N k_l 7 Me\+ ! N=N 9 ...... - 7 + 7 - - 7 + 9 9 + 9 kl 2 k2 k II 2 ~ 19 ...... 19 ...... 19 ...... \ H -196- (see Appendix Mechanism B) if k 1 > > k2( 9), and a deuterium isotope effect as observed. =2 Dimerization of 9 with 7 may be similar to the known facile acid catalyzed dimerization of 1,1-diazenes (diazenium ion plus 1,1-diazene followed by deprotonation). One resonance form of 9 is similar to a 1,1- diazenium ion with an anionic substituent 9a. J H3C \ /.. H3C .. N-N CH ~ .. / N-N .. 3 \ CH 3 19 -197- 0.0~-------~~-------~~-------~~-------, -0.5 ~----·-·---·-·-·---·-·---·--~·-----·-•-•._220°K 0 f--·- · - · - · -1 . -1.5 f~ ......... ·-·-·-·-·-----·-·-·-·-·-·........ ·-·-·-·...-......._ 2250K ._ - -2 .0f- .......................... <! c: ......J -3.0f- - - 230°K - ........ -2.5 f- - ............ ........... - ........ ........ •..... 235°K .......... -3. 5L---_ _ ___,I_ _ _ __.___ l _ _ _--=-1'="'~------=' 0 20 40 60 80 TIME (min) Figure 62. Unimolecular decay kinetics for decomposition kinetics of A. max = 474 nm species (30) from thermolysis of 18 (2MTHF). Table XXXVI. First Order Rate Constants for Decomposition of 30~ ~2-MTJ-IF. Temperature (OK) k 1 (A -1 sec-1) r2 220 4.68 X l0-5 0.999 225 1.09 x Jo-4 0.999 230 2.'35 x to-4 0.998 235 4.53 x Jo-4 0.999 -198- -7.0r--------------.-----------------.-----------------.--------------, 2- MTt-F Solvent "" • E0 = 16.2 :t 0 .5 KcoVmol ~ A = 11.8 ± 0.3 -8.5 ..:.:: c: ....J • -10.5L.___ _ _ 4 .2 4.3 --..J.__ _ _ _..L.-_ _ ___.__ _ _ _ _ , 4.4 4. 5 4.6 1000/T(K) Figure 63. Arrhenius plot of unimolecular decomposition of !..max= 474 species (30) from 18 in 2-MT'I-fF . -199- Scheme VI ~+N=N ~ - \ - 18 30 ~+N=N- -< 18 - 32 = (~) k 2 (30~ kl 33 ....... - 34 -200- Table XXXVIT. Qelative Rates and ~c;:f•s for 1,1-niazene llecompositions at _q40~ ()7qOJ<). 1, J-T)iazene ~+N=N -1 k rei 19.4~ 4 X 11)-9 19.1~ 1 X IO-~ 16.~Q 6 x 1o-6 1?.5~ 1 ~,./ """+N=N~ ~Pef. ~g (Pt?O). QThis work, intermediacy of 18 assumed, k = k2/K see text f?-~n'--n::· ). ~Ref. ~f (~.,e20). -201- The thermal decomposition of 1, l-diisopropyl diazene and formation of red-orange 30 Amax = !~74- nm at 90°K did not give interpretable kinetics as previously discussed for 1,1-diazene 7. The decay of 30 monitored at 4-74- nm ~ave unimolecular decomposition kinetics (Figure 62, Table XXXVI). The UV absorption of tetraisopropyl-~-tetrazene 20 does not grow in with thermal decomposition of 30. The major products from thermolysis of 1,1-diazene 18 are hydrocarbons 3'-, 33, and 34. This is consistent with a mechanism involving rapid equilibrium of 1,1-diazene 18 and tautomer 30 followed by rate determining thermal decomposition of 18 to afford hydrocarbon products expected from decomposition of 18 (Scheme VI) (see Appendix Mechanism C). The Arrhenius activation parameters for unimolecular decomposition of 30 Ea = 1.6.8 .±. 0.5 kcal/mol and log10 A= 11.~ .±. 0.3 (t.H:f = 15.8 kcal/mol, t.s=f = -5.7 e.u.) are consistent with the activation parameters expected for nitrogen extrusion from 1,1-diisooropyldiazene 18. For comparison the activation parameters for cyclic kinetically persistent 1, 1-diazene 5 are Ea = 19.1 .±. 0.4 kcallmol, log10 A = 1?.1 .±. 0.3 in TJiF solution while the activation parameters for 1, l-di-tert-butvldiazene 6 are Ea ~ 13 kcal/mol for an assumed log10 A= 13. The activation free ener~ies for 1,1-diazenes are compared in Table XXXVII. Matrix FT-IR Studies of Carbon Monoxide A group of infrared bands assigned to CO show photochemical behavior which mirrors the behavior of 1,1-dimethyldiazene 7 and its photochemical decomposition oroducts. Similar behavior was observed for CO bands -202- associated with H2''JN 3 and its prorfucts (see Chaoter 1). UV (VIS filtered) ohotolvsis of 13 0:1400, Ar, lOOK) (Figure ~4a) yields CO bands at ?142.0?, 21~7.~0, ?1'35.9'3, 21'34.75, anrl 213'2.95 cm-1 (Figure 64h) in addition to PT-I'R bands of l,l-diazene 7, dimethylaminoisocyanate ancl ethane (see P igure 44 ). Photolysis of 7 with visible light results in loss of two of the CO hands at ?14?.0/. (minor) and 2134.75 (major) and growth of two CO hands at 2117.~0 and ?1':12.9.5 cm-1 (Figure 64c). One CO band at 21'35.93 cm-l remains aooarently unchangerf as a shoulrfer on the 2137 .sw cm-l band. r::oincident with this transformation in the c:n stretch region, infrared bands clue to I, 1-diazene 7 ( a rna ior set of bands (r) M) and a minor set (Dm)) are lost while infared hands due to ethane and an unknown species (U) grow in. Further photolysis (A 340-410 nm) results in loss of FT -IR bands of (U) and growth of ethane bands. Coincident with this transformation, the CO band at 2112.9'5 cm-1 is lost while the CO hand at ?.137JW cm-1 grows in (Pi~ure 64d). Shifts of the C:O stretch frequency due to hydrogen bonding (e.~., H20, I\Jl-l'3 ), dipole-dipole and electrostatic interactions? I (e.g., 1\12) allows the qualitative assessment of the matrix environment of a CO-molecule complex (see C:haoter 1) fTable XXXVIII). The en hands at ?142.03 (minor) and ?134.75 cm-1 (major) correspond in intensity ratio and ohotochemical behavior to two sets of bands assi!:~ned to 1,1-diazene 7 (n~~' T)m). The shift of the minor CO band at 2142.0'3 cm-1 (COIT"\m) to higher energy relative to CO in an argon matrix (2138.25 cm-1) or CO in a nitrogen matrix (21 '39.69 cm-1) suggests a hydrogen bonding or -203- 0.80 0.75 0.70 0 . 65 R 0.60 b 0.55 s 0.50 0 0.£l5 r b 0.£l0 a 0.35 n c e b c 0.30 0.25 d 0.20 0. 15 0. 10 0.05 2150 2140 2130 2120 We.venumbers Figure 64. Successive FT-IR spectra of the CO stretch region (Ar, t0°K). (a) Refore ohotolysis of 13 (1:1400, Ar, lOOK). (b) After UV (VIS filtered) photolysis of 13 (l h) to form 7 and CO. (c) After VIS (470-610 nrn) ohotolysis (I h) of 7 to form ethane and (U). (d) After photolysis of fU) ~40-410 nm) fi h) to form ethane. -204- Table XXXVill. Matrix Isolation Carbon Monoxide (CO) Infrared Stretch .,:::'reouencies. Obs.~ (Ratio) CO/Matrix v CO (cm-1) Lit.E v CO (cm-1) (Ratio) C:O/ Ar (J :?000) ?138.25 2138.40 (1:1000) CO/N2 (l :?000) 2139.69 ?.139.69 (1:1000) CO/H?.O/Ar ":?:?000) ?14g.6o ?14~.~ (2: '5: 1OOO) 2143.3 (1:1:1000) ~O/NH3/Ar CO/Hi~"J/ Ar c ?140.90 C'.O/H.2/"J?/N2/ Ar c 2139.94 2138.97 2137.53 CO/(CH3 )2~N(T)M)/I\r <i 2134.75 COI(C'.H 3)2NN(T)rn )/ 1\r d 2142.02 CO/C2H6/N7/N2/ Ar -d CO/(U)/ Ar <i 2137.~0 2135.93 ?132.95 ~This work (!QOK). Q~ef. 71 (lOOK). ~Matrix ratio approximately (1:1:?000, 1\r). .2Matrix ratio approximately (1:1:1400, Ar). dipole-dipole stabilizing interaction (higher energy stretch) for this CO. The lower energy stretch (relative to free matrix) for the major CO stretch at 2134.75 cm-1 rco/nM) sug?.;ests a destabilizing dipole-dipole or electrostatic interaction for this CO. The orientation of CO relative to 1, t -diazene 7 could give rise to either a dipole-dipole stabilized (CO, nm) or destabilized fCO/DM) CO environment. A steric interaction with the methyl groups of 7 -?05- may be sufficient to prevent a stabilizing interaction with CO. Two C:O bands at 2137.80 and 2135.93 cm-1 mirror the behavior of bands due to ethane which grow in with ohotolysis of 7 and (U). The lower energy stretch frequencies of these CO's (relative to free matrix CO) suggest an intimate association of CO with ethane and N2 (possibly two l\J2's) in a matrix site. For comoarison, the stretchin~ frequency for CO in a rigid 3- methvl oentane matrix is ?132 cm-1. In addition to ethane, VIS ohotolysis of 7 affords infrared bands of a minor unirlentified soecies fU). A low frequency CO stretch at 2132.9 5 cm-1 grows in with (A 340-41 I) nm). fTJ) and is suhsequentlv lost with photolysis of (U) The low frequency of the CO band at 2132. q 5 suggests a destabilizing dioole-dioole or electrostatic interaction of CO with (U). A destabilizing dioole-dioole interaction of CO with the lone oairs of cis 1,2dimethylr!iazene 8c would be consistent with the observed CO stretch frequency. Interestingly, the FT -IR. bands of (U) correlate well with those of 8c. -'206- Summary The low temperature matrix isolation and direct spectroscopic characterization of 1, 1-dimethyldiazene 7 and 1, 1-diisopropyldiazene 18 have been described. The UV (VIS filtered) photolysis of carbamoyl azides l3 and 17 in a rigid medium (organic glass, 8QOK or Ar matrix, lOOK) provides a new general method for the photochemical generation of reactive 1, 1-dialkyldiazenes. This photochemical route is considered to proceed by the photoCurtius rearrangement of a carbamoyl azide to an aminoisocyanate followed bv photodecarbonylation of the aminoisocyanate to a 1, 1-diazene and CO. The 1, l-diazenes are purple showing structured absorption curves in the visible Amax = 556 nm for 7 and Amax = 504 nm for 18. 1,1-0iazene7 was independently generated by photolysis of (Z)-3,3-dimethyl-1-phenyltriazene1-oxide 16 (2-~~THF, 800K) to form 7 and nitrosobenzene. Comparison of the absorptions permits estimation of the extinction for 7 of 6 (M-1 cm-1). The characteristic N=r.J double bond stretches of the 1,1-diazenes are located at -?.07- 1600.% cm-l for 7 and 1600.92 for 18. The spectroscopic characterization of matrix isolated l, l-dialkyldiazenes is consistent with the theoretical (GVP,-~T) prediction for H2'\J"J 3 and the experimental characterization of 3 and kinetically persistent l, 1-diazenes 4, 5, and 6. Thermolysis of 7 and 18 yields deeply colored species 9 (A.max = 464 nm) and 30 (Xmax = 474 nm), respectively. These species are tentatively irlentified as the azomethinimine tautomers of the 1, t -diazenes witll n-hvrlrogens. Photolysis of 1, t -diazenes 7 and 18 at their n-iT* transitions in the visible also results in formation of 9 and 30 (organic ~lasses, ~0°K). The products of suhseauent ther111olysis of 9 and 30 are the 7.- tetrazene and hydrocarbon products expected from the respective 1,1diazenes. The kinetics for thermal decomposition of 9 and 30 are consistent with an equilil-)rium with the corresponding 1,1 -diazenes 7 and 18 followed by bimolecular dimerization of 7 to 2-tetrazene 19 and unimolecular nitrogen extrusion for 18. Photochemical decomposition of 7 (Ar, 10°'K) yielrls infrared bands of ethane anrl an unidentified sPecies (U) which is subsequently photolyzed to ethane. which mirror the The observation of several FT-I'R. stretches for CO photochemical behavior of 1, 1-diazene 7 and its photorlecomposition Products in an argon matrix (Ar, 1OOK) sug~ests an intimate association of r:o with these products in a rigid environment. -208- EXPERIMENTAL SECTION Melting points were determined using a Thomas-Hoover melting point apparatus and are uncorrected. Infrared spectra were recorded on either a Perkin-Elmer 257 or a Shimadzu I'R.