The Synthesis, Spectroscopic Observation and Chemical Reactivity of N-(2,2,6,6-Tetramethylpiperidyl)Ni trene Citation Hinsberg, William Dinan, III (1980) The Synthesis, Spectroscopic Observation and Chemical Reactivity of N-(2,2,6,6-Tetramethylpiperidyl)Ni trene. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/3R32-XC97. https://resolver.caltech.edu/CaltechTHESIS:07252016-112114482 Abstract 1,1-Dialkyldiazenes (aminonitrenes, N-nitrenes) unlike their more stable 1,2-dialkyldiazetle isomers (azo compounds) have not yet been isolated or detected by spectroscopic methods, but rather are assumed intermediates based on a substantial body of chemical evidence. The first direct observation of a 1,1- dialkyldiazene is described here. The visible spectrum at -78° of N-(2,2,6,6-tetramethylpiperidyl)nitrene (33) (λ max = 541 nm) provides experimental evidence on (1) the energy required for the n + π* electronic transition, and (2) the vibrational spacing of the first electronically excited state. The infrared spectrum at -78° ( 14 N= 14 N stretch at 1595 cm -1 ; 14 N= 15 N stretch at 1569 cm -1 ) provides evidence that the 1,1- diazene has considerable N=N double bond character in the ground state. The first kinetic study of the thermal decomposition of a 1,1-dialkyldiazene is described. The temperature dependence of the unimolecular rate (k 1 ) of fragmentation of 33 was examined in three different solvents and kinetic evidence for a direct bimolecular pathway for the formation of 1,1 "-azo-2,2,6,6-tetra- methylpiperidine 41 from 33 is provided. The activation parameters for the unimolecular fragmentations are log A = 11.6 ± 0.5, Ea = 16.9 ± 0.7 in n-hexane, log A = 13.7 ± 0.3, Ea = 20.0 ± 0.4 in Et 2 0, log A= 13.6 ± 0.3, Ea = 20.1 ± 0.4 kcal/mole in THF. Using computer simulation it is found that the curved portions of the tn A vs. time plots may be modelled as competitive unimolecular and bimolecular reactions (k obs = k 1 + k 2 [33]). In Et 2 0 at -16°, k 1 = 5.03 x 10 -4 sec -1 and k 2 = 5.0 x 10 -2 liter/mole-sec. Also reported are proton and carbon-13 nuclear magnetic resonance data for 33, along with the results of a preliminary study of its photoreactivity. 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): Dervan, Peter B. Thesis Committee: Unknown, Unknown Defense Date: 31 August 1979 Funders: Funding Agency Grant Number NSF UNSPECIFIED Petroleum Research Fund UNSPECIFIED Record Number: CaltechTHESIS:07252016-112114482 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07252016-112114482 DOI: 10.7907/3R32-XC97 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 9895 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 26 Jul 2016 22:48 Last Modified: 14 Jan 2025 23:18 Thesis Files Preview PDF - Final Version See Usage Policy. 20MB Repository Staff Only: item control page

THE SYNTHESIS, SPECTROSCOPIC OBSERVATION AND CHEMICAL REACTIVITY OF N-(2,2,6,6-TETRAMETHYLPIPERIDYL)NITRENE Thesis by William Dinan Hinsberg, III In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1980 (submitted August 31, 1979) i i To My Parents jii ACKNOWLEDGEMENTS I wish to thank my research director, Peter Dervan, for his advice, support and encouragement throughout this work. I would also like to thank all the members of the Dervan group, past and present, for their friendship; certainly that is one of the most important things I have gained at Caltech. Mike Ingle deserves a special thanks for all he taught me during my first years as a graduate student. I am indebted to the staff of the Caltech Chemistry Depart ment who were always generous with their time and helpful in cutting red tape, and to the technicians in the Chemistry Shops, without whose skill this research could not have been realized. A special thanks goes to Ms. Deborah Chester for typing this thesis. The financial support of this research by the National Science Foundation and the Petroleum Research Fund, and the receipt of an NSF Predoctoral Fellowship are gratefully acknowledged. Finally, thanks to my wife, Frances Houle, for her love and companionship. iv ABSTRACT 1,1-Dialkyldiazenes (aminonitrenes, N-nitrenes) unlike their more stable 1,2-dialkyldiazetle isomers (azo compounds) have not yet been isolated or detected by spectroscopic methods, but rather are assumed intermediates based on a substantial body of chemical evidence. The first dJrect observation of a 1,1- dialkyldiazene is described here. The visible spectrum at -78° of N-(2,2,6,6 - tetramethylpiperidyl)nitrene (~) (Amax = 541 nm) provides experimental evidence on (1) the energy required for the n + n* electronic transition, and (2) the vibrational spacing of the first electronically excited state. trum at -78° (14N=l4N stretch The infrared spec- at 1595 cm- 1 ; 14 N= 15 N stretch at 1569 cm-1) provides evidence that the 1,1diazene has considerable N•N double bond character in the ground state. The first kinetic study of the thermal decomposition of a 1,1-dialkyldiazene is described. The temperature dependence of the unimolecular rate (k 1) of fragmentation of ~ was examined in three different sol vents and kinetic evidence for a direct bimolecular pathway for the formation of 1,1 '-azo-2,2,6,6-tetra- v methylpiperidine !!_ from 33 is provided. The activa- tion parameters for the unimolecular fragmentations are log A= 11.6 ± 0.5, Ea = 16.9 log A= 13.7 ± 0.3, Ea = 20.0 13.6 ± 0.3, Ea ± ± 0.7 in _!!- hexane, 0.4 in Et20, log A= = 20.1 ± 0.4 kcal/mole in THF. Using computer simulation it is found that the curved portions of the tn A vs. time plots may be modelled as competitive unimolecular and bimolecular reactions (kobs = k1 + k 2 [ll]). 10- 4 sec- 1 and k 2 In Et 2 0 at -16°, k 1 = 5.03 x = 5.0 x 10-2 liter/mole-sec. Also reported are proton and carbon-13 nuclear magnetic resonance data for~' along with the results of a preliminary study of its photoreactivity. vi TABLE OF CONTENTS Page I . II. . INTRODUCTION ........ I I ........................ • 1 A. The Preparation of 1,1-Diazenes ........... 3 B. The Chemistry of 1,1-Diazenes ............. 5 C. Theoretical Studies of 1,1-Diazenes ...... 15 D. Experimental Strategy .................... 20 RESULTS AND DISCUSSION . .... .. ................ 26 A. The Synthesis of N-(2,2,6,6-Tetramethylpiperidyl)nitrene B. (~) ........... 26 The Electronic S~ectrum of N-(2,2,6,6Tetramethylpiperidyl)nitrene C. The Infrared Spectrum of N-(2,2,6,6Tetrarnethylpiperidy1)nitrene D. (~) ...... 28 (~) ...... 34 Electron Spin Resonance Spectra of N(2,2,6,6-Tetramethylpiperidyl)nitrene (33) . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 E. Nuclear Magnetic Resonance Spectra of N-(2,2,6,6-Tetramethy1piperidyl)nitrene F. (~) ........................... 42 Decomposition Kinetics of N-(2,2,6,6Tetramethylpiperidyl)nitrene G. (~) . . .... 57 Photoreactivity of N-(2,2,6,6-Tetramethylpiperidyl)nitrene (33): Pre- liminary Results ...............•....... 85 ·vii H. Attempted Syntheses of 3,3,7,7 Tetramethyl-1,2-diaza-1-cyclohepte ne (59) . . . . . . . . . . . . . . . . . . . ........ . ......... . 88 I . Summary . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . · .. . 9 6 III . EXPE RI MENTAL . • • . •. • •. . •• • •• . .••••.. . ••.• • •....• • 99 IV. REFERENCES • •..• •• •• . ... . ..••••• . •••.•••••...•.. 164 Part I INTRODUCTION - 1- A fundamental aim of mechanistic organic chemistry is to understand the pathways by which reacting mol ecul es proceed to products. Numerous indirect methods have been used to elucidate reaction pathways.l, 2 Among these are kinetic studies, the us e of isotopic tracers, thermochemical measutements, determinations of reaction stereochemistry , and trapping of intermediate species. A more direct method for studying the featur es of reaction pathways is the spectroscopic observat ion of reactive intermediates and the study of their chemical behavior. The subject of this thesis is the examination of one such intermediate, the l,l ~ dialkyldiazene. l,l-Dialkyldiazenes3 (aminonitrenes, N-nitrenes) ! unlike their more stable 1,2-dialkyldiazene isomers (azo compounds) l have not yet been isolated or detected by spectroscopic methods but rather are assumed intermediates based on a substantial body of chemical evidence. The chemical reactions of pres·umed 1, 1-dialkyldiazene intermediates sh ow behavior sugges ting that the reacting species is a singlet . For example, they do not exhibit the free R\ I N-N• .. • R Ia .... R\~=N- R-N=N-R R/ lb 2 -2- radical-like behavior characteristic of triplet ni t r enes and th ey accept a proton on the monovalent nitro gen to yie ld a fairly s table cation. nitrenes This is in direct contrast to other R-N·· (~) (eg. alkyl, aryl, carbethoxy and cyano- njtrenes), which are ground state triplets . 4 These results suggest that the 1,1-dialkyldiaz en e may be represented by the resonance structures la and lb and that the di polar form lh contributes importantly to the hybrid. The existence of 1,1-diazenes was fi rst proposed in 1893 by Michaelis and Luxemburg to explain the formation of 1,2 - diazenes rather than the "normal" tetrazene products in the oxidation of 1-allyl-1-aryl hydrazines. 5 (;_ ( I I N=N ----.. Ar In the past ) I N=N Ar N-NH 2 Ar three decades the chemistry of 1,1-dinzenes has been investiga ted extensively;3 some of the more important results and generalizations concerning 1,1-diazenes are discus sed below . -3- PreRaration of 1,1-Diazenes.3 Th ere exists a variety of independent methods by which 1,1-diaze nes may be generated (see Figure 1). Of these, perhaps the most versatile is the oxidation of 1,1-disubstituted hydrazines (!) . A wide variety of oxidants have been us e d, including mercuric oxide, 6 lead tetraacetate, 7 brornine,a manganese dioxide,9 t-butyl hypochlorite 1 0 and diethyl azodicarboxylate. 11 dialkyldiazenium ions (~) The neutralization of 1,1- is actually a variation ori this method in which a disubstituted hydra zi ne is oxidized in acidic media and tl1e r es ulting 'diazene is trapped as its protonated form; n e utralization results in the formation of products characteristic of l,l - diazen es. 12 of difluoramine with secondaryarnincs (~) The reaction to form l,l- diazenesl3 has not received much use, presumably due to the explosive nature of the fluorinated reagent. The thermal and photochemical cleavage of N-aminosulfoximines (!) is po t entially a useful means of generating l,l-diazenes. 14 However, the nature of the R-groups is severely restricted by the me thod of synthesis of Z (formed by trapping 1,1- diazcne s with dimethyl sulfoxide; see below). Th e decom- posit ion of 1,1-djsubstituted-2-sulfonylhydrazines (!) in basic protic media or of the analogo us sodium salts in aprotic media are versatile methods for generating 1,1- ~ + 10 2 -t R2 N=N 9 lorA +base Methods of Preparation of 1,1-Diazenes. 8 1 O 6 R2 N-H R2N=N-SMe 2 7 ~or bv NaN203 ~NF2 ~ R2N-H ... R2 N-N .base 5 R2N-NHS02R ry + R2 N=N ~ R2 N-NO 2)base ._. I) H+ Figure 1. Cu 3 CI3 _1_~ R2N-NH2 + R2N==NH x- I ~ I - 5- diazenes. 15 The thermolyses of sulfonylhydrazines! at 306-439° have recentl y been shown to yield products con - sistent with the intermediacy of 1,1-diazenes. 1 6 The re- duction of N-nitrosamines 9 with alkaline sodium dithioni t e! 7 lithium in liquid a~nonialB or ethyl diphenylphosphinite 1 9 in benzene solution has been shown to yie ld products derived from 1,1-diazenes. The reaction of sodium nitro- hydroxa mate 20 or N-benzenesulfonylhydroxylamine 21 with secondary amines has been used to generate 1,1-diazenes, apparently proceeding by initial generation of nitroxyl (IIN=O) which adds to the secondary amine. Sub sequent loss of the elements of wat er generates the diazene.3a Finally, the oxidation of 1,1 -dialkylhydrazines with cupric chloride in aqueous solution gave a complex, ~, wh ich released the 1,1-dialkyldiazene upon treatment with acid f ollow ed by neutralization.22 The Chemistry of 1 l-Diazenes. 3 1,1-Diazenes exhibit rich and diverse chemistry (see Fjgure 2); the exact mo de of stabilization taken is determined by internal factors (structure and substitution) as well as external factors such as the temperature, the nature of the medium and even the rate of addi tion of the r e actants . 3b The react ions of 1,1-diazenes may be divided into three types: (1) fragme ntations, (2) isomeri zati ons , and (3) bimolecular reactions. u 14 I 0 Fi gu r e 2. The Reac tion s of + N2 Hydrocarbon l,l ~ Diaze n es . 7 I R \ \R, II I \oMSO R I N==N-SMe2 R I N-N \ .. \+ N==N ~ R' R• I 12 J--R'" R" N-N ~ \ R R-N=N-R' R-N=N-R .... 4.,__- I 'N-NJ ,,,, R _@_ ~ R' I I 0\ -7- Fragmentation reactions with loss of molecular nitrogen are the decomposition processes of 1,1-diazenes whjch have bc:en studied in greatest detail. Depending on the structural details of the molecule, s ubstituent -assisted fragmentation, concerted fragmentation or a combination of both may occur. The substituent-assisted process was first observed by Busch and Weiss in 1900: 2 3 + This mode of fragmentation has also been observed in cyclic cases, with analogous products; for example:l 8 Ph Q N-NO Ph cc:: c:HPh + Ph It has been demonstrated that, under appropriate conditions, a variety o f substituents and substitutuon · patterns will suffice in making this a dominant decomposition pathway:l 6 , 2 5, 2 6 -8 - ()Ph HgO 65• .... I NH 2 N~CN ~ Kllft04 25 • QPh 0Ph ()Ph ~ .,NC CN NH2 -Q- .... c( 306 -439° I HNS02R J)CH~ ~ :=/ In general, the experimental results are in accord with a mechanism involving radical or biradical species. However, cases are knownl 6 which may involve competitive biradical and concerted pathways. Conc erted fragmentati on of 1,1-diazenes may occur in a number of cases. The stereochemical outcomes of certain decompositions have been in accord with those predicted on the basis of orbital symmetry considerations. 27 For exam- ple, McGregor and Lemal fo und that cis and trans-2,5d1methyl-3-pyrolline (~and~' respectively), when treated with s od i um nitrohydroxamate, yielded only the trans-trans and the cis, trans-hexadienes (}_?. and !.~) respectively. -9- -{?. · -er. . .. -N 16 18 These results are consistent with the disrotatory opening of the i ntermediate l,l - diazene.2B Analogous low-energy concerted pathways have been exploited, for example, in a study of ortho-xylylene ring-closure stereochemistry, 29 and as a mild method for the generation of ortho-benzyne. 30 Isomerizations of 1,1-diazenes generally take one of two forms: (a) rearrangement of the 1, 1-diazene !. to an isomeric hydrazone _!lor (b) rearrangement to a 1,2-diazene 11 or 14. The formation 6£ hydrazones is a general reac- tion for all 1,1-dia zenes which possess an alkyl C-H bond alpha to the nitrogen functionalit y . 3 Hydrazones are ob- served in such cases only when the 1,1 - diazene is generat~d in a protic medium; labelling experiments have demonstrated that the "nitrene" nitrogen ends up as the doubly bonded -10- nitrogen in the product:3l E~ )i N-1'N I Et batt \ ROH S0 2 Ph ... \ ~+ N= ~N __....... Et 1 N- 1 ~N I I H Et \\ CHCH 3 This isome riz ation has been the subject of intensive research, yet the mechanism remai ns unknown. 1 ,2-Diazenes and diaziridines have been ruled out as possible intermediates.32 At present the reaction is regarded as proceeding t h rough an az omethinimine3a 19 which is analogous to the _ __...,... H ~ I +N-N .-11 - - --- < N-N I ~ H 19 enol form of a ketone . Therefore, the function of th e protic so l vent may be to faci1it~te tautomerization of the 1 ,1 diazene. Ring expansions of presumed 1,1-diazenes to 1,2 diazenes occur in certain cases where the resulting azo linkage is incorporated in an aromatic sys t em. As an ex- ample, Rees and coworkers f oun d that leid tetraacetate oxidation of either 20 or 21 resulted i n formation of the - 1]- benzotriazine 22.33 co R I R .. CCN I I NH 2 N R R ~ gQ. ~N-NH2 .. ~N-N / N' R oCN~~ 22 t.( 21 1 ,2 -Diazenes have also been shown to arise from the 2 , 3 - sigmatropic r earrangements of ally l -s ub st i tuted 1,1 diazenes. The work of Michaelis and Luxemburgs provides an example of this (see above). Their early ob servations have been confirmed and the scope of the reaction broadened by the work of Baldwin and coworkers;3 4 they provide evidence that the reaction proceeds in a concerted ma nner , rather than by a dissociation-recombination route. -12- ~HzY ~N~ 63% ... 