Reactions of Multidentate-Ligand Copper(I) Complexes with Dioxygen Citation Lisensky, George Clark (1981) Reactions of Multidentate-Ligand Copper(I) Complexes with Dioxygen. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/8662-xh16. https://resolver.caltech.edu/CaltechTHESIS:10242025-212603541 Abstract Syntheses, isolation, and characterization of several copper(I) complexes with multidentate ligands (including macrocycles), and their reactions with dioxygen are reported. A binuclear dioxygen complex was obtained from 1, 1-difluoro-4, 5, 11, 12-tetramethyl-1-bora-3, 6, 10, 13- tetraaza-2, 14-dioxo-cyclctetradeca-3, 5, 10, 12-tetraenato copper(I) in dichloromethane solution at -79° C. In the absence of excess dioxygen up to 40% of regenerated copper(I) has been observed upon warming. At room temperature the stoichiometry O 2 /Cu is greater than 1.0 and increases with solvent basicity. At large triphenylphosphine to copper ratios the uptake is decreased and at low ratios the uptake is enhanced. The macrocyclic complex 5, 7, 7, 12, 14, 14-hexamethyl-1, 4, 8, 11-tetra-azacyclctetradeca- 1, 4, 8, 11-tetraene copper(I) perchlorate in the solid state or in dichloromethane solution at -79° C is ketonated in either of two positions. The 3, 5, 7, 7, 10, 12, 14, 14-octamethyl derivative is ketonated in only one position. Similar reactions are observed for the 4, 11-diene copper(I) complexes as well as partial oxidative dehydrogenation. 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): Gagné, Robert R. Thesis Committee: Unknown, Unknown Defense Date: 22 July 1980 Record Number: CaltechTHESIS:10242025-212603541 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:10242025-212603541 DOI: 10.7907/8662-xh16 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17731 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 27 Oct 2025 18:48 Last Modified: 27 Oct 2025 18:51 Thesis Files PDF - Final Version See Usage Policy. 59MB Repository Staff Only: item control page

REACTIONS OF MULTIDENTATE-LIGAND COPPER(I) COMPLEXES WITH DIOXYGEN Thesis by George Clark Lisensky In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1981 (submitted July 22, 1980) ii Acknowledgements The work reported herein has been guided by Bob Gagne, without whose encouragement and assistance it would not have been possible. I thank him for his support, his style, and the sociable group he has chosen to encourage. Discussion and interaction with other members of the Gagne group has been helpful, especially with Dave Marks who was always ready to listen to what hadn't succeeded most recently. I couldn't have worked with a better group of people. -- The support personnel at Caltech have been terrific, especially Jan an'd Sally in the analytical lab, and Erich, Sig, and Gabor in the glass shop who have put up with my frequent ASAP requests over the years. I also acknowledge the escape provided by our friends. The Putnams were always ready for another hand of Nerts and an endless succession of painting and wallpapering projects. Our groups at the First United Methodist Church gave us a bigger family to belong to. My wife, Carol, has been a strong source of support and there when I needed her. What more can I say? The littlest Lisensky, Laura, has done her best to be quiet during this writing. Lastly, I thank Henriette Wymar for typing this thesis, and whoever left this poem in the Caltech computer: iii CAT SUP BOTTLE, CATSUP BOTTLE •. NONE'll COME AMJ THEN 4 LOT'LL. O. NASH END X-RAY iv Reactions of Multidentate-Ligand Copper(!) Complexes with Dioxygen Abstract: Syntheses, isolation, and characterization of several copper(!) complexes with multidentate ligands (including macrocycles), and their reactions with dioxygen are reported. A binuclear dioxygen complex was obtained from 1, l-difluoro-4, 5, 11, 12-tetramethyl-1-bora-3, 6, 10, 13tetraaza-2, 14-dioxo-cyclctetradeca-3, 5, 10, 12-tetraenato copper(!) in dichloromethane solution at -79 ° C. In the absenc.e of excess clioxygen up to 40% of regenerated copper(!) has been observed upon warming. At room temperature the stoichiometry 0 2 /Cu is greater than 1. 0 and increases with solvent basicity. At large triphenylphosphine to copper ratios the uptake is decreased and at low ratios the uptake is enhanced. The macrocyclic complex 5, 7, 7, 12, 14, 14-hexamethyl-l, 4, 8, 11-tetraazacyclctetradeca-1, 4, 8, 11-tetraene copper(!) perchlorate in the solid state or in dichloromethane solution at -79° C is ketonated in either of two positions. The 3, 5, 7, 7, 10, 12, 14, 14-octamethyl derivative i's ketonated in only one position. Similar reactions are observed for the 4, 11-diene copper(!) complexes as well as partial oxidative dehydrogenation. V Reactions of MuJtidentate-Ligand Copper(!) Complexes with Dioxygen Table of Contents Page Chapter One o Introduction 1 A. Thesis Introduction 2 B. Copper-Containing Proteins and Diaxygen 3 C. Kinetic studies of Copper(I) Autoxidation 5 D. Claims of Small Molecule Dioxygen Complexes 7 E. Dioxygen Complexes of CobaJt(II) and Iron(Il) 9 F. strategies for study of Dioxygen Complexes 10 G. Known Macrocycle Oxidations •• 15 H. Summary 19 I. References 20 Chapter Two. Dioxygen Reactions of Two Copper(I) Multidentate Ligand Systems 28 A. Choice of Ligands 29 B. A Multidentate Mononuclear Copper(!) Complex 30 C. A Binuclear Dicopper(I) Complex 34 Do Synthesis and Characterization of the Binuclear Complexes 39 E o References Chapter Three. Diaxygen Reactions with Copper(!) Tetraazamacrocyclic Complexes 44 46 A. CuLBF2 49 B. CuLBF2 Reaction with Dioxygen at -79° 52 c. CuLBF2 Reaction with Diaxygen at 25° 56 D. Slow Dioxygen Reactions of Copper(!) with Similar Macrocycles 59 vi Table of Contents (continued) Page E. CuLBF2 Ligand Modification 63 F . • Cu(I)Me 6 -tetraene(ClO4 ) 63 G. Dioxygen Reactions of Cu(I)Me 6 -tetraene(ClO4 ) 67 H. Ligand Modification of CuMe 6 -tetraene 72 I. Dioxygen Reactions of CuMe 6 -diene(ClO 4 ) and CuMe8 -diene (ClO4 ) 79 J. Dioxygen Reaction with Ni(I)Me8 -tetraene(ClO4 ) 79 K. A Copper(!) Complex Which Does Not React with Dioxygen 80 L. Summary 83 M. Experimental 84 N. References 101 Chapter Four. Consideration of Mechanism and Summary 104 A. Mechanism Consideration 105 B. Thesis Summary 110 C. References 113 Appendix 1. structural Characterization of Dimeric µ-Carbonate and Tetrameric µ-Hydraxo Complexes Resulting from the Autoxidation of a Copper(!) Carbonyl Complex 115 Appendix 2. Oxidation state of Copper in Reduced Copper(!) Macrocycles and Their Adducts 126 Appendix 3. 13 C nmr Chemical Shifts of Diene and Tetraene Complexes 134 vii Table of Contents (continued) Page Proposals I. Synthetic Phosphorus Chemistry 136 • II. Polypyrrole Coated Electrodes as Diodes 145 m. Gas Chromatographic Resolution of Enantiomers 150 IV. Synthesis of Linked Macrocyclic Ligands 155 V. Substitution Kinetics of Five-Coordinate Complexes 163 1 Chapter One. Introduction 2 Chapter One. Introduction. As early as 1880, Bertholet observed that cuprous salts in solution react with atmospheric oxygen. 1 A peine filtrees, ces dissolutions se troublent aussitOt, en deposant un precipite ... En effet, il resulte de I 'absorption de l'oxygene de l'air; il continue a se former tant que la liqueur renferme de chlorure cuivreux. Of course, biological systems have utilized the interaction of copper with dioxygen for much longer than ch~mists have studied it. Copper is an essential element for most life forms since copper-containing protein and enzyme systems 2' 3 are prominent in the use and activation of molecular dioxygen. 4, 5 This chapter will first examine what little is known about the interaction of dioxygen with copper in copper-containing proteins. The results of others' k_inetic studies of copper(!) autoxidation follow. These studies indicate copper dioxygen species exist at least as intermediates in small molecules as well. The attempts by chemists to synthesize copper dioxygen complexes are then discussed. These have been a qualified failure, with few systems adequately studied and the oxidized products obtained not well characterized. The more successful syntheses of cobalt(Il) and iron(Il) dioxygen complexes are then examined. In particular, the strategies used to successfully study dioxygen complexes, such as steric control and low temperatures, as well as choice of solvent and redox potential effects are summarized. Lastly, since macrocyclic ligands will be used and copper(!) is known to activate dioxygen for substrate oxidations, the 3 known macrocyclic ligand oxidation reactions will be discussed. The first chapter is intended to provide the background needed to place the \ results reported in the fallowing chapters in perspective. Chapter Two will examine our initial studies with multidentate ligands which indicate macrocyclic systems must be used for these labile copper complexes. Chapter Three will report studies of a low temperature, binuclear copper dioxygen complex in solution, as well as ketonation of a macrocyclic ligand by copper dioxygen activation. Copper-Containing Proteins and Dioxygen Copper-containing protein and enzyme systems are prominent in the use and activation of molecular dioxygen. In humans copper is the third most abundant. metal, after iron and zinc. Cytochrome c oxidase, containing both iron and copper, is vital in the respiratory process in • reducing di oxygen to water, and as such is responsible for 90% of the oxygen consumption of the biosphere. 6 In arthropods and molluscs, the oxygen transport protein hemocyanin utilizes copper. 7 The blue copper enzymes laccase, ascorbate oxidase, and ceruloplasmin mediate substrate oxidation coupled with reduction of dioxygen to water. 8 , 9 The copper-containing amine oxidases and galactose axidase respectively oxidize primary amines and alcohols to aldehydes, coupled with the reduction of oxygen to peroxide. 9 Tyrosinase and dopamine-/3hydroxylase are monooxygenases which add one of the atoms of dioxygen to their respective substrates. 10 The above important copper proteins have one thing in common; all presumably involve dioxygen coordinated to copper, but little is known about the interaction or even the nature or 4 structure of the copper binding site. None of the above mentioned copper proteins has been studied crystallographically although plastocyanin11 (an electron carrier blue copper protein) and superoxide dismutase 12 , 13 (which contains both copper and zinc) have been examined. Plastocyanin shows imidazole nitrogen coordination from histidine residues as well as the sulfur coordination expected for a blue copper protein. Superoxide dismutase contains copper coordinated to four imidazole nitrogen atoms, one of which is deprotonated and also coordinated to the zinc atom. Coordination of at least some histidine-nitrogen atoms as fourid for these copper proteins is postulated for most of the others on the basis of electron paramagnetic resonance nitrogen superhyperfine splittings as well as titration evidence. Of special interest are the copper proteins which reversibly bind dioxygen: hemocyanin which functions biologically as an oxygen carrier, •and the enzyme tyrosinase which functions biologically to activate oxygen for substrate incorporation. In hemocyanin, subunits of two copper(!) atoms reversibly bind one dioxygen molecule. 6 Oxyhemocyanin is diamagnetic; electron paramagnetic resonance 14 and magnetic measurements 15 , 16 indicate strong antiferromagnetic coupling between pairs of copper(II) ions. Resonance Raman study of hemocyanin17 - 19 shows a new band at 749 cm -l indicating a peroxo complex upon addition of 0 1 6 The band appears at 708 cm- for 1 • 2 1 8 0 • 2 Use of un- symmetrically labeled dioxygen shows both oxygen atoms to be equivalent . 19 A second resonance Raman band at 267 cm-1 which does not shift with oxygen isotope labeling is assigned to a copper(II) imidazole • • 5 vibration. 18 Titration evidence also indicates imidazole coordination. 20 The monooxygenase tyrosinase very likely has a similar coordination environment. Treatment with one hydrogen peroxide per two copper atoms reduces the copper and gives a form which can then reversibly • d oxygen. 21 ' 22 The v1S1 •• ble spectrum 21 ' 22 of oxytyrosmase . bm is very similar to that of oxyhemocyanin, as are the electron paramagnetic 23 24 resonance spectra , of the nitric oxide derivatives. Additionally tyrosinase shows a two-electron reduction with E = +O. 36 V (pH 7. O) 25 for the enzyme but not for the apoenzyme. Note that the similar "soft" nitrogen ligand environments of hemocyanin and tyrosinase give both reversible oxygenation and activation of oxygen for incorporation into substrates. Application of these criteria to the choice of copper(!) complexes studied here has resulted in the use of soft nitrogen imine and pyridine groups to coordinate the metal. The use of oxygen by the copper proteins shows that given the correct ligand environment, copper dioxygen complexes can be synthesized and specific and useful ·substrate oxidations can be achieved. Kinetic studies of autoxidation of small molecule copper(!) compounds are examined next. Copper dioxygen complexes exist for these small molecules too, if only as fleeting intermediates. Kinetic studies of Copper(!) Autoxidation Few kinetic studies of the autoxidation of copper(!) complexes in non-aqu£ Jus solvents have been carried out. The autoxidation of copper(!) chloride in glacial acetic acid, 26 tetrakis (acetonitrile) copper(I)perchlorate in dimethylsulfoxide, 27 and various. substituted 6 copper(!) bis-i:iienanthroline perchlorate complexes in nitromethane, 28 , 29 was studied. and all were found to follow the rate law k[Cu(I)]2(O2]. Such a rate law implies a binuclear transition state, + fast + cu+ * fast products, and the twoCu + 0 ► CuO ===~ C~O , ) 2 2 2 electron reduction of 0 2 to H2 O2 • In contrast, most aqueous acetonitrile systems 30 - 38 follow a rate law k[Cu(I)][O 2]. Such a rate law could imply a mononuclear transition fast state, Cu+ + 0 ~ cue: , products, and the stepwise one2 electron reduction of 0 2 • However, the rate law is equally well accounted for 37 by the binuclear mechanism if a steady state approximation is applied to Cuot and k_ 1 «~[cu+]. The experi~ental observation of the two-electron oxygen reduction product (H O rather ) 2 than the one-electron reduction product (HO the reaction of aqueous Cu(I) with HO2 • 2 is also inconclusive since is nearly diffusion controlled. 39 · ) 2 The best evidence for the binuclear mechanism in aqueous solution has been 01::tained with a series of substituted copper(!) imidazole complexes, whose autoxidation kinetics indicate both paths being followed simultaneously. 4o, 41 Whether the reduction of 0 occurs in one- or 2 two-electron steps, the measured reaction stoichiometries are found to be two copper atoms per one dioxygen as shown in scheme I-1. 7 In general, further reactions with H20 2 are not observed or are determined to proceed at a much s1 ower rate, and the products in water are not isolated. Claims of Small Molecule Dioxygen Complexes Although kinetic data demonstrate the transient formation of dioxygen containing copper complexes, none has 'been well characterized. Even the oxidized products have not 'been well characterized in general. For example, addition of dioxane to autoxidized dimethylsulfoxide solutions of tetrakis(acetoni!rile} copper(I)tetrafluoroborate is reported to result in a precipitate analyzing (C, H, S, Cu) for an oxo species, C~O(DMSO}idioxane)(BF4} 2 , that is stable for several days at room temperature. 42 The only reported clean diox;gen reactions of copper(!} are those of the substituted bis-phenanthroline complexes in nitromethane which yield stable copper(II) coordinated-nitromethanide anion bis-phenanthroline complexes. 43 Published claims of actual nonprotein copper dioxygen complexes are m.sed on partially reversible 44 , 45 or irreversible 46 , 47 oxygen uptakes and color changes, 48 - 52 elemental analysis, 46 or changes in the infrared spectrum upon exposure to oxygen. 47 , 53 - 55 Both superoxo (1083 cm-1 , 1109 cm-1 ) 53 , 54 and peroxo (856 cm -l, 845 and 885 cm - 1 ) 47 ' 55 type complexes have been claimed solely on the m.sis of new infrared peaks. The 'best studied complex is that reported by Wilson (Fig. 1-1). 44 Solutions of the complex absorb one dioxygen per two copper atoms in dimethylsulfoxide, acetonitrile, dimethylformamide, pyridine, and 2, 6-lutidine. After heating (40° C) and degassing (N2 } the green solution 8 Fig. I-1. Wilson's complex44 N N .- to remove coordinated dioxygen, the complex is capable of reabsorbing O. 8 dioxygen per two copper atoms. Integration of the electron paramagnetic resonance spectrum of the oxygenated solution accounts for only 20% of the available copper and the signal is probably due to the irreversibly oxygenated component. complex is presumably diamagnetic. The apparent dioxygen The red compound in the solid state also absorbs dioxygen and turns greeno .In general, metal complexes are often irreversibly oxidized by dioxygen without forming isolable dioxygen complexes. In some cases the metal is oxidized and in others the ligand may be oxidized. Destructive oxidation is also common. In most, if not all, of these reactions peroxo and hydroperaxo intermediates are involved. 56 No small molecule copper dioxygen complex has been isolated as a pure solid. Identification of the products of oxygen reactions with copper(I) complexes, other than that they contain copper(II), has rarely been done successfully. However, col:alt(II) and iron(II) dioxygen complexes have 9 successfully been prepared. Examined next are the kinds of products oltained and the techniques that were successful in studying them. Dioxygen Complexes of Cooo.lt{II) and Iron{II) The air oxidation of cobalt(II) to cooo.lt(III) is often used to synthesize cooo.lt(m) compounds. By the choice of suitable ligands such as Salen57 , 58 or Acacen 59 and the proper conditions, the intermediate mononuclear peroxo and binuclear superoxo species (Scheme I-2) may be reversibly formed and/or isolated. B Scheme I-2 ~ 0 ,✓-o ~ i 0 2 ~ .:::::::==~ o~ 0 ~ ~~ ~ - - - - + com Products B B ~~Slow com B , I B The air oxidation of iron(II) complexes typically gives irreversible iron(III) µ-oxo species (Scheme I-3). By choice of the proper conditions Scheme I-3 /0 B O OCR:) ~~~~~-~ ~B~ ~ Fast 10 and/or a suitably modified porphyrin ligand, the reaction can be stopped (or the subsequent steps slowed) at the mononuclear reversible dioxygen complex~ Such is the case for hemoglobin and myoglobin. • t ere st 60 - 63 sine • e The "suitable" condi·t·ions are of part·icula r in some of the strategies used to attain the comlt and iron dioxygen complexes are applicable to the study of the copper(!) dioxygen interaction. Factors found to be important in previous studies of dioxygen complexes include steric prevention of reaction with a second metal atom as well as temperature, solvent, and redox potential effects. Strategies for Study ·of Dioxygen Complexes One strategy used for the iron complexes is to prevent the approach of a second iron atom. Hemoglobin and myoglobin contain isolated iron(II) porphyrins which cannot readily meet. Such is also true of a 1-(2-phenylethyl)imidazole carbonmonoxy heme complex in a polystyrene film prepared by Wang 64 in 1958. After removal of the CO the material bound dioxygen reversibly. (This experiment has been confirmed more recently). 