-435 spectrophotometer. Fourier trans- form infrared spectra (FT -JR) were recorded on a ~attson Instruments Sirius 100 FT -IR equipped with Star lab minicomputer data station and high resolution graphics terminals u11der a positive nitrogen purge at 0.25 cm-1 using l-JgCrlTe (~~CT) detector cooled to 770K u11less otherwise noted. Proton ~MR spectra were ohtainerl 011 either a Varian Associates EM-390, Jeol FX- 900 or Varian Associates XL-200 spectrometer. Chemical shifts are reported as parts per million fppm) downfield from tetramethylsilane (TMS) in c units and couoling constants are in Hertz (Hz). N~~R data are reported in this order: chemical shift; multiplicity, s = singlet, d = doublet, t = triplet, m = multiplet; number of protons; couoling constants. Nitrogen (15N) NMR soectra were obtained on either a Jeol FX-900 or Brucker WM-500 spectrometer. Chemical shifts are reported in parts per million (ppm) from 15~ nitromethane with a negative value indicating an upfield position. T1euterium (2H) NMR spectra were recorded on a Jeol GX-400 spectrometer. Chemical shifts are reported in porn from TMS. Electronic spectra were obtained using a Varian associates '219 spectrophotometer. Electronic spin resonance (ES'R.) spectra were recorded using a Varian Associates E-line spectrometer equipped with an Air Products and Chemicals l-Jelitran L TD-3ll 0 liquid helium transfer apparatus. Emission spectroscopy was performed with assistance of members of Professor H. B. Gray's group at Caltech using -209- a noncommercial spectrophotometer with a 250-watt xenon source and Hamamatsu ~-406 and R-955 photomultiplier tubes. 'Raman spectroscopy was performed with assistance of members of Professor S. I. Chan's research ~roup at Caltech using a Spex Industries model 14018 double monochromator equipped with 2400 line/mm holo~raphic ~ratin~s and a l-iamamatsu R-955 photomultiplier tube. Spectral slit widths were 3 cm-1. A Spectra-Physics model 170 argon iron laser was used directly or in conjunction with a Spectra-Physics 375 dye laser. Samples were prepared in _'5 mm O.T). quartz tubes immersed in a liquid 1\1'2 filled quartz finger dewar. Light scattered at 900 relative to the incident beam was collected. A Spex Industries SC-32 c;;~.A,.MP controller/de~ta orocessor was used for data manipulation. For analytical vapor phase chromatography (VPC), a Hewlett-Packard 5700" ~as chromatograph equipped with a Hewlett-Packard 18704A inlet splitter and flame ionization detector was used. Hydrogen was employed as the carrier and nitrogen was used as the makeup gas. Packed column analytical V°C was done using a Hewlett-Packard 5720A gas chromatograph equipped with a flame ionization detector and nitrogen carrier gas. instrument was used with 1/8 in. steel columns. This All quantitative VPC was accomplished with a Hewlett-Packard 3390.1\ electronic integrator. VPC response factors for aliphatic hydrocarbons were assumed to be 1.00 relative to n-alkanes. Ouantitative analysis of other products were corrected for detector response. For preparative VPC, a Varian 920 instrument equipped with a thermal conductivity detector and helium carrier gas was used. This instrument was used with 0.25 in. or 0. 375 in. aluminum packed columns and -210- Table XXXIX. VPC: Columns D isig;na tion Description Carbowax 20\A 10 ft. x 3/~ in. aluminum; 25% Carbowax 20M on 60/80 Chrom W-A \V-fVv1CS 20 ft. x 1/8 in. aluminum; 25% 2,4dimethyl-sulfolane on 80/100 Chrom PNAW Penn walt 6 ft. x 1/4 in. glass; 28% Pennwalt 223 M/100 Chrom R Sf:-'30 40 meter x 0. 2 rnm IT) fused silica capillary 15 meter x 0.3? mm IT) fused silica SE-54 capillary 3 ft. x 1/~ in. 60/80 sieve 5 A <;ieve 5 A adapted for use with 1/8 i11. packed columns. VPC columns are listed in Table XXXIX. with Identities of products were established by coinjection techniques authentic samples from preparative VPC and spectroscopic characterization. lsotooe compositions were determined from combination of mass spectroscopy and FT -IR. Gas pressures were measured from known volumes with an MKS type 221 capacitance manometer. Mass spectra (MS) were recorded on a nuPont 24-492'13 Mass Spectrometer. Combustion analyses were performed at the Caltech Microanalytical Laboratory. All reactions were run under a positive pressure of argon. Linde Ul-fP argon and Ul-IP nitrogen were used as received. passed through a -78oc trao prior to use. received. Ethane (Matheson) was CO (Matheson) was used as Methyltetrahydrofuran (2-MTHF) was distilled from calcium -211- hydride then distilled from sodium/benzophenone ketyl using excess sodium prior to use. Toluene, pentane and acetonitrile (CH3CN) were dried over type 4 A moler:ular sieves. Oiisopropyl amine was distilled from barium oxide. T)imethyl carbamoyl chloride (Aldrich) was used as received. Sodiumt-15N azide (Prochem q7 atom %) was used as received. (nPrCN) was distilled from P205. Oimethylamine dG-hydrochloride (KOR Isotopes, 99+% d6)110 was used as received. from calcium hydride. Butyronitrile 3-Methvlpentane was distilled Triacetin was dried over type 4 A molecular sieves. Para-dimethylamino benzaldehyde was treated with activated charcoal and recrystallized from ethanol. Norbornene was purified by distillation from sodium. Acetyl acetone was purified by preparative VPC (Carbowax 20M). Low Temperature Matrix Isolation Apparatus. The low temperature apparatus is of conventional design9 constructed on a mobile cart employing an Air Products J")isplex CSW-202 DMX-lE closed cycle helium refrigeration unit with rotating optical shroud)l6 The expander module is mounted on a rack and pinion gear assembly for horizontal positioning suspended from a threaded crank lift for precise vertical positioning of the optical portion of the rotating shroud. l-Hgh vacuum (< I0-7 mbar) was achieved with an Edwards Mk.2 f.'liffstackllq and measured with an Edwards Penning 8 gauge mounted on a '3" coooer manifold equipped with Veeco 123 brass bellows type high vacuum valves. Temperature control was maintained with an Air Products !)isplex-E heater/controller unit with an iron-doped gold/chrome! thermocouple (2-3000K ~ 0.1 K). Calibration was checked against LHe (4.20K) and LN2 (77 .50K) by immersion. The thermocouple tip is embedded in -212- Figure 65. Low temperature spectroscopic ce1150 for electronic (UV-VIS) spectroscopy and rotatable shroud. -?13- CELL END WINDOW VITON 0-RING PLATE\ [)aD [)al ~ BODY I ~ I I \~~ENE SET Ell() -o SCREW s· 4-40 X 16 SCREW Figure 66. GASKET Low temperature UV-VIS spectroscopic cell components.50 -214- RADIATION SHIELD GAS INTRODUCTION PORT /"-...... I \ " \~\ \ \ \ ,_./J VIEWING PORTS 4 EA. @90° WINDOW HOLDER Figure 67. Rotatable spectroscopic shroud for matrix studies. -2t 5- 1"0. D.Csl Figure 68. Inner window fe.g., Csl for matrix FT -IR) mounting assembly. Fi~re 69. ~ / VEE CO BELLOWS VALVE • MAGNETIC STIRRER LEAK DETECTOR ACCESS ~HELIUM BACKING/ROUGHING VALVE SAMPLE BULB TO ROUGH PUMP DIFFSTAO<~ EDWARDS PENNING HIGH VACWM GUAGE tl 3L SAMPLE BALLAST ACCESS PORT I -"' I N -+--MATRIX GAS INLET Schematic of low temoerature matrix isolation apparatus vacuum system and gas handling mani folrl (not to scale). SUBLIMATION SAMPLE INLET /" VEEOO BELLOtiS VALVE DISPLEX CRYOSTAT EXPANDER MODULE ABSOWTE PRESSURE---.. GUAGE -217- C1___.1__N01 I 0 I 0 KtA 0 W = WHITE R = RED Br = BROWN B =BLACK 81 =BLUE G =GREEN Y =YEU.OW y I : RELAY {K) LAMP : C2 : t'l)2 ~~~A~~--J_K~--p~B~I______~~~~-~:W~~W ®~--~CD WATER ACC FLOW SENSOR K1B ~ ....._..,....___R_ _ _®-=ir'@ ti=;,.....__B--e_G___. 208 VAC G 33K.Q ~ R 8 Dv ffi~ffi R t R R cp K1C AND DISPLEX COMPRESSOR 33 KU 20A TO 110 VAC {ABO\/E) WATER FLOW SOLONOID .....___;_R.;__..... ~ @ (j) R" ~3K.Q I lOA DPDT 110 VAC TERMINAL STRIP DIFFSTACK HEATER Figure 70. Schematic of cryostat/ diffstack cooling water protection circuitry. -218- the cold tip of the exoander module just above the window mount. Matrix gas mixtures were oreoared usin~ standard manometric techniques in a glass vacuum manifold equipped with high vacuum greaseless joints (Witeg) a11d double 0-ring Teflon stopcocks (Ace Glass). l 26 measureri with a Pennwalt ~ode! Manifold pressures were 1500 absolute pressure gauge (0-800 torr).l? l Matrix gas and gas mixture deposition rates were controlled with a Granville-Phillips variable leak valve.l22 Matrix gas was admitted through a Nupro l 2 5 brass bellows type high vacuum valve. Sublimation deposition and cryooumo deposition samples were co-deposited with matrix gas from a 1/4 in. 1.0. pyrex sublimation tube connected with a Cajon Ultra-Torr fitting from a low temperature bath. For I'R. spectroscopy an inner Csl window and '<Br outer windows (International Crystal Labs) were used. Photolysis was carried out through an outer Suprasil I (Amersil) window through the back of the inner Csi window with defocused light from a 1000 watt ozone free xenon source (Oriel).l20 The copper vacuum manifold, flex hoses, mobile cart, expander module positioning crank, and rack and pinion assemblies were constructed in the Caltech mechanical shops. The glass vacuum manifold was constructed in the Caltech glass shops. Carbamoyl Azide ((_5).49 To a solution of 33.5 g (0.3 mol) semicarbazide hydrochloride in 150 mL distilled water cooled in an ice bath was added dropwise with stirring a solution of 22.~ g (0.33 mol) sodium nitrite in 50 mL rlistilled water. After the addition was complete, three drops concentrated sulfuric acid was added and the stirring was continued for 10 min. in the ice bath. The solution was saturated with ammonium sulfate and extracted with -219- ether (4 x 50 mU. The ether was removed under reduced pressure and the resulting solid was recrystallized from hot ether affording 16.4 g (6396) 1.5 as large prisms; mp. q~oc Oit. 970C). 49 This recrystallized material was freshly suhlimed to a ooc cold finger (bath temp. 30-700C, 1.0 mm Hg) prior to use. 1\1\i-q_ (CJ)ClJ) <S 5.'24 (bs); UV (MeOH) Amax = 214 nm (log e: = 3. '3), A. = 266 nm Oog e: = 2); I'R. (CH3CN) 3450, 3350, 2155, 1724, 1600, 13n, 1?.?.'>, Ill'>, 864, 699, 564, 510, 455 cm-1; MS m/e 86.0533 (M+). Carbamoyl Azide 1.5N {1.5-1.5N). A solution of 100 mg (1.?. mmon carbamoyl azide (15) in .5.0 mL dry acetonitrile was stirred with an excess (typically 4 equiv) 320 mg (4.~ mmol) 1-15N-sodium azide for 48 to 60 h at room temperature in the dark. The mixture was filtered and the solvent removed in ~· The remaining solid was purified by sublimation to a ooc cold finger 0.0 mm !-ig) affording 70-75 mg (75%) 1.5-1.5N approximately 8096 15N labeled, 20% unlabeled. l5N NMR (CH2Cl2) o-140.0~(s),-266.3l(s) (from CH315"'102); IR (CH3CN) 3450, 3350, 21.55, 2145, 2125, 1724, 1600, 1322, 1225, 1218, 1200, 1115, 860, 695, 560, 510, 450 cm-1; MS mle= 87.0503 (M+), m/e = 87/86 = 46/t?.. Carbamoyl Azide-dJ. (1.5-d2)· Carbamoyl azide (500 mg, 5.8 mmol) was stirred with 400 equiv (?.. 3 mol) 42 mL D20 (99.8% D) for 48 h in the dark at room temperature. The excess D20 was removed in vacuo and the product purified by sublimation under reduced pressure to a ooc cold finger (bath temp. 30-700C, 1.0 mm Hg), affording 390 mg (76%) 15-~ as a colorless solid. IR (CH3CN) 3360, 2670, 2590, 2480, 2425, 2200, 2170, 1720, 1340, 1320, 1245, t2l8, 1200, 1192, 8'20, 73 .5, 695,658 cm-1; MS mle= 88.0656 -220- (M+), m/e =88/87/86 =?.2/2/1. Carbamoyl Azide-~ l5N 0_5-<f2-1.5N). The procedure for carbamoyl azide-d2 (l '5-d2) was followed with 50 m~ (0. 