27% 10% The bimolecular chemistry exhibited by 1,1-diazenes includes the formation of tetrazenes !!, addition to n-honds, and trapping by sulfoxides. The formation of tetrazenes is the most general reaction of 1,1-diazenes. Although tetrazenes are formally the head-to-head dimers of diazenes, it is likely that under most conditions such coupling does not occur and that the tetrazene pro~1ct arises from some alternative pathway.3a For example, the neutralization of solutions of 1,1-dialkyldiazenium ions leads to the formation of tetrazen es 1 2 probably by reaction of the diazene with a diazenium ion 5. Similarly, in -13- + - R2N=N + + R2N=NH e - - - R2 N=N - -......... TetH+ Tet ...... certain cases the intermediate diazene formed in the oxidation of a 1,1-disubstituted hydrazine might intercept a molecule of starting material yielding a tetrazane 23 which under the reaction conditions reacts further to yield tetrazene. Evidence for this mechanism has been obtained by Rees and coworkers.35 They have isolated from the lead tetraacetate oxidation of N-aminophthalimide a 90% yield of R2N-NH 2 ---~~ R N-N-N-NR 2 I I 2 ~ R N-N=N-NR 2 z HH 23 the corresponding tetrazane, which on further oxidation afforded tetrazene. Simple ·bimolecular dimerization of 1,1-diazenes probably occurs under certain circumstances, for example, in the decomposition of 1,1-dialkyl-2-sulfonylhydrazine salts in aprotic media.1sb Simple 1,1-dialkyl diazenes have not been found to undergo addition reactions to ~ bonds and to dimethyl sulfo x id~. Such behavior is exhibited only by diazenes sub- -14- stitutcd with electron-withdrawing (e. g ., acyl) groups and by diaz enes in which the lone pair on th e trivalent nitrogen is part of an aromatic system. Such substituents, by delo c ali zation of the a mi no lone pair electrons, apparently increase the elcctrophilic character of the "nitrene" nitrogen.3a Experimen tal evidence indic ates that the ad- dition of diazenes to alkenes is ~ven > 95% stereospecific, a t low alkene conccntrations. 36 By analogy to carbene chcmistry37 thi s is generally tak en as evidence that the 1,1-di azene r eac t s as a singlet and further more th a t the singlet is th e grou nd state. 1,1 - Diaz enes have b een s hown to add to both electron - rich and electron - poor olefin s . 36b , 3B Ree s and coworkers have s ucceeded in adding phthalimido nitrcn e to acetylenes ; th e r es ulting anti a romatic 1-Hazjr in es ~undergo a spontaneous rearrangement to 2-H i s omers 25 : 39 Et 0 cxr-N Il I + 0 I Et c<N-N~EI O 24 Et -15- Theoretical Studies of 1,1-Diaz e nes . Due to the high reactivity of the systems involved, solid experimental information regarding the properties of t h e N2 1I 2 i some r s , _2 6 , ?:.2, and · 2 ~ i s d i f f i c u 1 t to o b tain. 4 D Theoreticians h ave devoted much time to studies H I I N=N H H I N=N \ H 26 \ H 27 I H N-N• .. H 28o \t=N- ' I H 28b of the ~ 2 H 2 energy surf a ce, 11 I- 55 investigating the nature of the bonding, the ordering of states, the relative energi e s of th e three isomers, the mechanisms of i nt erconvers :i on of the isomers, the reactions of N2 H2 species with alkenes, and the energetics of N-H bond cleavage for th e isomeric diimides. The most important question concerning the electronic structure of 1,1-diazene 28 is of the nature of the ground electronic state. Is it a triplet like other nitrenes, or a singlet, as the chemical behavior of 1,1-diazenes implies? Tabl e 1 lists the results of a number of thea- retical investigations into this question. It is apparent after an examination of the results that, as the calculational methods become increasingly sophisticated ~.g., by -16 Table l Singlet-Triplet Energy Gap in H2N-N Calculational Method Ground State - S- T Gap * Refer ence 26 . 3 ST0-3G T 4 - 31G T 11.7 HF 5. 2 T SCF** 2.1 T 4-31G + CI 1.6 s 13.8 GVB-CI s *kcal/mole; both states at equilibrium geometries . **Restricted open shell - SCF. 49 64 54 50 59 60 employing larger basis sets and extensive configuration interaction) , the energy of the triplet increases relative to the singlet. This trend is not unexpected; several of the authors explicitly state that the calculational methods they employed te nd to overestimate the stabilities of trip l et s t a t es . ~9 , 6~ The two most recent sets of calculations indicate that thesinglet is the ground state.s9,GO The most recent of these calculations , that of Davis and Goddard, 60 is the most ~xtensive and will be dealt with here in some detail. From their study , the following picture of the parent 1,1-diazene emerges (see Figure 3a). The ground state is a singlet (S 0 ) of planar geometry , with the dipolar form 28b making a substantial contribution to the electronic structure of the 1,1-diazene. The -17triplet state (T 1 ) lies 0.6 eV (14 kcal/mole) ab ove the ground state and is also of planar geometry. The first excited single t (S 1 ) lies 2.2 eV (51 kcal/mole) above the gro und state. The configuration of S 1 corresponds to an n-n* exci ted state. Thus the calculations predict that the 1,1-diazene should absorb li ght at 2.2 eV (560 nm) and hence should be detectable by electronic absorption spectroscopy in the visible region. The N-N 0 bond length calculated for 3!_ (S 0 ) is 1.25 A, identical to the experimental value observed for the 1,2-isomer ~; 66 it is reasonable to expect that~ will exhibit a N=N stretching vibration in the infrared at a frequency similar to that observed in the Raman spectrum of ( 15 2 9 3.£. em- l) • 6 7 Casewit and Goddard have recently performed a set of ab initio calculations investigating the thermochemistry of 26 and 28 (see Figure 3b) .'~ 2 Their GVB-CI calculations give a N-H bond energy in the 1,1-isomer ~of 42.0 kcal/ mole and a N-H bond energy· in the 1,2-isomer ~of 71 .4 H \+ I N=N H 28 ~ I H . N=N + H· ......,..__ 26 -18- (a) T ~ ~ ~ ~ 'At 4 .8tV 'A2 1.6tV +* 3 A2 0 .6tV 'At H2 N-N C2v s, T, So (b) ~=N ... H· t \ I \ I I \ ;d/ t \ \ \ 71.4 \ I \ \ \ I .t 29.4 \ \ H \ N= N/ \ • \ H/ Figure 3 (a) GVB-CI calculation~ 0 for H2N - N (C 2 v symmetry); (b) GVB-CI c'lcul ation s 42 f or the H2 N2 energy surface and the diazenyl radical 11 - N=N·. -19 kcal / mole . Thus t he N-H bond. in l!!_ is weaker by "' 30 kcal/mole. These calculated bond energies may be use d to estimate tlte C-N hond strengths i n alkyl substituted 1,1 - and 1, 2-diazenes by application of a correction f or the diffe r ence between a N- CH 3 and a N-H bond (bas e d on D(HMeN - 11) - D(HMeN-Me ) = 18.5 kcal/mole) . Gsa manner , Casewit an d Goddard estimate the bond e nergy in 1,1-dimethyldiazen e 29 to be 23.5 kcal/mo l e. Application Me Me I \+ N=N --.- N=N +Me · .....- N=N I I I Me Me Me 29 In this 30 -20of this correction to the N-H bond ener gy calculated f or 26 lead s to a predicted value for the C-N bond strength in 1,2-dimethyldiazene lQ_ of 52.9 kcal/mole; this i s in excellent agreement with the observed activation ener gy of 52.5 kcal/mole for the thermal decomposition o f 30 . 68 b Pasta and Chiprnan 65 carried out calculations a t the ST0-3G level on the energetics of the homolytic breakdown of 29. They find the C- N bond energy to be 26.2 kcal/ mole, in accord with the estimates of Casewit and Goddard. Experimental Strategy The study of reactive intermediates has required the development of many specialized techniques . A number of el e gant methods have been applied to the study of such molecules: the use of superacid media for the study of carbonium ions , 69 the use of matrix i s olation techniques? 0 flash spectroscopy , 7l and reactant des i gn such that once the intermediate is formed further reaction is slow . 72 - 74 For the study of 1,1-dialkyl diazenes this last method app e ars to be the most desirable. It would allow the study of 1,1 - diazenes in s olution phase by using simple modifications of t he spectroscopic techniques commonly employ e d b y or ganic chemists, such as UV-visible, infrared and nuc lear magnetic resonance s pectroscopy. -21..: We were encouraged by work in the literature concerning the benefits derived from the synthesis of "persistent" radicals, 72 radicals which may or may not be stabilized by resonance and inductive effectsbutwhich are long-lived and therefore subject to spectroscopic inspection because the rate of bimolecular reaction has been drastically slowed down by a steric blockade (see Table 2). The per- sistent radicals, with a suitable arrangement of bulky groups, have half-lives varying from seconds to years whereas, under similar conditions, transient radicals would have half - lives of less than a millisecond. Since the persistent radicals can easily b e prepared in relatively high concentrations, their structural and chemical properties can be examin e d with an ease and accuracy impossible to attain with transi e nt radicals. The general concept of ster1cally induced persist ence has been exploited in cases involving non-radical intermediates; for example cyclobutadiene and cyclopentadienone, both of which undergo facile dimerization reactio~s, have been studied in the form of the suitably substituted derivatives 3J. 7 3a and 3 z.73b 31 32 -22Table 2 Radical Decny Lifetimes 72 Not Persistent log t 112 (sec) * - 4. 3 second order - 3.3 second order Persistent *Calculated at 25° Deca y K1net.ics Decay Kinetics 2.4 first order 3.3 first order and at radical concentrations of 10- 5 M. -2 3- In order to synthesize a "persistent" 1,1-dialkyldiazene two propertie s must b e incorporated in the cho s en structure. First, there must be no f a cile unimolecul a r fragmentation pathway, and s econd, there must be a st e ric blockade to prevent dimeri zation to tetrazene . The fi rs t of th ese pro pe rtj.es may b e controlled by approp ria te cho i c e of a- sub s tituent s . Some cas es a re known th a t aff o rd only t e traze ne at zso ,3 a , 7 4 wh e r e a s othe r s g iv e r<'-NH2_H_o_o_..,.._ "'---<' 25• Q-N=N-9 LTA hi gh yi e lds o f dec ompo s ition produc ts :3 a ,b, 24 Ph r\N-NH2 _LT_A-t)loo.-. l_( Ph 25. Ph I + -24- Presumably alky l and phenyl (but not benzyl) sub s tituted 1,1-diazenes are sufficiently stable to fra gmentation at room temperature and below th at tetrazene occurs. formation The second property, the ster ic blockade, may be controlled by the placement of suitably bulky groups such that the dimeriza tion ratu to tetrazene is drastically reduced. A si~p le chemical system which potentially has [o] Slow! Dimlfization Tetrozene the required properties outlin ed above is that of the tetramethylpiperidylnitrene ~· This system has the ?<t ~=N 33 - 2 5- added advantage that it lacks a-C ·H substitution , thus preventing any 1,1-diazene +hydrazone rearrangement. In an ESR i nvestigation of hydrazyl radicals Ingold and coworkers had cause to i nvestigate the products resulting from the oxidation of the 1 , 1-hydrazine 36 with t - butyl hyponitrite? 6 They reported the isolation of the hydro carbons 37 and 38 which are consistent with the intermediacy of the 1,1-diazene ~· The results of our investi- gations of this model system are discussed b e low . ~ ~Hz -36 · BuON•NOBu CeHe so• ... D + 72 % 28% 37 38 Part I I RESULTS AND DISCUSSION -26The Sy n thesis of N-(2 , 2,6,6-Tetramethyl nitrene Cll) · The synthesis of the potential 1,1 - diazene precursor 1-amino-2,2 , 6,6-tetramethylpiperidine (36) is straightforward , employing a standard nitrosation 6 of the commer cially availab le amine l! followed by lithium aluminum hydride red uction under forcing conditions: 75 M H M00 34 35 NaN02 ... oq. HCI I ___,_,.... LiAJH 4 £'2_0/ BufJ M I NH2 36 A characterization of the products resulting from the oxidation of 36 was fir s t undertaken . We find that oxi - dation of 36 i n ethyl ether with yel l ow mercuric oxide affords , in addition to 37 and ~, small amounts at 25° of the isomeric hydr o ~arbons 39 and 40 i n the following approximate ra t ios: M I NH 2 HgO Et20 I~ 37 s 38 s 39 25• Q-N=~ ~ -40 -27 - A typical absolute yield of the c9 hydrocarbons wa s 65 %. In addition an absolute yield of 20% of the corres ponding tctrazene 1,1'-azo-2,2,6,6-tetramethyl~iperidine ~~was found by proton NMR examination of the reaction product s . These products are, in general, consistent with the formation and subsequent decomposition of the corresponding 1,1-diazene ~· The trimethylcyclohexane iQ is a rather surprising product in that it does not appear to be de rived from the expected 1,5 - biradical. Possible mechanisms for its formation will be discussed later. Oxidation studies aimed at generating and detecting the 1,1-diazene were undertaken. The oxidation of 1,1 - disubstituted hydrazincs with !-butyl hypochlorite in the presence of a tertiary amine is known to proceed readily at low temperatures (-78°~: and to yield products consi s tent with the intermediacy of l,l-diazenes.3 4 We find that the addition of !-butyl hypochlorite to a stirred solution of 36 and triethylamine in anhydrous diethyl ether affords, i n addition to an insoluble white precipitate (identified by infrared spectroscopy as triethylamine hydrochloride) an intense purple solution whi c h is stable for hours at - 78° but decolorizes in minutes at 0°. Hydrocarbon products ll-!Q are observed -28in 14-24% yield i n a r atio o f 24:9.7:2 . 5:1.0 re s pectively .. >0< ~ t-BuOCI E~ -78· NH 2 I II N purple 36 33 ... - 37 38 by analy t ical vapor phase ch romatography (VPC). 39 40 Genera- tion of t his so l ution at -78°, fol l owed by filtration at -78° gives a clear, t ransparent purple solution . Significantly, treatment of t he unh indered analogue of 36 , 1-aminopiperidine, with .!_-butyl hypochlorite under identical conditions leads to no such colored solution. After warming, the corresponding tetrazene was detected as a reaction product: In the following sections the characterization of this colored intermediate and further investigations are described and disc ussed . The Electronic Spectrum of N-(2,2,6,6 - Tetramethylpiperidyl)nitrene (33). ~ The visible absorption spectrum of the colored - 29- ethereal solution described in the previous section was obtained at cell. ~ -78° using a low-temperature spectroscopic The spectrum reveals a structured absorption band with two maxima at 514 and 543 nm (Figure 4), remarkably close to the n · l> lT* electronic transition pre<.licted by Davis and Goddard for the parent system H2 N-N. 6 0 The detail in this electronic spectrum appears to be vibrational structure. Although the spectrum is not sufficiently resolved that we can read th e wavelengths to any great accuracy, a crude analysis shows the spa c ings between maxima to be 1040 ± SO cm - 1 • Qualitatively the overall appearance of the spectrum is in accord with that expected for a chromophore X-Y \vho!-> e X-Y separation is slightly larger in the excited state than in the ground state. 77 The energy difference between maxima in the electronic spectrum presumably corresponds to the vibrational level spacing in the excited state of the 1,1diazene chromophore. The position of an absorption that involves nonbonding electrons (n ~ n*) is particularly sensitive to the polarity and hydrogen bonding capabilities of the solvent. We find that the visible spectrum of the 1,1- . diazene ~ is subject to a blue shift \-Ji th increase in solvent polarity analogous to solvent effects on the -30 - 0 .5 0.4 <V l) c a 0.3 ...0 L 0 (./) ...0 <( 0 .2 0.1 0 .0 400 500 600 700 Wavelength (nm) Figure 4. eth er . Visible spectrum of~ at -78° i n diethyl -31- n + n* transition of the carbonyl group.7B When !-butyl hypochlorite is added to ~ in the presence of triethylamine in dimethyl ether at -78°, filtered, concentrated, and diluted with dichloromethane, a Amax at 541 nm is observed. However, if the dimethyl ether is replaced with isopropyl alcohol, a A at 526 nm is observed, a max shift of 15 nm (1.5 kcal/mole) to shorter wavelength (see Figure 5). An interpretation consistent with this result is that the shift to higher energy for the n + n* electronic transition in isopropyl alcohol results from stabilization of the ground state in the more polar solvent. Part of the blue shift could arise from the destabilization of the Franck-Condon excited state in the hydrogen-bonding solvent. 7 9 For comparison, the n + n* transition of acetone shifts from 277 nm in chloroform to 272 nm in ethano1, 7 8 a shift of 5 nm (1.9 kcal/mole). Due to the symmetry-forbidden nature of n + n* electronic transitions the molar extinction coefficients (Emax) observed are usually·quite low. For example, c max ~ 14 for the acetone n + n* transition in the vapor phase.B 0 In accord with our assignment of the observed transition as n + n*, we find that the extinction coefficient of 33 at 543 nm in diethyl ether (Emax) equals - 32 - ,,.\ r../ I I \ '-'\ \ \ 0 \ \ 0 .0 \ \ Q) c \- \ \ 0 (J) .0 <! \ \ \ \ \ \ \ ', 0 .0 400 500 600 700 Wavelength (nm) Figure 5. Visible spectrum of 33 at -78 ° in dichloro- methane, A 541 nm (---); in isopropyl alcohol, max Amax 526 nm ( - ) . -3318 ± 3 as measured by a combination of proton NMR and electronic absorption spectroscopy (see Experimental Section). Recently the electronic spectrum of a second 1,1dialkyldiazene has been recorded in our laboraties by Schultz. 81 Oxidation of 1-amino-2,2,5,5-tetramethyl- pyrollidine CQ) with !_- butyl hypochlorite in a manner identical to that described here yields an intense red solution, which is stable for hours at -7 8° but which decolorizes rapidly at room temperature. Low temperature ' 43 - absorption spectroscopy in dichloromethane at -78° reveals a structured absorption band due to 43, quite similar to that observed for 33, but with a A at 497 nm ·, in iso -max propyl alcohol this band is shifted to Amax 487 nm. For comparison, it is found that the isoelectronic ketones 44 -34- similar in structure to the diazenes ~ and ~ (!! and 45, respectively) display a similar change in >. max with - change in ring size; in ethyl ether the n -+ n* absorption of 44 shows a >. max = 305 nm, while that of 45 shows a >.max = 296 nm. The Infrared S ectrum of nitrene i eridyl)- (~). ~ When !