65 ' 66 A similar approach used successfully to obtain reversible oxygenation involves preparation of an imidazole modified silica gel to coordinate an iron porphyrino 67 To accomplish the same in solution requires some elegant synthetic organic chemistry. Collman's picket fence porphyrins 68 - 70 and Baldwin's cappBd porphyrins 71 sterically block one side of the porphyrin to decrease or prevent binding of the axial mse to that side. Dioxygen then binds almost exclusively on the bulky side where close approach of a second bulky iron porphyrin is prevented. The mononuclear dioxygen complexes with these modified porphyrins are stable 11 at room temperature in solution for several hours. A non-porphyrin macrocycle utilizing a similar idea also shows increased stability over the non-bulky macrocycle. 72 A last possible way to prevent dimer formation is by reaction of the solid state complexes with dioxygen. If the metal atoms are far enough apart in the crystalline material, if there is a vacant coordination site, and if the solid is permeable to dioxygen, such a procedure should work. Solid imidazole proto- and mesohemes do combine reversibly with molecular oxygen. 73 Evaporated carbonyl solutions of an iron porphyrin with attached imida~ole reversibly bind dioxygen upon removal of the CO. 65 Dioxygen 1-methylimidazole complexes of iron picket fence porphyrin lose dioxygen under vacuum. Upon exposure to oxygen the reaction is reversed. 61 , 74 While several cobalt(II) Salen c~mplexes react in the solid state reversibly with dioxygen, 57 iron(II) Salen complexes react in the solid state to give the µ -oxo compounds: 75 In applying these steric .strategies to copper(!), synthesis of modified polymers and blocking one side of a macrocycle by organic groups were not considered, due to the difficulties in synt.tB sizing plain macrocyclic complexes. Reaction of dioxygen with solid complexes has been investigated, but for most compounds reported here either no reaction occurs in the solid state or the same stoichiometry is determined as in solution at -79 °. The solution studies are generally easier since the extent of reaction of the solid is difficult to monitor. Lowering the temperature has been found to increase the likelihood of obtaining dioxygen adducts. Dioxygen binding to transition metals is generally exc:thermic, so both enthalpy and entropy effects increase the 12 thermodynamic binding at low temperature. The activation energy for 0-0 bond cleavage is fairly high, so lower temperatures slow µ-oxo formation. Even for the bis(l-methyl-imidazole)iron(II) complexes of tetraphenylporphyrin76 and mesoporphyrin IX dimethyl ester, 77 reversible oxygenation has been observed in dichloromethane at low temperature (-79 ° and -50° respectively). Many other non-hindered axial lase iron porphyrin complexes also give reversible dioxygen complexes in various solvents at low temperatures. 78 - 81 The cobalt(II) Acacen complexes in dimethylformamide give reversible oxygen chemistry at -20° C, but at room temperature a continuous slow uptake of oxygen over a period of days is observed indicating oxidation of the ligand. 59 Use of the low temperature strategy to examine manganese(Il) tetraphenylporphyrin complexes has successfully resulted in the observation of reversible manganese dioxygen complexes at -79 °. 82 - 85 Application of the low temperature strategy to copper(!) has proved to be helpful and in one ca~e resulted in stabilization of a binuclear copper-dioxygen .complex in solution (see Chapter Three). To reduce chances of irreversible oxidation, aprotic solvents are usually used and ligands with easily lost protons avoided. Such solvents simulate the hydrophobic environment of the proteins. Although all heme-dioxygen complexes seem more sensitive to oxidation in hydroxylic solvents and in acids than in aprotic media, no thorough study has been made. 86 Protons presumably resuH: in the reaction, H+ M-02 ~ M(ill) + H02 - further products. The presence of easily lost protons on unsubstituted imidazole may be the cause of the much greater stability of the 1-methylimidazole porphyrin dioxygen complexes. 13 Tetraphenylporphyrinbis( imidazole )iron(II) reacts instantaneously and irreversibly (0. 21 ± 0. 05 0 /Fe) even at -78° in toluene, 77 but the 2 1-methylimidazole complexes give reversible adducts. 76 Similarly iron(II) complexes of mesoporphyrin IX dimethyl ester reversibly bind oxygen at -50° in dichloromethane in the presence of 1-methylimidazole, but are irreversibly oxidized if imidazole is used. 77 Even the capped iron porphyrins, which have a five-hour half life at room temperature with 1-methylimidazole in benzene, are oxidized instantaneously if imidazole is used instead. 71 Cobalt porphyrin complexes behave similarly, giving mononuclear dioxygen complexes with 1-methylimidazole that are stable for several days at room temperature, but decom87 posing irreversibly. and instantaneously with imidazole. While it has not yet been proven that imidazole acts as an acid to effect these drastic stability changes, such an effect seems reasonable and it is generally agreed that easily dissociable protons should be avoided when studying dioxygen complexes. The application of the aprotic solvent strategy to the copper (I) complexes studied herein is of even greater importance. In hydroxylic solvents copper(!) disproportionates to copper(II) and copper(0). In nitromethane the somewhat easily lost solvent proton from nitromethane would appear to account for the observed 0. 25 0 2 /Cu stoichiometry (see Chapter Three}. Another solvent effect is the enhancement of dioxygen binding by more polar solvents. While methylimidazole cobalt(II) protoporphyrin IX dimethyl ester in toluene requires an oxygen pressure of 417 Torr for half oxygenation at -23 °, only 12. 6 Torr are required in dimethylformamide. 88 (0 is also more soluble in toluene than dimethyl2 14 formamide.) 88 A similar effect is observed for bis(pyridine )iron(II) tetraphenylporphyrin in solution at -78°. Under one atmosphere of dioxygen complete adduct formation occurs in dichloromethane, only a small amount of adduct is formed in toluene, and the amount in diethyl ether is intermediate. 76 Thus the observed effect is that the polar dioxygen complexes are more favored in higher dielectric constant media. In agreement with electron transfer from metal to dioxygen 89 to give Co(ID)-o; is the observation that the binding constants are proportional to the cotalt(II) to cobalt(III) oxi93,tion potential. 9 o, 91 Thus the easier it is for cobalt(!!} to be oxidized to coba.H:(ill}, the stronger the bond in Co(III)-o;. Very positive reduction potentials may result in lack of r~activity, such as found for cobalt(II} porphyrins at 25 °. Very negative reduction potentials could give outer sphere electron transfer or irreversible adduct formation. An example of the latter is provided by pyridine chromium(II) tetraphenylporphyrin. The chromium(III) oxidation state is so favored (reduction of chromium chloride tetraphenylporphyrin in dimethylsulfoxide is at -0. 86 V vs SCE) that the binding of dioxygen is irreversible. 92 The most stable copper(!} complex reported here (highest Cu(II) to Cu(I) reduction potential} fails to react with dioxygen and may be an example of the opposite extreme (Chapter Three). Presumably there is an ideal potential range for reversible adduct formation for each metal. The reduction potential of copper in tyrosinase, which forms reversible dioxygen adducts, is +O. 36 V. 15 In addition to oxygen adducts, reaction with dioxygen can lead to ligand oxidation. Examined next are the oxidations macrocyclic compounds have · been found to undergo. An advantage of macrocyclic ligands is that they tend to stay intact and coordinated to the metal after oxidation. Known Macrocycle Oxidations Oxidative dehydrogenation of coordinated amine ligands to form imine linkages is a common macrocyclic response to oxidizing agents in general. 93 For example, the diimine shown in Scheme I-4, is successively oxidized to the triimine and then the tetraimine. Scheme I-4 \ C: H The strength of oxidizing agent necessary depends on the metal ion. With the above macrocycle, the iron(II) complex is oxidized by 0 , 94 2 the nickel(II) complex is oxidized by HN03 , 95 ' 96 and the copper(II) complex, reported for the first time here ( see Chapter Three), requires a perchloric acid permanganate treatment. Many other amine- containing macrocycles, including several nickel(II) species in the presence of dioxygen, undergo oxidative dehydrogenation. Fig. I-2 shows an example where a further aromatization has occurred after oxidative dehydrogenation. 97 16 Fig. I-2Q Oxidative dehydrogenation of a Ni(II) complex in the presence of baseQ The reaction stoichiometry is 1. 0 0 2 /Ni(II). The mechanism for the oxidative dehydrogenation presumably involves higher oxidation states of the metal. Initial conversion of the M(II) to M(III), and subsequent electron abstraction from an N-H bond would yield a M(II) ligand radical species. This could then be oxidized to yield the observed product. Such a mechanism is supported by the decomposition of M(m) complexes in solution. 94 , 98 , 99 A second type of ligand reaction, which has been observed occasionally upon oxygenation of macrocyclic complexes, is ketonation (Fig. 1-3). In Fig. I-3C the ligand, of course, has also undergone oxidative dehydrogenation. While the product in Fig. I-3A could be 1 identified by spectroscopic means (v C=O =1670 cm - ), the product in Fig. I-3B required crystallographic analysis as the carbonyl and imine 1 vibrational frequencies appear to be degenerate (1678 cm - ). Ligand modification of B by adding a methyl group at the ketonation position blocks the reaction as expected, and only the slow reaction to cobalt(III) occurs. Ketonation of a copper(!) macrocycle has been observed here and is reported in Chapter Three, as well as blockage of that ketonation by a methyl group substitution. 17 .. I (B) l r--N coJJJ 2 ~ cc:F 0 ~ N Slow HVH 0 n( (C.) Fen~ N N_/ y 0 Figo 1-3. Ketonation by dioxygen reaction has been observed in these three macrocycies: (A), reference 100; (B), reference 101; (C) reference 102. 18 The proposed mechanism 1 OO involves the intermediate shown in Fig. I-4. Similar addition at the central carbon has been observedlOO,l0 3 ,104 for nitriles and acetylenes giving products like those in Fig. I-5, and making the proposed intermediate more plausible. The common structural feature of the macrocyclic ligands undergoing ketonation is the t3-diimine unit, -N=CR-CH2 -CR=N-. Copper(Il) has also _be_e n observed to catalyze ketonation of non-macrocyclic t3-diimines. 105 , 106 H Fig. I-4. Presumed intermediate in the ketona±ion reaction of /3• dilmines. R H H n >-=N;M:N:) u Fig. I-5. Nitrile and acetylene adducts of f3 -diimines. 19 Summary Formation of dioxygen complexes and/or activation of dioxygen for substrate oxidation occurs for some copper-containing proteins. Kinetic studies indicate copper dioxygen complexes exist for small molecules too, although so far only as intermediates. Few products have been identified from reactions of dioxygen with copper(I)o The present study involves the dioxygen reactions of multidentate, flexible, soft nitrogen ligand copper complexes. The aim was to adjust the ligand characteristics and reaction conditions such that interaction of dioxygen with copper was favored. If a copper dioxygen adduct could not be isolated, possible ligand oxidation was investigated since the enzyme reactions indicate tbat copper is capable of activating dioxygen toward substrate reactions. Chapter Two will discuss our early work on multidentate systems, beginning with a discussion of the choice of optimal ligands for this study of copper(!) complexes and dioxygen. 20 References 1. M. Berthelot, Ann. Chim. Phvs. V. Serie, !Q, 503-521 (1880). 2. T. Vanngard in "Biological Applications of Electron Spin Resonance," H. M. Swartz, J. R. Bolton, and D. C. Borg, Eds., Wiley Interscience, N. Y., 1972, pp. 411-447. 3. R. MalkinandB. G. Malmstrom, Adv. Enzymology,~' 177-244 (1970). 4. G. McLendonandA. E. Martell, Coord. Chem. Rev., 19, 1-39 """' (1976). 5. E. Ochiai, J. Inorg. Nucl. Chem., ll, 1503-9 (1975). 6. W. S. Caughey, W. J. Wallace, J. A. Volpe, and S. Yoshikawa, in "The Enzymes," P. D. Boyer, Ed., 3rd ed., Vol. 13, Academic Press, N. Y., 1976, pp. 299-344. 7. R. Lontie in "Structure and Function of Haemocyanin," J. V. Bannister, Ed., Springer-Verlag, N.Y.,1977, pp.150-155. 8. J. A. . Fee, Structure and Bonding, E_, 1-60 (1975). 9. B. G. Malmstrom, L. E. Andreasson, and B. Reinhammar, in Ref. 6, 3rd ed., Vol. 12, 1975, pp. 507-579. 10. W. H. Vanneste and A. D. Zuberbuhler, in "Molecular Mechanisms of Oxygen Activation," 0. Hayaishi, Ed. , Academic Press, N. Y. , 1974, pp. 371-404. -- 11. H. C. Freeman, Nature, 272, 319-324 (1978). 120 J. S. Richardson, K. A. Thomas, B. H. Rubin, and D. C. Richardson, Proc. Nat'l. Acad. Sci. U.S.A., TI_, 1349-1353(1975). 13. D. C. Richardson in "Superoxide and Superoxide Dismutases," A. M. Michelson, J. M. McCord, and I. Fridovich, Eds., 21 References (continued) Academic Press, N. Y., 1977, pp. 213-223. 14. T. Nakamura and H. S. Mason, Biochem. Biophys. Res. Comm., 3, 297-299 (1960). " 15. T. H. Moss, D. C. Gould, A. Ehrenberg, J. S. Loehr, and H. S. Mason, Biochem., 12, 2444-2448 (1973). """ 16. E. I. Solomon, D. M. Dooley, R.H. Wang, H. B. Gray, M. Cerdonio, F. Mogno, and G. L. Romani, J. Amer. Chem. Soc., ll, 1029-1031 (1976). 17. J. S. Loehr, T. B. Freedman and T. M. Loehr, Biochem. Biophys. Res. Comm., ~' 510-515 (1974). 18. T. B. Freedman, J. S. Loehr, and T. M. Loehr, J. Amer. Chem. Soc., ll, 2809-2815 (1976). 19. T. J. Thamann, J. S. Loehr, and T. M. Loehr, J. Amer. Chem. Soc., 99, 4187-4189 (1977). """ 20. B. Salvato, A. Ghiretti-Magaldi, and F. Ghiretti, Biochem., 13, """ 4778-4783 (1974). 21. R. L. Jolley, Jr., L. H. Evans and H. S. Mason, Biochem. Biophys. Research Comm., 46, 878-884 (1972). """ 22. R. L. Jolley, Jr., L. H. Evans, N. Makino, and H. S. Mason, J. Bio. Chem., 249, 335-345 (1974). """" 23. A. J. M. Schoot Uiterkamp and H. S. Mason, Proc. Nat'l Acad. Sci. U.S.A., 70, 993-996 (1973). ""' 24. A. J. M. Schoot Uiterkamp, H. Van der Deen, H. C. J. Berendsen, and J. F. Boas, Biochim. Biophys. Acta, lli, 407-425 (1974). 22 References (continued) 25. N. Makino, P. McMahill, H. s. Mason, and T. H. Moss, -- J. Biol. Chem., 249, 6062-6066 (1974). 26. P. M. Henry, Inorg. Chem., 5, 688-689 (1966). 27. G. Rainoni and A. D. Zuberbuhler, Chimia, 28, 67-70 (1974). 28. J. A. Arce, E. Spodine and W. Zamudio, J. Inorg. Nucl. Chem., "' """ 37, 1304-1305 (1975). """ 29. J. A. Arce, E. Spodine and W. Zamudio, J. Inorg. Nucl. Chem., 38, 2029-2031 (1976). """ 30. H. Nord, Acta Chem. Scand., 9, 430-437 (1955). 31. R. D. Gray, J. Amer. 32. A. Gunter and A. D. Zuberbuhler, Chimia, 24, 340-342 (1970). "' Chem. Soc., 91, 56 (1969). """ """ 33. A. D. Zuberbuhler, Helv. Chim. Acta, 53, 473-485 (1970). """ 34. A. D. Zuberbuhler, Chimia, """23, 416-418 (1969). 35. A. D. Zuberbuhler •in "Metal Ions in Biological Systems," Vol. 5, H. Sigel, Ed., Marcel Dekker, N. Y., 1976, pp. 325-368. 36. A. L. Crumbliss and A. T. Poulos, Inorg. Chem., !1, 1529-1534 (1975). 37. I. Pecht and M. Anmr, J. Chem. Soc., Ser. A, 1902-1904 (196'8). 38. A. L. Crumbliss and L. J. Gestaut, J. Coord. Chem., .§_, 109-111 39. Y. N. KozlovandV. M. Berdnikov, Russ. J. Phys. Chem., il, 338-340 (1973). 40. A. D. Zuberbuhler, Helv. Chim. Acta, 59, 1448-1458 (1976). """ 41. M. Guntensperger and A. D. Zuberbuhler, Helv. Chim. Acta, ~' 2584-2594 (1977). 42. H. Gampp and A. D. Zuberbuhler, Chimia, 32, 54-56 (1978). ~ 23 References (continued) 43. J. Zagel, E. Spodine, and W. Zamudio, J. Chem. Soc . .Ihlton, 85-87 (1974). 44. M. G. SimmonsandL. J. Wilson, J.C.S. Chem. Comm., 634-636 (1978). 45. S. J. KimandT. Takizawa, J.C.S. Chem. Comm., 356-357 (1974). 46. C. S. Arcus, J. L. Wilkinson, C. Mealli, T. J. Marks, and J. A. Ibers, J. Amer. Chem. Soc., .,...,.._ 96, 7564-7565 (1974) . 47. C. E. Kramer, G. Davies, R. B. Davis, and R. W. Slaven, J.C. S. Chem. Comm., 606-607 (1975). 48. J. E. Bulkowski, P. L. Burk, M. F. Ludmann, and J. A. Osborn, J.C.S. Chem. Comm., 498-499 (1977). 49. A. H. Alberts, R. Annunziata, and J. M. Lehn, J. Amer. Chem. 99, 8502-8504 (1977). -Soc., """ 50. D. M. L. Goodgaine, M. Goodgame, and G. W. R. Canham, Nature, ,,.._,...,... 222, 866 (1969). 51. L. Graf and S. Fallab, Experentia, ~' 46-47 (1964). 52. M. Munakata, S. Nishibayashi, and H. Sakamoto, J.C.S. Chem. Comm., 219-220 (1980). 53. K. J. Oliver, T. N. Waters, D. F. Cook, and C. E. F. Rickard, Inorg. Chim. Acta, ~' 85-89 (1977). 54. J. H. Darling, M. B. Garton-Sprenger, and J. S. Ogden, Faraday Symposia Chem. Soc., ~' 75-82 (1973). 55. H. Sakurai, A. Yokoyama, and H. Tanaka, Chem. Pharm. Bull. (Tokyo), ll, 2373-2385 (1970). 24 References (continued) 56. F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry," Wiley, N. Y., 4th ed., 1980, p. 156. 57. C. Floriani and F. Calderazzo, J. Chem. Soc. Ser. A., 946-953 (1969). 58. M. R. Paris and D. Aymes, B. S. Chim. Fr., 1431-1444 (1976). 59. A. L. Brumbliss and F. Basolo, Science, W, 1168-1170 (1969); J. Amer. Chem. Soc., 92, .,...,.._ 55-60 (1970) . 60. F. Basolo, B. M. Hoffman, and J. A. Ibers, Acc. Chem. Res., 8, 384-392 (1975). "' 61. J. P. Collman, Acc. Chem. Res., 10, 265-272 (1977). """ 62. J. W. Buchler, Ang. Chem. Int. Ed. Engl., U, 407-423 (1978). 63. B. R. James, A. W. Addison, M. Cairns, D. Dolphin, N. p. Farrell, D. R. Paulson, and S. Walker, in "Fundamental Research in Homogeneous Catalysis," Volume 3, M. Tsutsui, ed., Plenum Press, N. Y., 1978, pp. 751-772. 64. J. H. Wang, J. Amer. Chem. Soc., §.Q, 3168-3169 (1958). 65. C. K. ChangandT. G. Traylor, Proc. Nat'l. Acad. Sci. U.S.A., 70, .,...,.._ 66. 2647-2650 (1973) . C. K. Chang and T. G. Traylor, J. Amer. Chem. Soc., 95, ""' 8477-8479 (1973). 67. O. Leal, D. L. Anderson, R. G. Bowman, F. Basolo, R. L. Burwell, Jr., J. Amer. Chem. Soc., il, 5125-5129 (1975). 68. J. P. Collman, R. R. Gagne, T. R. Halbert, J. C. Marchan, and C. A. Reed, J. Amer. Chem. Soc., ~' 7868-7870 (1973). 25 References (continued) 69. J. P. Collman, R. R. Gagne, C. A. Reed, T. R. Halbert, G. Lang, and W. T. Robinson, J. Amer. Chem. Soc., .,..,... 97, 1427-1439 (1975). 70. G. B. Jameson, G. A. Rodley, W. T. Robinson, R. R. Gagne, C. A. Reed, and J. P. Collman, Inorg. Chem., !1, 850-857 (1978). 71. J. Almog, J. E. Baldwin and J. Huff, J. Amer. Chem. Soc., 97, ,..,,.._ 227-228 (1975). 72. J. E. Baldwin and J. H. Huff, J. Amer. Chem. Soc., .,..,... 95, 5757-5759 (1973). 73. A. H. Corwin and S. D. Bruck, J. Amer. Chem. Soc., .,..__ 80, 4736-4739 (1958). 74. J. P. Collman, R. R. Gagne, and C. A. Reed, J. Amer. Chem. Soc., ~' 2629-2631 (1974). 75. R. H. Niswander and A. E. Martell, Inorg. Chem., !1, 2341-2344 (1978). 76. D. L. Anderson, C. J. Wechsler, and F. Basolo, J. Amer. Chem. Soc., 96, 5599-5600 (1974). """ 77. J. Almog, J. E. Baldwin, R. L. Dyer, J. Huff, and C. J. Wilkerson, J. Amer. Chem. Soc., .,..,... 96, 5600-5601 (1974) . 78. C. K. Chang and T. H. Traylor, J. Amer. Chem. Soc.,~' 5810-5811 (1973). 79. G. C. Wagner and R. J. Kassner, J. Amer. Chem. Soc., ~' 5593-5595 (1974). 26 References (continued) 80. W. S. Brinigar and C. K. Chang, J. Amer. Chem. Soc., .,.._., 96, 5595-5597 {1974). 81. W. S. Brinigar, C. K. Chang, J. Geibel, and T. G. Traylor, J. Amer. Chem. Soc., .,.._., 96, 5597-5599 (1974) . 82. B. Gonzalez, J. Kouba, S. Yee, C. A. Reed, J. F. Kirner, and W.R. Scheidt, J. Amer. Chem. Soc., 97, 3247-3249 (1975). """' 83. C. J. Weschler, B. M. Hoffman, and F. Basolo, J. Amer. Chem. Soc., .,.._., 97, 5278-5280 (1975). . 