58 mmol) carbamoyl azide 05d2) was followed with 50 mg (0. 58 mmol) carbamoyl azide 15N (15-15N) affording cS 30 m~ (59%) as a colorless solid. 15N NMR (D20) -140.10 (s), -l62.~2 (s); IR (CH3CN) 3360, 2670, 2590, 2480, 2425, l170, ?.14_5, 21'25, 1720, 1340, 1320, 1245, 1218, 1200, 1192, 8?0, 735, 695, 658 cm-1; M'5 m/e = ~9.06?.7 (\~+), m/e = 89/88/'1,7 = 23/5/1. Ammonium Azide.l ?.8 To a paste of 6.5 ~ (0. J mol) sodium azide in 6.5 mL distilled water was added 40 mL benzene. The slurry was cooled to -50C in an ice-salt bath. To this was addecf dropwise with stirring 2.7 mL (0.05 mol) concentrated sulfuric acid. Once the addition was complete, the benzene layer was decanted, dried briefly over anhydrous sodium sulfate and added to 40 mL anhydrous ether at -5oc. Dry ammonia gas was bub'bled through this solution at -50C yieldin~ an insolu'ble precipitate. The product was filtered, washecf with anhydrous ether anc! dried in a desicator yieldin~ 4.5 g (7 5% ), 1330C sublimes (Ht. 13 30 su!:>limes).ll8 IR (KP,r) 3110, 20?0, 1620, 1500, 446, 640 cm-1. Dimethyl Carbamoyl Azide (13).4 7 To 14.0 g (0.2'2 mol) sodium azide in 25 mL dry acetonitrile was added with stirring under argon 10 mL (0.11 mon dimethyl carbamoyl chloride (21) in 10 mL dry acetonitrile at room temperature in the dark. The mixture was stirred for 24 h or until complete as determined hy IR (CH3CN) (vco, 1740 cm-l of the carbamoyl chloride reolaced by 'VCO, 1685 cm-1 of carbamoyl azide) or VPC (15m SE-54 60°C). -221- The mixture was filtered and the acetonitrile removed in vacuo. --- The resulting clear oil was distilled under reduced pressure yielding 11.7 g (92%) (13) as a colorless oil; bp 53-550C (5 nm Hg). Preparative VPC (Carbowax 20M, l800C, 1800C injector) afforded analytically pure 13. NMR (CDCl3, 220C) o 2.97 (s, 3l-I), 2.92 (s, 3H); IR (CCl4) 2930, 2445, 2167, 2145, 1697, 1485, 143~, 1409, 1390, 1270,1230, 1135, ll19, 1060,963,725,647 cm-1; UV (MeOT-J) Amax= ?28 nm Oog e: = 3.7), A= 250 nm Oog e: =2.7); MSm/e = 114.05466 (M+) (theor. 114.05415). Dimethyl Carbamoyl Azide 1.5N (13-1.5N). To 0.25 g (3.8 mmon 1-15Nsodium azide in 2 mL dry acetonitrile under argon was added 0.40 g (3.72 mmon dimethyl carbamoyl chloride (21) via syringe. The mixture was stirred in the dark for ?4 h at room temperature. The mixture was filtered and the acetonitrile removed under reduced pressure affording 0.41 g (97%) 1J-15N as a colorless oil. afforded 0 2.97 Preparative VPC (Carbowax 20M, 18ooc, 1800C injector) pure analvticallv (s, 3H), 2.92 (s, 0 -141.40 (s), -266.0'2 (s); 148.5, 1436, 1408, 13~5, NMR 3H); IR (CC14) 2930, ?395, 2162, 2135, 2117, 1690, 1265, 1217, 1200, 1130, 111.5, 1060, 960, 720 645 cm-1; MS m/e = 115.05149 (M+) (theor. lt5.05119). 1,1,4,4-Tetramethyl-2-tetrazene (19).5g To a solution of 10.0 g (0.17 mon 1,1-dimethyl hvdrazine in 100 mL anhydrous ether was arlded with stirring under argon 40.0 ~ (0.18 5 mol) yellow mercuric oxide in small portions to maintain reflux of the ether. Once addition of the mercuric oxide was complete the mixture was refluxed an additional 20 min and then cooled -/.22- on ice. The black precipitate was removed by filtering through celite and sodium sulfate and washed with ether. The light yellow filtrate was dried over sodium sulfate and the ether removed under reduced pressure. The remaining yellow oil was distilled under reduced pressure affording 5.6 g (62%) 19 as a colorless oil; bp 44-450C (30 mm Hg). Preparative VPC fCarbowax 20M, 1350C, l500C injector) afforded analytically pure material. 1\J\AR. (CnC13) o 2JD (s) ( lit. (C:OC13) o 2.~3 (s)) ;58 IR (film) 3015,2980 , noo, 1475, 1425, 1260, 1245, 1145, 1095, 1042, 1005, 900, 827 2~15, ?83~, cm-1; IR ( Ar, 1:700, l0°K) 30?2, 3020, 301 O, 2980, 297 5, 2890, 2~80, 2870, 2830, n95, 2n5, 1505, 1475, 1465, 1450, 1438, 1400, 1282, 1245, 1145, 1095, 1040, 1035, 1013, 100'>, 900, 828 cm-1; UV (3-methy1 pentane) Amax= ?7R nm Oog e: = 4.1). Photolysis of Tetramethyl-2-tetrazene (19). A mixture of freshly VPC purified tetrazene in argon (1:700) was r!eposited at 5 mmollh over 2 h at 20°K. Photolysis with two C<;-7 -54 UV transmitting visible cut out filters for 1 5 min. resulted in loss of the I'R. bands of the tetrazene and formation of new bands consistent with products resulting from combination/disproportionation of two dimethyl amino radicals to form dimethyl amine h> NH, 3340 cm-1) and N-methyl methylenimine (v =CH 2, 3030 cm-1) and (vN=C, 1663 cm-1) lit. 0658 cm-1)129 together with the following product bands: 3340, 3030, 2950, 2910, 2865, 2835, 2795, 1663, 1480, 147 .5, 1458, 1445, 1400, 13~0, 1230, 1'225, 1155, 1150, 1055,1030, 1027,950,808,795, 770, 735 cm-1 ( lit. N-methy1methylimine IR (Ar, 20 K)129 3030, 3005, 2894, 276~, 1658, 1472, 1467, 1439, 1399, 1218, 109~, 1045, 1023, 948, 93~ -223- cm-1 I. Trans-1,2-dimethyl Diazene (8t))09 To a suspension of 5.64 g (26 mmol) yellow mercuric oxide in 10 mL distilled water at ooc was added slowly with stirring a solution of 1.80 g ('30 mmon 1,2-dimethyl hydrazine in 15 mL distilled water. An additional 1.51 g (7 mmon mercuric oxide was added and the mixture stirred for an additional 2 h at room temperature. The product 1,2-dimethyl diazene was distilled from the reaction mixture under reduced pressure '50 mm Hg) '.Vith mild heating to 5QOC and collected in a trap at -78oc. The product was further purified by trap (-780C) to trap (-1 %°C) distillation throu~h a Orierite column at high vacuum yielding 0.82 g (64%) 8t as a colorless oil. o 3.62 (s); lit. NMR (d6- NMR (d6-acetone) acetone) l 09 0 3.63 (s) trans ( 3.51 (s) cis!. Dimethyl Carbamoyl Chloride-d(; (2I-<i6).47 Hexadeuteriodimethyl- ammoniurn chloride (KOR. Isotopes, 99+% cl6) 1.9 g (45 mmon was converted to the free amine by stirring with 4.0 g sodium hydroxide in 8.5 mL dd H20 and 50 mL dichloromethane in an lee bath. After 15 min the dlchloromethane laver was separated, dried over potassium hydroxide pellets and added clropwise with stirring over 45 min to 12 g 020 mmol) phosgene (CAUTION: extreme poison!) in 30 mL dichloromethane under argon at -1 ooc in an ice/salt bath. After 1 h the excess phosgene and 3/4 of the dichloromethane were removed in ~ and trapped at -7~0C over potassium hydroxide pellets. To the concentrated reaction mixture was added 25 mL dry ether to precipitate 1.5 g of hexadeuteriodimethylammonium chloride for filtration. Evaporating tl)e filtrate in ~ left 2.3 g dimethyl carbamoyl chloride-d6 -224- (21-<16) which was distilled in vacuo; bp 60-6 t oc (17 mm Hg) yielding 1.95 ~ (76%) as a colorless oil. I'R. (film) 2210, 2140, 2120, 2080, 1740, 1638, 1420, 1370, 1320, 1280, 1220, 1115, 1060, 94 5, 800, 660, 640 em- l. Dimethyl Carbamoyl Azide-~ (1 '3-d6). 4 7 To a suspension of 1.4 g (2? mmon sodium azide in ~ mL drv acetonitrile under argon was added with stirring 1.65 g (14. '5 mmon dimethyl carbamoyl chloride-d6 in 3 mL dry acetonitrile. The mixture was stirred for 24 h at room temperature in the dark or until jud~ed complete by VPC (SE-54 at 600C). The mixture was filtered and the solvent removed in vacuo. nistillation afforded a colorless oil 1.51 g (87%); bp 55-560C (8 mm Hg). Preparative VPC (Carbowax 20M, V~0°C, V~0°C injector) afforded analytically pure 13-~. ?50C) 6 2.91 (s, '3H), ?.92 (s, '3H); 2H NMR (CDCl3, IR (film) 2240, 2215, 2150, 1685, 1390, 12'30, 1115,1060,1010, 91.'5, 825, 72'5, 615 cm-1; li.~S m/e= 120 (M+), mle= 1211120/119 = 6/97/2. Methylhydrazine-formaldehyde Hydrazone (10). Methyl hydrazine- formaldehvde hydrazone (IQ) was prepared by the method of Muller.lll To a solution of 14.5 ~ 001 mmon methylhydrazine sulfate in 30 mL distilled water was added 40 mL 5 N sodium hydroxide. To this solution was added slowly with stirring 7 mL 37% aqueous formaldehyde. The mixture was stirred 1 h and then saturated with sodium hydroxide. The mixture was extracted with ether (? x 50 mU and the ether removed in vacuo. Recrystallization from acetic acid/petroleum ether afforded 2.5 g (46%) colorless needles; mp 121-123oc (lit. 121-1230C))11 NMR (CDCl3) 6 10.92 (bs, 1H), 6.33, 5.91 (Af\ quartet, 2H, J = 11.5 Hz), 2.72 (d, 3H, J = 4 -225- Hz). (Z)-3,3-0imethyl-1-phenyltriazene-1-oxide (l6).40a To a solution of 9. 36 g (90 mmon nitrosoberr~ene in 150 mL 9 5% ethanol cooled to -soc in an ice/salt bath was added dropwise with stirrin~ a solution of 1.80 g (30 mmol) l, 1-dimethylhydrazine in 40 mL 95% ethanol. After the addition was completed, the reaction mixture was stirred an additional 45 min at -50C and then slowly allowed to warm to room temperature. At approximately ooc the reaction turned from green to bright yellow. The solvent was re:noved at room temperature in vacuo. chromatographed on 150 g The remaining brown oil (5.7 silica gel eluted with CH2Cl2. g) was After aooroximately 100 ml forerun, a 300 ml fraction containing 5.4 5 g of bright yellow solid 1, '2-diphenvldiazeneoxide was collected. The next 700 mL contained ~.75 g 16 as a light yellow oil. Vacuum distillation bp. 80-810C (0.5 mm l-fz) afforded 3.'24 g (65%) 16 as a yellow oil. NMR (CT1Cl3) o 8.2-7.9 (m, ?l-f), 7.6-7.3 (m, 3l-f), 3.15 (s, 6l-f) (lit. ~MR40a fCDCl3) 0 8.'2-7.9 (m, 2l-f), 7.6-7.3 (m, 3l-f), 3.15 (s, 6l-f)); I'R.(CC14)3080,3010,'2970,2900,2850,2815, 2785,2047, 19~2, 179.5, 1750, 1590, 1470, 1420, 1390, 1295, 1250, 1220, 1165, 1135, 1110, 1063, 1025, 1015, 9?0, &25, 680, 640 cm-1; Amax= 326 nm (log E = 3.9); UV (EtCH) Amax= 227 nm (log e: = 3.8). Diisopropyl Carbamoyl Chloride (22).83 To a solution of 3.3 mL (4 5 mmoD phosgene (CA.UTION: extreme poison!) in 25 mL dry toluene at -10°C under argon was added droowise with stirring 10.0 g (99 mmol) dry diisopropyl amine in ?0 mL dry toluene. The mixture was stirred for 8 h at -1 0°C to 0°C and warmed to room temperature. After stirring an additional 4 h at room -226- temperature the mixture was cooled to ooc and the precipitated diisopropyl amine hydrochloride was filtered. The filtered solids were washed with dry toluene and the toluene removed in vacuo. The residual solid was recrystallized from dry pentane (hot) yielding 3.15 g (86%) colorless crystals; mp 570C (lit. 570C).sn N~~R fCf)Cl3, 240C) o 4.54 (sept, lH, J = 6.8l-fz), 3.5~ (seot, 1H, J = 6.~ Hz), 1.35 (d, 6H, J = 6.8 Hz), 1.22 (d, 6H, J = 6.g Hz); I~ fCCl4) 3000, ?970, ?9?.0, 2870, 1740, 1465, 1453, 1417, 1380, 1386, 1342, 1'31 5, 1?.67' 1no, 1197' 1150, t 130, 1 t 03, 1023, 900, 870, 76g, 609 cm-1. T)iisopropyl Carbamoyl Azide (17). To 5.0 g (77 mmol) sodium azide in t 5 mL dry acetonitrile was added with stirring under argon 3.30 g (20 mmol) diisoorooyl carhamoyl chloride (22) in 5 mL dry acetonitrile. The mixture was stirrerl at room temperature in the dark for 24 h or until complete by IR.. The mixture was filtered and the acetonitrile removed in vacuo. Vacuum distillation afforrled 2.