-butyl hypochlorite is allowed to react with 36 in dimethyl ether in the presence of triethylamine and the reaction mixture is filtered, concentrated, diluted with dichloromethane at -78° and introduced into a low temperature infrared cell at ~-78° the infrared spectrum shows a strong absorption at 1595 cm- 1 that disappears when the solution is warmed to 25° (see Figure 6a). This is suggestive of a N=N double-bond stretching frequency (see Table 3). In order to test this assignment, applica- tion of Hooke's LawBS allows an approximation of the stretching frequency change which should be observed upon synthesis of the appropriate 1,1-diazene ~· 14 N= 1 5N isotopically labelled The v(1 4 N=1 4 N)~(1 4 N=1SN) ratio calculated in this manner is 1.0171. Thus, the predicted N=N stret - ching frequency for the labelled 1,1-diazene 46 is 1568 cm- 1• - 35- ..... , ' ,--, \ I I I \ /\ \ I \ I I ,~ \."'' l I\ JI l 1595_} 1569---' v 1700 1600 I I v 1500 1700 1600 1500 Frequency ( cm- 1 ) Figure 6 . (a) R2 ll.tN=l4N; (b) R2 14N=l s N. a t - 78° after warming to 25° (---) . At - 7 8 o (--) ; -,36- Table 3 Reported N=N Stretching Frequencies Compound v (N=N) (cm- 1 ) Reference 1576 82a 1563 82b 1545 83 1575 84 H I N=N CH 5 I - 3 7- Successive treatment of 2,2,6,6-tetramethylpiperidine 34 with sodium nitrite-lSN, lithium aluminum hydride -34 and !-butyl hypochlorite/triethylamine afforded the l abel led 1,1-diazene ~, the infrared sp ectrum of which showed no absorption at 1595 but rather a new absorption a t 1569 cm-1, a shift of 26 cm-1 consistent with the ass ignment to a N=N stretch for the 1,1-diaze ne 33 (see Figure 6b). These results suggest that there is consider- able double-bond character in the 1,1 -dial kyldiaze n e N=N bond, remarkably close to a 1,2-diazene isomer. This is in c omplete accord with the calculat ions of Davis and Goddard.6° A comparison of these IR spectra to similar data rece ntly obtained on the homologous N-(2,2,5,5-tetramethylpyrollidyl)nitrene (il) is instructive.S l It was found that, in di ch loromethan e solution at -78° , the diazene 4 3 exh ibits a N=N stretching frequency at 1638 cm- 1 , and s ubstitut1on with nitrogen-IS lower s this vibration to -381612 cm- 1 ; thus the effect of 6 ~ 5 ring contra ction is an increase in v(N=N) of 43 cm-1. Interestingly, we find that in dichloromethane the ketones of similar struc t ure (.±_! and iil show shifts of the same magnitude in the C=O stretching frequency: ~N A- X+'A.N=N 33 43 ~- t::. v(N=N) = 1595 = 4 3 em- 1 ~ v( N=N) = 1638 45 ~ v(C=O) = 1690 +t::. = 40 cm- 1 ~ v(C=O) = 1730 A possible explanation of this effect in ketones has been put forward by Coulson.BG As the C-CO-C bond angle de- creases, the carbonyl carbon has greater p-orbital character in the orbitals of the ring with a consequent increase in the s-orbital character of the C-0 a-bond. This in turn increases the force constant of the carbonyl group and causes an increase in the frequency of the carbonyl stretching vibration. Perhaps a similar mechanism is operating in the 1,1-dialkyldiazenes. -39- El ectronS in ResonanceS ectra of N-(2,2,6,6-Tetramethylpjpe ridyl)nitrene (~~). The multiplicity of the 1,1-diazenc ground state anJ the singlet-triplet gap have been the subject s of a number of theoretical and indirect experimental studies. As was discussed above, both theory and experiment indicate that the ground state should be a singlet. Direct experi- mental velification of this by electron spin resonance (ESR) studies would be desirable. Based on the known zero-field parameters of alkyl- and arylnitrenes, the fjrst-derivative ESR spectrum o f a triplet 1,1-dialkyldiazene would be expected to exhibit a line in the regi on of 5000 - 8500 gauss for a microwave f requency of ~ 9 GHz~ 7 A 0.06 molar solution of 1,1-diazene 3 3 in chloroform was prepared , degasseJ and frozen at 77°K in the cavity of an ESR spectrometer . We find no features in the spectrum which can be assigned to a triplet 1,1-diazene, even when operating th e spectrometer at high sensitivity (see Figure 7). The only feature in the spectrum is an absorption at ~ 3300 G due to a monoradical species . At 77°K in the solid phase this absorption has the app ea rance of a distorted triplet (see inset , Figure 7). After warming the sample to 25° and allowing the diazene to decompose, a solution ESR spectrum was recorded at - 40- Fi gure 7. ESR spectrum of a 0.06 M solution of 33 at 77°K in chloroform, recorded at high sensitivity at 9.234 GHz. ~ 3300 gauss. Inset: expansion of the region at - 4] - 00 1 (!) 0 N _t <D ( /) (/) :::> <t (.!) l() 0 ...J - ~ 0 ...J w u: 0 -42~ 270°K. The signal remains in undiminished intensity , being a well-resolved 1:1 : 1 triplet with a= 16 gatlSS. This monoradical is most likely 2,2,6,6 - tctramethylpiperidyl-N -oxyl, based on a compari s on with its published ESR spectrum . 88 The distorted appearance at 77°K is probably due to viscosity effects.B9 nitroxide is unclear. The origin of this A control experiment demonstrated that the starting material 36 contained no paramagitetic impurities. Nuclear Magnetic Resonance Spectra of N-(2,2,6 6 - Te trame thylpiperidyl)nitrene (~). Proton nuclear magnetic resonance (NMR) spectra of the diazene 33 were obt ained in de ule rochloroform. The 1,1-diazene was first partially purified by low tempera ture (-82°) column chromatography on basic alumina. A representative spectrum (shown in Figure 8) indicat e s th e presence of a substantial amount of tetrazene 41 along with absorptions at o 1.15 and 2.15 ppm which inte- grate in a 2:1 ratio; these new absorptions are assi gned as the methyl protons and the ring protons, respectively, of the 1,1-diazene ~· Warming the sample to 25° result s in complete decolorization of the sample and the disappea rance of the signal s at 1.15 and 2.15 ppm, while the tetrazene signals increase and new signals app ea r in the -4 3- spectral region where the four hydrocarbons ~-i.Q_ absorb (Figure 9) . These NMR results indicate that under some conditions the unimolecular decomposition of 33 and the bimolecular dimerization to tetrazene 41 are competitive. Since by addition of an internal standard (CH 2 Cl 2 ) the absolute concentration of 1,1-diazene can be measured, a direct determination of its molar extinction coefficient was possible using proton NMR spectroscopy (see above). In addition the absolute yields and. mass balance of the hypochlorite oxidation reaction could be determined by the addition of dichloromethane internal standard to the crude reaction mixture. Integration of the NMR spectrum of the mixture then allowed calculation of the yields. Carbon-13 NMR spectra of 1,1-diazene solutions, obtained in dichloromethane-d 2 at -80°, exhibit four signals which may be attributed to the diazene 33 (Figure 10). Thermolysis of the sample at 25° results in the disappearance of the 1,1-diazene signals and the appearance of new absorptions due to the tetrazene i l (Figure 11). Identical experiments using hexadeuteroacetone as solvent demonstrate that no absorptions were obscured by the CD 2 Cl 2 quintet (see Figures 12 and 13). The carbon-13 chemical shifts of the diazene ~ and assignments are summarized in Figure 14, along with the chemical shifts -44- Figure 8. Proton NMR spect rum at -55° D = 1 , 1 -d iazen e, T ~ tetraz ene . of~; - 45- 0 CD/ ~ u I ul ....I ------0 \ I ..... C\J -l{) / (\J u N :I: u -46- Fi gure 9 . Spectrum after warming sample to 25° (spectrum recorded at -55°); T = tetrazene, II= hydrocarbons. 0 t\1 Q) ~, -- -=c:mave .nMJF:· ,- ( .() l() -I t\1 0 C\1 :r: (.) -48- Figure 10. Carbon-13 NMR spectrum at -80° of a CD 2 Cl 2 solution of diazene ~; D = diazene, A= !-butyl alcohol, M • dimethyl ether, N = 1-amino - 2,2,6,6-tetramethyl piperidine, E = triethylamine, T = tetrazene. -490 c_ 0 C\1 ct \ UJ z ' ' c -------== z 0 v 0 <D I ct 0 I 0 Q 0 ~ 0 v 0 !.e -50 - Figure 11. Carbon-13 NMR spectrum at -80° of a CD 2 Cl 2 solution of diazene ~ thermo1yzed at 25°; see Figure 10 for explanation of symbols. - 51 - 0 w / _ 0 N ...... <l \ w.........._ 0 v 2:./ z 0 ID :E <l / 0 CD 0 0 0 ~ 0 !'.: 0 !e -52- Figure 12. Carbon-13 NMR spectrum at -80° of a hexa - deuteroacetone solution of diazene 33; see Figure 10 for explanation of symbols. -53- '· I 0 ----~--------------------~--- 00 0 ~ 0 ~ -54- Figure 13. Carbon-13 NMR spectrum at -80° of a hexa- de ut e roacetone solution of diazene 33 thermolyzed at 25°; see Figure 10 for explanation of symbols. -55- w zI -56- 18.1 \ 17.6 /40.2 ~'27. 6 0 \ 4.4 ~in COCI 3 22 .9 \ \ / 40.5 H~ 56.8 ~,255 36 in COCI 3 30.0 AA-70.6 I \ \16.6 Me Me ~in DMSO Figure 14. Carbon-13 NMR chemical shifts and assignments for diazene 33; shifts are reported as ppm downfield from tetramethylsilane in o units. - 57- of some model compounds for comparison . Th~ carbo n chemical shifts o f 33 undoubtedly reflect the contr1butions of a variety of shielding effects , includin g sub stituent inductive effects, steric effects and electric fie ld effects.9ob Decom osition Kinetics nitrene (~). ~ To date there i s no experimental info rmati on concerning the rates and activation parameter s of the the rmal reactions of 1,1-diazenes. Since such data are important to both theorists and experimentalists, a study of th e decomposition kinetics of 1,1-diazene ~was underta ken. The diazene used for kinetic studies was purified by column chromatography at -82°, using deactivat ed basic a lumina and dimethyl ether-propane mixtures as th e eluting solvent. The chromatography and subs e quent addition of excess triethylamine we re ne.c essary to obtain reproducible kinetics. The decay kinetic s of ~ were studied in ethyl ethe r in the temperature range of -1.2° to -21.4° by mon i torin g the optical density of the purple solution l i ter) at 541 nm as a fun ction of time . c~ 10- 3 moles/ The disappear - ance o f ~ was strictly first order at higher t empe r atur e s (-1.2° to - 11.2°) becoming a combination of first and -58- higher order kinetics as the temperature was low e red (-13 . 3° to -21.4°). Plots of inAversus time at the lower temperatures exhibit a curved segment at short times followed by a linear segment at longer times (Figure 15). First order rate constants were taken to be the slopes of the linear portions of these plots; the rate constants from nine kinetic runs in ethyl ether are listed in Table 4. Table 4 First-Order Rate Constants in Ethyl Ether TemEerature (°C) - 1.2 - 3.6 - 6.2 - 9.0 -11 . 2 -13.3 -16.0 -18.9 -21.4 k 1 (sec- 1 ) 4.28 X 10- 3 3.14 X lQ-3 2.03 X 10- 3 1. 4 2 X I0-3 1. 01 x l o- 3 0 . 763'x lo-3 0.482 x Io- 3 0.324 X 10- 3 0.226 x lo- 3 These data are plotted in Arrhenius form in Figure 16, yielding the indicated activation parameters. The ob- served activation energy for the unimolecular thermal decomposition of 1,1-diazene 33 of 20.0 ± 0.4 kcal/mole is substantially lower than that observed in the decomposi- - 59 - Figure 15. R.n A versus time plots for the decomposition of 33 in ethyl ether . -60- 0 I() N 0 0 N -., c: 0 -.... LLJ z ~ 0 0 (/) !!2 "' ( /) LLJ :l ~ LLJ q rt) ' 0 0 •• • •• • •• • •• •• • •• ••• • • •• • ••• •• • •••• •• ••• • ••• •• • • 0 i 0 tri I 0 I() ., .c ct .s ~ (/) I 0 ,..: I 0 ~ .... 3.6 3.7 F1 Fl 3.8 Fl .. 3.9 [OJ R G 4.0 kcal/mole log A = 13.7 ! 0 .3 1000/ T ( •K-') [OJ Solvent Ea = 20.0 ! 0.4 Et 2 0 Arrhenius plot of diazene first order decomposition k inetics -8.0 -7.0 R in ethyl ether (R = 0.999). Figure 16. In k -6.0 -5.2 I ..... 0'1 -62 - tion of s imilarly substituted 1,2-d ialkyl dia zenes. For instance, for trans-azo - t er t - butane ~ the observed enthalpy of activation is All~ = 42 . 2 ± 0.8 kcal/mole .9 1 48 This is in accord with the theoretical results of Casewit and Goddard 42 and those of Pasta and Chipman .65 In order to asses s the effect of solvent on the unimolecular de compo sit ion rate of 33, k]netic studies wer e carried out in n-hexan e and in t etrahydrofuran. Analysis of the kinetic data in a manner identical to that e mployed in the ethyl ether kinetics affords the first order rate constants listed in Table 5 (hexane) and Table 6 (THF). The Arrhenius plots of data are shown in Figure 17 (hexane ) and Figure 18 (THF). A comparison of the unimolecular decomposition rates of 33 in the three solvents at the same temperature demons trates that the rate is sensitive to the nature of the solvent and increases with decreasing solvent polarity ( see Tab le 7). The observed r a te dep e n- dence on solvent polarity is consistent with a polar -6 3- Table 5 First-Order Rate Constant s in n-Hexane TemEerature (oC) ~ 1 . Csec- 1 ) 4.34 X 10- 3 3.39 X 10- 3 2 . 31 X 10- 3 1. 69 X 10- 3 l. 08 X 10- 3 0.752 X 10- 3 - 8.0 -10.4 -13 . 4 -16 . 0 -18.9 -22.0 Table 6 F~rst-Order Rate Constants in Tetrahydrofuran TemEerature (°C) 4.2 + 2.6 - 0.2 - 2.6 - 5.6 - 7.4 -10.4 + k 1 (sec- 1 ) 5.24 X 10- 3 4.25 X 10- 3 2.85 X l0-3 2.07 X 10- 3 l. 36 X Io-3 1.10 x lo- 3 0.683 X 10- 3 (R = 0.999). Figure 17 . -8.0 -7.0 3.7 l'l 3.8 = Fl 3.9 El 1000/ T (°K -·) El log A 1 0.7 kcal/mole Solvent l'l 4.0 11.6 ! 0.5 Ea = 16.9 ! .!!.-C 6 H l4 Arrhenius plot o f diazene first order decomposition kinetics inn-hexane In k -6.0 -5.0 I +:>- 0\ 3.6 1000/ T 3.7 (•K - •) 0 .4 3.8 13.6 ! 0 .3 20.1 ! Ea : log A : Solvent THF 3.9 keal/mole Arrhenius plot of diazene first order decomposition kinetics in -8.0 -7.0 tetrahydrofuran (R = 0.999). Figure 18. In k -6.0 -~.0 VI I a- -66- Table 7 Solven t Effects on the Decomposition of 33 at -l0°C Solvent Et k 1 (sec- 1 ) k n - Hexane Ethyl Ethe r Tctrahydrofuran 30.9 34 .6 37.4 3. 4 9 x lo- 3 1. 23 X 10-3 0.73 X 10- 3 4.8 1.7 1.0 rel b.G#(kcal/mole) 18.3 18.9 19.1 ground state decompo s ing by a less polar transition state . Such an effect has been observed in the decompositions of 1,2-dia1kyldiazenes by RUchardt and coworkers. 9 2 For the three solvents examined, tn k correlates with Et reasonably we1193 (Figure 19). slope of 0.24 . The indicated line ha s a This is substantially larger than the slopes observed in the az oalkane studies of Rfichard t . For example, they find that the unimolecular decomposition of 49 at -22 ° gives a linea.r correlation of tn k with Et with a slope= 0 . 16.92a Presumably this reflects a stronger 49 s olvent interaction with the dipolar 1,1-diazene ground state a s compared to the relatively non-p~lar 1,2-diazene ground state . 30 32 Et 34 Et 2 0 36 38 tn k versus Et for the unimolecular decomposition of 33 at -10. 0° . -7.0 -6.5 .n.-C 8 H14 The indicated line has a slope of 0.24. Figure 19 . In k -6 .0 -5.5 I I -...J 0\ -68The Arrhenius par a me ters for the three solvents a re summarized in Table 8 , along with the calculat ed enth a lpi es and entropies of activation. Table 8 Summary of Activation Parameters* Solvent Ea nSf nllf (kcal / mole)(k c al / deg - mol)(kcal/mol) l og A -7 . 3±2.8 n - Hexane 11.6 ± 0.5 16.9 ± 0.7 Ethyl Ether +2.3 ± 1 . 4 13 . 7 ± 0.3 20.0±0.4 Tetrahydrofuran 13.6 ± 0 . 3 20.1±0.4 +1.8 1: 1.4 *Error limits represent one standard deviation . 16.4 ± 0 . 7 19.5 ± 0.4 19 . 6±0. 4 One possible caus e for the curva tu re observed in the ! n A versus time plots at the lowe r temperatures studied is a dimerization of the 1 , 1-diazene 33 in competition with unimo1ecular de c omposition . On the basis of the proton NMR results described above this is a reas onable possibility. Usi ng c omputer simu lation , we find that the curved in A Hydrocarbon• ~+ 1 2[~] '-/.=N ----~ Tetrazene 33 vers u s time plots of the results in ethyl ether may be modeled as competitive unimo lecular and bimolecular re- -69- actions Ck b = k 1 + k 2 L33]). 0 s -In the first set of simulation s the values of the unimolecular rate constants k 1 were constrained to the experimental values (determined in the manner described above) within experimental error (±S i ). The bimolecular rate constants were varied until a satisfactory fit with the experimental d~ta was obtained. We find that an adequate fit may he obtained using the same bimolecular rate constant at all four temperatures that were simulated. The rate constants are given in Table 9 and the fit between experimental and calculated values is illustrated in Figure 20. Table 9 Rate Constants Used in First Simulation; see Figure 20. Temperature (°C) k 1 (sec;;. 1 ) k 2 (t/mole-sec) -13.3 -16.0 -18.9 -21.4 7.45 x Io-~· 5.03 X 10- 4 3.17 x lo- 4 2.18 X 10- 4 5.03 X l0-2 5.03 X 10- 2 5.03 X 10- 2 5.0 .3 X 10- 2 log k 1 = 13.7 - 20.0/0 -70- Figure 20 . Results of simulation of kinetics in ethyl ether, using the rate constants given in Table 9. 1.0 2,0 Axl01 0 25 50 TIME 75 • EXPERIMENTAL. -CALCULATED 100 (MINUTES) 125 150 -.....) 