84. B. M. Hoffman, C. J. Weschler, and F. Basolo, J. Amer. Chem. Soc., .,.._., 98, 5473-5482 (1976) . 85. R. D. Jones, D. A. Summerville, and F. Basolo, J. Amer. Chem. Soc., .,..,,..,,.._ 100, 4416-4424 (1978) . 86. J. H. Wang, Acc. Chem. Res., ...... 3, 90-97 {1970) . 87. D. V. stynes, H. C. stynes, J. A. Ibers, and B. R. James, J. Amer. Chem. Soc., 95, 1142-1149 (1973). """' 88. H. C. stynesandJ. A. lbers, J. Amer. Chem. Soc., 94, """' 5125-5127 {1972). 89. B. M. Hoffmann, D. L. Diemente and F. Basolo, J. Amer. Chem. Soc., 92, 61-65 (1970). """' 90. M. J. Carter, L. M. EngeThardt, D. P. Rillema, and F. Basolo, J.C.S. Chem. Comm., 810-812 {1973). 91. M. J. Carter, D. P. Rillema, and F. Basolo, J. Amer. Chem. Soc., ~' 392-400 (1974). 92. S. K. Cheung, C. J. Grimes, J. Wong, C. A. Reed, J. Amer. Chem. Soc., 98, 5028-5030 (1976). """' 27 References (continued) 93. C. J. Hipp and D. H. Busch in "Coordination Chemistry," Vol. 2, A. E. Martell, ed. , American Chemical Society Monograph f 174, Washington, D. C., 1978, pp. 435-444. 94. V. L. Goedken and D. H. Busch, J. Amer. Chem. Soc., .,.._,.._ 94, 7355-7363 (1972). 95. N. F. Curtis, J.C.S. Chem. Comm., 881-883 (1966). 96. N. F. Curtis, J. Chem. Soc. Ser. A, 2834-2838 (1971). 97. E. G. Vassian and R. K. Murmann, Inorg. Chem., §_, 2043-2046 (1967). 98. E. K. BarefieldandM. T. Mocella, J. Amer:. Chem. Soc., 97, ~ 4238-4246 (1975). 99. D. C. Olson and J. Vasilevskis, Inorg. Chem., 1Q, 463-470 (1971). 100. M. C. Weiss and V. L. Goedken, J. Amer. Chem. Soc., .,.._,.._ 98, 3389-3392 (1976). 101. B. Durham, T. J~ Anderson, J. A. Switzer, J. F. Endicott, and M. D. Glick, Inorg. Chem., ll, 271-278 (1977). 102. D. P. Riley, Ph.D. Thesis, The Ohio State University, 1975, as reported in Ref. 93. 103. K. Bowman, D. P. Riley, D. H. Busch, and P. W. R. Corfield, J. Amer. Chem. Soc. , .,.._,.._ 97, 5036-5038 (1975) . 104. D. P. Riley, J. A. Stone, and D. H. Busch, J. Amer. Chem. Soc., ~' 1752-1762 (1976). 105. C. A. Sprecher and A. D. Zuberbuhler, Ang. Chem. Int. Ed., ll, 189 (1977). 106. F. L. Uroo.ch, U. Knopp, and A. D. Zuberbuhler, Helv. Chim. Acta, .,.._,.._ 61, 1097-1106 (1978) . 28 Chapter Two. Dioxygen Reactions of Two Copper(!) Multidentate Ligand Systems 29 Chapter Two. Dioxygen Reactions of Two Copper(!) Multidentate Ligand Systems Extreme lability, facile disproportionation, and air sensitivity are difficulties which must be overcome when studying reactions of copper(!) complexes. The first two govern the choice of suitable ligands; the last simply requires Schlenck, vacuum-line, and inert atmosphere glove box techniques. Choice of Ligands Both copper(!) and copper(II) are substitution labile. Complexes of monodentate and even bidentate ligands lead to solutions containing several species. Use of polydentate ligands, including macrocycles, inhibits these multiple equilibria and increases the chance of correlation between solution and solid state structures. The composition of the active species is an important question that is difficult to answer for copper oxidation catalysts in general. Polydentate ligands are expected to stay coordinated to the copper even after oxidation has occurred, thus simplifying the problem of copper lability. Many copper(!) complexes disproportionate rapidly to copper(II) and copper(0). 1 Ligand environments having "hard" saturated amines and/or an enforced square planar coordination environment are known to destabilize copper(!) with respect to copper(II), as shown by electrochemical studies. 2, 3 Soft ligands, including sulphur, phosphorous, aromatic amines (pyridine, pyrazolate) and imines tend to favor copper(!). The imidazole ligands found for copper in hemocyanin and tyrosinase, as well as the positive reduction potential· for copper(II) to copper(!) in 30 tyrosinase, suggest that soft nitrogen ligands and flexible· geometries will favor the desired reactions with dioxygen as well as reduce the tendency toward disproportionation. Prior to the start of this study, few multidentate nitrogen ligand copper(!) complexes had been isolated, 4 , 5 except for the substituted 1, 10-phenanthroline, 5 2, 2 '-bipyridyl, 5 and hydrotris(l-pyrazolyl)borate complexes. 6- 9 Only one nitrogen donor copper(!) macrocyclic complex had been isolated. 2, lO The oxygenation studies thus required synthesis of new copper(!) multidentate nitrogen ligand complexes. This chapter discusses the dioxygen reactions of two related multidentate ligand complexes. One is mononculear and the other binuclear (Fig. II-1). Cu Cu CO Cu 0 HN .,.., Fig. II-1. The mononuclear, ~ and binuclear, B, complexes whose oxygen reactions are examined in this chapter. Mu1tidentate Mononuclear Copper(!) Complex i The mononuclear compound, A, .,.., is the copper(!) carbonyl complex of 1, 3-bis(4-methyl-2-pyridylimino)isoindoline. This tridentate ligand is prepared by the condensation of 1, 2-dicyanobenzene with 2-amino-4- 31 methylpyridine. 11 The synthesis and dioxygen reactions of this compound have appeared elsewhere (Ref. 12 included as Appendix 1), but are summarized here. The three coordinate copper(!) complex could not be isolated from solution.. Several attempts to incorporate fourth ligands other than carbon monoxide· were unsuccessfµl. Upon expo- sure to dioxygen, light-yellow suspensions of the carbonyl complex in benzene gradually turn brown. Crystals of the oxygenation products were obtained by slow evaporation. The stoichiometry measured in benzene on a vacuum line indicates 0. 5 ± 0.1 dioxygen consumed _and 0. 91 ± 0. 05 carbon monoxide evolved per copper atom. This stoichiometry is consistent with the formation of a dioxygen bridged dimer, 2Cu(L)CO + 0 2 - (CuL) 20 2 + CO. Further examination indicates the reaction is much more complicated. Crystallographically identified products include a carbonate-bridged dimer (Fig. II-2) and a hydroxy-bridged tetramer (Fig. II-3) in which one-quarter of the ligand molecules have been oxidized to phenolate anions. Large amounts of non-crystalline material have not been identified. Although no dioxygen compounds have been isolated, it is likely that they are involved as intermediates. A plausible mechanism for the formation of the bridging carbonate involves intramolecular insertion of coordinated carbon monoxide into coord~ted dioxygen (Scheme Il-1). Formation of the tetrameric species Scheme n-1 0 II LCu-O-C-OCu L 32 Fig. II-2. The carbonate-bridged dimer from oxygenation of A. " 33 ~ ~v ----- • ~?)3 ~ ~~ .. I; 'i •.' Fig. II-3. The hydroxy-bridged dimer from oxygenation of A; the partial occupancies of the four phenolate anions shaded in the figure sum to one. 34 (CuL) 3 (CuLO)(µ-0H) 3 may also involve a dioxygen intermediate, with copper promoted activation of coordinated dioxygen leading to ligand hydroxylation (Scheme II-2). Scheme II-2 H 0 .cu-oI ~ N -OH I Several important observations concerning this system may be summarized. 1) stoichiometry measurements indicated possible dioxygen complex formation but such was not observed crystallographically. 2) Copper(!) can reduce dioxygen by four electrons resulting in 0-0 bond cleavage, even in an aprotic solvent like benzene. 0-0 bond fission had not been crystallographically shown previously for any copper(!) dioxygen reaction. 3) This work is the first structural characterization of any copper(!) autoxidation product. Binuclear Dicopper(I) Complex ~ A binucleating version of 1, 3-bis( 4-methyl-2-pyridylimino) isoindoline can be synthesized by treatment with hydrazine, yielding 1, 4-bis(4-methyl-2:..pyridylamino)phthalazine. The intermediates of the reaction are probably similar to those observed for phthalimide 13 35 (Scheme II-3). The much less soluble binucleating ligand without the Scheme II-3 methyl substituents has been used to form binuclear copper(II), cobalt(II) and recently, nickel(II) complexes. 14 - 16 In these cases the metal may be four, five, or six coordinate with bridging and terminal halides, hydroxides, or waters completing the coordination. The present ligand coordinates two copper(!) atoms and a product, B, can " be isolated by addition ~f pyrazolate anion (Scheme II-4). Pyrazole· was chosen to complete the coordination environment both to stabilize the coppern) oxidation state and because binuclear copper(!) complexes with a bridging pyrazole had recently been observed to be threecoordinate. 17 Molecular models show that the three-coordinate dicopper(I) complex provides a good fit for dioxygen in a binuclear cis geometry. X-ray structural characterization of the dicopper(I) complex is in progress by R. E. Marsh. Oxygen stoichiometries determined at room temperature give O. 25 ± O. 01 dioxygen taken up per copper(!) atom. The techniques used were stoichiometry in dichloromethane by difference using a Toepler 36 Scheme II-4 HN Cu MeOH ► NaOMe Cu HN 0 37 pump, in propylene carbonate using a Warburg manometer, and in the solid state (again by difference with a Toepler pump). The solution uptakes were rapid (see Fig. Il-4) and presumably controlled by the dissolution of dioxygen. Uptakes were performed in quadruplicate. A stoichiometry of O. 25 dioxygen per copper atom corresponds to simple oxidation of copper(!) to copper(Il) with the reduction of dioxygen to water, hydroxide or oxo species. With such a stoichiometry there can be no ligand oxidation and no dioxygen complex isolated. The infrared spectra show few changes upon oxygenation and the 180 2 16 and 0 2 products are identical. Magnetic susceptibility of the oxygenated solid, µcorr= 1. 60 {3M. per copper (25°, Faraday balance), indicates a copper(II) product as expected. Jf a dioxygen complex is involved, it could be more stable at low temperature (see Chapter One) . . Thus the stoichiometry was measured in dichloromethane at -78 ° C. By difference using a Toepler pump no reaction could ·be detected. The same solutions then gave the expected uptake at room temperature. Since such an observation is also consistent with a reversible reaction at low temperature, a constant pressure measurement in dichloromethane at -78 ° was undertaken. Again no uptake was observed, indicating lack of oxygen reactivity at low temperature. One explanation for this lack of reactivity involves the pyrazolate ligand. Not only would dioxygen binding be expected to be enhanced at low temperature, but so would the pyrazolate binding. If the first step of the dioxygen reaction involves dissociation of the pyrazolate, low temperatures could shut off the dioxygen reaction. 38 u &l IJ a.o :s ~ct Q. ! QQ A7 .. a.-. 0 f O,lf Q3 o.i. 0.1 Fig. Il-4. Dioxygen uptakes of propylene carbonate solutions of B at constant pressure (atmospheric) as a function of time. Three t!'oncentrations are shown; all took up 0. 25 O2 /Cu. 39 The reduction of dioxygen to water or hydroxide requires a proton source. That proton source could be the ligand itself. Consequently the dicopper(I) binucleating ligand pyrazolate tetrafluoroborate complex was treated with sodium methoxide to give the anion of the binucleating ligand and loss of the tetrafluoroborate anion, ~ (Scheme II-4). The stoichiometry determined in propylene carbonate at constant pressure is unchanged (0. 26 ± 0. 02 0 2 /Cu). In dichloromethane, the deprotonated compound was much less soluble. However the original slurry dissolved rapidly upon addition of dioxygen and the uptake as measured by difference wa~. 0. 30 ± 0. 05 0 2 /Cu. In summary, it should be noted that, 1) this binuclear system apparently effects the four-electron reduction of dioxygen, and 2) dissociation of the pyrazolate appears to be necessary for the reaction to occur. Both the mononucleating and binucleating systems have been complicated by having even one dissociable ligand. A more ideal system would have no dissociable ligands and a vacant coordination site. Macrocyclic complexes which meet these requirements are discussed in the next chapter. Experimental: Synthesis and Characterization of B .,...,.,.,....,...,'-"""~W-:,-,,,,._,,"""""~~"'-"'---~-t.,...h__a~lazwin..,...e....... 1, 3-bis( 4-methyl-2pyridylimino) isoindoline (17. 0 g) prepared by the method of Siegl, 11 was dissolved in methanol (300 ml) with heating. Hydrazine hydrate (10 ml, Aldrich) was added, and the solution refluxed for ten minutes. After sitting at room temperature overnight, the solution was reheated for one hour during which time the precipitate redissolved, the solution 40 got darker, and additional ammonia was driven off. Upon cooling to room temperature the yellow needles were isolated by filtration. Yield: 11.7 g, 66%. Anal. Calcd. for C20 H18 N6 : C, 70.16; H, 5.30; N, 24.54. Found:C, 69.90;H, 5.25;N, 24.95. There are at least two isomers of the free ligand. The melting point of the product was usually 98-99 ° C but occasionally material with melting point 174-175° was obtained by the same procedure. Above the melting point of the lower melting isomer the compound resolidifies and then remelts at the higher melting point. Curiously, heating the product above the higher melting point and then cooling results in the lower melting isomer. This cycle can be repeated several times with no obvious degradation. The two isomers have slightly different infrared spectra. At the mass spectrum probe temperature both are converted to the same isomer and the mass spectra ol:tained are identical with large peaks at 341.1 and 342.1 mass units. (The ligand has mass 342.16). Which isomer reacts with copper(!) to give the complex is not determinable since they interconvert in solution and the equilibrium would continually produce more of the reacting isomer until the reaction went to completion. (The copper complexation is somewhat slow). starting with either isomer gave identical products with copper(!) and pyrazolate. The proton magnetic resonance spectra of the starting mononucleating ligand and the expanded binucleating ligand are compared in Fig. II-5. id~lamino )phthalazine pyra- ---~...,...,._.,.__~w"""""""'~~w-"""',...__,""""""""'"""~ zolate tetrafluoroborate. 1, 4-bis( 4-methyl-2-pyridylamino)phthalazine (2. 00 g, 5. 8 mmol) and pyrazole (0. 40 g, 5. 8 mmol) were dissolved 41 I ---- - - -;--·--- ---· - ----,I i - -·- - ' !' • -- - -·-- ---- . ---- ---- ' • . . • ----:--:,- ,- - - i ---·--- .i ----· ------ - - --· ------ta - I .( ! I i ' i e I I c----- ··. -- ·- · ·---· . ... ' fH~: . f -·,i - ·- -·-·-- - -- - · I I I I . . • i . C I - - -· --- i __ I . i I ------ ~~ 10. 9 - r--s Fig. II-5. Proton magnetic resonance spectra of the free ligands of A and Bin dimethylsulfoxide -de· Assignments are as indicated anci chem1cal shifts are downfield from TMS. All peaks shown integrate as 2 except the half-size peak f in the bottom spectrum (the methyl group signals are not shown and integrate as 6H). 42 in hot acetonitrile (150 ml) in an oxygen-free glove box. Upon addition of Cu(CH 3 CN)4 BF4 (3. 68 g, 1.17 mmol) prepared by the published method, 18 the solution turned a dark red. Addition of triethylamine (2 ml) to deprotonate the pyrazole resulted in the formation of an orange precipitate over a period of an hour. The solution was allowed to cool slowly to room temperature after which the product was isolated by suction filtration (3. 29 g, 90% yield). Anal. Calcd. for C23H21 N8 C~BF4 : C, 44.32; H, 3.40; N, 17.98. Found: C, 44.75; H, 3.45; N, 18.25. The infrared spectrum shows 1 vibrations of the BF; group (1050 to .1100 cm- ) and conductivity measurements in acetonitrile indicate at 1:1 electrolyte. amino htha]azine anion Y:razolate. A slurry of dicopper(I) 1, 4-bis- (4-methyl-2-pyridyl amino)phthalazine pyrazolate tetrafluoroborate (1. 0 g) was stirred overnight in methanol (200 ml) in an oxygen-free glove box in the presence of an excess of sodium methoxide. The deprotonated ligand complex was recovered by filtration and washed with methanol. Yield: 0. 63 g, 73%. The infrared spectrum of the product shows no vibration that can be assigned to the tetrafluoroborate anion and conductivity measurements in acetonitrile indicate a neutral compound. The proton magnetic resonance spectra of the free ligand and both copper complexes are compared in Fig. 11-6. 43 ... ·- t -·- ----- . - '' • : - .- - -- . .. ... ·- ·· . - - ! --- ·· .• -- --- 'I • • •• • • - - ·- •• l? - - - - I - • • • ,~~U~ w1,~t," j~, l,.:1,).. ,. 'IO"\T1,~,t'f~ ,,1(1,W'T~ 0 :::,.I I Cu 0 r~,~~ ' I . .. ---r· i ---:-- • -- -- -- ·---- --- - ·· ·- • •. .. . - - ·- i ' . . --·· --- . ·--- -- .. . f ····· --- ' --; . - .. -- i. ,-r-;--- •i· I 10 . . 9. ' -7 Fig. II-6. Proton magnetic resonance of the free ligand of B (bottom); the ligand dicopper(I) BF4 pyrazolate complex, B(center); and the ligand anion dicopper (I) pyrazolate complex, C (top)."' All peaks shown integrate as two protons except those otherwise marked. The solvent is dimethylsulfoxide-<ls, the methyl groups are not shown (integration 6H), and the chemical shift is downfield from TMS. 44 References 1. F. A. Cotton and G. W. Wilkinson, "Advanced Inorganic Chemistry," Wiley-Interscience, N. Y., 4th edo, 1980, pp. 798-821. 2o D. C. Olson and J. Vasilevskis, Inorg. Chem., !Q., 463-470 (1971). 3. G. S. PattersonandR. H. Holm, Bioinorg. Chem., .1, 257-275 (1975). 4. R. Osterberg, Coord. Chem. Rev., 12, 309-347 (1974). 5. F. H. Jardine, Adv. Inorg. Chem. Radiochem., 17, .,._ 115-163 """ (1975). 6. M. I. Bruce and A. P. P. Ostazewski, J.C.So Dalton, 2433-2436 (1973). 7. M. R. Churchill, B. G. DeBoer, F. Jo Rotella, Oo M. Abu Salah, and M. I. Bruce, Inorg. Chem., !1, 2051-2056 (1975). 8. C. S. Arcus, J. L. Wilkinson, C. Mealli, T. J. Marks, and 9. J. A. Ibers, J. Amer. Chem. Soc., 96, 7564-7565 (1974). """ C. Mealli, C. s. Arcus, J. L. Wilkinson, T. J. Marks, and J. A. Ibers, J. Amer. Chem. Soc., 98, 711-718 (1976). """ 10. G. A. Melson, Ed., "Coordination Chemistry of Macrocyclic Compounds," Plenum Press, N. Y., 1979. 11. W. O. Siegl, J. Org. Chem., il, 1872-1878 (1977). 12. R. R. Gagne, R. S. Gall, G. C. Lisensky, R. E. Marsh, Inorg. Chem., 18, 771-781 (1979). Included here as Appendix 1. ""' 13. H. D. K. Drew and H. H. Hatt, J. Chem. Soc., 16-26 (1937). 14. L. K. Thompson, V. T. Chacko, J. A. Elvidge, A. B. P. 45 References (continued) Lever, and R. V. Parish, Can. J. Chem., .,....,... 47, 4141-4152 (1969) . 15. A. B. P. Lever, L. K. Thompson, and W. M. Reiff, Inorg. Chem., .,....,... 11, 104-109 (1972) . 16. J. A. Doull and L. K. Thompson, Can. J. Chem., .,....,... 58, 221-228 (1980). 17. R.R. Gagne, R. P. Kreh, and J. A. Dodge, J. Amer. Chem. -- -Soc., 101, 6917-6927 (1979). 18. B. I. Hathaway, D. G. Holah, J. D. Postlethwaite, J. Chem. Soc., 3215-3218 (1961). 46 Chapter Three. Diaxygen Reactions with Copper(!) Tetraazamacrocyclic Complexes 47 Chapter Three. Dioxygen Reactions with Copper(!) Tetraazamacrocyclic Complexes While both. of the ligands in Chapter Two were multidentate, it was necessary to add an additional liga.ndin order to isolate a copper(!) complex. Dissociation of this additional ligand either because of ]ability or the requirement for a vacant coordination site complicated the results and led to a variety of products. Tetradentate N4 -macrocyclic complexes can solve the problem of dissociable ligands but only one such copper(!) complex had previously been .. 