~ g (82%) 17 as a colorless oil; bp. 55-560C (1.0 mm Yg). Preparative VPC (Carbowax 20M, analvtically pure material. uwoc, 1800C injector) afforded NMR. (CT)Cl3, ?30C) o 3.80 (m, ?l-i), 1.?4 (m, l2H); IR. (CC14) 2990, 2960, 29?0, 2870, ?430, 2140, 1682, 1465, 1450, 14?5, 1375, 1365, 1305, 1n1, 1220, 1?.05, 1150, 1130, 1037, 936, 882, 850, 725, 608 cm-1; UV (MeOH) Amax= 230 nm (logE =3.5); MSm/e=170(M+),m/e= 171/170 = 8/86; Anal. calcd. for C7H 14N40: C, 49.40; H, 8.29; N, 32.92. Found: C, 49.38; l-f, 8.?5; N, 32.81. Oiisopropyl Carbamoyl Azide-l5N 07-l5N). To 0.26 g (4.0 mmol) 1l5N sodium azide in ? mL dry acetonitrile was added 0.61 g (3.7 mmon diisooroovl carbamoyl chloride (22) in ?. mL dry acetonitrile. After 36 h at -227- room temperature the mixture was filtereci and the solvent removed in~· 13ulb-to-bulh distillation at hi~?;h vacuum (1.0 mm Hg) afforded 0.41 !?; (79%). Preparative VPC (Cabowax 20M, 1800C, 1800C injector) afforded analytically pure 17-15N. NMR (CT)Cl3, 230C) 6 3.80 (m, ?.H), 1.24 (m, 12H); 15N NMR (CH2Cl2 vs. CH315N02) 6 -142.30 (s), -265.70 (s); IR (CCl4) 2990, 2960, 2920, ?.870, 2430, 2140, 2117, 1682, 1465, 1450, 142.5, 1375, 1365, 1305, 1281, 1218, 1?00, 1150, 1130, 1037,936,882,850,725,608 cm-1; MS m/e= 171 (M+), m/e = 172/171/170 = 8/100/3. Diisopropyl Nitrosamine (23))6 (T)A.NGER.: Cancer suspect aRent!) To a solution of 10.1 g (O.lO mol) diisopropyl amine in 100 mL 6% hydrochloric acid was added with stirring at 70-750C a solution of 15 g (0.22 mon sodium nitrite in 80 mL water dropwise. The mixture was heated for 24 h and then cooled in an ice bath. The cooled mixture was saturated with salt and extracted with ether (3 x 50 mU. The combined ether extracts were dried over sodium sulfate and the ether removed in vacuo yielding t I.? g 23 as a yellow solid. Fractional distillation 87-880C (18 mm H~?;) (lit. 87-880C, 18 mm)26 gave a yellow oil which solidified on cooling, 9.9 g (85%). NMR (CT)Cl3) o 4.95 (sept, lH, J = 6.6 Hz), 4.19 (sept, 1H, J = 6.6 Hz), 1.47 (d, 6l-f, J = 6.6 l-Iz), 1.16 (d, 6H, J = 6.6 Hz). 1,1-Diisopropyl Hydrazine (24). t 14 To a refluxing solution of 5.0 g (43 mmon diisopropyl nitrosamine (23) in 50 mL absolute ethanol under argon was added small chunks was continued in order to maintain a molten globule for ?.0 min. nuring this time, the yellow color of the nitrosamine had faded and a white precipitate had formed. The mixture was cooled on ice and 150 mL of -228- deoxygenated distilled water was slowly added to quench the excess sodium metal and dissolve the sodium ethoxide. The resulting cloudy solution was extracted (2 x 50 mU with ether and dried over sodium sulfate under argon. The ether was removed in vacuo and the hydrazine distilled 110-1320C Oit. 130-137.0(:)26 affording 3.75 g (65%) 24 as a colorless oil. l, t ,4,4-T etraisopropyl-2-tetrazene (20). t, l ,4,4-T etraisooropyl-2- tetrazene (20) was prepared by mercuric oxide oxidation of the freshly prepared t, 1-diisopropyl hydrazine by the procedure of Effenberger. 112 NMR. (C::T)C13) 0 1.14 (d, 24H, J = 7 Hz), 3.92 (sept, 4'1-i, J = 7 Hz) (lit. NMR. 0 1.14 (d, J = 7 Y-{z), 3.92 (sept, J = 7 Hz)) ;112 UV(heptane) A max 7.91 nm (log e: = 4.05). Low Temperature Electronic Spectroscopy. The inner cell used for electronic spectroscoov in organic glasses and low temperature solutions was designed by Hinsberg.50 The cell body is constructed from OFHC copper, with stainless steel tubing soldered to it. The cell bodv is nickel plated. The cell windows are <;uprasil I (A mersil) 117 ground to size. Tl)e seals between the windows and the cell body are made with Viton 0-rings. The path length of the cell is 10.0 mm and the volume is approximately 4 mL. Two Suprasil (A mer sin outer windows on the rotating optical shroud are employed for electronic spectroscooy and photolysis. T)ilute solutions of freshly VPC ourified samples are prepared in freshly distilled dry solvents and freezepumo-thaw-degassed through three cycles. Sample solutions are introduced into the cell through 18 gauge Teflon tubing (Alpha Wire Corp.)l U~ which extend to the outside of the optical shroud through two Cajon Ultra-Torr -229- Table XXXX. Or~anic Glasses T~ (OK)~,~ Finger Dewar Ootical Cell Comments ?.-MTl-JF J'90 Yes Yes The best overall 1-'-'1ethvloentane J' ~ 'i Yes l\!o Crystallizes poor solubility 1-Prooanol J' 110 Yes >900K Fractures violently n-J:\ utyroni trile J' 100 Yes No Crystallizes Triacetin J' 210 Yes No Fractures Methyl methacrylate J' 110 Yes No Solvent Poly( methyl methacrylate) Yes Freons No Possibly as solvent mixture Crystallizes Acetonitrile J' 130 Yes No 1\~ ethvlcyclohexane J' 90 Yes Variable Crystallizes (undependable) ~Tal<en to ben~ 1012 co. QQ.ef. 107, 177. -230- fittings. Sample solutions of the carbamoyl azides in a glass forming solvent are slowly drawn into the cell under positive argon pressure with syringe suction to prevent bubble formation. As the cell is cooled to form a glass, more sample is drawn in due to contraction of the solvent (2-MTHF contracts approximately 22% going to 77 K). Stress fractures in the organic glass often occur unexpectedly during the course of an experiment destroying optical clarity. With :nest solvents other than 2-MTHF stress fractures occur at or near 77-800K, This can result in the fracturing of the inner cell windows with devastating effects. Less severe fractures can usually be remedied with annealing to 20-300 above the fracture temperature followed by recooling to a temperature just above the fracture point. Photochemical transformations were carried out with defocused light through an outer Suprasil window and monitored by electronic spectroscopy at appropriate time intervals. Electronic spectra were recorded with the aid of a black cloth to exclude light from the spectrophotometer sample compartment. Temperature control was typically .:!:.0.50K with the tip of the thermocouple just above the cell taken to be representative of the cell temperature at equilibrium. FT -IR Spectroscopy. A Mattson Instruments Sirius 100 sample compartment cover was modified to permit access of the Oisplex optical shroud. In order to maintain an efficient nitrogen purge of the spectrometer and cell compartment as clear vinyl cover was taped to the optical shroud with vinyl tape and affixed to the modified sample compartment cover by means of a detachable Velcro seal cemented around the perimeter of the access hole. A positive dry nitrogen gas pressure was produced from a -231- regulated high pressure liquid nitrogen dispenser dewar. For a typical matrix isolation FT -IR experiment, the Displex apparatus, gas handling manifold and flexible connecting line to the matrix gas cylinder regulator are checked to be He leak tight and exhaustively outgassed with frequent warming with a heat gun for 24- h prior to gas admission establishing a vacuum of better than 1 x 10-6 mbar. The carefully purified dry sample vapor is admitted through a greaseless Teflon joint and the desired gas mixture in the desired matrix host gas is prepared by standard manometric techniques. The gas mixture (typica11y 1:700-1:3000) is a11owed to equilibrate for 24- h prior to deposition. If the guest molecule is photoreactive, the glass portion of the gas handling system is wrapped with aluminum foil to exclude light. The Displex compressor is switched on and the temperature controller adjusted to the desired deposition temperature (200K for argon and 150K for nitrogen matrices). Once the inner Csl window temperature has cooled to 770K, the high vacuum valve isolating the cold head from the diffusion pump is closed to prevent back-streaming of diffusion pump oil to the optical window. The sample window is allowed to equilibrate for approximately 1 h at the deposition temperature prior to deposition. During this period the background interferrogram (empty cell) for the experiment is accumulated after the appropriate nitrogen purge time to achieve a stable baseline. To start a deposition, the manifold stopcock leading to the GranvillePhillips variable leak valve is opened together with the high vacuum isolation valve (Nupro) to the deposition port on the cold head. The deposition rate is -232- adjusted with the Granville-Phillips leak valve to the appropriate flow rate and checked periodically. deposition begun. The optical shroud is then rotated and the (For a sublimation or cryopump deposition the sample would have been equilibrated at the bath temperature by this time and the sample isolation valve would be cracked open to allow co-deposition with matrix gas just prior to rotation of the optical shroud.) Deposition progress is monitored by FT -IR at appropriate intervals. Once deposition is complete the window temperature is slowly lowered to ~ lQOK and the spectrum of the matrix isolated material recorded. If deemed necessary, annealing of the matrix can be achieved through several 10-30-lOOK temperature cycles followed by cooling to lQOK, (Note that a reproduceable alignment of the sample window is necessary for a constant baseline.) Photolysis is carried out through an outer Suprasll window through the back of the Csl inner window. The photolysis beam is defocused (incident intensities of approximately 50-100 mW per cm2) to prevent excessive heating of the sample matrix as evidenced by an increase in the inner window temperature and "boil off" of the matrix. Progress of photochemical transformations is monitored by FT -IR at appropriate time intervals. A reproduceable alignment of the sample window and infrared beam is necessary for a constant baseline. FT-IR Spectra Workup. Work-up of FT -IR absorbance spectra was achieved using Mattson Instruments FT -IR software. All spectra for a given experiment were ratioed and processed using a background interferrogram of the IR cell (or IR cell and deposited matrix gas) recorded prior to sample -233- rleposition. All absorl)ance spectra were carefully corrected using interactive suhtraction software for interfering residual atmospheric impurities not removed in the nitrogen purge ( e.g., 'f-I20 (g), C02 (g;)1. ( 'Y?O (s)1 resulting from ice formation on the Ice liquid bands nitrogen cooled 'YgCdTe (VICT) detector over the course of an experiment were also judiciously corrected. P.aseline inclination or baseline rolls were carefully corrected with appropriate baseline correction software. The narrow half- banrl widths of matrix isolated species k 7. cm-1) make these corrections straightforwarrl. T')ifference FT -I~ spectra are presented demonstrate the effects of a given transformation. (a.) to clearly ~ubtraction factors refer to the formula A. - a. B = C, where A. and B are the absorbance spectra to be subtracted and r. is the resulting difference spectrum. H2N'\J Product Analysis. Carbamoyl azide ftypically 1. 50 mg) was placed in a 5 mm on auartz tube fitted with a vacuum stopcock. To this was transferred 300 11 L ?-"v1THF (or appropriate dry solvent) freshly distilled from sodium/benzophenone ketyl. The solution was freeze-pump-thaw-degassed for three cycles. The solution was thoroughly mixed and then immersed in a ~uprasil finger dewar filled with liquid nitrogen. Once the solution had glasserl it was subjected to photolysis for the requisite time period (typically 5 h photolvsis with 2-CS-7 -54 filters under carefully reproduced irradiation conditions resulted in 2. 