1-' -72A second set of simulations was performed , with the ' only constraint being that the derived rate constants k 1 and k 2 must strictly adhere to an Arrhenius relationship; that i~ !n k must be a linear function of 1/T. Both kr and k 2 were varied until an optimum fit was obtained . The r ate constants derived from this simulation are given in Table 10 along with th e Arrhenius parame ters genera t ed by these rate constants . The fit be- tween th e experimental a nd ca l culated data i s illustrated in Figure 21. Tab le 10 Rate Constants Used in Se cond Si mul ation; see Figure 21 . Temperature (° C) k 1 (scc- 1 ) k 2 (!/mole - sec) -13.3 - 16.0 -18.9 -2 1.4 7 . 25 X 10 - 4 5.03 X 10-11 3.30 X 10- 4 5 . 9 5 X 10- 2 5.16 X 10- 2 4.43 X 10 - 2 3 . 89 X 10- 2 log kl = 11.4 2.50 X - 17.4/0 Io- 4 log k2 = 4 .5 - 6.8/0 - 73- Figure 21. Results of simulation of kinetics in ethyl etl1er, using the rate constants gi ven in Table 10. .. 7 4- 0 !!2 0 Q (/) w 1- :::) z ~ ...1 0 UJ 1- ct ...1 ~ ~ z UJ ~ :::) 0: ...J Q. u w I() 1'- ~ w X 5 UJ I • 0 I() -75- A number of conclusions may be drawn from the result!? of these simulations. First, these results demonstrate that the experimental kinetic data·are consistent with and thus constitute permissive evidence for competitive unimolecular and bimolecular decomposition pathl'lays. This is the first kinetic evidence for a direct bimolecular pathway for the formation of tetrazenes from 1,1-diazenes, as opposed to the alternative indirect pathways discussed earlier (sec Introduction). Secon~ the results in- dicate that the temperature dependence of the dimerization rate is small. Although our data are not sufficient to derive a ccurate Arrhenius paran~ters, they do allow limits to b e plac e d on the activation energy. The results of the first simulation indicate an Ea ~ 0 for th e bimolecular process. The s e cond simulation yields an Ea ~ 7 kcal/mol e . The actual activation energy !or the process is probably somewhere between these values. An accurate determination of the Ea for the dimerization reaction will require kinetic studies in temperature and concentration r e gimes where the unimolecular rate is negligible compared to the bimolecular rate. As was mentioned earlier one of the C9 hydrocarbon products, 1,1,3-trimethylcyclohexane (40), does not appear to be derived from the 1,5-biradical SO. A -76 - likely s ource of this product i s the hexenyl radical 51 . Walling and Cioffari have demonstrated 95 that hydro carbons 38 and i.Q_ may arise from th e rearrangement of 51 followed by hydrogen atom abstraction. This rais es D -Q n0 D 50 51 38 39 40 the qu es tion of the or ig in of th e radical 51. A possibl e source is a monosubstituted diazene i~, an isomer of the 1,1 - di aze ne 33 . It is believed that rea ctions of mono - s ubstituted 1,2-diaz e nc s ~with oxygen proceed by hyd r o gen atom abstraction followed by deazotization to gi ve QH R· + + N1 + R-H 52 a (1,1- subst) b(l,2-subst) th e carbon-centered radica1 :96 ,9 7 0 1 + R- N=N- H _ _...,_ R- Na· + ·01 H _ ____...,_ R· + Na 53 -77- Evidence has been given for the formation of phenyl radical in the exposure of phenyldiazene (~, R "" C6 H5 ) to air 9 8 or ferric ion.9 7 Presumably the hydrocarbon radical R·, once formed, initiates a chain reaction resulting in the formation of R-H. The monosubstituted diazene ~ may conceivably arise in two ways: (a) by initial C-N bond cleavage of the 1,1-diazene 33 to the diazenyl biradical ~, followed by hydrogen atom migration to yield 52a or 52b (Figure 22a) or (b) by a direct rearrangement of 33 to 52b (Figure 22b) in a reaction which is similar to the Cope reaction undergone by amine oxides:99 To account for the observed products, other competitive pathways leading to the biradical 50 or the radical 56 are required. ~, ~, Con~.:eivably ~ and 38 may arise from ~; and !Q. may arise from g; and E, 38, and 39 may arise from 50 (see Figure 23). -78 c t q~- ... Qa· c t Q~H - c A2H (a) '==< .... 55 56 ~ 50 -52 51 N=N 33 Figure 22. Possible mechanisms for formation of ~; (2) biradical pathway; (b) direct pathway. Figure 23. 6 .,.. I 37 !56 38 _ _j 39 Sl 40 Possible precursors to the observed hydr ocarbon products li-iQ. 33 N=N q- c a c q c c 50 c I -..,J 1.0 -80- In order to gain information regarding the unimolecular decomposition pathways which the 1,1-diazene 33 follows, a study of the hydrocarbon product ratios as a function of solvent and temperature was undertak en. The 1,1-diazene 33 was generated and purified by colu1~ chromatography. Triethylamine and octane were added and the elution solvent was replaced with the solvent of interest (~-hexane, ethyl ether or tetrahydrofuran). Aliquots of the resulting solutions were thermolyzed at 0.0, -10.0, and -20 . 0° and the hydrocarbon products were analyzed by analytical VPC. The results of this study are given in Table 11. A quantitative interpretation of these results in terms of the mechanisms illustrated in Figures 22 and 23 will require a knowledge of the behavior of the radicals 51 and 56 in the three solvents of interest in the tem- perature range -20° to 0°. However, some qualitative conclusions may be drawn from these data. First, it is apparent that, in all three solvents, as the temperature is lowered the absolute yield of hydrocarbon decreases. This is consistent with the existence of a con~eting de composition pathway (e.g., tetrazene formation) which becomes increasingly competitive as the temperature is lowered and is in qualitative agreement with our kinetic - 81 Table 11 Hydrocarbon ProJuct Ratios* as a Function of Solvent and Temperature t> c: c Average Relative Yield** 40 38 39 37 n-Hexane 0.0°C -10.0°C -20.0°C 2 2 1 24 24 25 5 4 2 68 69 72 100% 85% 70% Ethyl Ether 0.0°C -10.0"C -2o.ooc 5 5 4 24 24 24 10 61 62 65 100% 77% 62% Tetrahydro f uran 10 0.0°C 11 -10.0°C -20.0°C 9 24 23 24 11 12 55 54 58 100% 87% 71% 9 7 9 * Va lues represent averages of duplication determi nations. **Relative to yields at 0.0°C equal to 100%. -82- results. Second, it can be seen that as th e solvent polarity is increased the relative proportion of trimethylcyclohexane 40 increases. At the same time the relative proportion of ol efi n~ decreases. This may simply reflect th e variation in the behavior of the radicals 51 and 56 or of the dia zenyl biradical 54 as a function of solvent. Alternatively, it may indicate a differential stabilization or destabilization of two alternate decomposition pathways of the diazene 33 with changing polarity of the medium. Finally, the results indicate that the product ratios arc not sensitive functi ons of temperature . In all three solvents a change of 20° in temperature results in va ria tion s of 4 % or less in the relative proportions of products. In most cas es the variations arc within experimental error. Thi s lack of te mperature sensitivty impli es th a t the competitive pathways leading to the various products 37-40 have very similar activation energies. As pointed out above, all four products cannot be rationalized with the intermediacy of the biradical SO alone; nor can they be exp l ained by th e intermediacy of monoradical 51 or 56 alone. The unirno1ecular decom- position of 1,1-diazene 33 must invo lve a minimum of two - 8 3- of these intermediates. For example, exclusive competi- ' tive formation of diazenes 52 and ~, which in turn lead L.O the radicals 51 and 56, can account for the ob- served products. It is not obvious, if this is the actual mechanism, why the product distributions should vary with solvent and temperature in the observed manner. Alternatively the products may be accounted for by competitive formation of the biradical SO and the 1,2diazene 52b: -84 - Again, an explanation for the observe d variati ons in product distribution with solvent and temperatur e i s not obvious. Inspection of Table 11 revea ls that the r elative a mount of tetramethylcyclopentane ~ is essentially unaffected by solvent and temperature variations, while the amount of olefin 37 is coupled to the amount of 38 and 40. Conceivably this reflects a solvent- and tempera- tur e-insensitive bifurcation of 1,1-diazene decomposition pathways. For example, the followin g scheme is consistent with the product data: A+ '-/=N .. 24% 76% ---- c -85- The se mechanisms are only speculation, however. Hope- fully a better knowled ge of the behavior of the radicals 51 and 56 as a function of solvent and tempera ture will allow more definite conclusions to be reached. Photoreactivi ty of N- ( 2 ,2,6,6-Tetramethylpiper i dyl)nitrene (~): Preliminary Results. A preliminary study of the photochemistry of the 1,1diazene 33 was performed. The photochemistry of 1,1- d ia zenes might be expected to resemble that of the i so electronic ketones 100 and the isomeric azo compounds, 101 with a-bond cleavage leading to radical and biradica1 s pec ies being the predominant mode of photodecomposition. We find that thel,l-diazene 33 is inert to radiation at 546 nm in ethyl ether in the n + n* region. How e ver, broad band irradiation of ethereal solutions of 33 at wavelengths greater th an 200 nm brings about r ea dy decomposition of the diazene as evidenced by a r ap id bleaching of the solution. VPC analysis of the photolysed solution reveals the formation of the C9 hydroc ar bons ll, ~' and 39 in 97% absolute yield (Figure 24). An upper limit of < 0.5% can be placed on the amount of trimethylcyclohexane 40 present. - Figure 24 . \ Ratio• >200nm D T<-5s•c 37 - 72 38 7 39 97 •1o Absolute Yield 21 0 QD Photochemical behavior of 1,1-diu ~ e ne 33. 33 No Reaction I 00 0\ -87- The low photoreactivity of 33 in the n + n* state is ' rather surprising in view of the thermal lability of the 1,1-diazene. Based on our kinetic studies the n + n* state has sufficient energy for C-N bond cleavage. Ap- parently other photophysical processes (e.g., internal conversion, intersystem crossing followed by quenching of the triplet state) are more efficient than the photodecomposition process. We cannot rule out the possibility that the 1,1-diazene n + n* sjnglet state is being quenched by a cosolute. Evidently, irradiation of the 1,1-diazene 33 with ultraviolet light forms an upper excited state of the 1,1-diazene, possibly a n + n* state. Our attempts to detect absorption in the ultravjolet due to the 1,1diazene 33 have been unsuccessful because of interfering cosolutes which absorb strongly in that region (for example, the tetrazene 41 exhibits a Amax ~ 252 nm, E7800 in ethyl ether). The hydrocarbon ratios observed in this photodecomposition (Figure 24) presumably reflect the behavior of the biradical 50 at the temperature of the experiment (-59° to -72°). -88Attempted Syntheses of 3,3,7,7 - Tetramethyl-1 , 2 -d i az a - 1cycloh epten e (59) . • The thermochemistry and the fragmen t ation reactions of cyclic tertiary sub s tituted 1,2-djazenes have been the subjects of intensive r esearch for over a decade . ~ N=N 57 A N=N 59 For example, the thermolysis of compound ~was employed in an ear l y stereochemical study of 1,4 - biradicals.l0 2 A systematic study of nitrogen extrusion from compounds ~ (n = 4, 5, 6) has led to a better understanding of the ground and excited state energy surfaces of cyclic azo compounds and has revealed interesting thermal and photochemical reactivity patterns.l03 For example~ (n = 5) photoextrudes nitrogen with almost unit quantum efficiency while the homologous ~ (n = 6) exhibits a quantum yield of 0.008 for nitrogen elimination. The eight-membered ring homologue ~ (n = 8) has recently been synthesized and, i nt erestingly, the isolated compound was the trans i s omer . 92b -89- The seven-membered ring 59 in thi s series of azo alkanes is a desirable compound to have availabl e . A comparison of its reactivity to the other members of the series should prove enlightening. Our interest in this compound derives from the fact that ~ is a possible ring-expansion product resulting from the 1,1 - diazene 33 . 50 -90- In addition, since 59 is a precursor for the same biradi : cal species ~as may result from 1,1-diazene ~it would be informative to compare thermal and photochemical reactivities and product distributions of the two diazenes. Unfortunately, azoalkane 59 is unknown in the chemical literature. Our attempts to synthesize this compound are described below. The attempted syntheses of 59 are based on successful syntheses of the lower homologue 58 (n = 6), all of which employ an oxidative coupling of the nitrogen atoms in 2,5-diamino-2,5-dimethylhexane (~). Thus a synthesis of the homologous diamine .§..!. is required. This compound is readily available using the sequence outlined in Figure 25, where the key step is rearrangement of the diamide 62 to the protected diamine ~. 1 04 Our first attempts at ring closure of 61 utilized the method of Greene and Gilbert. These workers were suc- cessful in oxidizing 60 to 58 (n = 6) using hydrogen peroxide-sodium tungstate.lOS Application of this re- action to the oxidation of diamine 61 resulted in a complex reaction mixture in which none of the desired azoalkane could be detected by proton NMR and gas chromatography. A variant of this method employs an excess of hydrogen peroxide and the ring-closure product is iso- Figure 25. Br Cl NH~ cr NHt-BOC - NaOMe/MeOH 63 NHt-BOC THF I I) SOCI 2 NH2 SnCI4 C02 H Synthetic route to the diamine 61. t r-- HCI/EtOH /~'oLi ~Br + ~OLi CONH2 I 1-' I 1.!) - 92- lated as theN - oxide (or th e N,N' - dioxide) which is then de oxygenat e d with hexachlorodi si lane. 1 0 5 Treatment of 58(n=6) diamine ~with excess hydro ge n peroxide in the presence of sodiumtungstateagain r es ults in a complex product mixture. Proton NMR examination of this product mixture reveals no absorption s whi ch are con sistent with th e presence of ~' the corresponding N- ox ide or th e corres ponding N,N' - dioxide. The major component of this mi x- ture was isolated and tentatively identified as the di nitroalkane 64 on the basis of spec tral evidence . + Other Products -9 3- Our second attempt at ring closure of 61 utilized the procedure of Nelsen and Bartlett for the iodine pentafluoride oxidation of amines to azoalkanes.10 6 Treatment of the diamine 61 with IF 5 formation of a non-vol atile product. re ~ ulted in the Examination of the product mixture by proton NMR and gas chromatography did not reveal the presence of the desired azoalkane. A third approach to ~ was based on the method developed by the groups of Ohmel0 7 and Tirnberlake,lOB employing the oxidation of N,N'-dialkyl sulfamides to generate az oalkanes. RN-sn--NR H ' "2 H ' Such an approach has been success- Ox R R 'N-N' \I Oa ~ R-N=N-R 502 fully employed in the synthesis of a large number of acyclic azoalkanes, but there have been no reports of the fonnation of cyclic azoalkanes by this method. In order to determine if the synthesis of cyclic azoalkanes was feasible using this method the cyclic sulfamid e 65 was synthe s ized and allowed to react with !-butyl hypochlorite and base. The azoalkane 58 (n = 6) was detected - 94- via gas chromatography t o be the sole volatile or ganic S~CI2 §Q PJ .. ~ N 02~ ,N .. ~~ t -BuOCI KOt- Bu H -65 product. ~(n • 6) Applica t ion of this identical method to 66, the sulfamide derived f ro m diamine fil. results in the formation of a large n umber of products . Examination of Wony Product• 61 . ~) t-8uOCI K0t-8u !§. 59 the product mixture by uv-yrs absorption spectroscopy allows an upper limit of~ 0.5% yield of 59 t o be estimated . It is rather s u rprising that the azoalkane 59 so stoutly resists synthesis while the lower and higher homolo goues may be readi l y prepared . An examinati on of molec ular models indicates a p ossible explanation. The six - membered ring is ab l e to adopt a con formation in wh i ch there are no severe steric interactions involving the -9 5- four methyl groups. In th e seven-membered ring incor- porating a cis-azo linkage, s uch a strain-free conformation is not po ssib le; severe tran sannular me thyl-methyl or methyl-hydrogen interactions are present in all conformations. In the eight-membered ring containing a cis-azo linkage such transannular interactions are so severe that the trans-isomer is apparently the more s table. Although there are no severe steric interactions in the seven-membered ring containing a trans-azo linkage molecul ar models indicate th at such a compound will be hi ghly strain e d. Thus azonlkane 59 is expected to be a hi ghly strained mol ecule whether it be cis- or trans-. Presumably this lar ge amount of s train relative to cis~ (n = 6) and tr~~ - ~ (n = 8) is the source of the difficulties experienced in the synthe s i s of ~There are seve ral approaches to 59 which have not yet been tried but which may well prove s uccessful . One possible route is outlined below, startingwith the known 92 b diurethane 67. Another approach would involve the electrochemical oxidation of sulfamide 66. Electro- chemical methods h ave been employed to generate strained or ganic rnolecul es .l0 9 Bauer and Wendt have recently report e d th e synthesis of acyclic azoalkanes in hi gh yield by the electroch emical oxidation of the lithium -96- __..._a~~-N-XX -~0'"~ 59 .... --..--- - salts of t h e corresponding sulfamides. 