1 isolated. The reactions of dioxygen with several new macrocyclic complexes (Figo m-1) are reported herein. Of particular importance are CuLBF , 2 ! and Cu(I)Me tetraene(Cl0 8 ) , 4 !Q. CuLBF 2 , !, forms a binuclear dioxygen coinplex (Fig. m-2) at -79 ° in ·CH2 C~ solution. Cu(I)Me8 tetraene(Cl04 ), !Q, undergoes relatively clean ligand oxidation at -79 ° in dichlorometh~e to precipitate a ketonated copper(Il) complex (Fig. m-3). The dioxygen reactions of the remainder of the m~cro~0-j, 0~ I ~ F·J-0,.N~ " Fig. m-2. Binuclear dioxygen complex of CuLBF2 , 1. " Fig. m-3. Ketonation of CuMe 8 -tetraene(Cl04 ), 1.Q• 48 r8'9 r8'9 ~ I 9 Cu Cu :::;:::N::r Cu ......-:: ~,F · ~ 11 )::cu::( V 1 . ~,F X ::C X ::C y u 2 "' 3 "' 4 "' E,F A(? - "' ;a; ;;.-N cuX ~ N~ Cu ~ ~ 5 "' 7 "' 8 "' 6 "' 9 "' 10 """ Fig. m-1. Copper(!) tetraaza macrocycles whose oxygen reactions are examined in this chapter. 49 cycles in Fig. m-1 are discussed here in relation to the dioxygen reactions of CuLBF 2 , !., and Cu(I)Me tetraene{ClO ) , 4 8 !.Qo The corresponding copper(JI) complex was synthesized by the Schiff mse condensation of 2, 3-butanedionemonoxime with 1, 3-propanediamine, followed by treatment with copper(II) and boron trifluoride etherate (Scheme m-1). 2, 3 The copper(II) was then reduced electrochemically to copper(!). Electrochemical reduction was found to be both specific and practical and was used to prepare all copper(!) complexes discussed in this chapter and shown in Fig. m-1. CuLBF2 , 1, pre"' • cipitated ·f rom acetone solution upon reduction and Was"isolated by filtration. An important question is whether upon reduction the electron goes into the copper atom or into the ligand. Suffice it to say that the metal atom reacts like copper(!) with both carbon monoxide and dioxygen. This question is discussed further in Appendix Two. CuLBF 2 , !., is coordinatively unsaturated and reacts with a variety of fifth ligands. Both the four-coordinate nature of the reduced species and the five-coordinate nature of the carbonyl adduct have been verified by crystallographic analysis (Fig. m-4). 3, 4 CuLBF !., is , 2 thus a good candidate for study of copper(!) dioxygen reactions, being both a macrocycle with no labile ligands and coordinatively unsaturated. The four-coordinate CuLBF 2 , !., complex reacts readily with dioxygen in dichloromethane solution at room temperature, changing color from deep blue (€ = 10,000 at 6717 nm in acetone) 3 to brown. Immediate purging with nitrogen gas or immediate freeze pump thawing 50 Scheme m-1 . EJ= nc,04 BF.:3·Et2O ):N Cu ~N V ::c e- ~ ~ ~-:::✓-~ ., / ;: .• Fig. m-4. The four-coordinate CuLBF2 , 1 (bottom), and five coordinate CuLBF2 CO (top), as determined by crystallographic analysis. ,· , -~ 52 restores some but not all of the original color to give a dirty blue solution. While this observation is consistent with reversible dioxygen adduct formation, further investigation indicates such is not the case. CuLBF2 , !, Reaction with Dioxygen at -79° CuLBF2 ., !, forms a binuclear dioxygen complex (Fig. m-2) at -79° in dichloromethane. The mechanism in Scheme m-2 is supported by the following observations. 1) stoichiometry determinations by Scheme m-2 Products Products difference using the Toepler pump find 0. 50 ± 0. 06 dioxygen taken up per copper atom at -79 ° in dichloromethane solution. At low temperatures, reaction ks in Scheme IIl-2 predominates. 2) Reaction of deepblue dichloromethane solutions of CuLBF2 , !, with dioxygen at -79° gives a brown solution. If any excess dioxygen present is removed before warming to rooin temperature, the blue color reappears upon warming (Fig. m-5). As much as 40% of the original amount of CuLBF2 has been quantitatively observed by visible absorption spectroscopy (684 run in dichloromethane). Infrared spectra of the material so 53 1iOO 4100 Fig. ID-5. Visible absorption spectroscopy of CuLBF2 , 1, in dichloromethane. All samples are the same concentration in the same cell. Top: CuLBF2 , 1. Center: CuLBF2 , 1, in the presence of dioxygen, spectrum obtained at approxfmately 200° K. Bottom: CuLBF2 , 1, treated with dioxygen at 200°K, excess dioxygen removed, and sample warmed to 25 °. 54 treated verify CuLBF2 , 1, as a major component. The extent of reaction k4 in Scheme m-2 must be small as it does not fit the experimental observation of regeneration of CuLBF2 , !_. At room temperature, reaction k_ 3 followed by~ in Scheme m-2 becomes important. 3) No gas is released as the solution warms and regains its blue color. Reaction k_ 1 in Scheme m-2 is not observed. 4) The electron paramagnetic resonance signal 01::tained in frozen solution following reaction of CuLBF2 , !., with dioxygen at -79° in dichloromethane, or by quick freezing soon after addition of dioxygen at room temperature, does not account for the majority of the copper present (Fig. m-6). The binuclear complex (LBF2 )Cu-O2 -Cu(LBF2) proposed in Scheme III-2 as the major species present at -79° should be diamagnetic by analogy with oxyhemocyanin (see Chapter One), and is in agreement with the observed electron paramagnetic resonance signal. 5) Observation of reversible color changes and small electron paramagnetic resonance signals immediately after dioxygen introduction at room temperature indicate that rate of reaction~ in Scheme m-2 is competitive with~ at room temperature. A sample of CuLBF2 oxygenated at -79 ° in dichloromethane was filtered in a cooled Schlenck apparatus and the recovered solid dried in vacuo at low temperature. The room temperature infrared spectrum of this material is identical with that 01::tained from room temperature reaction products. If the product at low temperature was different from that at room temperature, it has decomposed on warming as a solid. Magnetic susceptibility measurements (25 °, Faraday method) 55 L ----, I ---~ 100 G ~ g = 2. 00 Fig. m-6. Electron paramagnetic resonance spectra in frozen solution._The samples contained the same internal capillary tube filled with a dilute octahedral chromium oxide sample in KCl. Large signal: 2. 3 mM Cu(II) LBF2 ClO4 in EtOH. Small signal: L 8 mM Cu(I) LBF2 in CH2 C~ after reaction with dioxygen at -79°C. 56 indicate a copper(II) product with µ = 1. 88 {3. M.. While the observations at low temperature were encouraging, the decomposition of the solid on warming implies the chance of isolating a dioxygen complex with this ligand is small. Investigation of the room temperature product was undertaken in hopes of identifying the decomposition products so that the ligand could be modified in such a way as to prevent decomposition. CuLBF2 , !., Reaction with Dioxygen at 25 ° The room temperature stoichiometry of the CuLBF2 , !., dioxygen reaction was measured in a variety of aprotic solvents. The values determined fall into four groups (Fig. ll-7). There is no correlation of solvent dielectric constant or dipole moment with stoichiometry. Excluding dichloromethane the groups are acidic, neutral, and basic aprotic solvents. 5 The reaction in nitromethane is much simpler than in the other solvents. In nitromethane, the room temperature stoichiometry is 0. 25 dioxygen per copper atom. Copper(!) has been oxidized· to copper(II) and the four-electron reduction of dioxygen to water, hydroxide, or oxo species has occurred. No ligand oxidation has taken place. The infrared spectrum of the product is that of CuLBF2 ClO 4 1 minus the perchlorate bands plus a very strong band at 1550 cm- • This absorption is assigned to the nitromethanide anion as found in autoxidation studies in nitromethane (see Chapter One), although an 1 expected7 additional strong band at 1270 cm- is not observed. Addition of a tenfold excess of tertiary amines increases the stoichio- 57 :2• 2 • r_-.:.::~ /Dimethylsulfoxide (2. 03) ~2. O·-Dimethylformamide (1. 97) ~~~~-.: ' Pyridine (1. 94) · 1.8 c Basic .- L6 -Acetonitrile (1. 57) L•_.:_ _: "Acetone (1. 54) t_·~~Propylene carbonate (1. 46) _ .:·. .. \'Tetrahydrofuran (1. 44) . hloromethane (1 . 17)- ·-:1.::· 2. oluene (1. 39) D1c l-· Neutral -i.o ~ o. 8 -0.6 ~o. 4 - . = -;_ -- ✓ Nitromethane (O. 25) -0. 2 -o. o- Fig. m-7. Dioxygen per copper stoichiometry of CuLBF2 , 1, in various solvents at room temperature. " Acidic 58 metry in nitromethane to 0. 5 dioxygen per copper atom. The different stoichiometries are not due to the increased basicity of the solution since addition of sodium hydroxide has no effect on the stoichiometry. Trialkylamines are known to react with hydrogen peroxide to form trialkylamine oxides which react with hydrogen peroxide to form hydrogen peroxide adducts. 6 In either case, both triethylamine and triethylenediamine (also known as 1, 4-diazabicyclo{2. 2. 2)octane) remove hydrogen peroxide and only two electrons rather than four are available from each dioxygen for copper cxidation. This results in a doubled oxygen uptake. Hydrogen peroxide must be present as an intermediate in the oxygenation of CuLBF2 in nitromethane. Excluding nitromethane, the stoichiometry values are all greater than 1. 0 dioxygen per copper atom. stoichiometries higher than 1. 0 dioxygen per copper require that oxidation of the ligand or solvent is occurring. The more basic the solvent, the more such oxidation occurs. Evidence that the . copper atom is involved in the oxygenation reaction, and that oxygen is activated toward substrate oxidation, is seen in. studies of the dioxygen reaction with CuLBF2 in the presence of triphenylphosphine. Triphenylphosphine forms a 1:1 complex with CuLBF2 as determined by Job plcts and by isolation of the solid adduct. Note that under the conditions of the experiment (room temperature dichloromethane solutions), triphenylphosphine alone does not react with dioxygen. With a 10:1 excess of triphenylphosphine to CuLBF2 , the oxygen reaction is almost stopped (uptake O. 08 ± O. 05 dioxygen per copper atom after 48 hours); with small relative amounts the uptake is normal; but in the region of one or two triphenylphosphine per CuLBF2 59 the uptake is enhanced (Fig. m-8). These observations can be explained at high triphenylphosphine to CuLBF2 ratios by blocking of the reaction at copper(!) by triphenylphosphine since the equilibrium is shifted such that there are almost never any four coordinate copper atoms to react with dioxygen. At the ratios around one to two, triphenylphosphine is acting as a substrate for some "Cu-O2 " species. Indeed, triphenylphosphine oxide has been identified in the infrared spectrum of the products. At low ratios CuLBF2 reacts as it does in the absence of triphenylphosphine. The longer the reaction is allowed to run and the small equilibrium concentration of uricomplexed CuLBF2 reacts, the more to the right and the higher the stoichiometry curve in Fig. III-8. Except for the product in nitromethane, oxygenation of CuLBF2 seems to give polymeric products as once they are removed from solution they are insoluble in acetone, ethanol, dichloromethane, dioxane, acetonitrile, and tetrahydrofuran. The product is soluble in dimethylsulfoxide and dimethylformamide but a satisfactory method of recrystallization has not been found. Attempts to remove the copper(II) atom or reduce it back to the copper(!) oxidation state to permit a useful nuclear magnetic resonance spectrum of the oxygenated product have not been successful. Slow Dioxygen Reactions of Copper(!) Complexes 1 and§. Two copper(!) macrocyclic complexes, 1 and§_, .were synthesized by D. M. Ingle 8 and found to react quite differently with dioxygen. The macrocyclic complex derived from 1, 4-diaminobutane, 3 (Fig. m-1), and that derived by addition of two boron difluoride "" 60 ,.s ., ~ • Z.~ho\l.-S 0... 0... 0 u ..._ ., 1.0 3' n F-'.F 4 ~ :j c:: ">-en ~ Q5" ):;:eu::C V j_ 2. -t-r-i.phe.nylphosP'ine. / c.oppe.r Fig. m-8. Dioxygen per copper stoichiometry of CuLBF2 , 1, versus triphenylphosphine per copper present. The different curves are for the same concentrations but different reaction durations. 3 61 bridges to the bis-dimethylglyoxime complex, ~ (Fig. m-1), react much slower than CuLBF2 , !, with dioxygen. Dichloromethane solutions of these complexes stay the deep-blue color of the copper(!) complex for over an hour in the presence of dioxygen and are blue well past the first equivalent of dioxygen consumed. The blue color is gone from CuLBF2 , !, solutions within seconds after the addition of dioxygen. Although the oxygen uptake is slower, the . amount of oxygen taken up is greater for 3 and 5 than for CuLBF2 , 1 (for example, see "" """' • A Fig. _III-9). •' Alternatively, the copper(II) macrocycles may be reacted with pctassium superoxide. The dimethylglyoxime and diaminobutane derivatives give the intense blue color of the copper(!) oxidation state on treatment with potassium superoxide, but Cu(II)LBF2 ClO4 gives the brown color of the oxygenation product. These observations require expanding Scheme m-2 to Scheme m-3. For the dimethylglyoxime and Scheme m-3 Products diaminobutane derivatives k_ 1 must be at least competitive with ~ for 62 2. LS / I LO I I I / I I I I QS I I / I I I I I I I 50 Fig. m-9. Dioxygen uptakes in dichloromethane solution at constant pressure (atmospheric) as a function of time. The solid curve is that found for CuLBF2 , 1, and the dashed curve is that found for the diaminobutane derivative, 3. Note that the reaction of 3 with dioxygen is much slower. "' .,... 300 63 the copper(!) complex to be observed. CuLBF2 , !_, Ligand Modification On the msis of other macrocycle oxidations (see Chapter One), ketonation was suspected as one of the macrocycle decomposition products. CuLBF2 macrocycle decomposition upon oxygenation involving the center carbon can be ruled out by the similar dioxygen reactions of the macrocycle synthesized from 2, 2-dimethyl-l, 3diaminopropane, 2 (Fig. III-1). The comparison between the diamino"' propane, 1, and diaminobutane, 3, derived macrocycles is somewhat "' "' questionable due to the differing reaction rates. Nevertheless, the lack of a dioxygen reaction for the macrocycle derived from 1, 1, 4, 4-tetramethyl-l, 4-diaminobutane, 4 (Fig. m-1), implicates the saturated "' carbon atoms next to the nitrogen atoms as the culprits. It has proved impossible to synthesize the 1, 1, 3, 3-tetramethyl-l, 3-dia~inopropane necessary to verify this hypothesis. 8 Different macrocyclic systems were considered for study since the CuLBF2 ligand substitution .possibilities were exhausted. The macro- cycle 5, 7, 7, 12, 14, 14-hexamethyl-l, 4, 8, 11-tetraazacyclotetradeca1,4, 8, ll-tetraene (shown as the copper(!) complex,!!, in Fig. m-1), has two desirable traits for an oxygenation study. The macrocycle has undergone all the oxidative dehydrogenation possible (see Chapter One) and further substituted macrocyles can be envisioned (vide infra). One drawback was that neither the copper(!) nor the copper(II) complex had 64 been synthesized previously. Condensation of four acetone and two ethylenediamine in the presence of two equivalents of perchloric acid yields the Me 6 diene macr·o cycle as the dihydroperchlorate (Scheme m-4), and has the advantage of giving only the trans diene isomer. 9 -ll (Template reactions of acetone with ethylenediamine complexes of nickel(m and copper(II) give several isomeric macrocyclic products). 9 , 12 - 17 Reaction of the ligand dihydroperchlorate with copper(II) yields the Cu(Il)Me 6 -diene(ClO4) 2 complex (Scheme m-4). As shown in Scheme m-4, a chemical oxidation procedure to convert Cu(Il)Me 6 -diene(ClO4 ) 2 to Cu(Il)Me 6 -tetraene(ClO4) 2 has now been found. Treatment of Cu(Il)Me 6 -diene(ClO4) 2 with excess permanganate in a perchloric acid solution oxidatively dehydrogenates the ligand and yields the salmoncolored copper diperchlorate complex. Cu(II)M.e 6 -tetraene(ClO4 ) 2 does not have N-H stretching bands in the infrared spectrum at 3220 cm-1 as does the Cu(Il)Me 6 -diene(ClO4) 2 , and the new imine groups result in 1 a band at 1656 cm- in addition to that found in Cu(II)·Me 6 -diene(ClO4 ) 2 at 1672 cm- 1 (Fig. m-10). The reduction of copper(II) to copper(!) is again done electrochemically in acetone, but the soluble Cu(I)Me 6 -tetraene{ClO4) requires careful addition of toluene to induce precipitation of the dark-red solid. The Me 6 -tetraene identity of the ligand is conclusively demonstrated by the proton magnetic resonance spectrum. Similar spectra have been 18 observed for [Ni(II)Me 6 -tetraene] (ClO4) 2 , [Fe(II)Me 6 -tetraene19 (CH CN) ] {ClO ) , 19 and [Co(III)Me -tetraeneBr ] (ClOJ. The 3 2 4 2 6 2 65 Scheme ID-4 lCu(CIO4)2 • ~ e- C :S Cu ~ Cu Me6 -diene (ClO4 ) 7 "' CuMe 6 -tetraene (ClO4 ) 8 "' (cro4)2 66 --~ ..... ~~ -: = .... ...;. . .;. • - - - ----' . ~- ~ ~ - - ;-t""' :!=t t:::: - ~ ~ ~ ~ . - - _:__ ._1:---::: - \·11 .:.. ·t-- ,___.__ -,-,---__; w~-- -:-.-. . ; '. -1----! ,..;.T- ; ~ -=~- ~ _;..... -:-+- ~ I ~ -= .- - ~ +;i-L~- - ' ' .: ' ' --~;. ~~!. -.. .;. ~ t,-,.;· ---t- '--+ ~ ....J..a. I ,~ I - ___. · ·--+--- ; -·..L ....; ...........::- __:_: --·: :~~~ . . ·' ' · :~ 3200 ,000 2100 1800 1700 .--.;..:.:~-- ~ 1600 : _t,- • 1500 1400 Fig. m-10. Infrared spectra in the regions 1300 to 1850 cm-1 and 2700 to 3600 cm •1 of Cu(II) Me 6 -diene (Cl 0 4 ) 2 (top) and Cu(II)Me6 tetraene (Cl 0 4 ) 2 (bottom). 67 copper(!) and nickel(II) proton magnetic resonance spectra are shown in Fig. m-11. The ·spectra are simpler than for the diene complexes; 20 oxidation of Me6 -di~ne to Me8 -tetraene removes the asymmetric atoms, and the chelate rings appear to undergo rapid flexing giving equivalent gem methyl groups. Dioxygen Reactions of Cu(I)Me 6 -tetraene(ClO4), !! Upon addition of dioxygen to dichloromethane solutions of Cu(I)Me 6 tetraene( ClO4), the orange-red solution turned green and a brown product precipitated in greater than 80% yield..by weight. Magnetic susceptibility measurements (25 °, Faraday method), indicate copper(II) was formed, as µ = 1. 6-1. 7 {3.M .. The oxygenated product produced in a fairly clean reaction is the result of ketonation of the macrocycle (Scheme III-5), based on the following observations. 1) The electron Scheme m-5 yy ~Cu~ AA paramagnetic resonance spectrum of Cu(II)Me6 -tetraene(ClO4 ½ in frozen propylene carbonate solution shows good resolution and superhyperfine detail (Fig. m-12). The same spectrum is also obtained 68 'YY e C (Ct04)2 c:Ni): < J PP"" ) d b e~c a d 0(I.) :E Q e Figure m-11. Proton nuclear magnetic resonance spectra of Ni(II) Me6 -tetraene (ClO4 ) 2 (top) and Cu(I) Me 6 -tetraene (ClO4 ) (bottom). (I.) :E f-, C Cl) C d a b 0 ~ .,, I.) 69 Fig. ID-12. Electron paramagnetic resonance spectrum obtained in frozen propylene carbonate solution of Cu(II) Me6 -tetraene (ClO4 ) 2 or Cu(I) Me6 -tetraene (ClO4 ) reacted with dioxygen under a variety of conditions (see text). 70 in frozen 16 o.i and 1 t>2 treated propylene carbonate solutions of Cu(I)Me6 -tetraene(ClO4 ), ~' as well as frozen propylene carbonate solutions of the product of the reaction of dioxygen with solid .,.._ 8 and the precipitate from the reaction of dioxygen with 8 in dichloromethane solution at -79 °. Thus the " coordination around the copper atom has not changed and the product under a variety of conditions is uniform. 2) The infrared spectrum of the oxygenated solid Cu(I)Me 6 -tetraene(ClO4 ), _!!, or the precipitate upon oxygenation in dichloromethane at -79 °, has new bands at 1580 1 and 1620 cm- in comparison with Cu(II)Me 6 -tetraene(ClO4 ) 2 • The Me 6 ligand, eve;n after having been previously treated with permanganate and perchloric acid, has been further oxidized! Both of these 1 bands shift 5-10 cm- when 18 0 2 is used instead of normal abundance dioxygen (Fig. m-13). (The observed frequencies as well as the shift upon isotopic substitution are low for a pure carbonyl vibration. The carbonyl vibration must be part of a larger vibration both decreasing the frequency and lessening the shift upon isotopic substitution). 