99 % decomposition of starting material as determined indepenrlently by I"R.). For thermal product analysis the resulting blue-violet glass of 3 was then allowed to warm slowly to room temperature. Upon removal from the ??OK rlewar the glass quickly turned opaque yellow-orange -234-- as it softened (characteristic of 1 Amax= 386 nm and the Amax= 4-80 nm species) followed by vi~orous gas evolution. The solution quickly became colorless and in the -50 to ooc temperature ran~e there was again vigorous gas evolution together with formation of an insoluble white + precipitate (N'f-f4-N~) from the colorless solution. Thermal and photochemical products were determined as follows: The gaseous products 1.0, N2 and H2 were identified by VPC coinjection witl--t authentic samples (sieve 5 A column 250C, He carrier, thermal conductivity rletector, injected in l atm C2H~) by mass spectral analysis or quantitated by Toepler pump analysis. For Toepler pump analysis, the gas mixture was oumped from the sample tube and condensable products trapped at 7701(. The total ~as PV was rletermined and the mixture was then passed throuf!;h a hot tube of CuO (31 0°C) to oxidize CO to C02 and H?. to H20 which were trapped at -78° and the N2 content determined by PV difference followed by C02 PV determination. Ammonia content was determined for l)oth the volatile (NH3) and non+ volatile (NH4- N3) components hy the indophenol method.ll5 The volatile components fo the reaction were vacuum transferred (770K) to 10 mL saturated boric acid solution (degassed) and allowed to warm to room temperature. This solution was tansferred to a 50 mL volumetric flask. 15 mL saturated boric acid, 5 mL chlorine (Cl,) saturated distillated water and 5 mL 8°-h Phenol solutions were added with swirling. The mixture was heated on a steam bath for 3.0 min and cooled on ice. After cooling to 5°C, 5.0 mL of a '3.0 M sodium hvdroxide solution was added and the mixture diluted to -235- volume. After 5 min the intense blue color of the indophenol was fully developed. The absorbance at 625 nm was measured relative to a reagent blank containing the appropriate volume of '--"'THF. The ammonia content was determined from a calibration curve preoared using ammonium chloride as a standard. The non-volatile ammonium azide residue in the sample tube was dissolved in l.5 mt distilled water and determined as above. Hydrazine content was cietermined independent of a volatile r..JH 3 determination as the his azine of para-dimethylamino benzaldehyde.Jl5 The volatile components of the reaction were vacuum transferred (770K) to a mixture of 5.5 mt 0.1 N H2S04 and 0.5 mt of a solution para-dimethyl amino benzaldehyde (40 mg/mt in 2.0 N H2S04) (degassed). Upon thawing, the orange color of the bis-azine appeared and was fully developed in 15 min at room temperature. The absorbance at 4 58 nm was measured relative to a reagent blank contining the appropriate volume of 2-MTHF. The hydrazine content was determined from a calibration curve prepared using hydrazine sulfate as a standard (carbamoyl azide does not interfere). Formaldehyde was determined by the Hantzsch reaction.Jl3 The reaction mixture was added to 4 mL of reagent (15.0 g ammonium acetate, ~00 11L acetic acid, 200 l.1 L freshly purified acetyl acetone in 100 mL distilled water). The mixture was heated in a sealed tube to 370C for 1 h. a~Jsorbance The at 412 nm was measured relative to a reagent blank containing the aporooriate volume of 2-MTHF. The formaldehyde content was determined using a molar extinction coefficient of 8000 at 412 nm for the oroduct diacetyl dihydroleutidine (detection limit 2 X 1o-8 g H2CO). -236- Photochemical products were determined in a similar fashion with visible light irradiation with a CS-1-7 5 filter (> 34-0 nm) or the combination CS-1-75 and two CS-3-70 filters (> 500 nm) until the blue-violet color of the 1, 1-diazene was gone (typically 4--5 h) prior to warming the glass for product analysis. Unfiltered photolysis (A. > 250 nm) was also carried out for the appropriate time period (typically 5 h) with no color development. Matrix Isolation of H2NN (3), H2NN15 (J-15N), D2NN (3-d 2 ), D2NN15 (3-d2_15N). The respective carbamoyl azides 15 (freshly sublimed) were cryopumped from a -10 or -2ooc ice/salt bath and co-deposited with matrix gas (UHP argon at 200K, UHP nitrogen at 150K) at 5-10 mmol gas per hour over approximately 6-4- h, respectively. The deposition progress was closely monitored by FT -JR. These deposition conditions gave consistently transparent matrices with well resolved bands and minimal atmospheric impurities (C02, H20). The window temperature was slowly lowered to 10°K (1 0 min) and the infrared spectrum of the starting carbamoyl azide recorded after the appropriate nitrogen purge time (30-4-5 min typically). The matirx was photolyzed with defocused light from a 1000 watt xenon source through an outer Suprasil window and through the back of the inner Csl window using 2-CS-7-54- UV transmitting, visible (4-10-680 nm) cut out filters. Typical photolysis times for 80% conversion of starting carbamoyl azides were 90-120 min. Photolysis of the resulting 1,1-diazenes was carried out with CS-1-75, 2-CS-3-70 filters (> 500 nm) for 120-180 min carefully monitored by FT -JR. Photolysis of trans-1,2-diazenes was carried out with either monochromatic light (380 2:. 10 nm) from a monochromator or with CS-5-58 and CS-7-51 filters (360-41 0 nm). Matrix Isolation of 1,1-0imethyldiazene (7) and 1,1-Dimethyldiazene15N (7-15N). A. 1:1400 gas mixture of the respective dimethyl carbamoyl azicie (freshly purified bv preparative VPC) in argon was allowed to eauilibrate in the foil wraooed gas manifold of the T)isplex apparatus. The mixture was deoosited (?OOK) at 10 mmol/h over approximately 4 h (100 mm gas mixture, 37 mmol). The Csl window temperature was lowered to 1OOK and the sample soectrum recorded. Photolysis was carried out throu~h an outer <;uorasil window through the back of the inner Csl window with defocused light through 2-CS-7-54 UV transmitting visible cut out filters for 60 to 90 min. Photolvsis progress was monitored periodically by FT -IR. Photolysis of the product diazenes was carried out in a similar fashion through CS-1-75, CS-4-96 and C<;-3-70 filters (466-610 nm) for approximately 61) min or C5-1-75 and CS-~-61 (> 590 nm). Photolysis of product (TJ) from photolysis t,l-dimethyldiazene 7 was carried out with CS-5-58 and CS-7-51 filters (360-4 10 nm) for aoproxi mate! y 60 min. Matrix Isolation of 1,1-niisopropyldiazene (18) and 1,1-0lisopropyldiazene-l'>N 0&-15N). The respective carbamoyl azides (freshly purified by oreparative VPC) were cryopumped from a - '300C ice/salt bath and codeoosited with ar~on matrix gas (UYP argon) (?QOK) at l 5 mmol gas per hour over aoproximately 4-5 h (150-175 mm gas, 55-65 mmol, respectively). The CSJ window was slowly lowered to lOOK anci sample spectrum recorded. Photolysis through the back of the Csl inner window with 2-CS-7-54 filters was carried out for 90-150 min anci carefully monitored by FT -IR. Photolysis of the product diazene was effected with defocused visible light from C:orning filters CS-1-75 and CS-4-96 (360-610 nm) for 4-6 h and carefully monitored bv FT -I'R. Kinetics. ?-~THF glasses of the respective diazenes were prepared in the low temperature electronic spectroscopy cell as described above. cell temperature was slowly The raised to effect diazene decomposition. T)ecompositions were monitored using either the repetitive scan program of the C:ary 219 spectrophotometer (multiple absorptions) or using a timed delay single wavelength monitor. All data prior to temperature equilibration (J'30 min) were discarded. Typically the reactions were monitored through several half-lives and reported rate constants are derived from a conventional linear least squares analysis. First-order decays were checked at two starting concentrations through a minimum of three half-lives and in all cases gave identical rate constants within experimental error. -239- REFE'R.E"'CES AND NOTES (1) (a) ~aird, N. C.; 13aird, N. C.; <;wenson, J. 'R.. Can. J. Chern. 1973, 2.,!_, 3097. (b) Ibid. 1973, 21., 3301. Rarr, R. F. Wettermark, G. (c) ~1erenyi, G.; Chern. Phys. 1973, 1, 340. (d) Wagniere, G. Theor. Chim. Acta 1973, 31, 269. (e) Lathan, W. A.; Curtiss, L.A.; Hehre, W. J.; Lisle, J. B.; Poole, 1. A. Prog. Phys. Org. Chern. 1974, l.!_, 1975. (f) Winter, N. W.; Pitzer, 'R.. M. J. Chern. Phys. 1975, 62, 1249. (g) Vasudevan, '<.; Peyerimhoff, S. n.; Buenker, 'R.. J.; Kammer, W. F. ChE-rn. Phys. 1975, ?_, 137. fh) Ahlrichs, 'R..; Staemmler, V. Chem. Phvs. Lett. 1976, 'F, 77. (i) liowell, 1. M.; J<irschenbaum, L. J. J. Am. Chem. Soc. 1976, 98, 877. (j) Baird, N.C.; Wernette, n. A. Can. J. Chern. 1977, ~' 350. (k) Skanke, P. N. Chern. Phys. Lett. 1977, 47, ?59. (1) F\aird, N. C.; Kathpal, l-1. B. Can. J. Chern. 1977, 55, 863. (m) Cimiraglia, 'R..; 'R.iera, J. M.; Tomasi, 1. Theor. Chim. Acta 1977, 46, 2?3. (n) navis, 1. H.; Goddard, III, W. A. J. Am. Chern. Soc. 1977, 99,7111. (o) 13igot, R.; Sevin, A.; nevaquet, A. Ibid.1978, 100,2639. (p) Pople, J. A.; Krishnan, 'R..; Schlegel, l-1. 13.; Binklely, J. S. Int. J. Ouantum. Chern. 1978, ll, 545. (q) Parsons, C. A.; T)ykstra, C. E. J. Chern. Phvs. 1979, 21., 30?5. (r) Pasto, 0. 1.; Chipman, 0. M. J. Am. Chern. Soc. 979, 101, 2?90. (s) Pasto, 0. J. Ibid. 1979, 101, 6852. (t) Casewit, C.; Goddard, W. A... III. Ibi<i. 19~0, 102, 4057. (u) Kemper, M. J. H.; Ruck, H. M.• Can. J. Chern. 1981, 59, 3044. (?) (a) Foner, S. N.; Corey, 'E. J.; Hudson, 'R.. L. Mock, W. L. J. Chern. Phys. 1958, 28, 719. (b) J. Am. Chern. Soc. 1962, 84, 685. (c) -240- 'R.osengren, 1<.; Wiber~, ~.; Pimentel, G. L. J. Chem. Phys. 1965, 43, .507. (d) Rachhuber, 'fi.; Fischer, G. Angew. Chem. Int. Ed. 1972, l.!, 829. (e) Willis, C.; P,ack, R. A. Can. J. Chern. 1973, 2..!_, 3605. (f) f\ondybey, V. E.; Nibler, J. W. J. Chern. Phys. 1973, 58, 2125. (g) Wiberg, N.; Fischer, G.; Bachhuber, H. Chern. Ser. 1974, 107, 1456. (h) 1=\ack, R. A.; Willis, C.; Ramsay, D. A. Can. J. Chern. 1974, 52, 1006. (i) Carlotti, M.; Johns, J. W. C.; Trombetti, A. Can. J. Phvs. 1974, 5?., 341). (j) Wiberg, N.; Fischer, G.; f\achhuber, H. Angew. Chern. Inti. ~d. 1976, !2, 385. (k) Frost, T). C.; Lee, s. T.; McT)owell, C. A.; Westwood, N. P. C. J. Chern. Phys. 1976, 64, 4719. (1) Wiberg, N.; Fischer, G.; f\achhuher, H. Angew. Chern. Inti. J:d. 1977, ~' 780. (m) Foner, s. N.; 'fiudson, 'R.. L. J. Chern. Phys. 1978, 68, 3162. Wiberg, N.; Fischer, G.; f\achhuber, H. (n) 7. Naturforsch. 1979, 34b, 1~85. (o) Neudorfl, P. S.; Rack, R. A.; T)ouglas, A. E. Can. J. Chern. 1981, 59, 506. (3) For a review of diimide chemistry: 'fiunig, S.; Muller, 'fi. R.; Thier, W. Angew. Chern. Inti. F.d. 1965, !!