1 10 It is possible t hat employment of electrochemical methods may result in fewer side reactions in the oxidation of 66 than were observed in the chemical oxidation. liOMe/MeOH R- N- SO - N-R I 2 I H H ----.--~ -·- ~ ,. N--N ----••~ \ I R-N==N- R 502 We have been successful in generating and studying the first persistent 1,1-dialkyldiazene . The electroni c spectrum of N-(2 , 2 , 6,6-tetramethylpiperidyl)nitrene (33) exhibits a structured absorption band in the visible -97region (A max = 541 nm in dichloromethane) assigned as an n + ~* absorption on the basis of extinction coefficient and solvent shifts. The infrared spectrum of 33 ex- hibits a strong absorption at 1595 cm-1 in dichloromethane which is assigned ~s a N=N stretch on the basis of isotope shifts. The ESR spectrum of 33 demonstrates that the 1,1-diazene ground state is a singlet. The proton NMR spectrum of ~ in CDC1 3 has absorptions at o 1.15 and 2.15 ppm, assigned to the methyl protons and ring protons, respectively, of the 1,1-diazene. The carbon-13 NMR spectrum of 33 in CD 2 C1 2 shows absorptions at o 81.0, 39.7, 16.6, and 29.3 ppm, assigned to C-2, C- 3, C-4 and CH 3 , respectively, of the 1,1-diazene. A study of the decomposition kinetics of 33 in ethyl ether allows the derivation of the following activation parameters: log A= 13.7 ± 0.3, Ea = 20.0 ± 0.4 kcal/mole. Computer modeling of the kinetics in ethyl ether demonstrates that the experimental data are consistent with competitive unimolecular and bimolecular decomposition pathways. The unimolecular decomposition rate of 33 is sensitive to solvent, the rate decreasing with increasing polarity; at -10° krel(THF) = 1.0, krel(Et 2 0) = 1.7, k re 1 Cn-hexane) = 4.8. - A preliminary photochemical study -98- demonstrates that under our expe rimental conditions the n + n* state of 33 is unreact1ve. Ilowever, broad-band irradiation at A > 200 nm results in ready decomposi- tion A complete photochemical and photophysical of 33. study of the 1,1-diazene ~ still remains to be done, ]ncluding measurement s of emission, sensitization experiments and investigations of the nature of th e upper excited states of 33. Part III EXPERIMENTAL -99- A) Apparatus: Low-Temperature Spectroscopic Cells and Glassware. ~ The IR and UV-VIS cells arc designed for use in conjunction with an Air Products WMX-lA Vacuum Shroud and LC-1-110 Cyro-Tip Refrigerator.lll The windows for the vacuum shroud are sodium chloride (purchased from International Crystal Laboratories 112 ) for IR spectroscopy and Suprasil I (ground to size from blanks purchased from Amersil, Inc. 1 13) for UV-VIS spectroscopy. Samples are introduced into the cells through 18 gauge Teflon tubing (purchased from Alpha Wire Corp. 114 ) by applying suction with a syringe. The cell body of the UV -VIS spectroscopic cell is constructed from OFliC copper, with stainless steel tubing soldered to it. The body is nickel-pJated and has a thermocouple well with set screw. The cell windows are Suprasjl I and were ground to size from blanks purchased from Amersil, Inc.ll 3 The seals between the windows and the body are made with Vlton 0-rings. The cell path length is 10.0 millimeters, the volume ·is approximately four millimeters. The cell body of the IR cell is OFHC copper with stainless steel tubing solde red in place. The cell windows are cesium bromide (purchased from International -100- Two 0.5 mm lead s~acers are Crystal Laboratoriesl 1 2 ). used to give a pathlength of 1.0 mm . The spectroscopic cells were constrLtcted in the Caltech Instrument Shop. The low - temperature filtration funnel and chromatography column were co1tstructed in the Caltech Glassworking Shop. B nth ese s and Procedures Melting points were determined usi~g a Thomas-Hoover Melting Point Apparatus and are uncorrected. Infrared spectra were recorded using a Perkin-Elmer 257 Infrared Spectrophotometer. Proton nuclear magnetic resonance (NMR) spectra were obtained on a Varian Associates A-60A or EM- 390 Spectrometer . Chemical shifts are reported as parts per million downfield from tetramethylsilane in 6 units and coupling constants are in cycles per second (li z). Proton NMR data are reported in the order: chemical shift; t multiplicit~, s = singlet, d = doublet, = triplet, m = multiplet; nulliber of protons; coupling cons tant; assignment. Carbon - 13 NMR spe ctra were recorded on a Varian Associates XL-100 Spectrometer; carbon NMR chemical shifts are reported as parts per million downfield from tetramethylsilane in 6 unit s . -101- Figure 26. shroud. Low temperature spectroscopic cells and - 10 2- Fi gures 27 and 28. Shroud end plate and s hroud window. - 1 0 3- DIAMETER CIRCLE DIAMETER .125~~ FOUR HOLES ~ DRILL THRU SHROUD ENOPLATE TWO REQ'D STAINLESS STEEL DIME NSlONS IN INCHES - 104- t--H-SHROUD WINDOW 1 TWO REQ D SUPRASIL I DIMENSIONS IN INCHES -105- Figures 29, 30, 31, and 32. Construcd1on drawings for low temperature UV-VIS spectroscopic cell. -106- 60 en (!) ~ ..J ..J w ; ~~-:----0=- , ~ w z w ·---s l________J g: 0 ~ ~ 0 0 z ~ ..J ..J ~ r---- w 1~ ..J a. 0 z lAJ A "' 0 A~ ~ w 0:: 0 U) 1 .04& --+-Hfl- BOLT CIRCLE 1.375 DIA DEEP .875 • 1.015 GROOVE-- :: ORI LL THRU DRILL .188 ALL DIMENSIONS IN a TAP 0-80 INCHES ~ _ ...r·•ea FOR .050 DIA STAINLESS TUBING .270 I I -...J 1-' 0 OFHCCOPPER ONE REQ'D CELL BODY T.C. DEEP I DRILL i6 a I l t:.313 1.563 FOUR HOLES DRILL THRU a TAP 4-40 UNF VIEW A CROSS- SECTION - 1 08 - 1.375 DIA BOLT .063~~ DRILLED HOLE FOUR HOLES DRILL THRU t END PLATE TWO REQ'D STAINLESS STEEL ALL DIMENSIONS .. INCHES - 10 9- CELL WINDOW TWO REQ'O SOPR.SIL I -110 - Fi gure 33, 34, 35, and 36 . Construction drawings for low temperature IR spectroscopic cell. \ SCREW 4-40 'l 4 0P'i"'IM'' o•nw••a••• 3. ~LD~\ Oe•·•••c• IR CEL.L. GASKET LFRDNT CaBr WINODWl \_LE40 SPACER \_SPACER \ACK PLATE CtBr WIHOOW I ...... ...... ...... - 11 2 - I 4-28 UNF TUBING .0~ .._,_...J f -t 1.830 1.3!58 1.152 ct:t:=::l ll22 t _1 .197 .197 .098 -i DRILL THRU IR CELL HOLD£,_ ONE REQ'D OFHC COPPER - 11 3- Ill I II a TAP 4-40 I 1.308 BACK PLATE 1 ONE REQ D OfliC COPPER -114- .039~~ SPACER ONE LEAD REQ ' O -115- Figure 37. Low temperature chromatography column. - 1 1 6- THREE SUPPORT RODS EVACUATED JACKET .,..._ 6 .0 __.,.. 10.0 28.5 ..__ 12.0 -4---+----..._2.0 3.0 f 14.0 MEDIUM FRITTEO ~f 24/40 DIMENSIONS IN CM DISC - 117- Fi gure 38. Low temperature filtration funnel . - 11 8- l14/20\ r-0----.. 7.0 -4--+-- 6 .0 --t---i~l MEDIUM FRITTED DISC 3.5 l 14/20_/ 3.0 ! DIMENSIONS IN CM -119 Analytical vapor phase chromatography (VPC) wa s pe rformed on a Hewlett - Pa ckard 5700A Gas Chromatograpl1 equipped with flame ionization det ector and Hewl e tt Packard 3370B Digital Inte g rator; nitrogen was u se d as carrier gas. For preparative VPC a Varian Associates 920 Gas Chromatograph was employe d using helium as carrier gas. The chromatography columns used in this wor k are l is t e d in Table 12. De tector response for hydrocarbons was assumed to be 1.00 relative to n-oc tane. Quantita- tive analyses of all other compounds were corre c t e d f or de tector response. Dibutyl ether was distilled from sodium. Ethyl ether and tetrahydrofuran were dist i lled from sodium ben zo phenone ketyl. Triethylamine, diethylamine and diisopropyl- amine were distilled from barium oxide. Chloro fo rm was distilled from phosphorous pentoxide. Deuterochloroform used in obtaining diazene NMR spectra was passed through a short column of basic alumina. Acetone-d 6 and methylene chloride-d 2 were dried over 4A molecul ar sieve s . Propane and dimethyl ether used in low-temperature chromatography were dried over 4A mol ecular sieves. ! - Butyl alcohol wa s distilled from calcium hydride. Pentane and hexane were s ha ken with concentrated sul f uric acid, saturated -120Table 12 VPC Columns Column Designation Pennwalt 233 Pennwalt 223 Carbowax 400 B, a a, a UCW-98 SF-96 SF-96 Description 5' x 1/4" glass, 28 % Pennwalt 223 on 80/100 Chrom R 10' x 1/8" stainless ste e l; 28% Pennwalt 223 on 80/100 Chrom R 10' x 1/8" stainless steel; 10% Carbowax 400 on 100/120 Chrom P A/W DMCS 10' X 3/8" aluminum; 25 % a,a'-oxydipropionitrile on 60/80 Chrom P 20' x 1/8" stainless steel; 10% a,a'-oxydipropionitrile on 100/120 chrom P A/W DMCS 10' x 3/8" glass; 25% UCW - 98 on 60/80 Chrom W 10' x 1/8" stainless steel; 10% SF-96 on 100/120 Chrom P A/W 10' x 3/8" aluminum; 25% SF-96 on 45/60 Chrom A -121 sodium bicarbonate, saturated sodium chloride solution, dried over calcium chloride and distilled from cal c ium hydride or lithium aluminum l1ydride. Lead tetraacetate was placed under vacuum for 12 hrs to remove acetic acid. ! - Butyl hypochlorite was washed with 10% sodium carbona te, water, dried over calcium chloride and distilled under nitrogen. 1-Amino-2,2,6,6-tetramethylpiperidine was always purified by preparative VPC (Pennwalt 223, 180°) immediately prior to use. Electron spin ~esonance (ESR) spectra were obtained using a Varian Associates E-Line Spectrometer. Elec- tronic spectra were obtained using a Cary 14 or 14M Spectrophotometer. Mass spectra were recorded on a Dupont 24-492B Mass Spectrometer. Elemental analyses were performed at the Caltech Microanalytical Laboratory. Unless otherwise indicated, reactions were carried out under a positive pressure o~ dry nitrogen. 1-Nitroso-2,2,6,6-Tetramethylpiperidine (~). A modification of the procedure of Overberger and coworkers was used. 6 A solution .~ ~ 0.8 M in HCl was pr epared by addition of 14 ml concentrated hydrochloric acid to 200 ml water and placed in a three -necked 1 1 round-bottomed flask equipped with magnetic stirrer, -122reflux condenser, addition funnel and thermometer. To this was added slowly with coolin g 23.7 gm (0.152 mole) of distilled 2,2,6,6-tetramethylpiperidine. After dissolution of the solid the flask wa s heated to 75° and a solution of 42 . 1 g (0.61 mole) sodium n itrite in 150 ml water was added over 30 mins. Heating was continued for 96 hrs during which time a yellow oil formed. After cooling to room temperature 100 ml ethyl ether was added to the reaction mixture which was then transferred to a separatory funnel and the layers sepa rated. The aqueous layer was extracted twice with ether. The combined ethereal extracts were wa s hed with 100 ml 10 % HCl solution, 100 ml saturated sodium bicarbonate solution, 100 ml of saturated sodium chloride solution. The ethereal extract was dried (Na 2S0 4 ), concentrated, and distilled giving 25.1 g (89%) of~: bp 114 - 117° at 15 torr (1it.ll5 91-92° at 12 torr); IR (CC1 4 ): 2925 (C-H), 1455 (N=O, CH2), 1375, 1360 cm- 1 (gemdimethyl); NMR (CDC1 3 ) o 1.90-1.50 (m,6), 1.64 (s,6H, C~3) , 1.40 ppm (s,6H,CH3); UV (Et20) 395 (e:69.5), 238 nm ( e: 4500) (lit.llS 397 (e:86), 238 ( e: 4500). 1 - Amino-2,2,6 ,6 -TetJTamethylpiperidine (~) A modification of the procedure of Roberts and In gold was used. 7 5 A slurry of 5.0 g (0.13 mole) of lithium aluminum hydride in 150 ml 1:1 ethyl ether/~- -123butyl ether was stirred in a 500 ml tltree-necked roundbottomed flask equipped with addition funnel, magne tjc stirrer, thermometer and reflux condenser topped with a s till head. To this was added dropwise a solution of 12.5 g (0.071 mole) 1-nitroso-2,2,6,6 - tetramethylp iperidine (~) in 25 ml ethyl ether. After stirring for an additional 30 mins solvent was distilled until the internal temperature re ached 95°. Heating was continued for four hrs, at which time the reaction mix wa s cooled to 0° and hydrolyzed by careful addition of a large excess of water. The hydrolyzed mixture was transferred to a separatory funnel and the layers separated. The aqueous layer was extracted (2x75 ml) with ether and the combined organic extracts were extracted (3xl00 ml) with 10% HCl solution. The coniliined acid extracts were made strongly basic with 20% sodium hydroxide solution and were extracted (3xl00 ml) with ether. These ethe- real extracts were combined and extracted with 100 ml of saturated sodium chloride solution. The ethereal extracts were dried (Na 2 S0 4 ) , c oncen trated and distilled giving 9.1 g (86%) 180°): of~ (95% pure by VPC, Pennwalt 223, bp 82-85° at 20 torr (lit. 7 5 80-83° at 20-21 torr ); IR (film) 3350, 3250 (NH 2 ), 2960, 2925 (C - H), 1370, 1360 cm- 1 (gem-dimethyl); NMR (CDC1 3 ) o 2.8 -124(s,2H,N!:!_ 2 ), 1.50 (s,611, ring methylenes), 1.06 ppm ( s , 12H, - C!.:!_ 3 ) ( 1 it . 7 5 2 . 8 , 1. 4 6, 1. 0 1) . l-Amino( 1 5N)-2,2,6,6 - Tetramethylpiperidine (~). A solution of 0.66 ml concentrated HCl in 6 ml H2 0 was prepared and placed in a 25 ml round-bottomed flask , equipped with a serum cap, reflux condenser and stirrer. m~gnetic To this was added with cooling 1 .1 7 g (8.3 mmol) of 2,2,6,6-tetramethylpiperidine, followed by a solution of 1.00 g Nal 5 N0 2 (97.2 atom%, Prochem) in 5 ml of water. The resulting solution was heated to 80° and concentrated HCl was added dropwis e with stirring until the solution was slightly acid to pH paper. Heating at 85° was continued for 96 hrs, during which time a yellow oil formed . The mixture was allowed to cool, the organic product was extracted into ether, and the resulting organic layer was washed with 10% HCl so lution, saturated sodium· bicarbonate solution and saturated s odium chloride :· Sol ution and dried (Na 2 S0 4 ). This was concentrated affording 1.12 g (79%) of lnitroso(15N) - 2,2,6,6 - tetramethylpiperidine (one spot by TLC [CIIC1 3 , silica gel] with R£ identical to authentic 1n itroso-2,2 ,6,6 -te trame thylpiperid i ne). The labelled n itrosamin e (1.12 g, 6.55 mmole) was dissolved in 25 ml - 12 5of 1:1 ethyl ether/~-butyl ether and combined with 499 mg (13.1 mmol) of lithium aluminum hydride in a 50 ml round -bottomed flask equipped with stir bar, thermometer and reflux condenser topped with a still head. Solvent was distilled from the reaction fla s k until the internal temperature reached 95°; heating was continued for four hrs . The r e action mi xture was cooled and worked up in a manner similar to that used for th e unlabelled compound. The crude hydrazine was purified by a high-vacuum distillation (trap-to - trap) at room t e mperature followed by preparative VPC (Pennwalt 223, 180°) to yield 176 mg (17%) of 68: IR (film) 2960, 29 25 (C - H), 1370, 1360 cm-1 (gem-dimethyl); MS M+ = 157 (calc'd. 157); from the relative intensities of the m/e 156 and m/3 157 peaks a 97 ± 2% isotopic purity i s calculated. 1,1' - Azo-2,2,6,6-Tetramethylpiperidine (41). The method of Roberts and Ingold was used.75 A soluti on of 638 mg (4.09 mmole) of 1-amino-2,2,6,6tetramethylpiperidine (~) and 614 mg (8.41 mmole) of diethylamine in SO ml anhydrous ethyl ether was pla c ed in a 250 ml three-necked round-bottomed flask equipped with dropping funnel and magnetic stirrer. The solution was cooled to 0° and a solution of 2.05 g ! 2 (8.1 mmol e ) -126 in 100 ml ether was added dropwise with stirring until the iodine color persisted. The reaction mixture was filtered to remove precipitated dicthylammonium iodide. The solution was washed three times with SO ml wat er, once with SO ml saturated aqueous sodium chloride and dried (Na 2 S0 4 ). for one hr. ~-pentane This was concentrated under high vacuum The non-volatile residue was dissolved in and cooled to -78° which resulted in precipi- tation of a white solid . The solvent was removed and the compound was recrystallized from pentane at -78° to yield 124 mg (20%) of!!= mp 42-44° (lit.75 45-47°); IR (CHC1 3 ) 292S (C - H), 1460 (CH 2 ), 137S, 1360 cm-l (gem-dimethyl); NMR (CDC1 3 ) 6 1.26 ppm (s,24H, C~ 3 ) (lit. 7 5 1.60 (s,12H, ring protons), 1.57, 1.28); UV (Et 2 0) 287 ( £3400), 2S2 nm ( £7800). 2,6-Dimethyl-2-heptene (~). A modification of the procedure of Starr and Eastman was used.ll6 A solution of 4.31 g (28.0 mmol) of 2,2,6,6-tetramethylcyclohexanone in 400 ml pentane wa s placed in a quartz photochemical reactor , deoxygenated and cooled to 0°. The solution was photolyzed wi th a 450 watt Hanovi a lamp for 20 hrs. The re ac tion was followed by analytical VPC (SF - 9S, 140°). The photolysis mixture was concentrated , di s tilled und er re- - 127duced pre ss ure. The products were isolated by prepara- tive VPC (6,6, 65°), collecting the two major peaks of relative retention times 3.88 and 5.34 (~-pentane ~ 1.00). The earlier eluting peak was shown by proton NMR to be 1,1,3 - trimethylcyclohexane by comparison with an authentic sample (Chemsampco) : NMR (CDC1 3 ) o 1.8-1.1 (m,9H , ring protons), 0.90 (s,6H,C!:!, 3 ) , 0.80 ppm (d,3H,711z,CI-1C!:!.3) . The later el uting peak was shown by proton NMR to be identical with 2,6-dimethyl-2-heptene as reported by Starr and Eastman: 116 NMR (CDC1 3 ) o 5.14 (m,lH, C=C-!:!,), 2.0 (m,2H,C=C-CH 2 ), 1.70 (s,3H,C=C- C!:!, 3 ) , 1.60 (s,3H,C=C - C!:!, 3 ) , 1.6-1.1 (m,3H), 0.88 ppm (d,6H,6Hz,CH 3 CH-C!:!