3) Cu(I)Me 6 -tetraene(ClO.J irreversibly takes up 1. 25 ± 0. 05 dioxygen per copper atom in dichloromethane solution at -79 °. The same irreversible stoichiometry is obtained for the solid at room temperature. Elemental analyses of the oxygenated solids confirm both oxygen incorporation and the 1. 25 dioxygen per copper atom stoichiometry. (Calcd for CuMe 6 -tetraene(ClO4 ) + 1. 25 0 2 : C, 40. 09; H, 5. 89; N, 11. 69. Found: C, 39. 95; H, 5. 50; N, 11. 55). 4) Conductivity measurements show the product is a 2:1 electrolyte for a molecular 71 1700 1600 1500 1400 WAVENUMBER CM-1 Fig. III-13. Superimposed infrared spectra of the product with Cu(I)Me 6 -tetraene(ClO4 ). 16 1300 0 2 and 18 02 72 weight of about 490 mass units, determined by a plot of the equivalent conductance versus the square root of the equivalent concentration giving a slope of -608 in methanol and -425 in nitromethane. 21 The products are not bridging oxo or hydroxide species. The solubility in other polar organic solvents also supports non-polymeric products. ~igand Modification of CuMe 6 -tetraene (ClO 4 ), ~ For Me 6 -tetraene there are two possible positions for the ketone (Scheme m-5). The ketonation could be blocked by methyl substitution at those positions. Use of 2, 3-diaminobutane in place of ethylenediamine in an attempted macrocycle dihydroperchlorate synthesis gave no reactiono Similar template syntheses have previously been reported to be unsuccessful. 13 , 22 Template reactions of acetone with 1, 2-diaminopropane complexes of nickel(Il) and copper(II) give isomeric macrocyclic products. 13 , 22 - 24 These isomers could include the cis_and trans diene isomers observed for Me 6 -diene, and cis and trans isomers of the two additional methyl groups, not to mention the C-racemic and C-meso isomers associated with the two asymmetric carbon centers of the diaminopropane residues and the isomers associated with the amine nitrogen atoms (Fig. m-14). Fortunately, only the C-meso C-trans trans-diene (Ci) and C-racemic C-trans trans-diene (perpendicular C2 ) isomers have been observed (Fig. m-14, upper left). The C-trans and trans-diene configuration is confirmed by a study of the proton magnetic resonance 25 spectra of the nickel(II) complexes. The crystallographic space group requires the C-meso isomer to be centrosymmetric. A complete 73 ~AAAAAA C-racemic AA C-meso AA · AA AA Y'r;~~~ rr ✓~g ✓~ Fig. m-14. Possible isomers from the 1, 2-diaminopropane acetone condensation (N-H isomers ignored). 74 crystallographic structure has been determined for the C-racemic isomer. 23 , 26 Condensation of four acetone and two 1, 2-diaminopropane in the presence of two equivalents of perchloric acid slowly yields the Me8-diene dihydroperchlorate and has the advantage of giving the 3, 10-methyl substituted C-meso isomer exclusively (Scheme m-6), 9 , 23 , 24 as confirmed by comparison of the published proton magnetic resonance spectra of the orange nickel(II) complexes. 25 The C-meso dihydroperchlorate macrocycle will also yield C-meso copper complexes. Treatment of the purple Cu(II)Me8 -die•ne(ClO4 ) 2 complex with excess permanganate ·and perchloric acid gave salmon colored Cu(II)Me8 tetraene(ClO4 ) 2 • Reduction of the copper(II) compound to copper(!) was achieved electrochemically in acetone and the dark-red product isolated by careful addition of toluene. For Me 8 -tetraene, the gem methyl groups are no longer equivalent as can be seen in the proton magnetic resonance spectra of the copper(!) and nickel(II) perchlorate complexes (Fig. ID-15). Reaction of Cu(I)Me8 -tetraene(ClO4 ), !Q, as a solid or at -78° in dichloromethane solution with diaxygen, gave stoichiometries of 1. 25 dioxygen per copper atom and brown products, just as found for Cu(I)Me 6 -tetraene(ClO4 ), _!!. . The electron paramagnetic resonance spectra of oxygenated Cu{I)Me8 -tetraene{ClO4), !Q, produced under various conditions are also identical with those for oxygenated Cu{I)Me 6 tetraene{ClO4 ), ~- The infrared spectrum of the oxygenated Cu{I)Me8 tetraene{C1O4), !Q, contains only the band at 1580 cm -i and not the 1 mnd at 1620 cm-1. The 1580 cm- band shifts when 18O2 ·is used 75 Cu C-meso Me 8 -diene(Cl04 ) 9 Cu C-meso Me 8 -diene (C104 )2 Cu C-meso Me 8 -tetraene(Cl04 ) Cu C-meso Me 8 -tetraene (C104 ) 2 10 --- 76 g d r i I i ! i -- - -l- ---- f . I i I . _ _.( C I ( . - --- - ---- la I - ____ j _ _ ! . 1 ppm . ...0 e i N = ; If _ • • •~i d yy CI04 't:cu~ g 1 r J C t f~d 1 ppm e a e b Fig. m-15. Proton nuclear magnetic resonance spectra of Ni(Il) Me 8 tetraene (ClO4 )2 (top) and Cu(I) Me 8 -tetraene (ClO4 ) (bottom). 77 (Fig. m-16). As predicted there is only one possible site for ketonation in Me 8 -tetraene since only one band appears in the 1600 cm -l region. That band shifts with 18 0 2 substitution. For the macrocyclic ligands the mechanism in dichloromethane must include copper(!) being oxidized to copper(II), such that any given macrocycle is only oxidized once. This is necessary in order to be consistent with the stoichiometry. The presence of added copper(II) macrocycle as a substrate in no case increased the stoichiometry value. The two new bands observed in the infrared spectrum of oxygenated Cu(I)Me 6 -tetraene(ClO4 ) have a somewhat variable relative intensity. This agrees with only one ketone per macrocycle and a random choice between the two positions. The new band upon oxygenation of Cu(I)Me8 tetraene( ClO4), !Q, does not exhibit variable intensity. Ligand modification at the last remaining position of attack is difficult. A template synthesis using isobutyraldehyde is known to give the half condensed- ligand but not the macrocycle (Scheme m-7). 27 Scheme m-7 M= Cull Nill ' A synthesis using isobutyraldehyde, ethylenediamine, and perchloric acid was attempted. After reacting together for several weeks a small amount of product was attained which proved to be Me 6 -diene(HC1O4 ) 2 ! It does not seem likely that the last position can be blocked on the 78 1700 1600 1500 1400 WAVENUMBER CM-1 1300 Fig. m-16. Superimposed infrared spectra of the 16 0 2 and 180 2 product with Cu(I)Me 8 -tetraene(ClO4 ). 1200 79 macrocycle by this route. Dioxygen Reactions of Cu(I)Me 6 -diene(ClO4 ), 1, and Cu(I)Me8 -diene(ClO4 ), 9 Both copper(!) complexes .,.... 7 and .,.... 9 (Fig. ID-1) can be prepared by electrochemical reduction in acetone of the corresponding copper(II) complexes and the mono-acetone solvate recovered by filtration. At room temperature both solids absorb dioxygen while slowly changing from yellow to green to purple. Magnetic susceptibility measurements (25 °, Faraday method) identify a copper(II) product with µ = 1. 80 /3. M. for 1 and µ = 1. 76 (3. M. for ~- A significant change in the macrocycle can be seen by comparing the infrared spectra of the copper(II) complexes with the oxygenated copper(!) complexes. The N-H stretching bands have decreased in intensity in both. Two new bands have appeared near 1600 cm- 1 for Cu(I}Me 6 -diene(ClO4 ), 1, but only one band is present for Cu(I)Me8 -diene(ClO4 ), ~- The infrared evidence suggests that the diene ligands undergo both partial oxidative dehydrogenation and ketonation. Preliminary Examination of Dioxygen Reaction with Ni(I)Me8 -tetraene(ClO4) Ni(I)Me 8 -tetraene(ClO4 ) was synthesized by the nitric acid oxidation of Ni(II)Me 8 -diene(ClO4 ) 2 to the tetraene, followed by electrochemical reduction in acetonitrile. The black Ni(I)Me 8 -tetraene(C1O4) obtained on slow addition of toluene is extremely oxygen sensitive and reacts in acetonitrile-toluene solution with dioxygen at less than 5 ppm in our helium atmosphere glove boxes. The successful synthesis 80 involved electrochemical reduction in the glove box and transfer of the solution to a high vacuum line for precipitation of the solid. The nickel(!) complex reacts rapidly with dioxygen in the solid state, becoming brown within seconds after air exposure. The oxygenated 1 products have a new band at 1580 cm- in the infrared spectrum in comparison with Ni(II)Me 8 -tetraene(ClO4 ) 2 • This band shifts upon 18 0 2 substitution just as found for the copper(!) complex (Fig. m~l 7). - Examination of the nickel complexes was designed to give further proof of ketonation by the distinctive 13 C nuclear magnetic resonance chemical shift of a ketone carbon. Unfortunately, a substantial amount of paramagnetic material is present in the oxygenated complex and no nuclear magnetic resonance spectra could be 01::tained. Discovery of a Copper(!) Complex, §, Which Does Not React With Dioxygen A tetradentate ligand derived from the bis(pyridine-2-aldoxime) complex by addition of a boron difluoride bridge is closely related to the tetradentate macrocyclic complexes. The copper(!) complex, §_, (Fig. m-1), was isolated by filtration after the one-electron reduction of the copper(II) complex in acetonitrile. Successful copper(!) synthesis is confirmed by the proton nuclear magnetic resonance spectrum (Fig. m-18'). This copper(!) complex, 6, gave a most surprising "' reaction with dioxygen. It didn't. No uptake was observed by difference on the Toepler pump after 48 hours in dichloromethane. The visible spectrum of methanol, acetonitrile, and acetone solutions are not changed upon addition of dioxygen. Cyclic voltammograms of the copper(!) complex, 6, are unaffected by the presence of dioxygen. "' 81 18Q 2 1700 1600 1500 1400 WAVENUMBER CM-1 1300 Figo ID-17. Superimposed infrared spectra of the 16 0 2 and 180 2 product with Ni(I)Me,rtetraene(ClO4}. 1200 82 H 0 . I ~N I . I --· 9 -- --. _ I 8 I 71 Fig. m-18. Proton nuclear magnetic resonsnce spectra in dimethylsulfoxide-d6 of pyridine (2-aldoxime) (oxime H not shown), (bottom), and the copper (I) complex 6, (top). . The chemical shifts are downfield from TMS. All peaks integrate equally except for the largest "doublet" which integrates in both as twice the area of the others. 83 One possible reason no oxygenation was observed is that the ligand could not ''trap" the oxygen once it interacted with the copper atom. Triphenyl phosphine (1. 5 equivalents) was added as a substrate but no oxygen uptake was detected. Other reasons why no dioxygen reaction was observed stem from the presumably more tetrahedral coordination geometry. Perhaps the copper(!) atom is no longer coordinatively unsaturated. Electro- chemical measurements show a 50 mV shift under carbon monoxide. Using a solubility of carbon monoxide .. in acetonitrile 28 of 8. 7 x 10-3 M 2 yields a binding constant of 6. 9 x 10 (for the method of calculation see Ref. 29). Five coordinate complexes are formed with this ligand, although not as strongly as with CuLBF2 or CuMe 6 -tetraene(ClO4 ). A second reason is the potential itself. The copper(!) state may be so favored (E = +O. 04 V in acetonttrile vs. SHE) that no "redox" reaction with dioxygen occurs to give an adduct. The reduction potential for this complex is much more positive than that of the macrocycles examined here (see Appendix 2, Table 2). Summary In summary, studies of the CuLBF2 and Cu tetraene(ClO4 ) systems provide complementary information. A dioxygen containing binuclear intermediate can be frozen out for CuLBF2 , even if the reaction is not reversible and decomposition occurs on warming. Cu tetraene(ClO4 ) also has a low temperature intermediate which is only observed as a fleeting blue color. In return, the oxidation of the 84 macrocycle is more amenable to study. When one of the two available ketonation positions is blocked by the methyl derivative, only one of • the original 18 0 2 .shifting ~rared bands is observed. Indications are that ketonation is also involved for CuLBF Cu(I)Me -diene(ClO 6 4 ) , , 2 !, as well as in 7, .and Cu(I)Me -diene(ClO ,.., 8 4 ) , 9. ,.., also appear to undergo some oxidative dehydrogenation. The latter Preliminary indications are that the nickel(!) complexes will react similarly. In addition a more tetrahedral tetradentate ligand copper(!) complex has been found to not react with dioxygen. Experimental Proton magnetic resonance spectra were obtained on a 90 MHz Varian EM-390 spectrometer. Carbon-13 magnetic resonance spectra were oltained overnight in 12 mm sample tubes for CuLBF2 and its adducts using a Varian XL-100 spectrometer, and after several hours in 5 mm tubes for the ~iene and tetraene complexes using a Jeol FX-90Q spectrometer. Tetramethylsilane was present as an internal standard in all magnetic resonance samples. Electron paramagnetic resonance spectra were obtained on a modified Varian E-line spectro.meter equipped with a Hell-trans low temperature dewar. Spectra were obtained at liquid nitrogen temperatures. Room temperature magnetic measurements were determined on a Cahn Faraday magnetic susceptibility balance. Solution conductivity was measured using a Yellow Springs Instruments Model 31 conductivity bridge. Electrochemical measurements utilized a Princeton Applied Research 174A Polarographic Analyzer. Electrochemical syntheses were accom- 85 plished with a Princeton Applied Research model 173 potentiostat . equipped with a model 179 digital coulometer. Infrared spectra were attained via KBr pellets on Beckman Instruments IR-12 and 4240 spectrometers. Visible absorption spectroscopy measurements were recorded on a Cary-14 spectrometer. Operations requiring an inert atmosphere were performed in a Vacuum Atmospheres glove box filled with helium, or on a high vacuum line. Stoichiometry by difference. An accurately weighed 10-4 mole sample and a magnetic stir bar were placed in a 5 ml round-bottom flask attached to an 8 ml calibrated volume. After evacuation of the system 4 ml of freshly degassed solvent was vacuum transferred onto the sample. The sample chamber was closed off, the calibrated volume filled with a known pressure of dry gas (500 to 700 Torr) and the temperature recorded. For low temperature stoichiometry determinations a dry ice acetone slush was placed around the sample chamber at this point. The calibra~ed volume was then opened to the sample chamber and the oxygenation reaction begun. The dioxygen pressure was about 250 to 350 Torr and the sample was magnetically stirred overnight during the reaction. After the reaction the sample chamber was opened to a degassed Toepler pump, with the solvent being caught in an intervening liquid nitrogen trap. The remaining gas was collected in a 14 ml calibrated volume and the temperature recorded. The amount consumed is the initial amount minus the final amount recovered. A propagation of error analysis suggests that for the parameters used, an error of ±0. 03 to 0. 07 0 2 per Cu is expected. A disadvantage of this method is that any dioxygen reversibly 86 bound will be removed and collected as not having reacted. The amount of time for the reaction to go to completion is not known, so duplicate runs of varying duration were always performed. Any leaks in the system will result in too much gas being collected, or equivalently, too little gas being consumed. The most likely errors thus lead to low stoichiometry values of dioxygen consumed per copper atom. The solvents used for stoichiometry determinations were dried as indicated before vacuum transfer to the reaction vessel: CH C~ (CaH 2 ) , 2 CH N0 (CaC~), acetone (4A mol sieves, CaS0 3 2 tetrahydrofuran (Na, CaH CaH ) , 2 ) , 4 ) , 2 pyridine .(KOH), CH CN (4A mol sieves, 3 dimethylsulfoxide (NaOH, vacuum distillation), toluene (H S0 2 , 4 CaH 2 , distillation), propylene carbonate (as attained), dimethylformamide (4A mol s_ieves, vacuum distillation). stoichiometry at constant pressure was recorded using the apparatus shown in Fig. llI-19. Water from a constant temperature ooth at 25 ° was circulated around a gas burette, a pressure reference manometer, and the reaction vessel. The solvent and a stir bar were man eter Fig. m-19 . . Constant pressure oxygen uptake apparatus. placed in the reaction vessel. While stirring vigorously, dioxygen was forced through the line and out the spoon joint in order to saturate 87 the solvent with dioxygen. A known amount of sample was placed in the spoon, the dioxygen flow stopped, and the spoon placed in the apparatus. When the system was at atmospheric pressure and equilibrium, as judged by the even levels in the silicone oil-filled manometer, the initial volume on the gas burette was recorded and the spoon rotated so the sample fell into the solution. The Hg bulb was manually raised and the uptake measured by the change in volume necessary to maintain the constant pressure. Atmospheric pressure was recorded by a separate measurement. This method, while useful for mJ?asuring reversible oxygen uptake, requires compounds which do not react in the solid state. Solvents with low vapor pressure like propylene carbonate are to be preferred. Preparation of Compounds General electrochemical reduction methods. Platinum gauze working and auxiliary electrodes were placed in the 25 ml end compartments of a three compartment H-cell having medium porosity sintered glass frits between compartments. The pseudo-reference electrode, Ag, supporting electrolyte, solvent //, where the junction is again a glass frit, was located in the working compartment such that it was not between the working and auxiliary electrodes. Supporting electrolyte (0. 05 M tetrabutylammonium perchlorate, dried exhaustively in vacuo before· use) was used only in the auxiliary compartments.. If the reduced compound could be isolated by filtration, 0.1 M supporting electrolyte W?,S used in all compartments. In practice the compound to be reduced 88 was: added, a cyclic voltammogram on a platinum button electrode ottained to verify solvent purity and correct functioning of the reference electrode, and the electrolysis begun. The working compartment was vigorously stirred magnetically. The electrolysis was stopped when the current fell to less than 1 mA, the resting potential of the cell was past the wave, and the desired number of coulombs had been passed. Coulometry was performed for all reductions since occasionally the current would fall off due to coating of the electrode. A cyclic voltammogram of the reduced material was ottained to check for chemical stability during the course of the several hour electrolysis. All reductions were performed in an oxygen-free helium atmosphere glove box. 4 8-Diaza-3 To a warm solution of 100 g 2, 3-butanedione monoxime in 250 ml of ethanol was added 50 ml of 1, 3-propanediamine. The volume of the solution was reduced ,on a rotary evaporator at 60° until a precipiate began to form (at about 200 ml total volume). After cooling to room temperature, the solid yellow mass was broken up, thoroughly washed with ether until white, and air dried. Yield: 86 g. A proton nuclear magnetic resonance spectrum of the product shows virtually no oxime starting materialo Preparation of the product by the published method 3 consistently results in at least a 2:1 mixture of starting oxime to desired product as determined by integration of the proton nuclear magnetic resonance methyl group signals. 89 4, 8-Diaza-3, 9-dimethylundeca-3, 8-diene-2, 10-dione oxime oximato Co rchlorate rate. A warm solution of 25. 0 g Cu(ClO4) 2 • 6H2O in 100 ml acetone was added to a warm solution of 24. 0 g of the previous compound in 200 ml acetone. (The Cu(ClO4 ) 2 • 6H 2O was exhaustively dried in vacuo prior to use.) The solution was stirred and heated for a few minutes and then set aside to let cool. The product was isolated by suction filtration and washed with ether. Yield: 23. 