_, 271. (4) Sylwester, A. P.; T)ervan, 'P. l3. J. Am. Chern. Soc. 1984, 106, 4648. (5) For reviews see: (a) Henderson, R. A.; Leigh, G. J.; Picket, C. J. Adv. in Inon~. Chern. ~ Rad. 1983, 27, 197-292 and references therein. (b) Chatt, L. M.; Pina, C.; 'R.ichards, R. L., Eds. "New Trends in the C":hemistry of Nitrogen Fixation"; Academic Press: London, 1980 and references therein. (6) For reviews of 1,1-diazenes, see: (a) Lama!, T). M. In "Nitrenes"; -24-1- Lwowski, W., Ed.; Interscience: New York, 1970, Ch. 10. (b) Ioffe, B. V.; Kuznetsov, M.A. Russ. Chern. Rev. (Engl. Trans.) 1972, !L!_, 131. (c) Lwowski, W. In "Reactive Intermediates"; Jones, M.; Moss, R., Eds.; Wiley: New York, 1978, Ch. 6. (d) Lwowski, W. In "Reactive Intermediates"; Jones, M.; Moss, R., Eds.; Wiley, New York, 1981, Vol. 2. (e) L wowski, W. In "Reactive Intermediates"; Jones, M.; Moss, R., Eds.; Wiley: New York, 1985, Ch. 8. (7) For a review of 1,2-diazene chemistry see: Engel, P. S. Chern. Rev. 1980, 80, 99. (8) (a) Hinsberg, W. D.; Dervan, P. B. J. Am. Chern. Soc. 1978, 100, 1608. (b) Hinsberg, W. D.; Dervan, P. B. Ibid. 1979, 101, 614-2. (c) Schultz, P. G.; Dervan, P. B. Squillacote, M.; Ibid. 1980, 102, 878. Lahti, P.; Sylwester, A. P.; (d) Dervan, P. B.; Roberts, J. D. Ibid. 1981, 103, 1120. (e) Hinsberg, W. D.; Schultz, P. G.; Dervan, P. B. Ibid. 1982, 104-, 766. (f) Mcintyre, D. K.; Dervan, P. B. Ibid. 1982, 104-, 64-66. (g) Schultz, P. B.; Dervan, P. B. Ibid. 1982, 6660. (9) For reviews of low temperature matrix isolation see: (a) Cradock, S.; Hinchcliffe, A. J. "Matrix Isolation"; Cambridge University Press, Cambridge, 1975. (b) Dunkin, I. R. "Matrix Isolation Technique and Its Application to Organic Chemistry." Chern. Soc. Rev. 1980, 1. (c) Downs, A. J.; Peake, S. C. In "Molecular Spectroscopy"; Barrow, R. F.; Long, D. A.; Millen, D. J., Eds.; The Chemical Society: London, 1973, Vol. 1, 523. (d) Chadwick, D. M. Ibid., 1975, Vol. 3, 281. (e) Barnes, A. J.; Hallam, H. E. Quart. Rev. 1969, 23, 392. (f) Turner, J. -?'+2- J. Angew. Chern. lntl. T:.d. 1975, .!..:t_, 30'+. (g) Moskovitz, 1\~.; Ozin, G. A., l:.ds.; "Cryochemistry"; Wiley-Interscience: New York, 1976. (h) Chapman, 0. L. Pure & Appl Chern. 1974, '+0, 511. (i) 13arnes, A. J .; Orville-Thomas, W. J.; Muller, A..; Ganfres, R., T:.ds. "Matrix Isolation Spectroscooy"; D. Reirlel: Dodrecht, Yelland, 1981. (10) '5walen, 1. T).; Ibers, J. A.. J. Chern. Phvs. 1962, 36, 91'+. ()I) "JANAF Thermochemical Tables"; NSRDS-1\JBCOi-37 U.S. Government Printing Office: Washington, n.C., 1970. (1?.) (a) Willis, C.; F\A.ck, R. A.; Purdon, J. G. Int. J. Chern. Kinet. 1977, ~' 79,7. (b) Willis, C.; Back, R. A.; Parsons, J. M.; Purdon, 1. G. J. Am. Chern. Soc. 1977, 99, 4451. (13) fa) Benson, S. W. "Thermochemical Kinetics"; ?nd ed.; York, 1976. (h) 'Benson, S.W.; Phase Unimolecular Reactions"; 0'1\!eal, l-1. E. Wiley: New "Kinetic nata on Gas l\1SRl1S-N~S 1970, No. 21, 448. (lt.t) Benson, S. W.; O'Neal, l-1. E. NSRT)S-1\!P,S 1979, No. 21, 31. 05) For example: Baumgarten, l-1. T:..; Whittman, W. F.; Lehmann, G. J. J. Heteroc. Chern. 1969, ~' 333. (16) Nakagawa, K.; Konaka, R.; Nakata, T. J. Org. Chern. 1962, 27, 1597. (17) For example: Overberger, C. G.; Lombardino, J. G.; J-liskey, R. G. ~ Or g. Chern. 19 57, ?2, 8 58. (18) For example: (a) Caroino, L. A. J. Am. Chern. Soc. 1957, 79, 4427. (b) Lernal, n. M.; Rave, T. W.; McGregor, S. 0. Ibid. 1964, 86, 2395. (c) nervan, P. 13.; (lq) For example: Uyehara, T. Ibid. 1979, 101, 1076. 13umgardner, C. L.; Freeman, J.P. Tetrahedron Lett. -243- 1966, 5547. (20) (a) Anderson, D. J.; Gilchrist, T. L.; Horwell, l). C.; 'R.ees, C. W.; Yelland, M. Chern. Commun. 1969, 146. (b) Rees, C. W.; Yelland, M. Ibid. 1969, 377. (?1) Rusch, ~.; Weiss, B. Chern ~er. 1900, 33, ?701. (2?.) (a) For a review of one bond vs. simultaneous two bond scission in 1,2diazenes, see 'R.ef. 7. (b) nannenberg, J. 1.; Rocklin, n. J. J. Org. Chern. 1982, 47, 45?.9. (23) Freeman, 1. P.; Graham, W. H. 1. Am Chern. Soc. 1967, 89, 1761. f24) nuan, n. C:.; Oervan, 'P. f\. J. Org. Chern. 1983, 48, 970. (25) Lichter, 'R. L. Ph.T). Thesis, University of Wisconsin, ""aciison, 1967, as quoted in 'Q_ef. 6. (26) Lema!, T). M.; Men~er, F.; Coats, t:. A.. J. Am. Chern. Soc. 1964, 86, 2395. (:~7) Adams, n. J. C.; ~radbury, S.; Howell, 'D. C.; Keating, M.; 'R.ees, C. W.; Storr, 'R.. C. Chern. Commun. 1971, 828. (28) Baldwin, J. t:.; Brown, J. E.; Hofle, G. J. A.m. Chern. Soc. 1971, 93, 788. (?9) fa) Mcf\ride, W.; 'Kruse, '1-f. w. J. Am. C:hem. Soc. 1957, 79, 572. (b) McRride, w.; Bens, E. Ibid. 1959, B, 5546. (30) Anderson, n. J.; Gilchrist, T. L.; 'R.ees, C. W. Chern. Commun. 1971, 800. (31) (a) Atkinson, 'R.. S.; Atkinson, 'R.. S.; Rees, C. W. Rees, C. W. Chern. Commun. 1967, 1230. (b) J. Chern. Soc. fC) 1969, 772. (c) -244- Atkinson, R. S.; Malpass, J. 'R..; Skinner, K. L.; Woodthorpe, K. L. Chern. Commun. 1981, 54q. (32) Willis, C.; Back, 'R.. A.; Parsons, 1. \~. 1. Photochem. 1977, ~' 253. (33) Craig, "J. C.; Kliewer, \~. A.; ~hih, N. C. J • .A.m. Chern. Soc. 1979, l..Ql, 2480. (34) Willis, C.; ~ack, 'R. • .A..; Purdon, J. G. Int. J. Chern. Kinet. 1977, 2_, 787. (35) (a) Wiberg, N.; ~ayer, H.; ~achhul)er, l-f. Angew. Chern. Inti. "F:d., F.ngl. 1975, .J.!, 177. (b) Kroner, J.; Wiberg, N.; 13ayer, H. Ibid. 1975, ~' l78. (36) Willis, C.; Back, R. A.; Parsons, 1. M.; Purdon, J. G. J • .A.m. Chern. Soc. 1977, qq, 4451. (37) Wiberg, N.; Haring, l-i. W.; Vasisht, S. K. Z. Naturforsch 1979, 34b, 356. (38) No explanation for the extrernely low pre-exponential factors has been offered. From the experimental activation parameters kH/ko = 1.97 for gas phase decomposition of 1 at toooc. (39) Bock, v. H.; Kompa, K. L. !. fur .A.nerg. und Allge. Chern. 1964, '332, 23~. (40) (a) Hoesch, v. L.; Koppell, 13. Helv. Chim • .A.cta 1981, 64, 864. (b) l-ioesch, v. L. Ibid. 1981, 64, 890. ( 41) (a) Felix, n.; Eschenmoser, A. Muller, 'R.. K.; 1-Iorn, V.; Soos, 'R..; Schreiber, J .; Helv. Chim. Acta 1972, 55, 1276. (b) Annunziata, 'R..; Fornasier, 'R..; Montanari, F. J. Org. Chern. 1974, 39, 3195. -245- (42) Kawasaki, M.; Iwasaki, M.; Ibuki, T.; Takezaki, Y. J. Chern. Phys. 1973, 59, 6321. Kawasaki, M.; Iwasaki, M.; Tanaka, I. Ibid. 1973, 59, 6328. (43) Jacox, M. E.; Milligan, D. E. J. Am. Chern. Soc. 1963, 85, 278. (44) (a) Poppinger, D.; Radom, L. J. Am. Chern. Soc. 1978, 100, 3674. (b) Rauk, A.; Alewood, P. F. Can. J. Chern. 1977, 55, 1498. (45) (a) Lieber, E.; Minnis, R. L.; Rao, C. N. Chern. Rev. 1965, 65, 377. (b) Reichen, W. Ibid. 1978, 78, 569. (c) Lockley, W. J. S.; Ramakrishnan, V. T.; Lwowski, W. Tetrahedron Lett. 1974 30, 2621. (46) Lwowski, W.; deMauriac, R. A.; Murray, R. A.; Lieselotle, L. Tetrahedron Lett. 1971, .2_, 425. (47) Lwowski, W.; deMauriac, R. A.; Thompson, M.; Wilde, R. E. J. Org. Chern. 1975, 40, 2608. (48) Kuck, V. J.; Wasserman, E.; Yager, W. A. J. Phys. Chern. 1972, 76, 3570. (49) (a) Thiele, J.; Stange, 0. J. Liebigs. Ann. Chern. 1894, 283, 1. (b) Curtius, T.; Heidenreich, K. J. Prakt. Chern. 1895, 160, 454. (50) The low temperature cell was designed by Hinsberg. Hinsberg, W. D. Ph.D. Thesis, California Institute of Technology, Pasadena, CA, 1980. (51) Quartz optics and 20 em water (infrared) filter. (52) (a) We presume this band to be the 0,0 absorption band in the absence of an emission spectrum. (b) Calculated using ratio of force constants from Ref. 1n and the experimental v 14N= 14N of 1574.16 cm-1 for3. - '246- (53) For a discussion on the origins of solvent-induced shifts, see: l-luherfield, P.; Lux, M.S.; Rosen, D. J. Chern. Soc. 1977, 99, 6828. (54) (a) Robinson, G. W.; Frosch, R. P. J. Chern. Phys. 1963, 38, 1187. Ihid. 1962, 37, 196'2. (b) Dexter, D. L.; Fowlere, W. B. Ibid. 1967, 47, 1379. (55) Siebrand, W. J. Chern. Phys. 1967, 47, 2411. (56) fa) Ipoen, f:.P.; Shank, C. V.; Woerner, R. L. Chern. Phys. Lett. 1977, 46, '20. (b) Heritage, J.P.; Penzkofer, A. Ibid. 1976, 44, 76. (57) Miller, 'R.. C.; Lee, '1:. K. C. Chern. Phys. Lett. 1976, !t_l, 52. (5~) Watson, 1. S. J. Chern. Soc. 1956, 3677. f59) Matrix ratios were estimated by comparison of the asymmetric azide stretches rv 1\13) with a known matrix ratio and known deposition size for gas ohase mixture deposited dimethylcarbamoyl azide 13. (60) Photolysis of 15 was generally halted at "'80% conversion. Higher ' conversions gave variable low yields of 3 apparently due to secondary photolysis. (61) Mattson Instruments FT -IR interactive subtraction software. (62) (a) Thomas, S. G.; Guillory, W. A. J. Phys. Chern. 1973, 77, 2469. (b) Khoshkhoo, f-l.; Nixon, E. R. Spect. Chim. Acta 1973, 29A, 603. (63) Prochemical Limited, New Jersey; (64) Casewit, C. 97 atom % ( 1-15N-NaN31. Ph.f). Thesis, California Institute of Technology, Pasadena, CA, 19~'2. (65) In our hands, argon matrix isolation of 3 gave consistently superior and better reoroducible matrices than repeated attempts at N2 matrix -247- isolation of 3 and its products (1:1000 to 1:3000, N2, 10-200K). FT-IR bands for 3 in N2 matrix (1:2000, lOOK) are found at 2880.98, 2815.89, 1586.45, and 1018.50 cm-1. (66) Sodeau, J. R.; Lee, E. K. C. Chern. Phys. Lett. 1978, 57, 71. (67) Robin, M. B.; Hart, R. R.; Kuebler, N. A. J. Am. Chern. Soc. 1967, 89, 1564. (68) Craig, N. C.; Kliewer, M. A.; Shih, N. C. J. Am. Chern. Soc. 1979, 101, 2480. (69) These results were sporadic and not readily reproduced suggesting matrix site effects and/ or intensity dependent photolysis. (70) House, H. 0. "Modern Synthetic Reactions"; W.A. Benjamin, Inc.: Menlo Park, CA, 1972, p. 254. (71) Dubost, H.; Abouaf-Marguin, L. Chern. Phys. Lett. 1972, Q, 269. (72) Turro, N. J. "Modern Molecular Photochemistry"; Benjamin/Cumings, Inc.: Menlo Park, CA, 1978, Ch. 10. (73) Andrews, L.; Arlinghaus, R. T.; Johnson, G. L. J. Chern. Phys. 1983, 78, 6347. (74) (a) Huo, W. M. J. Chern. Phys. 1965, 43, 624. (b) Rosen, B. "Spectroscopic Data Relative to Diatomic Molecules"; Pergamon: New York, NY, 1970. (75) McLean, A. D.; Yoshimine, M. J. Chern. Phys. 1969, 44, 3467. (76) Species such as H·~ HCO·~ NH2·~ H2NCO·~ can be ruled out by comparison with their literature ESR signals. Cochran, E. L.; Bowers, V. A. (a) Adrian, F. J.; J. Chern. Phys. 1962, 36, 1661. (b) -248- Foner, S. N.; C:ochran, l:. L.; Bowers, V. A.; Jen, C. K. Phys. Rev. Lett. 19.58, J., 91. (c) Bosco, S. R.; Cirillo, A.; Timmons, R. B. J. Am. Chern. Soc. 1969, .2.!_, 3140. (77) Smith, T). 'R..; Pieroni, J. J. Can. J. Chern. 1965, 43, 876. (78) Phenyl diazenyl radical has been observed by ESR. Suehiro, T .; Tashiro, T .; Nakausa, 'R.. Chem. Lett. 1980, 1339. (79) No new infrared band in the region 2300 to 1600 cm-1 showing a 15N multiple was observed. An unidentified band at 1218.83 cm-1 grows in with VIS ohotolysis of 3. (~0) Ling, A. C.; Willard, J. E. J. Phys. Chern. 1968, 72, 1918. (81) Senthilnathan, V. P.; Platz, M.S. J. A.m. Chern. Soc. 1980, 102, 7637 and references therein. (82) The viscosity of 2-MTHF n is estimated to be greater than 3 x 1020 poise at 77. 