_3). Oxidation of 1-Amino-2,2,6,6-Tetramethylpiperidine . -36 with Mercuric Oxide. Characterization of Hydrocarbon Products . ~ A solution of 1.176 g · (7.54 mmo1) of 1-amino-2,2, 6,6 - tetramethy1piperidine (36) in 5 ml anhydrous ethyl ether was added with stirring to a suspension of 6.50 g (30.0 mmol) yellow me r c uric oxide in 10 ml ethyl ether at room temperatur e . After stirr1ng for five hrs t he reaction mi x ture was filtered through Celite and concentrated by distillation. The volatile products were - 1 28 -1 distilled out under high - vacuum (lo-s torr , room temp erature) and purified by preparative VPC (B,B, 65°) . Th e four C9 H18 isomers eluted in the followi n g ord er with the retention times relat ive to ethyl ether = 1 . 00: 1 ,1,3 - trimethylcyclohexane (1.71) , 1,1,2,2-tetramethyl cyclopentane (2.03), 2, 6 -d imethyl-1 - hc ptene and 2 , 6dimethyl-2 - heptene (2.44). The 1,1,3-trimethylcyclo- hexane was identified by comparison of its NMR and IR spectra and VPC retent ion time (BB , 25°) with those of authentic material (Chemsampco). The 1,1,2,2-tetra- methylcyclopentane was identified by its spectral properties: IR (CHC1 3 ) 2950, 2870 (C -H), 1460 (CH 2 ) , 1375, 1365 cm- 1 (gem-dimethyl); NMR (CDC1 3 ) 1.5 6 , ( s , 611, - C!:!_ 2 -) , 0 . 8 5 ppm ( s, 12H , C!:!_3 ) (1 it. 1 1 7 o 1. 54 , 0 .8 2); MS ,M + = 126 Cealed . 126). The 2,6-dimethyl-2- h ep tene was identified by comparison of its NMR and IR spectra and VPC retention t ime (BB, 25°) to authentic material: IR (film) 2960, 2920, 2870 (C-H), 1465, 1445 (CH 2 ), 1380 , 1365 cm- 1 (iso-propyl); NMR (see above) . The 2,6-dimethyl-1-heptene was detected as an impurity in the 2,6-dimethyl-2-heptene by its characteri stic NMR absorptio n at o = 4 . 70 ppm . The presence of the 1-isomer was confirmed using analytical VPC (B B, 25°; Carbowax 400, 25°) by coinj ection of authentic material - 129(Chemsampco). In separate expe riment s , t he total absolute yield of the C9 h ydrocarbon p rodtJ c t s was determined to be 65% by analytical VPC (SP-96 , 140°) and th e absolu t e yie ld of tetrazene 41 was det e rmin eti to be 20 % by proton NMR. Oxidation of 1-Amino-2, 2 ,6, 6 -Tetr ame t hy lp ] per idine 36 with t-Butyl Hypochlorite. General Proce dure. 3 4 Into a 25 ml round bottomed flask equipped with a magnetic stirrer and cooled to -78° was condensed 12-15 ml of dimethyl ether. To this was added l ~ amino- 2 , 2 , 6 , 6-tetramethylpiperidine ~ (typically 1-1.5 mmole s ; 150 - 250 mg) by syringe, followed by 1.0 mole-equivalent o f triethylamine. With rapid stirr i n g , 1.0 mole-equiva- l e nt of !-butyl hypochlorite was added in 10-20 tions over a five min period. por- ~i A pale purple color appeared almos t immedi a tely and was fully developed within t e n ruins. The reaction mixture was stirred at -7 8° for one hr, then transferred v ia 10 gauge Teflon tubing to a jacketed fi lter funnel precooled to -78° . The mixture was filtered under vacuum into a 25 ml three - necked flask precooled to -78 ° and equi pped with ma gn e tic stir bar, serum cap and gas Jnlet tube . further ma ni pulations performed on this filtered solution are described below. -130 - Preparation of Solutions of 1,1 - Diazene 33 for Visibl e Typically, 0.6-1.5 millimole of l - arnino-2,2,6,6tetramethylpiperidine ~was oxidized with !-butyl hypo chlorite in the manner described above. After filtra - tion the dimethyl ether solution was concentrated to ~ 2 ml and 4-6 ml of the desired solvent (ethyl ether, ~ - hex an e, dichloromethane or isopropyl alcohol) was chilled to -78° and added to the concentrated solution of 1,1-diazene by means of a stainless steel doubleended needle. After mixing, the solution was drawn into the cooled low-temperature UV -VIS cell and the spectrum recorded. Preparation of Solutions of 1,1-Diazene 33 for Infrared ~ Approximately 0.8 mmole of 1 -a mino - 2,2,6,6 - tetra methylpiperidine (~) was oxidized with !-butyl hypo- chlorite in the manner described above. After filtra- tion the sample was concentrated to ~ 2 ml and 4 ml of spectrograde dichloromethane (chilled to - 78°) wa s added vi a a double - ended stainless steel needle. was again placed under vacuum mins. c~ Dry nitrogen was admitted The flask 0.03 torr) for 30 to the flask, the sam- -131ple was drawn into the low temperature IR cell (precooled to -78°) and the spectrum recorded. The sam- ple was then removed from the cell, allowed to decolorize a t 25°, recooled to - 78°, drawn into the cell and the spectrum recordeJ. An identical procedure and scale was used for the N15 l abel led compound~Oxidation of 1-Amino-2,2,6,6-Tetramethylpipe ridine with !-Butyl Hypochlorite. (~) Determination of Hydrocarbon Yields b The general procedure above was used to oxidize 189 .9 mg (1.28 mmole) of 1-amino-2,2,6,6 -tetramethylpiperidine (36) with !_-butyl hypochlorite. Before transferring to the jacketed filter funne l 24.01 mg of n-octane and 27.47 mg of cyclooctane were added to the reaction mixture. After filtration and concentration of the reaction mixture to~ 2 ml, 2 ml of dry ethyl ether was added and the resulting solution stirred. The mixture was then allowed to warm to room temperature (decolorize), transferred to another flask and diluted to 3 ml total volume. This mixture was analyzed by analytical VPC (Carbowax 400, 25°). The results of the analysis are tabulated in Table 13. Analysis of the re- action mixture on Pennwalt 223 (170°) revealed that 67 mg (0.43 mmole) of~ remained unreacted. This gives 0 .0 74 1.98 2.19 2,6-dimethyl-1-heptene (39) ClZ) 2,6-dimethyl-2-heptene Total Absolute Yield: 0.637 0.257 . 1.65 (~) 1,1,3,3-tetramethylcyclopentane 1,1,3-trimethylcyclohexane 0 . 0242 1.000 Relative Area 1.55 1. 00 Relative Retention Time (~) n-octane Compound Table 13 14.8 % 9. 5 - 1.1 3.8 0.4 Absolute Yield (%) I I N (,.:1 1-' -133 a total hydrocarbon yield (corrected fo r unreacted s t a rting material) of 22% . In another expe riment, 236 mg (1 . SO mmole) of l.~ wa s oxid i zed with ! -butyl hypochlorite in the mann e r described above. ~ 78°, Aft e r filtration and c oncentration a t 3 . 6 ml of dry ethyl ether was added, followed by 33.9 mg o f cy clooctane. A vi sibl e spectrum of the r e - s ulting solution was recorded~ the solution had an optical density of 1.30 absorbance units at 543 nm. Th e filtrate weighed 137 mg and was identified as triethylamine hydrochloride (1.0 rnrnole, 66%) by comparison of its IR spectrum (CHC1 3 ) to authentic material. The ethereal s olution was allowed to warm to room temperature (decolorize) and the hydrocarbon yield was determined by analytical VPC (SF-96, 100° and Pennwalt 223 , 150°). The results of these analyses show a 24% abso- lute yield of hydrocarbons, and 0.38 mmoles of 1-ami no2,2,6,6-tetramethylpiperidin~ remaining. This corre- sponds to a 32% hydrocarbon yield , corrected for remaining starting material. If it i s assumed that the number o f moles of hydrocarbon is equal to the number of moles o f diazene i n the original solution , than an order - ofmagnitude estimate of the extinction coefficient can be made: - 1 34 - = A = E R.c (l.OO) 1. 30 ~ l3 t0.24)[Q.lfiTn- - o.ooro- Oxidation of 1-Amino - 2 , 2,6 , 6 - Tetra me thylpiperidine ochlorite. Yi e lds b ( ~) Dete rmination of Absolute Proton NMR S ectroscopy. Oxidation of 160 mg (1.03 mmol e) of 1-amino - 2,2,6,6 t c tramethylpiperidine (~_) cedure described above. was carried out usin g the pro- After filtration and concen - tration, 1.5 ml of deuterochloroform containing 15. 0 ~ R. dichloromethane was added and the s ample was pumped on at~ 0.03 torr at -78° for one hr. The sample was then tran s ferred to three 5 mm NMR tub es and stored in liqu i d nitrogen. Subsequent proton NMR spectro s copy at -7 0° on this crude reaction mixture resulted in det ec tion o£ NMR absorptions due to the 1,1-dia ze ne ~: a broad si n glet at o 2 . 15 and a sha.rp singlet at o 1.15 ppm (ratio o£ areas 1:2 re s pectively). Inte gration of t he NMR spectrum al lowed the c alculation of the following approximate abso lute yields from th e oxidation: !-but yl alcohol, 0.85 mmol (93 %); N-(2,2 , 6,6-t e tr amet hylpiper J dyl)nitrcne (~) , 0.20 mmole (19%); 1,1 ' -azo - 2 , 2 ,6, 6 - t e trame thylpiperidine (_!!_), 0. 0 35 mmole (6%). The spec trum also showed 0 .5 8 1~1ol es of 36 (57%) unr e acted. -1135 The estimated concentration of 1,1 - dia zene 33 is 0.1 2 moles/liter . The sample was allowed to decolori ze by warming to room temperature. The NMR spectrum of thi s thermolyzed sample indicated that all of the diaz ene wa s converted to tetrazene 11pon warming the s ample. (Inte gration showed the presence of 0.14 mmoles of tetra ze n e ~) . Preparation of Solutions of 1,1 - Diazene 33 for Carbon13 NMR S ectrosco The oxidation of 300 mg (1.92 mmole) of 1-amino-2,2, 6,6 - tetramethylpiperidine (~) was carried out using the procedure described above. After filtration at -78° , 5 . 0 g of dichloromethane - d 2 was added and the sam ple was placed under vacuum (0.03 torr) for two hrs. The sample was then transferred to a test tube precooled to - 78° and containing 150 ~i of tetramethylsilane. The sample was placed in a 12 mm NMR tube and the carbon 13 NMR spectrum recorded. Typically, spectra were re- corded at - 80°, using the following instrument para meters: 5000Hz spectral width, 0 . 80 second acqui s ition time, 5.0 second pulse delay, 55 microsecond pulse width, and accumulating 300 transients. Carbon - 13 NMR samples in he xadeuteroacetone were prepared in an identical manner, substituting 3.0 ml of acetone · d 6 for the di - -136 chloromethane-d 2 • After recording the spectrum the sam- ple was allowed to decolorize by warming to room temperature, and a spectrum of the decolorized solution wa s recorded at - 80°. Preparation of Solutions of 1,1-Diazene 33 for ESR ~· The oxidation of 122 mg (0 . 78 mmole) of 1-amino-2, 2,6,6-tetramethylpiperidine (36) was carried out using the procedure described above. After filtration and concentration at -78°, 1.5 ml of CHC1 3 was added and the sample was placed under vacuum (0.03 torr) for two hrs. A portion of this solution was placed in an ESR tube, degass ed (four freeze-pump .. thaw cycles), placed in the ESR cavity (precooled to 77°K), and the spectrum recorded. The sample was allowed to decolorize at room temperature and a solution ESR spectrum recorded at 270°K. The remainder of the solution of 33 (1. 25 ml) was diluted to 4.5 ml with ethyl ether and a visible spectrum showed the optical density of .this solution at 543 nm to be A= 0.31. From this the concen- tr a tion o f 1,1-diazene in the ESR experiment can be calculated to be: -137 4.5 X 0.31 ~ O l /l ' c = 1.25 x 18 x 1. 0- . 06 rna es 1ter A control experiment showed no ESR s j gnal in a so lution of 36 in chloroform. Low-Tern erature Column Chromate Partial Purifica- tjon of 1,1-diazene ~· Deactivated basic alumina was prepared by was hin g 150 g of Woelm Activity I neutral alumina three times with 150 ml of 12% sodium hydroxide solution and decanting. After wa s hing twa times with 150 ml distilled Nat er the alumina was washed three times with anhydrous methanol and dried for seven hours at 215°. The low - temperature chromatography column was .. , equipped with a 200 ml three-necked round-bottomed flask (equipped with serum cap and gas inlet). A me- chanica! stirrer was fitted so as to mix the solvent in the ja cket surrounding the column. The column was charged with 30 g deactivated hasic alumina and capped with a serum cap. The jacket wa s filled with aceton e and cooled to - 82° by adding liquid nitrogen and sti rrin g . The co lumn wa s wetted and rinsed throu gh with 40 ml of liquid propane (cool e d to -78°), added by me ans of a double-ended needle. The concentrated 1,1 -d iazene so lu - t io n (prepared by oxidizing ~ 200 mg 1 -amino-2,2,6,6 - -138tetramethylpiperidine (~) with !-butyl hypochlorite in the manner described above, filtering and concentrating to ~ 3 ml) was carefully placed on the column via double-ended needle and allowed to percolate onto the adsorbent . Elution was begun with 3:7 dimethyl ether/propane, us i ng a positive pressure of nitro ge n at the column head to force the solvent tltrough. When the colored band reached the bottom of the column, elution was halted by re l easing the nitrogen pressure at the column head; th e forerun was dr~~n off ihrough a double - ended needle into an evacuated 200 ml flask, cooled to -78°. Elution was resumed unt i l most of the colored band had bee n collected; this colored solution was drawn off into a 100 ml round bottomed flask equipped with serum cap, ma gnetic stirrer and nitrogen inlet and precooled to -78°. Approximately three rnl of this solution were removed and a llowed to warm to ro om temperature with evaporation of solvent. The residue was dissolved in CDC1 3 and examined by proton NMR to determine the degree of purification. By comparison of the volumes and optical densities of the colored solution before and after chromatography it was determined that only~ 40% of the 1,1-diaze ne 33 survives -139 - the chromatography conditions; dime ri zation to the tetrazene 41 apparently occurs on the alumina surface (see below). When proton NMR samples of 1,1 - d.iazene 33 wer e being prepared cyclopropane was s ubstituted for propan e , us ing identical proportions. Proton NMR samples were prepared by adding 1.5 ml of deuterochloroform containin g 15 p! of dichloromethan e to the chromatographed diaz e ne solution and placing the solution und e r vacuum (0.03 torr) .for at least two hrs to remove chromatography solvents. The resulting concentrated solution was placed in 5 mm NMR tubes and the NMR spectra were recorded at - 50°. Typically the chromatographed diazene solution contained substantial amounts of triethylamine, tetramethylaminopiperidine diazene (~). (~), tetrazene Ci!), and 1,1 - The chromatography removed ! -buty l alcohol, a large portion of unreacted tetramethylaminopiperidine ~, and trace impurities. The extent of purification was not always reproducible; on several occ asions the chroma to graphy gave complete separation of the dia zene 33 from tetrame thylaminopiperidine ~· Triethylamine could b e e liminated from the chromatographed dia zene by th e -14ausc of a.9 rather than 1.a equivalents of the tertiary amine in the oxidation reaction. It was possible on occasion to obtain solutions of diazene in which the only contaminant was the tetrazene in a mo le ratio of ~ 2:1 (tetrazene:diazene). Proton NMR spectroscopy -sao indicated that 1,1-didzene 33 was present in a concentration of ~ a.a3 mo le s/ on such a solution at liter (relative to dichloromethane internal standard present in a concentration of a.16 moles/liter; see Figure 8) . This corresponds to an absolute yield of diazene 4 %. of~ The sample was warmed to room te mpera- ture and allowed to deco1orize. analysis at ~had Subsequent proton NMR -sao indicated that ~ 6a %of the 1,1-diazene extruded nitrogen and form ed the hydrocarbon s 3a% had dimeri zed to tetrazene ~' sec lZ_- !Q_ \vhile ~ Figure 9. (There is a large uncertainty associated with these percentages, and ~hese valu es should be treated as semiquantitative; i.e.,the results indicate that, under th e conditions of the experiment, the nitro gen extrusion and dimerization pathway ~ a r e competitive . ) 141 Decomposition Ki n etics of 1,1 -Diazene 33 . Apparatu s and ~· Solutions for stuJying the decompo s it ion kineti cs of 1,1-diazene 33 were p r e pared in the following manner. Typically, ~ 200 rug of 1-amlno -2 , 2 , 6,6-tet r amethyl- piperidine (~)\vas oxidized with !_- butyl hypochlorite and the resulting so lution pu ri fied by l ow-temperature col umn c hroma tography using the procedure descri bed above . Immediately after the co lore d band elut ed from the co lumn 100 ~! of dry t riethylamine was added to re - move any adventitious ac id. This was necessary in order to obtain reproduc ib l e kinetics. After remova l of a 3 ml aliquot of the c olored so lut ion for NMR assay of purification, 10-25 ml of the freshly distill ed solvent (ethy l eth er, n-hexane or t etrahydrofu ran) , chi lled to -7 8°, was adde d by means of a double-ended needle , and the resulting solution was placed und er vacuum (0.03 torr) f or two hours at -78° to remove the chromatography so lvents . Th e solution was th en transfe rr ed to a SO ml centTifuge tube and stored in liquid nitrogen until use. The l ow - temperature spectroscopic ce ll was modified in the fol lowin g ways. The coolant wel l was capped with a fitti ng which allowed the circulation of cold l iquids through the well (see Figure 39). This fitting consisted -142- COPPER TUBING COOLANT WELL I I I I I I I I I r':...