5 g. tetraaza-2 14-dioxo-cyclotetradeca-3 5 10 12-tetraenato co er(II) perchlorate] Monodioxane, (CuLBF )iC1O 2 ) 4 • 2 C HO 4 2 • 30 g of the previous compound and 600 ml dioxane were heated to 70° C with stirring in a one-liter Erlenmeyer flask. 24 ml of freshly distilled boron trifluoride etherate was added slowly. The solution was left at 100° C for one hour with stirring, and then allowed to cool. The next morning the product was filtered off and washed twice with dioxane, twice with ether, and air dried. Yield 37. 5 g. Recrystallization was accomplished by dissolving in a minimum of boiling acetone, filtering hot, and adding a small amount of dioxane to the solution. Upon cooling, the product was recovered by suction filtration and washed with diaxane and ether. Because of the intense color, and the difficulty in determin_ing the minimum amount of solvent necessary, several smaller batches are to be preferred to one large ootch. Yield was about 65%. 90 [1, l-difluoro-4, 5, 11, 12-tetramethyl-l-bora-3, 6, 10, 13-tetraaza-2, 14-dioxo-cyclotetradeca-3, 5, 10, 12-tetraenato] Co ~ er I), 0.8 g CuLBF2 (ClO4 ) 2 •C4.HaO2 reduced by constant poten- tial electrolysis in acetone at out to -1. 0 V vs. Ag (or about -0. 7 V vs. SHE), such that the maximum current allowed was 30 mA. After the electrolysis was complete (n =1. 0), the precipitate was filtered from solution, washed 2 x 5 ml with oxygen-free ethanol, and vacuum dried. Yield: 0. 37 g. (1, l-diiluoro-4, 5, 11, 12-tetrameth 1-1-bora-3 6 10 13-tetraaza-2 14-dioxo-c~clotetradeca-3 Tri • 0. 0853 g CuLBF2 and 0.1145 g Ph3 P were dissolved in a small amount of O2 -free acetone. (The Ph3 P had been recrystallized from 95% ethanol). Upon slow addition of ether, yellow-green needles precipitated in 60% yield. Anal. Calcd. for CuC 11 N4H18 B02 F 2 • C18H15 P: C, 56. 83; N, 9.14; H, 5. 43; Cu, 10. 4. Found: C, 56.90; N, 8~95; H, 5.55; Cu, 11.6. The product is diamagnetic by magnetic susceptibility (25 °, Faraday method). 6 9-tetrameth lundeca-3 8-diene-2 10-dione r(I Perchlorate. 8 To 2 g acetic acid in 30 ml ----~""'-"'-.,-..~------~~--"""~"""""""'"""'"""' ethanol was added 3. 0 g (30 mmol) 1, 3-diaminopropane-2, 2-dimethylpropane. 30 To the hot solution was then added 6. 0 g 2, 3-butanedionemonoxime (60 mmol). After 15 minutes 5. 0 g Cu(ClO4 ) 2 • 6H 2O (15 mmol) dissolved in minimal hot ethanol was added. After heating an additional 30 minutes, the product was filtered from the cooled solution and recrystallized from hot acetone (150 ml) and dioxane (40 ml). Yield: 1. 5 g. 91 1, l-Difluoro-4, 5, 8, 8, 11, 12-hexamethyl-l-bora-3, 6, 10, 13• • tetraaza-2, 4-dioxo-cyclotetradeca-3, 5, 10, 12-tetraenato Copper(II) ~ 8 Boron trifluoride etherate (0.5 ml) was added to the previous compound (0.9 g) in bet diaxane (20 ml), and the solution heated for one hour. Upon cooling a purple oil separated, which slowly crystallized on standing for several days. 1, l-Difluoro-4, 5, 8, 8, 11, 12-hexamethyl-l-bora-3, 6, 10, 13- . tetraaza-2, 4-dioxo-cyclctetradeca-3, 5, 10, 12-tetraenato Copper(I) 8 , 2. ~ 0. 60 g of the corresponding copper(II) complex was reduced in acetone at out to -1. 0 V vs. Ag. After the ele·c trolysis was complete, the solvent was removed under vacuum, the residue dissolved in a minimum of hot acetone (8 ml), and diethyl ether (5 ml) added. The product was isolated by filtration of the cooled solution and washed with ethanol. Yield: 0. 25 g. ~~ The corresponding copper(II) complexes were synthesized as previously reported. 29 Reduction to copper(!) and isolation of the product 8 was similar to that of 1. "" was prepared according to the procedure of Liu and Liu. 31 2. 44 g of pyridine-2-aldoxime dissolved in a minimum amount of ethanol was added slowly to a solution of 2. 42 g of copper dinitrate trihydrate in 30 ml of water. Dark-green crystals formed and after a few hours were filtered from solution, washed with a small amount of water and vacuum dried. Calcd for C1 JJ 11N4O2 CuNO3 : C, 38. 87; N, 18. 99; H, 3.07. Found: C, 39~10; N, 18.70; H, 3.25. 92 Difluoro-bis-(pyridine-2-aldoxime)borate copper (Il) nitrate 1. 0 g of the previous compound was stirred in 100 ml of warm dioxane. 2 ml of BF8 Et2O were carefully added and the solution left to reflux for 1½ hours. The pale-blue product is insoluble in dioxane and was filtered off, washed twice with dioxane and twice with ether. Difluoro-bis-(Pyridine-2-aldoxime) borate copper(!), 6 was prepared by constant pctential electrolysis at -0. 5 V vs. Ag of the corresponding Cu(II) compound in acetonitrile containing 0.1 M tetraethylammonium perchlorate. The orange product precipitated during the electrolysis was collected by filtration, and washed once with ethanol. Calcd for C1:alJ10N4O2 BF2 Cu: C, 40. 65; N, 15. 80; H, 2. 84. Found: C, 40. 60; N, 16. 30; H, 3.15. 5, 7, 7, 12, 14, 14-Hexamethyl-l 4 8 11-tetraazacyclot:etradeca4, 11-diene Dihydroperchlorate, Me 6 -diene(HC1O4 ) 2 was prepared in a ~~ manner similar to the published procedure. To 1 L of acetone in a 2 L beaker was added 44 ml of ethylenediamine. The solution was stirred while 72 ml of 6~ perchloric acid was added dropwise over a halfhour periodo (Caution: Use of a safety shield is recommended for this and all perchlorate salt syntheses). After being stirred at room temperature overnight, the product was filtered from the dark-brown solution, washed with acetone until white (3x), and air dried after a diethyl ether wash. Yield: 106 g (73%). If the solution was left stirring for two days, 114 g was obtained. Anal. Calcd for C1 JI 32N4 • 2HC1O4 : C, 39092; H, 7.12; N, 11.64. Found: C, 40.35; H, 7.00; N, 11.40. 93 5, 7, 7, 12, 14, 14-Hexamethyl-l, 4, 8, 11-tetrazacyclotetradeca~ 4, 11-diene Copper{Il) Diperchlorate, Cu{Il)Me 6 -diene{ClO4) 2 • A slurry of 50. 0 g copper diacetate tetrahydrate, 110. 0 g Me 6 -diene(HC1O4 ) 2 , and 1 L of methanol in a 2 L Erlenmeyer flask were stirred at reflux for one hour, and then the solution was left to cool slowly to room temperature. Small red crystals were filtered from the purple solution, washed once with methanol, once with diethyl ether, and air dried. Yield: 114 g {92%). Anal. Calcd for C1 JIS!N4 Cu(ClO4 ) 2 : C, 35.40; H, 5.94; N, 10.32. Found: C, 35.55; H, 5.75; N, 10.30. 5 7 7 12 14 14-Hexamethy:l-l 4 8 11-tetraazac clotetradeca4, 11-diene Copper(!) Perchlorate Acetone, Cu(I)Me 6 -diene{ClO4 )• C3H6O, 7. 1. 0 g Cu{II)Me 6 -diene{ClO4 ) 2 was reduced by constant potential ~ electrolysis in acetone at out to -0. 8 V vs. Ag, such that the maximum current allowed was 20 mA. After the electrolysis was complete {n = O. 98), a yellow precipitate was filtered from solution and vacuum dried. Yield: O. 5 g {60%). Anal. Calcd for C1 JI32N4 ClO4 • C3H6O: C, 45.50; H, 7.64; N, 11.17. Found: C, 45.05; H, 7.85; N, 11.45. Acetone solvate vc=o is present in the infrared spectrum at 1707 cm -i. 5, 7, 7, 12, 14, 14-Hexamethyl-l 4 8, 11-tetraazacyclotetradeca1, 4, 8, 11-tetraene Copper{Il) Diperchlorate, Cu{Il)Me -tetraene(ClO) . 40. 0 g of Cu{Il)Me 6 -diene(ClO4) 2 and 25. 0 g sodium permanganate {Research Organic/Inorganic Chemicals) in 400 ml H2O were heated and stirred in a 4 L beaker. 150 ml HC1O4 ( 6'1¾,) was added dropwise over a half-hour period. The MnO2 formed, as evidenced by the brown frothy mess obtained, was reduced to aqueous Mn{II) by further heating. 94 Then the minimum amount of H2O necessary to dissolve any product present as precipitate at reflux was added and the boiling solution suction filtered hot. As the solution cooled, crystalline product was obtained. After being cooled in an ice bath, purple crystals were isolated by filtration and washed with 100% ethanol. If the product is relatively pure tetraene complex, the purple crystals will dry to a salmon-colored powder during the ethanol wash. The product should be checked for the presence of CuMe 6 -diene{ClO4) 2 , as indicated by the N-H stretching frequency at 3220 cm- 1 in the infrared spectrum. If any N-H band is seen, the material should be retreated as above, scaling the reaction to assume all product is CuMe 6 -diene(ClO4)2 • Note that the tetraene complex is more soluble in hot H2O than the diene complex, so less additional H2O will have to be added before filtering. Often two treatments (but never three) were found to be necessary. Treatment with greater quantities of sodium permanganate and perchloric acid did not seem to affect the tetraene/ diene ratio obtained but did result in the contamination of the product with sodium perchlorate. (Use of potassium permanganate rather than the sodium salt always resulted in product contaminated with white crystalline potassium perchlorate which gave poor elemental analyses even after multiple recrystallizations.) .A.11 permutations of CuMe 6 diene(ClO4) 2 , NaMn04 , and HC1O4 additions were tried with the yield of this procedure being superior. Yield: 50% from each treatmento Anal. Calcd. for C1 JI 28N4 Cu(C104) 2 : C, 35. 66; H, 5. 24; N, 10. 400 Found: C, 35.25; H, 5.05; N, 10.20. 95 5, 7, 7, 12, 14, 14-Hexamethyl-l 4 8 11-tetraazacyclotetradeca1, 4, 8, 11-tetraene Copper(I) Perchlorate, Cu(I)Me 6 -tetraene(ClO4 ), !! was prepared by constant potential reduction of 1. 4 g of the copper(II) complex in acetone at out to -0. 8 V vs. Ag, such that the maximum current allowed was 20 mA. After electrolysis was complete (n = 1. 08), the contents of the working compartment were filtered, and the 20 ml solution diluted to 50 ml with deoxygenated toluene. The flask was then scratched to induce crystallization. When the Cu(II)Me 6 -tetraene(ClO4 ) 2 was contaminated with small amounts of Cu(II)Me 6 -diene(ClO4 ) 2 , only oils were obtained. After a half hour, the solution was slowly diluted to 200 ml with toluene, and the dark red product filtered from solution and dried in vacuo. Yield: 0. 68 g (60%). Proton nuclear magnetic resonance indicates no supporting electrolyte is present. Anal. Calcd for C1 JI28 N4 Cu(ClO4 ): C, 43. 73; H, 6. 42; N, 12. 75. Found:C, 43.70;H, 6.30;N, 12.40. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8, 11-tetraazacyclotetra4, 11-diene Dihydroperchlorate, Me -diene(HC1O4 ) • To 1. 5 L of acetone in a 2 L Erlenm~yer flask was added 85 ml of 1, 2-diaminopropane (Aldrich). The solution was stirred while 109 ml of 6(Y)k perchloric acid was added dropwise over a half-hour period. After stirring at room temperature for one week, the product was suction filtered from the dark-brown solution and washed consecutively with acetone (3x), diethyl ether (lx), acetone (lx), and diethyl ether (lx). The white product was air dried.. Yield: 69 g (27%). Anal. Calcd for C18H3 aN4 •2HC1O4 : C, 42.44; H, 7.52; N, 11.00. Found: C, 42.60; H, 7.15; N, 10.80. 96 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8, 11-tetraazacyclotetradeca-4, 11-diene Copper(Il) Diperchlorate, Cu(II)Me8 -diene(C1O4 ) 2 • 25. 8 g copper diacetate monohydrate was dissolved in 1 L hot methanol. Upon addition of 60. 0 g Me8 -diene(HC1O4 ) 2 the solution almost solidified with purple precipitate. The heating was continued for half an hour and the product cooled to room temperature, filtered, and washed with methanol and diethyl ether. Yield: 65. 7 g (98%). Anal. Calcd for C1 s1J 36N4 Cu(C1O4 ) 2 :C, 37.87;H, 6.36;N, 9.81. Found:C, 38.20; H, 6.30;N, 9.70. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8, 11-tetraazacyclotetradeca-4, 11-diene Copper(!) Perchlorate Acetone, Cu(I)Me -diene(ClO ), i• 0. 90 g Cu(II)Me 8 -diene(ClO4) 2 was reduced by constant potential ""' electrolysis in acetone at out to -0. 8 V vs. Ag, such that the maximum current allowed was 20 mA. After the electrolysis was complete (n = 0. 97), a yellow precipitate was filtered from solution and vacuun . . dried. Yield: 0. 40 g (48%). Anal. Calcd for C18H36N4 CuC1O4 • C3 H6O: C, 47.63; H, 7.99; N, 10.58. Found: C, 47.45; H, 7.85; N, 10.40. Acetone solvate "c=o is present in the inf rared spectrum at 1 712 cm -i. 3, 5 7, 7, 10 12 14 14-Octameth 1-1 4 8 11-tetraazac clotetradeca-1, 4, 8, 11-tetraene Copper(II) Diperchlorate, Cu(I)Me -tetraene(C1O4 ) 2 • 40. 0 g of Cu(II)Me8 -diene(C1O4 ) 2 and 28. 0 g sodium permanganate in 500 ml H2 O were heated and stirred in a 4 L beaker. 200 nil HC1O4 (6~) was added dropwise over a half-hour period. The MnO2 formed, as evidenced by the brown frcthy mess obtained, was reduced to aqueous Mn(Il) with further heating (1 hour). The solution was then 97 cooled in an ice bath and the crude product filtered from solution. The product was recrystallized by dissolving in a refluxing solution of 2. 5 L methanol and 250 ml H2 O and filtering hot. As the purple solution cooled, salmon-colored needles precipitated. The product was filtered from solution, washed with methanol, and dried with diethyl ether. Yield: 11. 3 g (28%). The infrared spectrum has no 1 peaks between 3200 and 3300 cm- • Anal. Calcd for C18H32 N4 Cu{ClO4 ) 2 : C, 38.14; H, 5. 69; N, 9. 88. Found: C, 38. 30; H, 5. 75; N, 9. 70. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l 4, 8 11-tetraazac clot:etradeca-1, 4, 8, 11-tetraene Copper(!) Perchlorate, Cu(I)Me -tetraene{ClO4 ), 10, was prepared by a constant potential reduction at -0. 9 V ~ vs. Ag in acetone of 1. 31 g Cu(II)Me8 -tetraene(ClO4 ) 2 • Due to the very low solubility of the copper{II) complex, an initial cyclic voltammogram could net be ol:tained and currents during the electrolysis were small. The electrolysis cell was shaken periodically to aid dissolution. Due to the long time involved (about 9 hours), the auxiliary compartments were emptied and refilled with fresh solution after half the theoretical number of coulombs had been passed. After the electrolysis was complete (n = 1. 04), and no copper(II) complex was visibly present as a solid, a dark-red solid was isolated from the filtered solution by a procedure identical to that for Cu(I)Me 6 -tetraene(ClO4 ), !!• Yield: 0. 64 g {59%). Proton nuclear magnetic resonance spectra indicate no supporting electrolyte is present. Anal. Calcd for C18H32N4 CuC1O4 : C, 46.25; H, 6.90; N, 11.99. Found: C, 46.15; H, 6.70;N, 11.60. 98 5, 7, 7, 12, 14, 14-Hexamethyl-l, 4, 8, 11-tetraazacyclotetradeca4, 11-diene Nickel(II) Diperchlorate, Ni(II)Me 6 -diene(ClO) . 51. 5 g nickel diacetate tetrahydrate was dissolved in 750 ml methanol, 90. 0 g Me -diene(HC1O ) added, and the solution refluxed for one hour. 11 6 2 2 After cooling to room temperature, the yellow product was isolated by suction filtration, washed with methanol and dried with diethyl ether. Yield: 100. 0 g (99%). Anal. Calcd. for C1 sli32 N4 Ni(ClO4 ) 2 : C, 35. 72; H, 5.99; N, 10.41. Found: C, 36.30; H, 5.90; N, 10.20. 5, 7, 7, 12, 14, 14-Hexamethyl-l, 4, 8, 11-tetraazacyclotetradeca4, 11-diene Nickel(!) Perchlorate, Ni(I)Me 6 -diene(ClO4) was prepared by a scaled-up literature preparation 32 in a Schlenck apparatus under argon. 5. 0 g Ni(II)Me 6 -diene(ClO4 ) 2 in 250 ml deaerated CH3 CN was added to 1. 25 g Na metal in 150 g Hg. After stirring for one hour, the solution was filtered and evaporated to dryness. The black crystalline product from the filtrate was washed with 250 ml deaerated H2 O, vacuum dried, and transferred to the glove box. Yield: 3. 55 g (87%). Anal. Calcd for C1 sli32N4NiC1O4 : C, 43.81; H, 7.35; N, 12.77. Found: C, 43.70; H, 7.15; N, 12.95. 5, 7, 7, 12, 14, 14, -Hexamethyl-1 4 8 11-tetraazacyclotetradeca1, 4, 8, 11-tetraene Nickel(II) Diperchlorate, Ni(II)Me 6 -tetraene(ClO4) 2 • 50 ml of cone HNO3 (7~) was diluted to 400 ml with H2O, 100. 0 g Ni(II)Me 6 -diene(ClO4 ) 2 added, and the mixture refluxed for about half an hour until no more brown oxides of nitrogen were given off. Upon cooling in an ice b.ath, yellow crystals were filtered from solution and washed with acetone and diethyl ether. Yield: 71. 0 g (72%). The 99 -1 infrared spectrum has no peak at 3180 cm . Anal. Calcd for C1 JI 28N4 Ni{ClO4) 2 : C, 35.99; H, 5.28; N, 10.49. Found: C, 35.85; H, 4.80; N, 10.35. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8 deca-4, 11-diene Nickel{Il) Diperchlorate, Ni(Il)Me 8 -diene(ClO) . 25. 0 g nickel diacetate tetrahydrate was dissolved in 400 ml methanol, 35. 0 g Me 8 -diene(HC1O4) 2 added, and the solution refluxed for 1½ hours. After cooling to room temperature, the orange product was isolated by suction filtration, washed with methanol, and dried with diethyl ether. Yield: 32. 2 g (83%). Anal. Ca1cd for C18H36N4 Ni(ClO4) 2 : C, 38.19;H, 6.4l;N, 9.90. Found:C, 38.50; H, 6.15;N, 9.70. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8, 11-tetraazacyclotetradeca-1 4 8 11-tetraene Nickel{ ~ Di erchlorate Ni II)Me -tetraene- 30 ml cone HNO3 (70%) was diluted to 150 ml with H2O, 20. 0g Ni(Il)Me8 -diene(ClO4) 2 added, and the mixture refluxed for 20 minutes until no more brown oxides of nitrogen were given off. Upon cooling, large, bright-yellow crystals were filtered from solution and washed with acetone and diethyl ether. Yield: 17. 8 g (89%). The infrared spectrum has no peak at 3180 cm-1. Anal. Calcd for C18 H32N4Ni(ClO4) 2 : C, 38.46;H, 5.74;N, 9.97. Found:C, 38.50;H, 5.80;N, 9.90. 3, 5, 7, 7, 10, 12, 14, 14-Octamethyl-l, 4, 8, 11-tetraazacyclotetra- ~ was prepared by electrochemical reduction of 1. 34 g Ni(Il)Me8 - tetraene(ClO4 ) 2 in CH 3CN. Due to the high solubility of the Ni(Il) complex, the first 80% of the electrolysis was performed under 100 constant current conditions of 20 mA. The remainder was constant potential electrolysis outto -1.15 V vs. Ago After the electrolysis was complete (n = 1. 06), the working solution was filtered and transferred in the glove box to a Schlenck filtering apparatus. The sealed apparatus was attached to a high vacuum line, the volume of the solution reduced to 10 ml, and 125 ml of dry degassed toluene (CaH 2 treated) added by vacuum transfer. After warming to room temperature, a black precipitate was filtered off and the filtrate evaporated overnight. The sealed apparatus was then returned to the glove box and the filtered precipitate retained . .. Yield: 0. 57 g (52%). Addition of toluene in the glove box resulted in precipitation of a tan nickel(Il) complex. 101 References 1. D. C. Olson and J. Vasilevskis, Inorg. Chem., !.Q, 463-470 (1971). 2. R. R. Gagne, J. Amer. Chem. Soc., 98, .,..,,.._ 6709-6710 (1976) . 3. R. R. Gagne, Jo L. Allison, R. S. Gall, and C. A. Koval, J. Amer. Chem. Soc., filt, 7170-7178 (1977). 4. R. R. Gagne, J. L. Allison, and G. C. Lisensky, Inorg. Chem., !1, 3563-3571 (1978). 5. . W. Huber, "Titrations in Nonaqueou.s Solvents," Academic Press, N. Y., 1967, p. 70. 6. A. A. Oswald and D. J. Guertin, J. Org. Chem., ~' 651-657 (1963). 7. M. J. Brooks and N. Jonathan, J. Chem. Soc. Ser. A, 1529-1532 (1968). 8. R. R. Gagne and D. M. Ingle, unpublished observations. 9. N. F. Curtis, Coard. Chem. Rev., ....-. 3, 3-47 (1968) . 10. N. F. Curtis and R. W. Hay, J.C. S. Chem. Comm., 524-5 (1966). 11. A. M. Tait and D. H. Busch, Inorg. Syn., !!!, 2-7 (1978). 12. N. F. Curtis, J. Chem. Soc., 4409-4413 (1960). 13. M. M. Blight and N. F. Curtis, J. Chem. Soc., 3016-3020 (1962). 14. N. F. Curtis, Y. M. Curtis, and H. K. J. Powell, J. Chem. Soc. Ser. A, 1015-1018 (1966). 15. M. F. Bailey and I. E. Maxwell, J.C. S. Chem. Comm., 908-910 {1966); J.C.S. Dalton, 938-944 (1972). 102 References (continued) 16. T. Ito and K. Toriumi, Chem. Lett. (Tokyo), 1395-1396 (1978). 17. R. R. Ryan, B. T. Kilbourn, and J. D. Dunitz, J.C.S. Chem. Comm., 910 (1966); B. T. Kilbourn, R. R. Ryan, and Jo D. Dunitz, J. Chem. Soc. Ser. A, 2407-2412 (1969). 18. V. L. Goedken and D. H. Busch, Inorg. Chem., !Q, 2679-2682 (1971). 