50K and approximately 5 x 1011 poise at 900K) 07 The viscosity decreases approximately an order of magnitude per 0.50K greater than 920K. Assuming Fick's law applies, the diffusional rate constant (kl)) can be estimated (8'3) (a) Hasserodt, TJ. Chern. F'ler. 1968, 101, 113. (b) Gold-Aubert, P.; Gysin, E. Yelv. Chim. Acta 1961, 44, 105. (84) Hanson, J. t:.. Unpublished results. (85) Rao, C. "1. 'R..; Goldman, G. K.; Bala-Subramanian, A. Can. J. Chern. -249- 1960, 3~, ~508. (~6) (a) White, D. 'K.; Greene, P. n. J. Org. Chern. 1978, 43, 4530. (b) Stowell, J. C. J. Org. Chern. 1967, 3?., 2360. (~7) Lambert, 1. P.. "Organic Structural Analysis"; Macmillan, Inc.: New York, NY, 1976. (~8) Preliminary experiments suggest that the major bands of 7 (DM) are more efficiently photolyzed with long wavelength visible light (> 600 nm) than tl)e minor bands (nm). This would favor assignment of the bands (T)m) to another species. (89) Ackerman, M. N.; Craig, "J. C.; Isberg, R. 'R..; Lauter, T). M.; Tacy, '1:. P. J. Phys. Chern. 1979, 83, 1190. (90) ll..~illigan, n. E.; Jacox, M. E. J. Chern. Phys. 1967, 47, 5146. Adrlitional weak bands observerl for (CH3•) at 328~ and 736 cm-1 not ot,serverl here. (q 1) ll.~u1tiole matrix CO stretches resulting from photo-decarbony1ation of ketenes have been noted by others. 'Reference 9?. (92) Sheridan, 'R.. Private communication. (93) This unpublisl)ed result has been communicated by others. Reference 9?. (94) (a) Goddard, J.D.; Schaefer, III, H. F. J. Chern. Phys.1979, 70,5117. (b) Gelbart, W. M.; Elert, M. L.; Heller, 0. F. Chern. 'R.ev. 1980, 80, 403. (c) Schaefer, HI, H. F. .A.ccts. Chern. 'R.es. 1979, J1, 288. P>uck, H. ~. 'R.ec. 'R.ev. 1982, 2, 193. (e) l3uck, H. M. Ibid. 1982, 223. (d) z., -250- (9 5) For recent evidence for sequential photochemical bond cleavage in 1,2-diazenes see: Holt, P. L.; "AcCurdy, K. E.; Adams, J. S.; 13urton, K. A.; Weisman, R. ~.; Engel, P. S. J. Am. Chern. Soc. 1985, 107, 2180. (96) Singh,~. 1. Am. Chern. Soc. 1969, 21_, 3670. (97) Padwa, A. "1,3-Dipo1ar Cycloaddition Chemistry"; Wiley- Interscience: New York, NY, 1n4, Vols. 1 and 2. (98) Farina, P.R.; Tieckelmann, 'fi. J. Org. Chern. 1973, 38, 4259. (99) Mc13ride, W. R.; Kruse, 'fi. W. J. Am. Chern. Soc. 1957, 79, 572. 000) Kuznetsov, M. A. Russ. Chern. Rev. 1979, 48, 563. flO!) (a) T)oba, T.; 103, 339. Ingold, K. U.; (b) Doba, T.; Siebrand, W. Ingold, K. U.; Chern. Phys. Lett. 1984, Siebrand, W.; Wildman, T. A. J. Phys. Chern. 1984, 88, 3165. (l 0?.) UV (VIS filtered) photolysis of more concentrated glasses gave proportionately greater absorbances of the red-orange Amax = 464 nm and A max =474 nm species. Glass temperatures greater than 85°K gave similar results. (1 03) For a review of monosubstituted J ,2-r!iazenes see: Kosower, E. M. Accts. Chern. Res. 1971, !!_, 193. 004) Fogel, L. D.; Steel, C. J. Am. Chern. Soc. 1976, 98, 4859. (105) Whitewav, S. G.; Masson, C. R. J. Am. Chern. Soc. 1955, 77, 1508. 006) (a) Leroi, G. F.•; ~wing, G. T:.; Pimentel, G. L. J. Chern. Phvs. 1964, 40, 2298. (b) ~entsen, J. G.; Wrighton, M. S. J. Am. Chern. Soc. 1984, I 06, 4041. -251- 007) Fischer, G.; Fischer, t.. Mol. 'Photochem. 1977, ~' 279. f108) Soecies such as H2~·~, HN•,2 ·N3~ l-l-N=C=O~, ·N=C=0,.2 HO-C=N~ HOO·,_g etc., can be ruled out by comparison with literature matrix infrared data. fa) Milligan, n. 'E.; Jacox, M. E. J. Chern. Phys. 1965, ll- ~, lj.lj.5P; (b) McCarty, M.; Robinson, G. W. J. A.m. Chern. Soc. 1959, ~' l!-l!-72. (c) Reference '-c.; (d) Milligan, n. E.; Chern. Phvs. 1967, l!-7, 5157. (e) Jacox, M. E.; Jacox, M. E. Milligan, D. E. J. Ibid. 1964, l!-0, 2ll-57. (f) Jacox, M. E.; Milligan, 0. E. J. Mol. Spect. 1972, ll-2, l!-95. (109) (a) 'R.enaud, R.; Leitch, L. C. Ackerman, M. N.; Can. 1. Chern. 1954, 32, 5ll-5. Oobmeyer, 0. J.; Hurdy, L. C. (b) J. Organomet. Chern. 1979, U~2, .561. 010) KOR Isotopes, Cambridge, MA.; >99% d6 hexadeuteriodimethyl- ammonium chloride. (Ill) Muller, t..; Runde!, w. Chern. ~er. 1957, 90, 1299. (lt2) Effenberger, F.; Fischer, P. Tetrahedron 1970, 26, 3029. 013) Nash, T. (1 ll!-) Mcintyre, ~iochem. 1953, 55, l!-16. n. K. Ph.O. Thesis, California Institute of Technology, Pasadena, CA., 19~ '3. (115) Streuli, C. A. "The Analytical Chemistry of Nitrogen and its Compounds"; Wiley-Interscience, New York, NY, 1970, Pt. 1. ( 116) Air Products and Chemicals, Inc., Advanced Products DepartmentCryogenics, P, 0. Box 2802, Allentown, PA 181 0_5. (ll7) Amersil, Inc., 685 R-amsey Ave., '1-iillside, NJ 07205. -252- ( 118) Alpha Wire Corp., 711 Lidgerwood Ave., Y:lizabeth, NJ 07207. ft19) Edwards l-Hgh Vacuum, Manor Royal, Crawley, West Sussex, RH 102LW, UK. (120) Oriel Corp., 15 Market St., 45tanford, CT 06902. (121) Pennwalt Corp., Wallace ~ Tiernan Div., 25 Main St., 'F3elleville, NJ 07109. 022) Granville-Phillips Co., 5675 E. Arapahoe Ave., Boulder, CO 80803. 023) Veeco Instruments, Inc., Terminal Dr., Plainview, NY 11803. 024) International Crystal Labs, 107 Trumball St., P. 0. Box 228, Elizabeth, NJ 07206. (125) Arcadia Valve and Fitting Co., 521 N. First Ave., Arcadia, CA 91106 (126) Ace Glass, Inc., Vineland, NJ 08360. 027) Angell, C. A.; (128) Frierson, W. J. Inorg. 45yn. 1946,1, I%. 029) Hinze, J.; Curl, Jr., R. F. 1. Am. Chern. Soc. 1964, 86, .5068. ~are, J. M.; Sare, E. J. J. Phys. Chern. 1978, 82, 24. -253- APPENDIX Kinetic Derivations n = 1,1-Diazene = Tautomer (azomethinimine) p = Product (tetrazene) H = Hydrocarbons T Mechanism fA) T) T 21') p Assuming steady state in ( f) 1 - d( nl dt k _ ( T) 1 = ( T)) = ( kl - d( T) dt = k -1 ( T1 - = k -l ( T) (I - (Di(k 1 + 2 k (n1) 2 k I ( T) ) +_2,k (D) 2 k 1 k -l ( T} k1 + 2 k 2 (n) kl k1 + 2 k2[ 1)1 ) -254- = = -d( Tl dt Mechanism (B) T p Assumins; stea<1y state in ( n) d( nl dt = k -l ( T) k 1( 0) k (nl 1 = k -l ( T) - k 2( D1( T) ( T)j = k _ ( T) 1 k + k 1_ Tl 1 2 -d( Tl dt = k -l ( Tl - k - k ( o)( T) 2 ( 1 k-I ( T) k 1 + kiT) ) =0 + ki Tl ( kpl k 1 + kiT) ) -255- -cf(T~ dt = k_ = k_ 1 1 (T) [I-(~:: (T1 ( 2 k 2 (T1 ) kl + k 2 (Tl k2(T)) k I_T] 2 J -dr T1 cft if kt << k2( T1 then - d( T~ dt = 2 k_ (T1 1 if kt >> k?.( T) then -d( T) dt = 2 (~;I) k ( T) 2 ?. Mechanism (C) n kl T k_t k?. J) Assuming steadv state in ( f)~ H + N2 -256- k ( n) -d(T) dt -d( Ti dt = ( T) -1 k 1 + k2 = k_ (T) 1 = k_1(T) = k_1 = if k 1 > > kz = (I k_lk2 k 1 + k2 - k1(T)) ( - klk-1 ) ( T) k 1 + k2 k1 k1 + ( T) k ) 2 rr1 -257- CHAPTER 3 15N NMR. Spectrum of a l,l-Diazene N-(2, ?,6,4-Tetramethylpiperidyl)nitrenel -258- l.5N NMR Spectrum of a l,l-Diazene. N-(2,2,6,6Tetramethylpiperidyl)nitrenel Peter B. Oervan,•2 Michael E. Squillacote,3 Paul M. Lahti,4 Alan P. Sylwester, and John D. Roberts* Contribution No. 6?.94 from Gates and Crellin Laboratories of Chemistry, California Institute of Technolo~y, 0 asadena, California 91125 Abstract: The low-temperature 1 '>N NMR spectrum of the 1,1-diazene, N(2,2,6,6-tetramethylpiperidyl)nitrene (l) is reported. The 15N double- and mono-labeled 1,1-diazenes la and lb were synthesized. The nitrene and amino nitrogens of 1 have resonances in dimethyl ether at -90°C at 917.0 and 321.4 pprn, respectively, downfield from anhydrous 15NJ-I 3, affording a chemical-shift difference of 595 ppm for the directly bonded nitrogen nuclei. The chemical shift of the ring nitrogen is consistent with an amino nitrogen whose lone pair is largely delocalized. The large downfield shift of the nitrene nitrogen is consistent with a large paramagnetic term due to a lowlying n-1T* transition. -259- INTRODUCTION 1,1-Diazenes (aminonitrenes, N-nitrenes) unlil<e their more stable 1,2diazene isomers (azo compounds) are usually not isolated or detected by spectroscopic methods, but rather are assumed intermediates based on a substantial body of chemical evidence.5 'Recently, the synthesis and direct . .. R/ .. / N=N R observation of persistent6 1, 1-diazenes, N-(2,2,6,6-tetramethylpiperidyl}nitrene (1)7 and N-(2,2,5,5-tetramethylpyrrolidyl)nitrene8 were reported. The infrared an electronic spectra, and kinetics of decomposition of these l,l-diazenes7 ,8 allowed the first comparison of experiment with theory on the nature of the bonding ann the relative energies of the states of the parent 1,1-diazene (H2N-N).9 J 5N NMR spectroscopy has proven to be a sensitive probe of the electronic environment of nitroR;en nuclei.l 0 With the availability of persistent 1,1-diazenes, we have obtainec1 the first 15N magnetic resonance spectrum of a 1,1-diazene. The low-temperature 15N NM'R. spectrum of N(2,2,6,6-tetramethylpioeridyl)nitrene 17 reveals the different electronic environments of the contiguous nitrogens in the 1,1-diazene. We find the chemical-shift difference for the amino and nitrene nitrogens to be 595 ppm. Their assignments were determined from the spectra of the 15N double- and -260- mono-labeled 1,1-diazene, la and lb. The nitrene nitrogen of the 1,1-diazene has a 15N resonance in dimethyl ether at -900C 917 ppm downfield form anhydrous 15~H3 and represents the most highly deshielded neutral nitrogencontaining organic compound now known. RESULTS AND DISCUSSION Successive treatment of phorone 2 wit~ 15NH4Cl/sodium hydroxide and hyrlrazine hydrate/potassium hydroxide afforded 15N-labeled (9 5+ %) 2,2,6,6-tetrarnethylpiperidine Nitrosation 3.11 (Nal5No2, 95%) and reduction with lithium aluminum hydride gave the double-labeled precursor, N-amino-(2,2,6,6-tetramethylpiperidine) (r-.Ja15N02) and reduction of 4.7 ,12 Similarly, nitrosation unlabeled 2,2,6,6-tetramethylpiperidine 5 afforded the mono-labeled precursor 6 (Scheme I). Scheme I o=C G_J·NH, Q='"~: 2 4 1a QH 5 Q_J NH 5 6 2 Q· N='5~: 1b -261- Oxirlation of 4 with nickel peroxide in dimethyl ether affords upon filtration and concentration (-780C) a deep purple solution of the persistent 1, 1-diazene 1 whose electronic spectrum ( Amax = 543 nm (CH3)20 ), infrared spectrum and thermal stability are known.7 The 15N N\1'R. spectrum of this purple solution (.rl M) at -900C contains four doublets (Figure la). The two doublets, IS 419.5 (J = 6.4 Hz) anti 164.6 (J = 6.4 liz), can be assigned to the tetrazene 7 which has been independently synthesized and characterized. 7 The pair of doublets at 917.0 (J = 15.5 liz) and 321.4 (J = 15.5 Hz) were shown by 15N_15"'J decoupling to be directly bonded nitrogen nuclei. Upon warming to ooc, the purole color disappears and the doublets at 917.0 and 321.4 disaopear concomitant with an increase in the resonances of tetrazene 7, the dimerization product of 1.7 These doublets are assigned to the 1, 1-diazene la. Oxidation of the mono-labeled precursor 6 in dimethyl ether at -780C affords singlets at IS 917.