~--r_, I I / - ... CELL I ') : I \ 1 I I .... ..... I '-----" Figure 39. Low temperature spectroscopic cell with modifications for kinetic studjes. - 14 3 - of a two - ho l e rubber s topper and two l en gth s of 1/4" copper tubin g , and wa s connect ed vla l atex tubing to a Little Giant Model 1 immersion pump 1vhi c h provided rap id circulation of co ld l iq uid s th ro u gh the coolant well. The outer cell windows were wrapped with heatin g tape in order t o prevent water condensatJon durin g extended kinetic run s. Two different constant-low tempe rature ba th s were used. For k inetic studies ahove - 1 7°, a Forma Sci- entific Mode l 2095 Refrigerated Bath was use d, emp lo ying e thylene glyco l - wat e r as t he bath l iquid above 0° and methanol as th e bath liquid below 0°. For kin etic s tudie s below - 17° a constant-low temperature bath constructed in these l aboratories (design bas ed on that o f Gunn 1 18) was employe d (see Figure 40). Basically th e bath utilizes a re servoir of l iq uid nitrogen, in th erma l contac t with the b ath liquid, as a re f r igerator. Temperature r eg ul ation i s effected by u se of an immersion h eat e r connec ted to a Ye llow Sp rin gs Instr uments Mode l 63RC Thermoregulator and Model 63 3 Probe. Methanol was employed as th e bath liquid. Temperatures were determined by means of an iron con s t antan thermocouple imbedd e d i n the body of the spec tro sc opic cell and connecte d to a Fluke Mod el 803B/ - 144 - STIR MOTOR Cu V ROO / ----·" ........... ~THERMISTOR PROBE PUMP HEATER CELL THERMOREGULATOR Fi gure 40 . Constant low-temperature circulat i n g bath . -145- AG differential voltmeter which allowed the determination of temperature to the nearest 0.2°. A typical kinetics run was carried out in the following manner: the spectroscopic cell was flushed at least four times and filled with freshly distilled solvent. The circulating pump was immersed in the cold constant-temperature bath and the cell allowed to stand until a constant temperature was reached. A 30 ml syringe was filled with dry nitrogen and was used to expel solvent from the cell and flush it with nitrogen . The previously prepared solution of 1,1-diazene ~ was carefully drawn into the cell. The cell was sealed and placed in the spectrophotometer cell compartment and the spectrophotometer was actuated, recording the absorbance at 541 nm as a function of time. The cell temperature was monitored continuously until a constant equilibrium temperature was established periodically thereafter. c~ 4 mins) and All data obtained prior to temperature equilibration were discarded. Typically the reaction was followed through ten half-lives; the data were analyzed by plotting the llogarithm of the absorbance versus time. ter In cases where spectrophotome- stability was the limiting factor, kinetics were followed over three half-lives and the data lvere analyzed using the method of Guggenheim.119a In all -146- case s the reported rate constants are der]ved by conven tional linear least-square analysis. are estimated to be accurate to ± The rate con s tants 5% or better, based on repetitive determinations . In order to confirm that a truly first-order proces s was being observed (rather than a pseudo- first-ord e r process) the following control experiment was carried out in each solvent used in this kinetic study : A por- tion of the stock solution of 1,1-diazene li was removed and the rate constant for diazene decomposition was determined . Another portion of the stock solution was removed and diluted with solvent to a volume ~ 1.5 t]mes the original volume . The rate constant for dia- zene decomposition was determined for this dilut ed sol ution and compared to that for the undiluted solution; in all cases the values \vere identical within experimental error (± 3% or better) . In each solvent the diazene decomposition kinetics were determined using at least two and usually three ind ependently prepared stock solutions of 1,1-dia ze n c 33 . There were no differences observed between the different stock solutions. -147 Determination of the Extinction Coefficient of 1,1Diazene 33. ~ The oxidation of 239 mg (1.53 mmol e ) of 1 -amino2,2,6,6-tetramethylpiperidine (36) with ~-butyl hypo chlorite was carried out using the general procedure descr ibed above, but substituting trime thylamine for triethylamine. Low-temperature chromatography was carr i ed out as above, using dimethyl ether/cyclopropane as the solvent system. Deuterochloroform (3 ml) was added to the chromatographed diazene s olution and the solution was placed under vacuum (0 . 03 torr) for two hours to remove chromatography solvents. The resulting concentrated solution was transferred to a graduat e d t es t tube at -78°. To 2.60 ml of this s olution was added 25.4 mg of dichloromethane. Proton NMR analysis of th is solution at - 50° showed the concentration of 1,1 - diazene ~to be 0.017 ± 0.002 moles/l i ter. To the remaining 1.40 ml of this solution was added su fficie nt dry ethyl ether to give a final volume o f 4.5 ml. A visible absorp tion , spectrum of this solu - tion gave an absorbance at 543 nm of 0.096 ± 0.005 absorbance units . . From this the molar extinction _ coefficie nt a t 543 nm may be ca l culat ed: -148- e: A 0.096 = i.C = ------------~~~-- = 18 ± 3 c1. o) co. ol7) egg) Photochemical Study of 1,1-Diazene ~· The oxidation of 1-amino-2,2,6,6 -te tramethylpip eridine (36) was carried out as described above and the resulting solution of 1,1-diazene ~was purified by low - t emperature column chromatography. To the solu- tion of chromatographed 1,1-dia ze ne was added 4.2 ml of dry ethyl ether and 8.3 mg of n - oc tane. An elec- tronic spectrum of the resulting stock solution was r ec orded, indicating an optical density of 0.38 at 543 nm; this corresponds to a concentration of 1,1 diazene 33 of 2.2 x 10- 2 mole s /liter. A 2.2 ml por- tion of the stock solution wa s dilut e d to 4.0 ml with ether (O.D. at 543 nm = 0.20) and irradiated with a 1000 watt focused beam xenon arc lamp operating at 760 watts, filtered through a neodymium nitrate/cupric chloride s olution filterl20 which passed 31% of the li ght at 546 nm (filter bandwidth at half-maximum 20 nm); the sample temperature never exceeded - 67°. After 180 mins irradiation time the O.D. of the solution at 543 nm wa s 0.165 (an 18% decrease). The sample was the n allowed to stand in the dark at ~ -72 °; after -149180 mins the OD at 543 nm was 0.145 (a 12% decrease). The remaining 2.0 ml of stock solution was irradiated with the same xenon lamp, filtered only through 7 . 5 em of distilled water. The sample tempera- ture never exceeded -59° . After 60 mins the sample was completely bleached. The sample was removed and was analyzed by VPC (Carbowax 400 , 25°). The results of this analysis are given in Figure 24. An effort was made to detect by analytical VPC (SF-96 , 80°) any 3,3 , 7,7-tetramethy1-1,2-diaza-lcycloheptene ~ (the 1,2-diazene isomeric to 1,1diazene ~) which could have been formed in the photo decomposition. No peaks consistent lvith the presence of 59 were found, using the retention time of the homo logue~ (n = 6) as an estimated retention time of 59. Computer Simulation of N-(2,2,6 6-Tetrameth nitrene (il) Decomposition . Kinetics. The MSIM4 Stochastic Mechanism Simulator developed by Houle and Bunker was used for this simulation. 94 Approximate values of the rate constants were obtained from the experimental kinetic data. These values were refined by trial and error until a satisfactory fit of the simulation to the experimental data was obtained. - 150 Due to the uncertainties in the experimental da t a , th e rate con s tants derived from th es e simul a tions have estimated uncertajnti e s on the ord e r of 20-30 %. Th e results are given in Tables 9 and 10; the fit s are shown in Figure 20 and 21 . Thermal Decomposi tion of 1,1-Diazene ll · Determin a t1on of the Effect of Solvent and Temperature on Hydrocarbon Product Ratios. ~ The oxidation of 1.4-2.1 mmoles of 1 - amino-2,2,6,6 tetramethylpiperidine (~) procedure described above . was carried out using the After purjf]cation by low - temperature column chromatography, SO ~t of triethyl- amine was added along with 2-3 ml of the solvent of ' interest (~ - hexane, ethyl ether or tetrahydrofuran) and 10.0 pt of n-octane as internal standard. The solu- tion of 1,1 - diazene ll was placed und e r vacuum (0.03 torr) for two hrs to r e move cHromatography solvents. Th e resulting solution was placed in six 5 mm x 100 mm bR s e -washed Pyrex tubes (capped w]th serum caps, thorou ghly flushed with dry nitrogen and precool ed to -78°) . The samples were then thermoly zed at the appropriate temperatur e (0 . 0°, - 10.0° or -20.0°) u s ing an ice/water or a Forma bath; the samples we re -151- allowed to completely decolorize. After thermoly sis the samples were subjected to analytical VPC ( 68 , 25°). The results are tabulated in Table 11. These results are the average values resulting from two samples at each temperature. Each s ample was analyzed twice to ensure reproducibility. 2,2,6,6-Tetramethylheptanedioic Acid (~) . The method ~f Creger was used.l0 4 a In a 1 1 three-necked round-bottomed flask equipped with drop ping funnel, reflux condenser anrl ma gne tic stirre r was placed 310 ml (0.50 moles) of a 1.60 M solution of n butyllithium in hexane . To this was added with cool- ing and stirring 50.6 g (0.50 mole) of diisopropylamine in 80 ml dry tetrahydrofuran . The cooling bath was removed and stirring was continued for 30 mins. After recooling to 0° , 22.0 g (0.25 mole) of isobutyric acid was added via syringe with stirring. The mixture was warmed to room temperature and stirred until the so lid dissolved. To this was added, with cooling, 25 . 2 g (0.125 mole) 1,3-dibromopropane. After stirring at room temperature for 20 minutes, an excess of water was added, followed by the addition of 10 % HCl until the aqueous layer was strongly acidic. The layer s -152were separated and the solve nt removed under vacuum from the organic layer. The aqueous layer was extracted twice with dichloromethane and these extracts were com bined with the residue from the organic layer. The combined extracts were washed three times with 10% sodium hydroxide solution. These basic extracts were acidjfied with c oncentrated hydrochloric acid, resulting in the formation of a \.,rhite precipitate. soljd was extracted into dichloromethane. The The solu- tion was washed with saturated sodium chloride solution and dried (Na 2 S0 4 ) . The solution was filtered and concentrated. Petroleum ether was added to the residue to precipitate, in three crops, 13.2 g (48%) of 69 as co 1or 1 e s s cry s t a 1 s : mp . 16 7 - 16 9 ° ( 1 i t . 1 2 1 16 8 - 16 9 . 5 ° ) ; IR (CHCl3) 3600-2300 (C-H,COO-H), 1700 (C=O), 1475 (CH 2 ), 1408 (C-0-H bend), 1385, 1365 (gem-dimethyl) 935 cm- 1 (0-H bend); NMR (CDC1 3 ) 6 11.5 (broad,2H,C0 2 H), 1.45 (m,6H,-CH 2 ), 1.16 ppm (s,l2H,-C~ 3 ). 2,2 ,6,6 - Tetramethylheptanediamide (~). Thionyl chloride (SO ml) and 12.20 g (0.056 mole) of 2 , 2,6,6-heptan edioic acid heated at reflux for two hrs. (~) were combined and Th u thionyl chlorjde was removed under aspirator pressure and the crude - 153dia ci d chloride was added to 50 m1 anhydrous ammonia at -78° dropwise with stirring. The cooling bath was removed and the ammonia allowed to evaporate. The residue was washed twice with 10% sodium hydro xide so lution, and the remaining solid wa s recrystallized from absolute ethanol to yield 9 . 35 g (78%) co 1 or 1 e s s cry s t a 1 s : mp 19 4 - 19 5 o ( 1i t . 1 2 1 of~ as 19 3 - 19 4 o ) ; I R: ( n u j o 1) 3 3 9 5 , 319 5 ( NH 2 ) , 16 5 5 ( C=0) , 16 2 5 c m- 1 (COHN 2 ); NMR (CD 3 0D) 6 1.45 (m,6H,-CH 2 ) , 1.15 ppm (s, 1 21-1 ,C!:!_3). 2,6-Dimethyl-2,6-bis(!_- butoxycarbonylamino)heptane (§l_). The method of Baumgarten and coworkers was em ployed. 104 b A 500 ml three-necked round bottomed flask was equipped with a magnetic stirrer and reflux condenser . In it was placed 250 ml of !_-butyl alcohol and 9.35 g (0.044 mole) of 2,2,6,6-tetramethylheptane diamide (~). The mixture was heated at reflux until dissolution occurred. The heat was removed and 1.0 ml of tin tetrachloride was added. Lead tetraacetate (39.0 g, 0.088 mole) was added in one portion with s tirring. for 17 hr s. Heating at reflux with stirring was continued The !_-butyl alcohol was removed under vacuum and the residue was dis so lved in ethyl ether, - 1 54 ·· decanted and washed three times with 10% s odium c a rb o~ nate solution, once wjth saturated sodium chloride s olution and dr ied (Na 2 SO~). This was concentrated afford - ing a white solid which was recrystallized from hept ane to y i e 1 d 12 . 04 g ( 76%) of ~: 2970 (C-H~1 , 1700 (C=O), 1390, 1365 crn - 1 (gem - dimethyl); (s,2H,N - ~), NMR (CDC1 3) 6 4 . 35 1.44 (s,18H,O-C(C~ 3 ) 3 ), Anal. N, 7.81. I R ( CH C1 3) 34 ¢0 ( N- H) , 1.75 - 1.35 1.25 ppm (rn,6H,-C~ 2 -), (s ,l 2H,C~ 3 ). Calcd. for C19H3aN204: C, 63.65; H, 10.68; Found: C, 63 .60; H, 10.66; N, 7 . 90 . 2, 6- Diamino- 2, 6- Dimethyl h eptane t Dihyd rochlor ide (2.Q_) . The procedure of Baumgarten and coworkers was employed.104b A solution of 12.04 g (0.0336 mole) of 2 ,6-dimethyl -2 ,6 -bis(_!-but oxycarbonylarnino)heptane (~) in 750 ml ethyl alcohol was prepared and cooled to 0° . Hydroge n chloride gas was bubbled through the solution for two hrs. The ethanol \vas ·removed under vacuum and the remaining solid was dissolved in the minimum amount of ethanol and precipitated with ethyl ether to yield 6.86 g (88%) of~ as white crystals: mp > 270°; IR (nujol): 3240 (N - H), 2400, 2240 cm - 1 (C - NH 3+); NMR (CD 3 0D) 6 1.61 (m,6H,-C!i 2 ), 1.35 ppm (s,12H,-C!:!_ 3 ). An analytical sample was recrystallized from ethanol -155acidified with aqueous hydrochloTic acid. Anal. Calcd. for C9 H2 ,,N 2 Cl 2 ·11 2 0: H, 10.52; N, 11.24. C, 43.37; Found: •.C, 43 . 17; 11, 10.44; N, 11.25. 2,6-Diamino-2,6-Dimethylheptane (61). A solution of 992 mg (4.29 mmole) of 2,6 - diamino 2, 6- dimethyl heptane di hydrochloride (_70) in 5 ml anh~drous methanol was placed in a 25 ml round bottomed flask equipped with sidearm and magnetic stirrer. To this was added with cooling and stirring 556 mg (10.3 mmole) of sodium methox ide in 10 ml methanol. The methanol was removed under vacuum and the res]due •was stirred twice with 10 ml portions of dichloromethane. The dichloromethane extracts were filtered under n]trogen and the solvent was removed under vacuum, leaving a yellow oil which was purified by preparative VPC (Pennwalt 223, 175°) to yield 444 mg (66%) of 61 as a co 1 or 1 e s s o i 1 : I R ( f i 1m) 3 3 4 0 , 3 2 6 0 ( NH 2 ) 29 5 0 ( C- H) , 1595 (N-H), 1465 (CH 2 ), 1380, 1360 (gem - dimethyl), 1205 (C-N), 855 cm- 1 (N-H); NMR (CDC1 3 ) o 1.32 (s,6H, -CH2 ), 1. 20 (s ,4H,NH 2 ) , 1.10 ppm (s,l2H,C_'i 3 ). Attempt ed mass spectrometry gave a parent peak at m/e = 159 (calcd. 158), apparently arising from ion-molecule reaction s -156 within the spectrometer. Th~ sample was quite sensitive to exposure to the atmosphere, becomin g cloudy within a few minutes of even brief exposure. Oxidation of 2,6-Diamino-2,6-Dimethylheptane H (~) with Peroxide. The method of Greene and Gilbert was employed.1os A solution of 6.86 g (0.0297 moles) of 2,6-diamino-2,6dimethylheptane dihydrochloride (lQ) was dissolved in 75 ml anhydrous methanol and cooled to 0°. A solution of 3 . 53 g (0 . 0653 mole) sodium methoxide in 10 ml methanol was added dropwise with stirring . The methanol was removed and the residue washed twice with ether. The ethereal extracts were filtered under nitrogen and concentrated and the residue was dissolved in 120 ml de gassed water. Sodium tungstate dihydrate (0.50 g, 0.0018 mole) was added. To this was added dropwis e with stirring 13.6 g (0.12 mole) 30% hydrogen peroxide. Stirring was continued for one hr, during which a blue solid formed. The mixture was extracted with chloroform, the organic layer washed with JO% hydrochloric acid, water and dried (MgS0 4 ). Proton NMR analysis of the residue showed a complex product mixture. None o f the absorptions were consi s tent with the presence of the de- -1 57sired 3,3,7,7-tetramethyl - 1,2-diazacycloheptene Noxide, the N,N'-dioxide or the corresponding azo com pound 59 (using the six - membered rJng analogues as models) . 10 5 The major product could be isolated (by the addition of heptane to the crude product mixture and filtration) as a solid, mp 65-67°, which was tentatively identified as 2,6-dinitro-2,6 - dimethylheptane (~) on the basis of spectral evidence: IR (CHC1 3 ) 2950 (C - H), 1535 (N0 2 ) , 1465, 1395, 1370, 1350 cm- 1 • NMR (CDC1 3 ) 6 1.88 (m,6H,-C!:!_ 2 ) and 1.55 ppm (s ,l 211 , Ct!