19. V. L. Goedken and D. H. Busch, J. Amer. Chem. Soc., .,...,,.._ 94, 7355-7363 (1972). 20. L. G. Warner, N. J. Rose, and D. H. Busch, J. Amer. Chem. Soc., fil!_, 703-4 (1967); ~' 6938-6946 (1968). 21. R. D. Felther and R. G. Hayter, J. Chem. Soc., 4587-4591 (1964). 22. M. M. Blight and N. F. Curtis, J. Chem. Soc., 1204-1207 (1962). 23. N. F. Curtis, D.. A. Swann, T. N. Waters, and I. E. Maxwell, Inorg. Chem., !U_, 4588-9 (1969). 24. N. F. Curtis, J.C.S. Dalton, 863-866 (1973). 25. T. Ito and D. H. Busch, Inorg. Chem., ll, 1770-1772 (1974). 26. D. A. Swann, T. N. Waters, N. F. Curtis, J.C.S. Dalton, 1115-1120 (1972). 27. D. A. HouseandN. F. Curtis, J. Amer. Chem. Soc.,~' 1331-1334 (1964). 28. D. H. Busch, unpublished observations. 29. R. R. Gagne, J. L. Allison, and D. M. Ingle, Inorg. Chem., !!!, 2767-2774 (1979). 103 References (continued) 30. E. B. Fleischer, A. E. Gebala, A. Levey, and P. A. Tasker, J. Org. Chem., ~' 3042-3044 (1971); M. S. Newman, D. H. Busch, G. E. Cheney, and C. R. Gustafson, Inorg. Chem., 11, ,.._,.._ 31. 2890-2893 (1972). C.H. LiuandC. F. Liu, J. Amer. Chem. Soc., .,...,.. 83, 4169-4172 (1961). 32. D. C. Olsen and J. Vasilevskis, Inorg. Chem., ~' 1611-1621 (1969). 104 Chapter Four. Mechanism Consideration and Summary 105 Chapter Four. Mechanism Consideration and Summary Consideration of Mechanism The actual mechanism of the ketonation reactions cannot be determined from the data attained. Nevertheless it is appropriate to examine oxygen activation mechanisms proposed for other metals in light of the current reactions. Two such catalytic mechanisms will be examined here. The first concerns oxygenation of hindered phenols by (bis-(3-salicylideneaminopropyl)amine)coba1t(II) and dioxygen. The second is that of the P-450 oxygenase -cytochromes. Recently Nishinaga and coworkers 1- 5 have examined the role of Co(Salpr) dioxygen complexes in the oxidation of hindered phenols (Scheme IV-1). Of particular interest is the reactio~ stoichiometry of 1. 25 0 2 /Co and isolation of crystalline organo-peroxy Co(III) complexes in some solvents. (In methanol the reaction is catalytic with the corresponding hydrqperoxide ottained at -10°, and the p-quinol attained on warming). 3- 5 The same stoichiometry of 1. 25 0 per 2 metal atom was observed for the dioxygen reactions of CuMe 6 -tetraene(Cl0 and CuMe -tetraene(Cl0 ) 4 8 4 ) . A peroxy intermediate has been proposed for the ketonation of cobalt macrocycles (Scheme IV-2) and is a likely intermediate for the copper macrocycle ketonations. Thus Co(Salpr) consumption of 1. 25 0 /Co to give a peroxy compound 2 could have the same mechanisms as the first part of the Cu(I) macrocycle oxygenation reactions. 106 Scheme IV-1 OH KJl T)J • ~+1.250 + 2 ¥ R 'c.(-,._NH r~ I\ ' ~ OMe CHfl2 0 OMe 107 Scheme IV-2 H Nishinaga and coworkers have proposed Scheme IV-3 as the 2 3 mechanism. , In this mechanism dioxygen incorporation is initiated Scheme IV-3 Co(II) + 0 2 ;= Co(Ill)o; slow Co(III)o; + Ph0H Co(m)OOH - Co(Ill)0H + ½02 .t Co(III)OOH + Ph0• fast Ph0• + Co(II) - - . Ph0-Co(III) fast Co(III)0H + Ph0H , Ph0-Co(Ill) + H20 Ph0-Co(Ill) + 0 2 - Ph(0H)OOCo(III) through hydrogen abstraction by Co(III)0; from the phenol. The phenaxy radical was detected by electron paramagnetic resonance. The mechanism is also supported by the Co(III)OOH obtained as well as half the normal amount of Ph(0H)OOCo(III) and a stoichiometry of 1. 0 0 2 /Co when Co(N-Me Salpr) derivatives are used instead of 108 Co(Salpr). Co(III)(N-Me Salpr)OOH is apparently stable and the lefthand part of Scheme IV-3 does net then apply. 2, 3 While a mechanism like Scheme IV-3 seems to fit the oxygenation of hindered phenols, it is difficult to apply it to the copper(!) macrocycle oxygenation regardless of the similar stoichiometry and products. The basic problem is that the macrocycle is attached to the copper and intramolecular reactions would occur faster than intermolecular reactions. It is also net clear how the cobalt(III) phenolate is oxygenated in the published mechanism to give the peroxy compounds, nor if this applies in any way to the ketonation of copper· macrocycles. The observed ketonation appears to be the ~roduct of electrophilic attack on the macrocycle CH2 groups. Electrophilic attack derived from diaxygen has been best studied in the P-450 axygenase cytochromes, and the mechanism is shown in Scheme IV-4. 6, 7 Scheme IV-4 Fe(III)(RH) is meant to indicate binding of the substrate to the enzyme in the vicinity of the Fe(III), as shown in Fig. IV-1. 109 RH Fig. IV-1. Substrate binding of cytochrome P-450. The most controversial step, postulation of an "oxene" as the active species, is justified as follows. S-ll 1) The reaction pattern is similar to the reactions of carbenes giving insertion into C-H bonds with retention of configuration, and addition to alkenes to produce a threemembered ring. 2) For a large number of enzymatic hydroxylations of aromatic compounds the substituent at the position which is eventually hydroxylated frequently shifts to an adjacent carbon ("NIB shift"), indicating an arene oxide intermediate and migration as the keto tautomer. 3) Treatment of iron(m) and chromium(III) porphyrins with 2 iodosylbenzene gives iodobenzene and a M(V)=0 - - M(IV)-O:' - M(Ill) O species. This reactive intermediate was. found to epoxidize cis-olefins and hydroxylate alkanes. lO-l 3 Application of a cytochrome P-450like mechanism to the oxygenation of copper(!) macrocycles again is fraught with difficulty. Perhaps an intermediate like Cu(I) + 0 2 - Cu(II)-o; ++ Cu(III)-o; or 2Cu(I) + 0 2 - Cu(II)-o; -Cu(II) - 2Cu(II)-O- .+ 2Cu(III)=O applies and the ligand itself is the proton source. The most obvious lesson in examining these mechanisms is that the focus should be shifted from the metal diaxygen species itself to the 110 reduced reactive oxygen species. Determination of the true structural identity of the oxygen derivative that attacks the macrocycle, the mechanism by which the active form of dioxygen is generated, and the mode by which the oxygen is incorporated into the macrocycle are questions that have net yet been examined. Summary The original object of this study was to synthesize stable copper(!) macrocyles, react them with dioxygen, ·and identify the products. These objectives have been attained. A binuclear copper dioxygen species has been identified in solution at low temperature. Ketonation of several copper(!) macrocycles has been observed, as well as blockage of ketonation by macrocycle substitution. A hydroxylated ligand copper(I) autoxirlation has been identified crystallographically. Prior to the work of R. R. Gagne, J. L. Allison, myself, and D. M. Ingle, only one copper(!) N4 -macrocycle had been isolated. Our collective work has greatly increased this number. The synthetic hurdle ·of preparing fairly large quantities of pure copper(!) .materials required for the studies reported here has been overcome. Along the way, several new geometries have been observed crystallographically, including square-planar, square pyramidal, and whatever geometry the binuclear complex described in Chapter Two turns out to have. As expected, the !ability of copper complexes is a problem. Dioxygen reactions of copper(I) multidentate complexes with additional 111 dissociable ligands are compucated by ligand dissociation. Macrocyclic ligands stay coordinated, but in return, the metal cannot be removed to ol:tain nuclear magnetic resonance spectra of the products. There must be coordinatively unsaturated copper(!) to react with dioxygen, as shown by the triphenylphosphine inhibition and enhancement of .dioxygen per. copper atom taken upo Several pitfalls in studying dioxygen reactions should be recalled. The measured dioxygen per copper stoichiometry does not necessarily agree with the amount found in the isolated solids, as observed for the mononuclear complexes in Chapter Two. Reversible color changes do not necessarily mean reversible oxygenation, as found for CuLBF2 in Chapter Three. It has been shown that Scheme m-5 is applicable to copper(!) dioxygen reactions, where the "decomposition" reaction includes oxygen incorporation into substrates. The oxidation of CO to co;, ligand hydroxylation, n1etal involved oxidation of triphenyl phosphine to triphenylphosphine oxide, and ketonation of a macrocyclic ligand have all been identified. Reduction of dioxygen to water, hydroxide, or ox:o species has been observed in conjunction with oxidation of copper(!) to copper(II). The tetrameric hydroxide bridged species produced upon dioxygen reaction of the mononuclear ligand in Chapter Two is the best verified, but stoichiometries of O. 25 dioxygen per copper atom can mean little else. Available protons, as in nitromethane solution, favor this simple redox reaction. There is a large aprotic solvent dependence of the stoichiometry on msicity. Ligand or solvent oxidation appears quite 112 common for copper(!) systems under aprotic conditions. Further work on this problem could go in several directions. More mechanistic information, including identification of the observed intermediate for the ketonation of the Cu(I) tetraene complexes, could be obtainedo Low temperature solution infrared experiments should be very useful in following the reactions, if precipitation of the product in the infrared cell is not a problem. Investigation of oxidation of added substrates could be carried out. Further ligand modification to prevent any ketonation may finally yield a solid room temperature .. stable copper dioxygen complex. In conclusion, it should be stated once more that copper(!) dioxygen products are very reactive species indeed. Q,) C) a -e 0 11.l ~ I I I I I I 400 500 600 nanometers I 700 I Fig. IV-2. Spectroscopically observed intermediate in the reaction of dioxygen with Cu(l)Me6 -tetraene(ClO4 ) in acetone solutiono 113 References 1. A. Nishinaga, K. Nishizawa, H. Tomita, and T. Matsuura, J. Amer. Chem. Soc., ~' 1287-1288 (1977). 2. A. Nishinaga, H. Tomita, and T. Matsura, Tet. Letters, 2893-2896 (1979). 3. A. NishinagaandH. Tomita, J. Mol. Catyl.,1, 179-199(1980). 4. A. Nishinaga, K. Watanabe, and T. Matsuura, Tet. Letters, 1291-1294 (1974). 5. T. Matsuura, K. Watanabe, and A. Nishinaga, J.C. S. Chem. Comm., 163-164 (1970). 6. J. T. Groves, Adv. Inorg. Biochem., !, 119-145(1979). 7. V. Ulrich, Angew. Chem. Int. Ed. Eng., ll, 701-712 (1972). 8. G. A. Hamilton, "Molecular Mechanisms of Oxygen Activation," O. Hayaishi, Ed., Academic Press, N. Y., 1974, pp. 405-457. 9. C. K. Chang and D. Dolphin, ffBioorganic Chemistry," E. E. van Tamelen, Ed., Academic Press, N. Y., 1978, pp. 37-80. 1 0. J. T. Groves, T. E. Nemo, and R. S. Myers, J. Amer. Chem. Soc., 101, 1032-1033 (1979). """""" 11. J. T. Groves, W. J. Kruper, T. E. Nemo, and R. S. Myers, J. Mol. Catyl., .......7, 169-177 (1980) . 12. J. T. Groves and W. J. Kruper, Jr. , J. Amer. Chem. Soc. , 101, 7613-7615 (1979). "'"'-'""" 13. C. K. Chang and M. S. Kuo, J. Amer. Chem. Soc., !Q!, 3413-3415 (1979). 114 Appendices 115 !Reprinted from Inorganic Chemistry, 18, 771 (1979).j Copyright C 1979 by the American Chemical Society and reprinted by permission of the copyright owner. Contribution No. 5771 from the Division or Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 9 I 125 Reaction of Molecular Oxygen with Copper(I). Structural Characterization of Dimeric µ-Carbonato and Tetrameric µ-Hydroxo Complexes Resulting from the Autoxidation of a Copper(I) Carbonyl Complex ROBERT R. GAGNE,• ROBERTS. GALL, GEORGE C. LISENSKY, RICHARD E. MARSH, and LAURINE M. SPELTZ Recei~d May 12. /978 Condensation of 1,2-dicyanobenzcne with 2-amin<r4-methylpyridine yields the tridentate ligand l ,3-bis(2-( 4-methylpyridyl)imino)isoindoline, HL, l. Reaction or HL with Cu(OAcl, gives Cu(L)OAc, 2. while HL plus CuCI in the presence of triethylamine and carbon monoxide gives diamagnetic, presumably four-coordinate Cu(l)=ntaining Cu(L)CO. 3 (vco 2068 cm·'). This carbonyl complex, Cu(L)CO, reacts with dioxygen in dry benzene or toluene to give a variety of products, two of which were characterized by crystal structure determinations. A carbonate-bridged dimeric complex (Cul),(µ-CO 3), .C, in which the carbonate group presumably results from metal-promoted oxidation of CO via 0 2, crystallizes in the space group PI with a• 11.420 (!) A., b = 16.479 (I) A, c = 10.954 (!) A., a= 90.98 (1) 0 , P • I 13.83 (1) 0 , -r = 101.71 (1) 0 , and Z = 2. The structure was solved using 4578 reflections and gave R(F) "' 0.058. Each Cu(II) is approximately square planar, being coordinated to three nitrogen atoms of the tridentate ligand and to one oxygen atom of the bridging carbonate group. A hydroxyl-bridged tetrameric complex (CuL)iCuLO)(µ-OHl,, 5, crystallizes with 2.5 benzene molecules and one water molecule per tetramer in the space group Pl with a= 14.586 (!) A. b = 14.925 (I) A.. c = 20.781 (2) A., a = 105.48 ( I )0 , P • 105.48 (I ) 0 , -r '"' 80.24 ( I ) 0 , and Z = 2. An R(F) of 0.065 was obtained using 9334 reflections. The molecule contains two square-planar and two trigonal-bipyramidal Cu(Ii) atoms, each being coordinated to three nitrogen atoms of the tridentate ligand and linked in a linear array by thr~. bridging hydroxyl groups. In addition a single oxygen atom, located on a pyridyl a carbon of the tridentate ligand, is distributed nonstatistically over the four tridentate ligands. This oxidation of the ligand presumably results from copper activation of dioxygen. l.atroduction Several copper-containing proteins ultilizc molecular oxygen in rather varied physiological processes. 1 ◄ Thus, hemocyanin,H tyrosinase.' cytochrome oxidase, 9 and laccase 10- 12 all react with 0 2, possibly forming intermediate copper-dioxygen complexes. Little is known about the nature of the active sites of these proteins or about the formal oxidation states, relative locations, or ligand environments of the copper atoms; indeed, it is not even known whether these proteins share common structural features . The principal information concerning the nature of the copper-dioxygen interaction derives from the resonance Raman spectrum of oxyhemocyanin; 1H 6 a band at 742 cm- 1 has been attributed to peroxide-like, presumably coordinated, dioxygen. However, there is no indication whether 0 2 is bound to one or both of the copper atoms comprising this active site. The other ligands bound to copper are, most probably, imidazole nitrogens of histidyl residues. Our present understanding of the reactions of 0 2 with relatively simple copper complexes does little to help clarify protein behavior. Most Cu(II) complexes do not react with 0 2• No well-characterized dioxygen complexes derived from Cu(I) are known, 17 although claims of 0 2 adduct formation have appeared. 11-26 Treatment of Cu(!) complexes with 0 2 usually results in irreversible oxidation to Cu(II), but oxidation products have not been fully characterized nor have kinetic studies yielded much mechanistic information. 27 In an effort to better understand the nature of the cop-per-dioxygen interactions in proteins, we are studying relatively simple mononuclear-21-- 30 and binuclear31 Cu(!) complexes. As we have discussed elsewhere., 29 the use of polydentate ligands can minimize associations and dissociatioo in solution, making it more likely that solid-state crystal structures corrcspond to solution struc;tures and sometimes permitting more facile determination of solution mechanisms. Indeed, a good model for "blue" copper proteins which takes advantage of this feature of polydentate ligands has recently been reported. 32 We have examined copper complexes of the tridentate ligand l,3-bis(2-(4-methylpyridyl)imino)isoindoline. 1, which 1w a nitrogen ligand environment for Cu(!) similar to that suggested for bemocyanin. 1H6.ll.J,I Herein we report the synthesis of an apparently four-coordinate Cu(!) carbonyl complex of 1, as well as structural characterization of two oxygenation products. Results Preparatioll of die Cu(I) Carboayl Complex, 3 (Scheme (). The tridentatc ligand 1 is synthesized by the condensation of 2-amino-4-methylpyridinc with 1,2.<ficyanobenzene. Reaction of 1 with cupric acetate gives a mononuclear Cu(II) acetate comple,i Cal(0Ac), 2_ 3H 7 Reaction of 1 with cuprous chloride in the presence of triethylamine and carbon monoxide gives a Cu(!) carbonyl complex Cu(L)C0, 3 (vco 2068 cm- 1), as a light yellow powder. The carbonyl complex 3 is also 116 r!pft I. The carbonate-bridged dimer (CuL),(..,-CO 1), 4: top. stereoview or a single molecule. showing the numbering scheme used for the ligand; bottom, stereoview or the crystal structure. The thermal ellipsoids are drawn at the 50'!6 probability level. 11 obtained by reduction of the acetate complex l with sodium amalgam and subsequent treatment with CO. The carbonyl complex 3 can be recrystallized from toluene in a CO atmosphere. Recrystallization under nitrogen, however, results in disproportionation to Cu(II) and Cu(0), implying that the coordinated CO is labile and that the three-coordinate complex is quite unstable. Attempts to produce the three-coordinate Cu(!) complex without a fourth ligand (e.g., via CuCJ and J) result in _a dark insoluble product. Attempts to incorporate other ligands besides carbon monoxide, namely acetonitrilc, tetrahydrofuran, or pyridine., similarly gave dark insoluble products whose infrared spectra showed no indication of the desired coordination.31 These dark products may be polymeric. The carbonyl complex 3 is only slightly soluble in common organic solvents. In order to increase its solubility, the methyl substituents on the pyridine rings were replaced by tut-butyl groups. However, this modified Cu(!) carbonyl complex (vco 2067 cm· 1) showed only a slightly increased solubility. Reactioll of tbe Cu(I) Carbonyl Complex 3 with Oxygea. Upon exposure to dry 0 2, light yellow suspensions of the carbonyl complex 3 in dry solvents such as benzene gradually turn brown. Reddish brown solids could be isolated from these suspensions, but low solubility and an unchanged infrared spcctrum 31 made characterization difficult. In an attc:mpt to define the nature of the oxygenation rcactioo, we obtained crystals directly from the reaction solution as follows . A small amount of the carbonyl complex 3 was dissolved in a large amount of benzene open to the atmosphere. The solution was allowed to evaporate slowly, and dark red crystals formed. An X-ray diffraction analysis of these crystals showed a dimcric ..,-carbonate complex (Culh(µ-CO 1). 