() and 419.5 and no oeaks at 165 and 321 (Figure lb). This fact allows the assignment of the q 17 ppm resonance to the nitrene nitrogen of the 1, 1-diazene chromophore. The 15N NMR parameters for the 1,1-diazene la and the tetrazene 7 are presented in Table I. The large NOE for N2 in la suggests almost full nitrogen-proton dipole-dipole relaxation for this nitrogen. The nitrene nitrogen N 1 was found to have a smaller NOE than N2 and, as evidenced by the response in a signal intensity to changes in pulse delay, a spin-lattice relaxation time, T 1, which is insensitive to magnetic field strength and much shorter than the T 1 of N2. Thus a soin-rotation relaxation mechanism seemingly predominates for the nitrene nitrogen. This may reflect the different electron distributions -262- la Q-·=·-Q lb 900 800 700 600 500 400 300 200 100 ppm rei. to NH3 Figure 1. Fourier transform 1.5N NMR spectra of 1,1-diazene 1 and tetrazene 7 in dimethyl ether at -900C (30 deg pulse angle, 20 sec repetition time, NOE suppressed). (a) Spectrum of Ia and 7 from the nickel peroxide oxidation of 200 mg of 4 at 9.04 MHz, 250 transients. (b) Spectrum of lb and 7 from the nickel peroxide oxidation of 100 mg of 6 at 50.7 MHz, 190 transients. The assignment of the resonances are described in the text. -263- Table I 15Nl Compound Q.. N=!"!:- Q-•=·-Q oa J~ NOF: 0 J NOE 15N2 917.0 15.5 -0.2 321.4 419.5 6.4 -2.0 164.6 6.4 15.5 -5.0 -5.0 7 ~T)ownfield from external anhydrous ammonia, at 250C usini a 1:4 CH 315N02:cn2Cl7. mixture as a secondary standard at 380.7 ppm. E.Error in coupling constants is .:!:.0.6 Hz. Error in chemical shift is .:!:.0.5 ppm. between the two nitrogens, since it has been shown that an electron distribution leading to large chemical shifts will lead to large spin-rotation interactions.13 The deshielding of the nitrene nitrogen can be attributed to the importance of the paramagnetic term as represented by mixing of the ground state with excited states which results in changes induced by the external field in the electronic wave functions.l4, 15 Experimental correlations of 15N chemical shift with the energy of transitions involving low-lying excited states, in particular n-11'* transitions, are well documented.l4 The magnitude of the downfield shift of the nitrene nitrogen in l is consistent with a very large paramagnetic term due to the low-lying n-11'* transition ( Amax = nm in (CH3)?.o).7 543 -264- The amino-nitrogen of Ia N? has a 15N chemical shift of 321 ppm comparable to the shifts of nitrones, nitrates and nitro compounds (280-380) ppm 10 which lack non-bonding electrons on nitrogen. These molecules are formally N-oxicles of imines, nitrites, and nitroso compounds (330-900) ppm 10 and are consirleral)ly more shielded, even though they are also associated with low-energy n-TI* transitions. Thus, it appears t'"tat non-bonding electrons on nitrogen must he involved in the n- TI* absorption of the mole- cule in order for a large paramagnetic shift to be present for the nitrogen.l4c The amino nitrogen, N2, is not deshielded and hence its lone pair must be substantially delocalized into the empty p orbital of the nitrene nitrogen, Nl. This is consistent with both experiment (14N:14N, stretch, 1595 cm-1)7 and theory (GVB-CI)9 both characterizing the 1,1-diazene as having an N=N 7T bond. COI'IICLUSION In summary, the low temperature 15N NMR spectrum of the 1,1diazene, N-(?,?.,6,6-tetramethylpiperidyl)nitrene (l) is reported. The nitrene and amino nitrogens of 1 have resonances in dimethyl ether at 917.0 and 321.4 ppm, respectively, downfield from anhydrous 15NH3, affording a chemical-shift difference of 595 ppm for the directly bonded nitrogen nuclei. The chemical shift of the ring nitrogen is consistent with an amino nitrogen whose lone pair is largely delocalized. The large downfield shift of the nitrene nitrogen is consistent with a large paramagnetic term due to a lowlying n-TI* transition. -265- EXPERIMENTAL SECTION 15N nuclear magnetic resonance 05N NMR) data were recorded on a JEOL FX-900 spectrometer at 9.04 "~Hz (Figure Ja) ann a Brucker WM-500 spectrometer at .50.7 Ml-fz (Figure 1b). Chemical shifts are given in parts per million (ppm) cfownfield from anhydrous 15NH3 in o units and couplin~ constants in cycles per second (Hz). 15N chemical shifts were obtained using a 1:4 CH315N02:CD2Cl2 solution as a secondary standard. Proton NMR data were obtained on a Varian EM-390 spectrometer. Chemical shifts are given in ppm downfield from Me4Si in o units and coupling constants in Hz. For preparative vapor-phase chromatography (VPC), a Varian Aerograph Model 920 instrument equipped with a thermal conductivity detector and helium carrier gas was used. The VPC column was a 5 ft. x 0.25 in. glass, Pennwalt ?.23 amine pacl<in~ (Applied Sciences Laboratories, Inc.). All reactions were run under an argon atmosphere. 4-0xa-2,2,6,6-tetramethylpipericfine-l5N (8). Phorone 2 ( 10.0 g, 0.072 mmon and "'100 mg of sodium hydroxide were placed in a 25-mL pyrolysis tube, then frozen under vacuum. An ammonia solution was prepared by neutralizing 5.0 g (0.092 mmon of 15NH4Cl (Prochemicals Limited, New Jersey; 95.4 atom - %) with 8.0 g (0.20 mol) of sodium hydroxide in 6 mL of water. This solution was distilled under reduced pressure into the pyrolysis tube. The reactants were degassed, sealed uncfer vacuum, and heated in a stainless steel bomb to 135°C for 18 h. The tube was then opened; the contents dissolved in et her, dried (1('2C03) and concentrated affording ~.2 g of a yellow oil. Recrystallization from petroleum ether at -780C gave 7.0 g -266- (62%) of 8 as white needles, mp 37-3~0C (lit. mp 360C))7 NMq_ (CT)Cl3) o 1.'22 (d, 12, J = 2 Hz), 2.'22 (s, 4). 2,'2,6,6-Tetramethylpipericfine-l5N (9). A total of 7.0 g (0.45 mol) of 8 were stirred with 10 g of potassium hydroxide, 9.3 mL of hydrazine hydrate, and 2.0 rnL of water in 67 g of triethylene glycol at reflux for 2 h. The product was distilled (bp 96-112°C), extracted with ether, dried (~a2S04) affording 4.5 g (72%) of a clear liquid 9, bp 145-1470C (lit. 151-1520C at 7.50 mm Hg))1 ~MR (CJ)Cl3) o 1.12 (d, 12, J =2Hz), 1.3 (m, 4), 1.6 (m, 2). l-Nitro~2,2,6,6-tetramethy1pipericfine-l5N2 (10). A solution of 1.5 g (0.0 11 mon of 9 in 13 mL of a 6.8% solution of aqueous hydrochloric acid was heated to 95oc. To this was added 1.5 g (0.021 mol) of l'l.la15N02 fProchemicals Limited, ~ew Jersey; 95.5 atom - %) in 8 mL of water. The solution was stirred at 95oc for 48 h. The reaction mixture was allowed to cool and was extracted with ether, washed with 10% aqueous hydrochloric aciri, saturated aqueous sodium bicarbonate, saturated aqueous sodium chloride and dried (Na2S04). The ethereal layer was concentrated affording 1.2 g (64%) of a yellow oil 10. NMR (CDClJ) ol.40(d,6,J =2Hz),1.62(d, 6, J = 2 liz), 1.7 (m, 6).12 l-Amino-2,2,6,6-tetramethylpiperidine-1.5N2 (4). A solution of 1.2 g (7.0 mmol) of 10 in 4 mL of dry ether was added dropwise to 580 mg (15.3 mmoH of lithium aluminum hydride in 17 mL of 1:1 di-n-butyl ether:diethyl ether. The temperature was slowly raised to 95°C with distillation of solvent; the temperature was then maintained at 9 50C for 3 h. The slurry was cooled to ooc, excess lithium aluminum hydride was quenched with water, and 25 mL of -267- ether was added. The layers were separated and the aqueous layer was washed with ether. The combined ether layers were extracted with 10% aqueous hydrochloric acid. · The aqueous layer was made basic witl-t 20% aqueous sodium hydroxide. This was extracted with ether, dried (Na2SOtt), and concentrated affording 1.1 g of a clear oil. This was further purified by preparative VPC (Pennwalt, 1800C) affording a 70% yield of pure l-amino2,2,6,6-tetramethylpiperidine-l5N2 4. NMR (CT)Cl3) ol.05(d,J =2Hz, 12), l.5 (s, 6), 2.8 (broads, 2).12 N-(2,2,6,6-tetramethy1piperidy1lnitrene-l5N2 ( l a). To 20 mL of anhydrous dimethyl ether, cooled to -780C, was added 191 mg (1.9 mmol) triethylamine and 300 rng (1.9 mmon 1-amino-2,2,6,6-tetramethylpiperidine15N2 (4) with the aid of a syringe. To this was added 3.4 g (19 mmol) nickel peroxidel8 through a solid addition funnel with stirring over 5 min. The reaction mixture was stirred at -780C for 2 h, then transferred through a Teflon tube to a cooled f-780C) jacketed filter funnel, and filtered into a three-necked flask cooled to -780C. The clear purple filtrate was concentrated (J'l M) and transferred into a 10-mm NMR tu'be for lowtemperature (-9QOC) 15N N~R studies. -268- REFERENCES AND NOTES (1) The authors are ~rateful to the National Science Foundation for generous support. (2) Camille and Henry Dreyfus Teacher Scholar 1978-1983. (3) National Science Foundation Postdoctoral Fellow. (4) National Science Foundation Predoctoral Fellow. (5) For reviews of 1,1-diazene behavior, see: (a) Lema!, D. M. In "Nitrenes"; Lwowst<i, W., ed.; Interscience: New York, 1970, Ch. 10. (b) Joffe, B. V.; l(uznetsov, M. H. Russ. Chern. Rev. ('=.ngl. Trans!.) 1972, i!., 131. (6) Griller, n.; Ingold, K. U. Accts. Chern. Res. 1976, ~' 13 and references cited therein. (7) (a) Hinsberg, III, W. D.; Oervan, P. ~. J. Am. Chern. Soc. 1978, 100, 1608. (b) Hinsberg, III, W. D.; Oervan, P. B. Ibid. 1979, 101, 6142. (8) Schultz, P. G.; Oervan, P. !=',. J. Am. Chern. Soc. 1980, 102, 878. (9) (a) Hayes, L. J.; Billingsley, F. P.; Trindle, C. J. Org. Chern. 1972, 37, 3n4. (b) Baird, N.C.; Barr, R. F. Can. J. Chern. 1973, 21, 3303. (c) Lathon, W. A.; Curtiss, L. A.; Hehre, W. J.; Lisle, J. B.; Pople, J. A. Prog. Phys. Org. Chern. 1974,1, 175. (d) Ahlrichs, R.; Staemmler, V. Chern. Phys. Lett. 1976, 37, 77. (e) Baird, N. C.; Wernette, n. A. Can. J. Chern. 1977, 2.2, 350. (f) Davis, J. H.; Goddard, III, W. A. J. Am. Chern. Soc. 1977, 99, 711. (g) Casewit, C. J.; Goddard, III, W. A. lhid. 1980, 102, 40 57. 00) (a) Levy, G. C.; Lichter, R. L. "Nitro~en-15 Nuclear Magnetic -269- Resonance Spectroscopy"; John Wiley & Sons: New York, 1979. (b) Witanowski, M.; Webb, G. A. "Nitrogen NMR"; Plenum Press: New York, 1973. (c) Witanowski, M.; Rep. NMR Spectrosc. 1977, Stefaniak, L.; Webb, G. A. Ann. z, 117. 01 l Leonard, 1\J. J.; "'7ommensen, 1:. W. J. Am. Chern. Soc. 1949, Zl, 2808. (12) Roberts, 1. R.; Ingold, K. U. J. Am. Chern. Soc. 1973, 95, 3228. (13) (a) Ramsey, "1. F. "Molecular ~earns"; Oxford University Press: 1956, p. 164. (b) neverell, C. Mol. Phys. 1970, 'i4, 64. (}4) (a) Herbison-l:vans, D.; Richards, R. E. Mol. Phys. 1964, ~' 19. (b) P.aldeschwieler, J. D.; Randall, E. W. Proc. Chern. Soc. 1961, 303. Gil, V. M. S.; Murrell, 1. 1\J. Trans. Faraday Soc. 1964, 60, 248. (c) Lambert, J. B.; Roberts,J. n. J. Am. Chern. Soc. 196.5, 87, 4087. (d) Anderson, L. 0.; Mason, 1.; van Bronswijh, W. J. Chern. Soc. (A) 1970, 296. (e) Mason, J.; van Bronswijh, W.; Vinter, J. G. J. Chern. Soc., Perkins II 1977. (15) The paramagnetic shielding term, op, of the Poole chemical shift theory t 6 can be approximated as being inversely proportional to flE, the mean excitation energy for all excited states. Low-lying n-n* excited states will dominate this term affording the largest deshielding contribution to a p· fJ 6) (a) Karplus, M.; Pople, J. A. 1. Chern. Phys. 1963, 38, 2803. (b) Poole, J. A. T)icuss Faraday Soc. 1962, 34, 7. (17) Francis, F. J. J. Chern. Soc. 1927, 2897. (1 ~) Nickel peroxide was prepared and the oxygen content determined by -270- the method of Nakagawa.l9 Activation of nickel peroxide prepared in this manner with 6% sodium hypochlorite solution yielded a dark solid with an activity of 3.9 to 4.5 x to-3 g atom oxygen/g nickel peroxide. 09) Nakagawa, 'K.; 'Konaka, 'R..; Nakata, T. J. Org. Chern. 1962, 27, 1597.