_3) • In another experiment, 372 mg (1.62 mmoles) of 70 was neutralized using sodium rnethoxide as above. The diamine 61 was dissolved in 3 ml of degassed di stilled • water containing 60 rug (0.2 mmole) of sodium tungstate dihydrate. oxide in water (335 over 60 mins. A 30% solution of hydro gen per~t, 3.3 mmoles) was added dropwise After stirring for another 20 mins the mixture was extracted with dichloromethane and the organic layer was washed with 10% hydrochloric acid, saturated sodium bicarbonate solution, saturated sodium chloride solution, and dried (Na 2 S0 4 ). After filtra- tion and removal of solvent the residue was subjected to high vacuum (l0- 5 torr) distillation at room - !58temperature. Examination of both the volatile and non- volatile fractions by proton NMR and VPC (UCW-98, 115°) revealed no absorptions or peaks consistent with the presence of the desired azo compound ~· Oxidation of Z,6~Diamino-Z,6-Dimethylheptane (~) with Iodine Pentafluoride. A modification of the method of Nelsen and Bartlett was used. 10 6 A solution of 88 ~2 (O.Z9 g, 1.3 mmole) of iodine pentafluoride and 3.0 ml of dry pyridine in 10 ml dry dichloromethane was prepared in a Z5 ml .round bottomed flask equipped with magnetic stirret and addition funnel. ring at To this was added dropwise with stir- -zoo a solution of zzg mg (1.45 mmole) of Z,6- diamino-Z,6-dimethylheptane (£!) ~urified by preparative VPC; Pennwalt ZZ3, 175°) in 4.0 ml dichloromethane. The mixture was stirred for one hr at by one hr at 0°. -zoo followed Water was then added, the mixture was transferred to a separatory funnel and the layers separated. The organic layer was wa~hed twice with 10 % hydrochloric acid, twice with 10% sodium thiosulfate, once with saturated sodium · b:icarbonate•, • once · with sa turated sodium chloride solution and dried (Na2S04). This was concentrated and distilled (lo-s torr) at room temperature. Proton NMR examination of the vola- -159tile fraction showed nothing but traces of dichloromethane. Proton NMR e xamination (CDC1 3 ) of the non- volatile residue showed a multiplet centered at 6 1.55 and three overlapping singlets centered at 1.05 ppm. Examination of the non-volatile residue by VPC (UCW-98, 115°) showed no peaks in the region expected for 3,3,7,7-tetramethyl-1,2-diaza-1-cyclo heptene member~d (~_) (based on the retention time of the six- ring analogue). An infrared spectrum (CHC1 3 ) of the non-volatile residue exhibited absorptions at 2960, !465, 1375, and 1355 cm- 1 • 3,3,6,6-Tetramethyl-2,7-Diaza-1-thiacycloheptane-l,lDioxide (~) . A modification of the procedure of Ohme and coworkers was employed.l0 7 A 500 ml three-necked round- bottomed f lask was equipped with a reflux condenser, addition funnel and magnetic stirrer and charged with 200 ml of dry chloroform. Sulfuryl chloride (0.29 ml, 3.5 mmole) was added , the flask was cooled to 0° and 1 . 1 g (14 mrnole) o f dry pyridine was added dropwise with stirring; the flask was then briefly warmed to room temperature. After cooling to - 50°, 0.50 g (3.5 - 160mmole) of 2,5-diamino-2,5-dimethylhexane (60) in 10 ml chloroform was added. The reaction mixture was allowed to slowly warm to room temp e rature and stirred for 19 hrs. After one hr at reflux the flask was cooled and an excess of water was added. The layers were sepa- rated and the organic layer was washed twice with 5% hydrochloric acid, once with saturated sodium bicarbonate solution, once with saturated sodium .chloride solution and dried (Na 2 S0 4 ). This was concentrated and ch romatographed on 30 g Activity I silica ge~ , eluting first with 150 ml chloroform, followed by 200 ml an· hydrous methanol. The · rnethanol fraction was concentrated and the residue sublimed (85° at 0.02 torr) to yield 0 . 134 g (19%) of~ as colorless crystals: IR (CHCl3) 3380, 3260 (N - H), 2970, 2920 mp 189 - 190°; (C-H ~ , 1415, 1380, 1365 (gem-dimethyl), 1325, 1160, 1130 crn- 1 ( >S 0 2 ) ; NMR (CDC1 3) o 4 . 35 (s,2H,N-H), 1.88 ppm (s,l2H,C~ 3 ); cal san~le Anal. N, 13.58. (s,4H,-C~ 2 - ) , MS,M+ = 206 Cealed. 206). 1.30 An analyti- was recrystallized from toluene. Calcd. for C8 H18 N2 S0 2 : Found: C, 46.58; H, 8.79; C, 46.90; H, 8 . 66; N, 13.75. -1613,3,6,6-Tetramethyl-1,2 - Diaza-l-Cyc1ohexene n ; 6) . (~, The method of Timberlake and coworkers was employed.1oa A solution of 92 mg (0 . 44 mole) of 3 , 3,6,6- tetramethyl-2,7-diaza-l-thiacycloheptane · 1,1-dioxide (~) in 10 ra1 dry !_-butyl alcohol was placed in a 25 ml round bottomed flask . To this was added 101 mg (0.90 mmol) of potassium _!-butoxide, and the mixture was stirred briefly. To this was added 98 mg (0 . 90 mmole) of !_-butyl hypochlorite. The mixture was stirred for 12 hrs at room temperature. After removal of the reaction solvent under vacuum the residue was dissolved in dichloromethane and subjected to a high vacuum (10- 5 torr) distillation at room temperature. Analysis by VPC (UCW-98, 125°) indicated the 1,2diazene 58 (n ; 6) as the sole volatile organic product. Dioxide (66). ~ This compound was prepared in a matter identical to that described above for 65. From 930 mg (5.88 mmoles) of 2,6-dimethyl -2 ,6-diaminoheptane (61_) (p ur~ :- . fied by VPC; Pennwalt 223, 175°) was prepared 213 mg (16.5% yield) of~ as white crystals: sublimes at 95° at 0.025 torr; mp. 225-226°; IR (CHC1 3 ) 3380, 3270 -162(N-H), 2950 (C-H), 1450, 1400, 1380, 1365 (gem-dimethyl), 1305, 1135 (>S02); NMR (CDCl3) o 4.40 (s,2H,N - t!), 2.01.6 (m,6H,-C!:!_ 2 ), 1.30 ppm (s,l211,C!:!_ 3 ); MS,M+ = 220 Cealed. M+ = 220). Anal. Calcd. for C9 H20 N2 S0 2 : N, 1 2 . 71 . Found : C, 49.06; H, 9.15; C, 4 8 . 74 ; H, 8 . 89 ; N, 1 2 . 7 7 . ~ octane-1,1-Dioxide (66) with t-Butyl Hypochlorite . • A solution of 218 mg (1.94 mmoles) of potassium t-butoxide in 15 ml dry !-butyl alcohol was placed in a 50 ml round bottomed flask. To this was added 213 mg (0.97 mmole) of 66 as a suspension in 10 ml !-butyl alcohol. After stirring at 30° for 30 mins, 211 mg (1.94 mmole) of t-butyl hypochlorite was added slowly; stirring at 30° was continued for 23 hrs. concentrated at 35°. This was The residue was subjected to distillation at room temperature (lo-s torr). The volatile fraction was examined by VPC (UCW-98, 125°) and was a complex mixture of at least 11 compounds. Examination of the volatile fraction by visible absorption spectroscopy showed no distinctive features in the region around 380 nm (azo n + n~ region). Examina- tion by proton NMR showed that the bulk of the nonvolatile fraction was unreacted starting mat*rial, with -163- traces of other compo unds present . -164REFERENCES 1. Alder, R. W. ; Baker, R. ; Brown, J. M. "Meehan ism in Organic Chemistry," Wiley, New York, N.Y., 1971. 2. Isaacs, N.S. "Reactive Intermediates in Organic Chemistry," Wiley, New York, N.Y., 1974. 3. For reviews of 1, 1-dia zenes, see: (a) Lemal, D.M. in "Nitrenes," Lwowski, W., ed., Interscience, New York, N.Y., 1970, Chapter 10; (b) Ioffe, B .V.; Kuznetsov, M.A. ~uss. Chern. Rev., 1972, 41, 131; (c) Lwowski, W. in "Reactive Intermediates," Jones, M, and Moss, R., eds., Wiley, New York, N.Y., 1978, Chapter 6. 4. (a) Wasserman, E.; Smolinsky, G.; Yager, W.A. J. Am. Chern. Soc. 1964, 86, 3166; (b) Smolinsky, G.; Wasserman, E.; Yager, W.A. J. Am. Chern. Soc. 1962, 84, 3220; Smolinsky, G.; Snyder, L.C.; Wasserman, E. Rev. Mod. Phys. 1963, 35, 576; Wasserman, E.; Murray, R.W.; Yager, W.A.; Trozzolo, A.M.; Smolinsky, G. J. Am. Chern. Soc. 1967, 89, 5076; (c) Wasserman, E.; Barash, L.; Yager, W.A. J. Am. Chern. Soc. 1965, 87, 2075. 5. Michaelis, A.; Luxemburg, K. Chern. Ber. 1893, 26, 2174. -165 6. For example, Overberger, C.G.; Lombardino, J . G.; Hiskey, R.G., J. Am. Chern . Soc. 7. _!_Q~, 79,6430. For example, Atkinson,R.S . ; Rees, C.W. Chern . Cornrn. 1967, 1230. 8. For example, Urry, W.H.; Kruse, H.W.; McBride, W. R. J. Am. Chern. Soc. 19 57, 9. 79, 6 56 8. For example, Bhatnagar, I.; George, M.V. J. Org. Chern. 1968, 33, 2407. 10. For example, Baumgarten, H.E .; Wittman, W.F.; Lehmann, G.J. J. Heteroc. Chem . 1969, · 6, 333. 11. Koch, K.-H.; Fahr, E. Angew. Chern . I nt. Ed . Engl. 19 70, 9, 6 34. 12. McBride, W.; Kruse, H.W. J. Am. Chern. Soc. 1957, 79, 572. 13. Baumgardner, C.L.; Martin, K.J.; Freeman, J.P. J. Am.Chern. Soc. 1963, 85, 97 . 14. (a) Anderson, D.J.; Gi~christ, T.L; Harwell, D.C.; Rees, C.W.; Yelland, M. Chern: l.omm. 1969, 146; (b) Rees, C.W.; Yelland, M. Chern. Cornm. 1969, 377. 15. For example: (a) Carpino, L.A . J. Am. Chem. Soc. 1957, 79 , 4427; (b) Lernal, D.M.; Ra ve, T.W.; McGregor, S.D. J. Am. Chelll. Soc. 1964, 86, 2395. -16616. Dervan, P.B.; Uyehara, T. J. Am. Chern. Soc . 1979 , 101, 2076. 17. Overberger, C.G.; Lombardinoi J.G.; Hiskey, R.G. J. Org. Chern. 1957, 22, 858. 18. Overberger, C.G.; Lombardino, J.G.; Hiskey, R.G. J. Am. Chern. Soc. 1958, 80, 3009. 19. Cadogan, J.I.G.; Thomson, J.B . Chern. Comm. 1969, 770. 20. Lemal, D.M .; Rave, T.W. J. Am. Chern. Soc. 1965, 8 7 J 39 7 . 21. Rave, T.W., Ph.D. Thesis,University of Wisconsin, Madison, Wisconsin, 1965, as quoted in reference 3a. 22. Boehm, J.R.; Balch, A.L.; Bizot, K.F.; Enemark, J.H. J. Am .Chern. Soc. 1975, U7, 501. 23. Busch, M.; Weiss, B. Chern. Ber. 1900, 33, 2701. 24. Overberger, C.G.; Valentine, M.; Anselme, J.P . J. Am. Ohern. Soc. 1969, 91, 687. 25. Overberger, C.G.; Herin, L.P. J. Org. Chern. 1962, 27, 417. 26. Overberger, C.G.; Marks, B.S. J. Am. Chern. Soc . 1955, 77, 4104. 27. Woodward, R.B.; Hoffman, R. "The Conservation of Orbital Symmetry," Academic Press , New York, N.Y. 1970. -16728. Lema!, D.M.; McGregor, S.D. J. Am. Chern. Soc. 1966, 88, 1335. 29. Carpino, L.A. Chern. Comrn. 30. Campbell, C.D.; Rees, C.W. J. Chern. Soc., (c), 19~, 494. 1969' 752. 31. Lichter, R.L., Ph.D. Thesis, University of Wisconsin, Madison, Wisconsin, 1967, as quoted in reference 3a. 32. Lerna1, D.M.; Menger, F . ; Coats, E.A. J. Am. Chern. Soc. 86, 2395. - -1964, --.,-. 33. Adams, D.J.C.; Bradbury, S.; Harwell, D.C.; Keating, M.; Rees, C.W.; Storr, R.C. Chern. Comm. 1971, 828. 34. Baldwin, . J ~E.; Brown, J . E.; HBfle, G. J. Am. Chern. Soc. 1971, 93, 788. 35. Anderson, D.J.; Gilchrist, T.L.; Rees, C.W. Chern. Comm. 1971, 800. 36. (a) Atkinson, R.S.; Rees, C.W . Chern. Comm. 1967, 1230; (b) Atkinson, R.S.; Rees, C.W. J. Chern. Soc. (C) 1969, 772. 37. Gaspar, P.P.; Hammond, G.S. in "Carbene Chemistry," Kirmse, W., ed . , Academic Press, New York, N.Y. 1964, Chapter 12. -16838 . Anderson, D.J.; Gilchrist, T.L . ; Harwell, D.C.; Rees , C.W. J. Chern. Soc. (C) 39. !P70 , 576 . Anders on, D.J.; Gilchrist, T.L.; Rces, C.W. Chern. Cornrn. 1969, 147. 40. For examples of experimental work on N2 H2 , see: (a) Rosengen, K. ; Pimentel, G.C . J. Chern. Phys. 1965, 43, 507; (h) Trombetti, A. Can . J . Phy s. 1968, 46, 1005; (c) Back, R.A . ; Wjllis, C.; Ramsay, D.A. Can. J. Chern. 1974, 52, 1006; (d) Wiberg, N.; Fischer , G.; Bachhuber , H. Angew. Chern. Int. Ed . Engl . 1976, 15, 385; (e) Willis, C.; Back, R.A.; Parsons, J.M.; Pardon, J.G., J. 41. A~ Chern. Soc. 1977, 99, 4451 . Pars ons, C.A.; Dykstra, C.E., submitted to J . Chern. Phys. 42. Casewit, C.; Goddard, W. A., to be published. 43. Lehn, J.M.; Munsch, B. Theoret. Chirn . Acta 1978, 12, 91. 44. Schaad, L.J.; Kinser, H.B., J . Phys. Chern. 1969, 73, 1901. 45 . Wong, D.P.; Fink, W. H.; Allen, L.C . J. Chern. Phys. 1970, 52, 6291. 46. Ra dom, L. ; Hehre, W.J.; Pople, J.A. J. Am . Chern . Soc. 1971, 93, 289. - 16947. Baird, N.C .; Sw6nsen, J.R. Can. J. Chern. 19 69 , 51, 3097. 48. Hariharan, P.C.; Pople, J.A . . Theoret. Chim . Ac ta, 1973, 28, 213. 49. Baird, N.C.; Barr, R.F. Can. J. Chern. 19 7 3, 51 , 330 3. SO. Wagniere, G. Theoret. Chim. Acta, 1973, 31, 269. 51. Merenyi, G.; Wettermark, G.; Roos, B. Chern. Phys. 1978, 1, 340. 52. Vasadevan, K.; Peyerimhoff, S.D.; Buenker, R.J.; Kammer, W.E. Chern. Phys. 1975, 7, 187. 53. Winter, N.W.; Pitzer, R.M. J. Chern. Phys. 1975, 62, 1269. 54. Ahlrichs, R.; Staemmler, V. Chern . Phys. Lett. 1976, 37, 77. 55. Howell, J.M.; Kirschenbaum, L.J. J. Am. Chern. Soc. 1976, 98, 877. 56. Camp, R.N.; Epstein, I.R.; Steel, C. J. Am. Chern. Soc. 1977, 99, 2453. 57 . Cimiraglia, R.; Riera, J.M.; Tomasi, J. Theoret. Chim . Acta, 1977, 46, 223. 58. Peric, M.; Buenker, R.J.; Peyerimoff, S.D. Can. J. Chern. 1977, 55, 1533. 59. Baird, N.C.; Wernette, D.A. Can. J. Chern. 1977, 55, 350. -17060. Davis, J.H.; Goddard, W.A. J. Am. Chern. So c . 19 7 7, 99, 71]1. 61. Pople, J.A.; Krishnan, R.; Schlegel, H.B.; Binkley, J.S. Int. J. Quant. Chern. 1978, 14, 545. 62. Talaty, E.R.; Schwartz, A.K.; Simons, G. J. Am . Chern. Soc. 1975,97,972. 63. Ditchfield, R.; Del Bene, J.E.; Pople, J.A. J. Am. Chern. Soc. 1972, 94 , 703 . 64. Lathon, W.A.; Curtiss, L.A.; Hehre, W.J.; Lisle, J .B.; Pople, J .A. Prog. Ph,ys. Org. Chem. 1974, 11, 175. 65. Pasta, D.J.; Chipman) D . M. J. Am. Chern . Soc. 1979, 101, 2290. 66. Car1otti, M.; Johsn, J.W.C.; Trombetti, A. Can. J. Phys. 1974, 52, 340. 67 . Bonybey, V.E . ; Nibler, J.W • .J. Chern. Phys. 1973, 58, 2125. 6 8. (a) Benson, S. W. "Thermochemical Kinetics," 2nd e d . , Wi 1 e y , New Yo r k , N . Y • , 1 9 7 6 ; ( b ) Benson , S.W.; O'Neal, H.E. "Kinetic Data Ori Gas Phase Unimol ecular Reactions,'' NSRDS-NBS No. 21, 1970, p. 448. -17169. Olah, G.A.; Olah, J.A. in · ~carbonium Ions," Vol. 2, Olah, G.A., ed., Wiley, New York, N. Y. 1970, Chapter 17. 70. Fo r examp le, see: Chapman, O.L.; Mattes, K.; Mcintosh, C.L.; Pacansky, J.; Calder, G.V.; Orr, G. J. Am. Chern. Soc. 1973, 95, 6134; Pacansky, J.; Bargon,J. J. Am. Chern. Soc. 1975, 97, 6896. 71. Porter, G. "Techniques of Organic Chemistry ," Vol. 8, Wiley, New York, N.Y., 1963, p. 1055. 72. Griller, D.; Ingold, K.U. Accts. Chern. Res. 1976, 9, 13. 73. (a) Garbisch, E.W.; Sprecher, R.F. J. Am. Chern. Soc. 1969, 91, 6785; (b) Masamun e , S.; Nakamura, N.; Suda, M.; Ona, H. J. Am. Chern. Soc. 19 73 , 95, 8481. 74. Overberger, C.G.; Palmer, L.C.; Marks, B.S.; Byrd, N.R., J. Am. Chern. Soc. 1955, 77, 4100. 75. Roberts, J.R.; Ingold, K.U. J. Am. Chern. Soc . 1973, 95, 3228. 76. Malatesta, V.; Ingold, K.U. 1974, 96, 3949. ~·A~. Chern. Soc . -17277 . Turro, N.J. "Modern f'.1olecular Photochemistry," Benjamin, Readin g , Mass., 1978, pp. 91-96. 78. Calvert, J.G.; Pitts, J.N. " Photochemistry," Wiley, New York, N.Y., 1966, p. 257. 79. For a discussion on the origins of solventinduced blue shifts, see Huberfie l d, P.; Lux , M.S.; Rosen, D., J. Am. Chern. Soc. 1977, 99 , 6828. 80. Ref. 78, p. 377. 81. Schultz, P.G.; Dervan, P.B., to be publish e d. 82. (a) We s t, W.; Killin gswor th, R.B. J. Chern . Phys_:_ 1938, 6, 1; (b) Spin.ltcr, L.; O'Brien, D. H.; Unter ein er , G.L.; Rus h, W. A., J. Org. Chern. !_1~, 83. 30,3278 . Craw ford , R . .J.; Mishra, A.; Dummel, R.J ., J. Am. Chem. So c . 1966, 88, 3959. 84. Ackermann, M.N.; J. Am. Chern . Soc. 85. Ell~nson, 19.~, J.L.; Robison, D.H. 90, 7173 . Silverstein, R.M.; Bassler, G. C. "Spectrome tri c Identification of Organic Compounds," Wiley, New York, N. Y., 1967, p. 66. -17386. (a) Coulson, C. A.; Moffitt, W. Phil. Mag. 1949, 40, 1; (b) Dolphin, D.; Wick, A. "Tabulation of Infrared Spectral Data," Wil ey , New York, N.Y., 1977, p. 175. 87. Wasserman, E. Prog. Phys. Org. Chern. 1971, 8, 319. 88. Hoffman, B.M.; Eames, T.B. J. Am . Chern. Soc. 1969, 91, 2169. 89. Wertz, J.E .; Bolton, J.R. "Electron Spin Resonance," McGraw-Hill, New York, N.Y., 1972, pp. 215-218. 90. (a) Jones, A. .J.; Hassan, M.M.A. J. Org. Chern. 1972, 37, 2332; (b) Wehrli; F.W.; Wirthlin, T. "Interpretation of Carbon-13 NMR Spectra," Heyden and Son Ltd., New York, N.Y., 1976. 91. Martin, J.C . ; Timberlake, J.W. J. Am. Chern. Soc. 1970, 92, 978. 92 . (a) Schulz, A.; Riichardt, C. Tet. Lett. 1976, 3883; (b) Duisrnann, W.; Ruchardt, C. Chern. Ber. 1978, 111, 596. 93. Reichardt, C. Angew. Chern . Int. Ed. Engl . 4, 29. 1965, -1 7 4- 94 . Computer simulation performed usin g the MSIM4 Program; Houl e , F .A.; Bunk e r, D.L., Program No. 293, Quantum Chemistry Program Exchange, University of Indiana, Bloomington, Ind. 95. Walling, C.; Cioffar i , A. J. Am. Chern. Soc. 1 9 72, 94, 6059. 96. Kosower, E.M. Accts. Chem. Res. 19 71, 4, 193 . 97. Nicholson, J.; Cohen, S.G. J . Am. Chern. Soc. 1966, 88, 2247. 98. Heesing, A.; Kaiser, B. Tet. Lett. 1970, 2845. 99. See DePuy, C.H.; King, R.W. Chern. Rev. 1960, 60, 431. 100. Reference 77, Chapter 13. 101. Engel, P.S.; Steel, C. Accts. Chern. Res. 1973, 6, 275. 102. Bartlett, P . D.; Porter, N.A. J . Am. Chern. Soc. 1968, 90, 5317. 103. Engel, P.S.; Hayes, ~.A.; Keifer, L.; Szilegyi, S.; Timberlake, J.W. J. Am. Chern. Soc. 1978, 100, 1876. 104. (a) Creger, P.L . J. Am . Chem. Soc. 1967, 89, 2500; (b) Baumga rten, H.E.; Smith, H.L.; Stakli s , A. J. Org. Chern. 1975, 40, 3554 . - 17 5- 105. Greene, F .D.; Gilbert, K.E. J. Org. Chem. 1975, 40, 1409. 106. Nelsen, S.F.; Bartlett, P.D . .J. Am. Chem . Soc. 1966, 1, 137. 107. Ohme, R.; Preuschho£, H.; Heyne, H.-U. Org . Synth . 1972, 52, 11. 108. Timberlake, J.W.; Hodges, M.L.; Betterton, K. Synthesis 109. 1972, 632. For example, see Vellturo, A.F.; Griffin , G.W . J. Am. Chern. Soc. 1965, 87, 3021.. 1.10. Bauer, R.; Wendt, H. An~w. Chern. Int. Ed . Engl . 1978, 17, 370. 111. Air Products and Chemicals, Inc., Advanced Products Department, P. 0. Box 538, Allentown, PA 18105. 112. International Crystal Laboratories, 107 Trumbull Street, Elizabeth, NJ 07206 113. Amersil, Inc., 685 Ramsey Avenue, Hillside, NJ 07205. 114. Alpha Wire Corp., 711 Lidgerwood Avenue, Elizabeth, NJ 07207. 115. McKay, D.; Waters, W.A.; J. Chern. Soc. (C) 1966, 813. 116. Starr, J.E.; Eastman, R.H. J. Org . Chern. 1966, 31, 1393. -176 117. Beard, J., Ph. D. Thesis, Stanford University , Stanford, Ca l ifornia , 1 969 . 118. Gunn, S.R. Rev. Sci . Instr. 1962, 33, 880. 119. (a) Gugge nheim, E.A. Phil. Mag . 1926, 2, 538; (b) Daniels, F.; Mathews, J.H.; Wi lliams, J.W.; Bender, P.; Alberty, R.A. " Experi mental Physical Chemistry," McGraw-H i ll, New York, N.Y., 1966, p . 134 . 120. Calvert, J.G.; Pitts, J.N. "Photochem is try," Wiley, New York, N.Y., 1966, p. 739. 121 . Kulp, S . S.; Fish, V.B . ; Easton, N.R. Can. J. Chern. 1965, 43, 2512.