4 (Scheme I). A reaction involving atmospheric carbon dioxide seems to be precluded, since crystals of 4 were also obtained under COi-free conditions. From the same CO,frcc bcnzcnc solution a second product of the reaction of Cu(L)CO, 3, with 0 2 was also obtained in crystalline form. A crystal-structure determination showed this compound to be a tetramcric complex (Cul)i(CulO)(µ-OH) 1, 5 (Scheme I), which crystallizes with 2.5 molecules of benzene and one molecule of water per tetramer. De9aipooe of tbt Carbonate-Bridged Dimer, 4. Stcrca;copic views of the carbonate-bridged dimer 4 and of the arrangement of the molecules in the crystal arc shown in Figure I; details of the bridging system arc given in Figure 2. The coordination about the copper atoms is approxima tely square planar, each copper being coordinated to three nitrogen atoms of the tridentate ligand and to one oxygen atom of the bridging carbonate group. The Cu-N distances to the pyridine nitrogen atoms NI and N5, which arc equal within experimental error at 2.020 (3) A, arc appreciably longer than those to the pyrrole nitrogen atoms N3, at 1.885 (4) A. The cis bond angles at the copper atoms range from 89 to 94° and the trans angles from 159 to 167° (Table I). The square-planar coordination is tetrahedrally distorted with the dihedral angles between the N3-Cu-N5 and NICu--0 planes, or between the N3-Cu-Nl and NS-Cu--0 planes, being 22° for ligand A and 24 ° for ligand B. (These 117 Table I. Coordination Angles ( deg) 0 (Cut),(i&-CO,), 4 O1-CUI-NlA N5A-CUI-N3A N3A-CUI-NIA Ol-CUI-N5A 90.7 90.0 90.1 93.9 O2-CU2-N5B NI B-cu2-N3B N3B-cu2-N5B O2-CU2-Nl B 92.8 90.5 89.3 91.3 N3A-CUl-01 NI A-CUI-N5A N3B-cu2-02 N5B-cu2-NIB 167.3 159.1 166.9 162.4 N3A-CUI-NIA N5A-CUI-Ol O1-CUI-Nl A N3A-Oll-N5A 90.1 93.0 98.5 88.8 (CuL) , (CuLO)(i,-OH),, 5 N30-CU-4-NID 89.2 N5D-cu4-03 93 .5 O3-01-4-N ID 98.9 N3I>--0>4-N5D 89.9 Ol-CUI-N3A NI A-CUI-N5A O3-CU-4-N3D NII>--0>4-N5D 15!.I 156.6 155 .2 152.3 N3B-cu2-02 O2-CU2--0l N3B-cu2-0l 128.9 106.5 124.6 N3C-Ol3--02 O2-CU3--03 N3C-CU3--03 150.0 96.2 113.8 NIB-cu2-02 NI B-CU2-N38 NIB-cu2-0l N5B-cu2-02 N5B-cu2-0l N5B-cu2-N3B N5B-cu2-NIB NIC-013-02 NIC-Ol3-N3C NIC-013-03 N5C-Ol3-02 N5C-Ol3-N3C N5C-Ol3-03 NIC-013-NSC 93 .3 88.6 85 .9 91.0 88.9 91.0 174.9 92.1 89.9 90.6 88.5 89.1 89.3 179.2 o ud 's UC 0.1-0.2° . N3A;;sl ✓ 11.11111,.o,, 1·0i81l/2 ~ Hl.HOI N5A---Cut=N1A 03 110.0781 1.22~\\ L912 C21T°0110.l7'l 1.302\ t306 ,urn02 <!"A CC).049,I 1-0.HI) -~~ .. -----,;N18 ~~2 202!>) N58 ~N3B I-O.t621 ....___.. to.VOi F",pre Z. Bond distances and atom numbering scheme of the coordinated atoms for the carbonate-bridged dimer 4. Esd', arc about 0.004 A for the Cu--0 and Cu-N distances and about 0.006 A for C-0 distances. Numbers in parentheses arc deviations, in A. from the best planes through the four coordinating atoms. angles would be 0° for a planar geometry and 90° for a tetrahedral geometry.) Alternatively, the distortion can be described by the deviations from the best planes through the four coordinating atoms, as in Figure 2. This tetrahedral distortion is achieved not by bending of the tridentate ligand but rather by raising the copper atom and the coordinated oxygen atom out of the plane of the tridentate ligand. Atom Cu I is 0.360, 0 I is I.I 26, Cu2 is 0.305, and 02 is 1.039 A. above the respective planes of the three coordinated nitrogen atoms. Whatever the description, the distortion serves to relieve the steric interference between the coordinated oxygen atom and the hydrogen atoms on Cl and C20; these 0-H distances range from 2.26 to 2.46 A.. The carbonate group has been shown by crystallographic analyses to coordinate to transition metals as a monodentate,l9 bidentate,"° bridging bidentate,•1.•2 binuclear bridging tridentate, 43 and trinuclear bridging tridentate44 ligand. In the present structure ◄, the carbonate group is bridging bidentate. Two steric effects govern the relative orientations of the CuL moieties and the bridging bidentate carbonate group: first, the sp2 bydridization of the oxygen atoms suggests that the copper atoms will lie in the plane of the carbonate group with C2 I--O--Cu angles of about 120°; second, when the plane of the carbonate group is approximately perpendicular to the planes of the tridentate ligands, ligand-ligand repulsion will be minimized. With these constraints, three conformations are possible: a symmetric arrangement with the ooncoonfinating oxygen atom directed away from the metal atoms; a second symmetric arrangement with the noncoordinating oxygen atom directed inward between the metal atoms; and a nonsyr:nmetric pinwheel arrangement. The firsi symmetric arrangement is lcnown 41 but is clearly ruled out in the present case because of the required close contacts between the tridentate ligands. The second symmetric arrangement bas yet to be observed crystallographically for any transition metal. The nonsymmetric arrangement is also lcnown 42 and is the observed configuration for 4. In the nonsymmetric carbonate coordination, the relatively short Cu2-01 (2.407 A.) and Cul-03 (2.637 A.) distances are not due to a fifth coordination but are simply a consequence of the geometry of the carbonate group. The carbonate group is nearly planar as expected. Within the carbonate group, the C21-03 distance of 1.255 (4) A. is much shorter than the other C-0 distances of 1.304 (3) A., indicating that the double bond is not equally shared. The O2-C21-0 I angle opposite the shortest bond, 114.7° , is the smallest; the O3-C21-02 angle is 123.8° and the O1-C2l-03 angle is 121.5°. The C21--0--Cu angle: is 107.5° for Cul and I 03.2° for Cu2. The intermolecular packing (Figure I, bottom) features the intimate pairing of planar ligand groups that is typical of large aromatic systems. The spacing between adjacent parallel ligands is about 3.5 A.. Descripdoa of the Hydroxyl-Bridged Tetramer, 5. The triclinic unit cell contains two tetrameric molecules, five benzene molecules, and two molecules of water. A single tetramer is shown in Figure 3, top, and the arrangement of molecules is shown in Figure 3, bottom. Hydroxyl bridges join the copper atoms together so as to form an interleaving arrangement of CuL moieties, with successive parallel ligands rotated by approximately 90° with respect to one another in an alternating sequence that suggests a variable capacitor. This 90° rotation permits a systematic sandwiching of the pyridine ring of one ligand with the pyrrole rings of the neighboring layers. The stacking distance between adjacent ligands is fairly uniform at about 3.5-4.0 A.. The tetrameric molecule contains two five-coordinate and two four-coordinate Cu(II) atoms; some details of the geometry about the copper atoms are shown in Figure 4. For a given tridentate ligand the distances from the copper atoms to the pyrrole nitrogen atoms are shorter than those to the pyridine nitrogen atoms, as was found for ◄. The Cu-N distances involving the five-<:oordinale copper atoms in 5 are longer than tha;e involving the four-coordinate copper atoms in 4 and 5, by an average of 0.033 A.. The outennclit copper atoms, Cul and Cu4, are each bonded to three nitrogen atoms of the tridentate ligand and to one 118 Flswe 3. The hydroxyl-bridged tetramcr (CuL)i(CuLO)(11-0H)i, S: top. stcreoview of a single molecule (the disordered oxygen atoms 04 (see text) arc shaded); bo1tom, stcreoview of the crystal structure (the disordered oxygen atoms arc not shown). .....~ 3 / ~ A for Cu4, and -1.643 A for 03; and (3) the dihedral angles between the N 3-Cu-N 5 and N 1-Cu-O planes or between the ,_ri#N,o•.- NtO~c..4....... ,o.4Z41 ,.au'\. • '03 l-0.40JI ,,..-\:;-;;gaAN3C N5C \ /o,'I,~ ) ~c..3 ' ' 1.93 ,, ~N1C 02 N38~~ ,,..~C...2~ NtB? \2.132 01(0.402! .....~ .- ~ / '--.; -N3A ,.._. 2.022c..1·,.a,.. ) •~~N1A Hl.418 •'II" 4. Coordination geometry in S. E.sd"s in the CIH) and Cu-N distanccs arc about 0.005 A. Number, in parentheses arc deviations, in A, from the best planes through the coordinating atoms of the two temunal, foor-<XJOrdinate copper atoms. The two central copper atoms have approximate trigonal-bipyramidal coordination, with NI and N5 in the axial positions. bridging hydroxyl oxygen atom in a distorted square-planar geometry similar to I.hat found in the bridging carbonate dimer, ◄, but the tetrahedral distortion is larger for the four-<:OOrdinate copper coordination in 5 than in ◄. The tetrahedral distortion in 5 is illustrated by: (I) the displaa:menlS from the best plane through the four coordinating atoms (Figure 4); (2) the displaccmenlS from the plane through the three coordinating nitrogen atoms of 0.401 A for Cul, 1.631 A for OJ, --0.468 N3-Cu-N I and N5-Cu-O planes of 36° . Besides relieving the nonbonded interactions between the oxygen atoms and the ligands, this increased distortion serves to increase the distance between successive ligands. The cis bond angles at Cu are all between 89 and 99° and the t,rans angles are between 151 and 157° (Table I). The two five-coordinate copper atoms. in contrast to the four-coordinate copper atoms. arc approximately in the planes of the coordinating nitrogen atoms (the deviations arc 0.003 A for Cu2 and 0.082 A for Cu3) and the ligands themselves are more nearly planar. The geometry about these copper atoms is best described as distorted trigonal hipyramidal. the equatorial planes containing two oxygen atoms and a nitrogen atom N3. The equatorial angles are I06.5, 124.5, and 128.9° for Cu2 and 96.2, I I 3.8, and I 50.0° for Cu3. In both cases the cis angles to the axial atoms NI and N 5 are between 88 and 93°; the trans angle is 179.2° for Cu2 and 174.9° for Cu3. The water molecules of crystallization. 05. apparently form weak hydrogen bonds to the nitrogen atoms N 2 of two ligands C and Din different molecules; the N-0 distances are 3.067 and 3.029 A. and the N-0-N angle is 134°. the five benzene molecules in each cell occupy voids in the structure (Figure 3, bottom) one of them lying on a crystallographic center of symmetry and the other four being in general positions. With only three hydroxyl groups bridging the four CuL + group$, the molecule would .be expected to be a monocation; in fact, it is a neutral species. The additional negative charge is contributed by an oxygen atom, 04, which is distributed at the CJ position among the four ligands in each tetramer and is always found on the end of the molecule when: overlap 119 of the ligands does not occur (sec Figure 3, top). Our least-squares refinement led to oa:upancy factors of 0.677 (10), 0.134 (5), 0.106 (5), and 0.070 (5) for this oxygen atom on the four ligands, for a total of 0. 987 (I 3) per tetra ma and within experimental error of the value 1.0 that corresponds to a neutral molecule. While we choose to designate the molecule as (Culh(CuLO)(µ-OHh, in any given tetramer molecule the oxygenated ligand LO can oa:upy any ooe o( the four positions with varying probabilities. The partial oxygen atoms 04 are especially interesting. The C I--04 bond lengths in the four ligands are as follows: A, 1.264 (5); B, 1.31 (2); C, 1.35 (3); D, 1.27 (4) A (the increasing values of the esd's reflect the decreasing values of the occupancy coefficients); the weighted average is 1.269 (5) A. This value is considerably shorter than found in normal C-0 single bonds, typical distances being 1.43 A for paraffinic systems•5 and 1.36 A for phenols.<14 However, it seems a reasonable value for a phenolate group, which we presume it to be. (We have found no example of a C-0 distance in a phenolate ion.) The negative charge on this phenolate group is apparently stabilized partly by weak bonds to the Cu atoms (the Cu--04 distances range from 2.93 to 3.11 A) and partly by hydrogen bonds from the bridging OH groups, as each of the partial 04 atoms has at least one OH neighbor within 2.8 A. The hydrogen ato·ms of the bridging OH groups, which were located with the help of difference maps, point in the general directions of the 04 atoms but the details of the hydrogen bonding arc not clear. The peculiarly uneven distribution of the phenolate group among the four ligands is not entirely senseless. One would expect a phenolated ligand to prefer one of the terminal, four-coordinate copper atoms over an inner, five-coordinate one, and the observed ratio of terminal-to-inner occupancy is 3: I. In addition, the CI position of ligand D is in close contact with one of the benzene molecules of crystallization. The distances from the site of 04D to the sites of the carbon atoms C33 and C35 arc impossibly short, at 2.49 and 2.51 A; accordingly, the presence of a phenolate group on ligand D must cause a displacement of the benzene molecule and, hence, a disruption in the crystal lattice. (Our least-squares refinement strongly suggests some disorder associated with this benzene molecule; sec Experimental Section.) There is- no such steric problem associated with ligand A, which shows the highest occupancy factor for the phenolate group. The preference for atom CI as the site of phenolation rather than C20, on the opposite side of the molecule, appears to be total. It is clearly dictated by the positions of the bridging OH groups, which in all cases are oriented so as to be able to interact with the 04 sites on CI but not on C20. The bridging Cu-0 distances (Figure 4) are, at first glance, bewilderingly irregular. At second glance, they make a good deal of sense. As expected, the Cu-0 bonds to the terminal, four-coordinate atoms Cul and Cu4 are the shortest. Then, because of conjugation effects, the two central bonds Cu2-02 and Cu3-02 are shorter than the penultimate bonds Cu2-0 I and Cu3-03. Finally, asymmetry between the two halves of the tetramer is caused by the unequal distribution of the phenolate oxygen atom 04 among the four ligands. Since this atom is found on ligand A most of the time, Cu I is less electron deficient than Cu4 and the Cu 1-0 I bond is longer than the Cu4-03 bond; by conjugation, Cu2-02 is longer than Cu3-02, and Cu3-03 is longer than Cu2-01 . The bridging Cu--0-Cu angles range from 120. l to 124.8°. The structure determinations of 4 and 5 have resulted in six independent determinations of the geometry of the tridentate ligand. and, assuming C,. (mm) symmetry for the ligand, in 12 independent measurements of most of the bond lengths and angles. Averaged values are shown in Figure 5, F",pn 5. Average values for the bond distances and angles for the tridentate ligand in 3 and 4. Most values are averages of 12 independent measurements, based on a presumed mm2 (C,.) symmetry of the ligand. The esd in a single such measurement, as obtained from our least-squares refinement, is about 0.008 A. or 0.6°; values in parentheses are the observed standard deviations about the average value. The esd in an awra~ distance is about 0.008(12 112), or 0.002 A. and in an angle is about 0.2° . and the individual values are available as supplementary material. Reaction ·Stoicbiometry. A suspension of Cu(L)CO, 3, in benzene was oxygenated on a vacuum line over a I 2-h period. and the-amounts of 0 2 consumed and of CO evolved were measured. Reproducible results were obtained. They indicated that 0.5 :l:: 0.1 mol of 0 2 were consumed and 0.91 :l: 0.05 mol of CO were evolved/mo! of complex 3. This stoichiometry is consistent with the formation of a dioxygen-bridged dimer. 2Cu(L)CO + 0 2 - (CulhO 2 + CO However, the oxygenation reaction appears to be irreversible, and neither infrared nor mass spectroscopy of the product suggested the presence of a dioxygen-bridged dimer. Discussion The copper(I) carbonyl complex Cu(L)CO, 3, prepared in this study joins a number of stable copper(!) carbonyl species that have been synthesized recently. These include (hydrotris(pyrazolyl)borato]copper(I) carbonyl, 6, and its dimethyl analogue (hydrotris(3,5-<iimethyl- I-pyrazolyl)borato ]copper( I) carbonyl, 1.•1•4 The expected tetrahedral geometry for these complexes has been confirmed by a crystal structure determination of the former.•' More recently, we have reported the synthesis, properties, and structure of a live-coordinate copper(!) carbonyl complex, 8, in which the Cu(I) has a C' co I N 8 n''-,11/ \ Cu " _j w Cu / '/ ~9 square-pyramidal environment. 2129 The mixed-valence Cu'Cu 11 complex 9 probably also contains a live-coordinate Cu(I) carbonyl group.31 Thus far, the insolubility of the present compound, 3, has precluded a crystal structure study, but a distorted square-planar geometry seems likely. The C-0 stretching frequencies for these carbonyl complexes are 2068 cm· 1 for 3, 2067 cm·• for the rut-butyl analogue of 3, 2083 cm·• for 6, 2066 cm·• for 7, 2068 cm·• for 8, and 2065 cm· 1 for 9; for hemocyanin ( I :2 CO/Cu), the observed value is in 120 ScbenM w liclleme D LC'IIC'o•O. --- the range 2040-2060 cm-t _io.st Thus, analogy with simple Cu(I) carbonyl complexes suggests any of a variety of possible metal stcrcochcmistries for hcmocyanin. Our present work is one example of a reaction between Cu(I) and 0 2 whose stoichiometry suggests the formation of a dioxygcn complex but which, upon detailed examination, must be far more complicated. The oxygenation reaction is irreversible, and we have not been able to isolate a dioxygcn species. Instead, we obtained from the reaction mixture crystals of a carbonate-bridged dimer, -4, and of a hydroxybridged tetramer, 5, in which one-quarter of the ligand molecules have been oxidized to a phcnolate, as well as large amounts of noncrystalline material which we have not been able to identify. We have no reason to doubt that the crystalline products arc representative of the products formed during oxygenation, although they .were obtained under slightly different conditions than were used for the measurements of the reaction stoichiometry. Although we have not ·been able to isolate any dioxygcn compounds, it is likely that they arc involved as intermediates in the various oxidation reactions. For example, a plausible mechanism for the formation of the 1<-carbonato complex -4 involves intramolecular insertion of coordinated CO into coordinated dioxygcn (Scheme II). In somewhat analogous reactions, carbonyl complexes of lr(I) and Rh(I) coordinated to tridentate phosphine ligands arc presumed to form mononuclear, five-coordinate carbonato complexes through dioxygen intcrmcdiates. 52..1 3 The overall stoichiometry of the reaction in Scheme II (cq 1) does not by itself account for the 2Cu(l)CO + 0 2 - (CulhCO 3 + CO (I) amount of CO liberated. As one of several processes going on it could help explain the stoichiometry of less than 1.0 CO/Cu liberated. Formation of the tctramcric species (Culh(CulO)(µ-0H) 3 may also involve a dioxygen intermediate, with copper-promoted activation of coordinated 0 2 leading to ligand hydroxylation (Scheme III). Hydrocarbon hydroxylations mediated by the cytochromes P, 50 (Fe) and by tyrosinasc (Cu) arc thought to proceed via coordinated 0 2, and macrocyclic ligand oxidations via 0 2 to give kctones, apparently promoted by cobalt, have been reported. s..ss Examples of hydrocarbon hydroxylations in which the oxidizing agent has been clearly demonstrated to be dioxygcn coordinated to a simple metal complex arc apparently unknown. Equation 2 agrees with the 2Cu(l)CO + 0 2 - CulO + Cul+ + OH- + 2CO (2) overall stoichiometry we observed for the oxygenation reaction; however. it would lead to the formation of equal amounts of the normal complex Cul+ and of the phenolated complex Cu(lO), whereas the tctramcr contains them in a 3: I ratio. It is quite possible, though, that charge balance, ligand-ligand overlap, and crystal forces could account for the selective crystallization of the tetramcric species; we emphasize again that crystalline material represented only a small portion of the reaction products. It is also conceivable that the small amount of phcnolated ligand necessary to produce the crystals resulted from an impurity in the starting material and that the bridging hydroxyl groups were formed by any of a number of mechanisms including simple protonation of a 1£-0XO species. Conclusions The carbonyl complex Cu(L)CO, 3, reac:ts with oxygen with a seductively simple stoichiometry. Characterizable oxidation products, however, indicate that the reaction is in fact very lllble 0. Summary of Crystal Data Cu1C..,HuN10CO, fonnula wt space group a,A b,A c,A a, deg 8, deg 'Y, deg V, A' z. d, g/r:m' Jl,Cffl- 1 unique data used R(F) 839.85 Cu,.C, 0 H•• N1004 · (C,H,),.,H,O 1839.01 PT pf ll.<420 (I) 16.479 (I) 10.954 (I) 90.98 (I) 113.83 (I) 101.71 (I) 1835 2 1.520 (calcd) 1.50 (3) (obsd) 19.20 4578 0.058 14.586 (I) 14.925 (I) 20.781 (2) 105.48(1) 105.48 (I) 80.24 (I) 4178 2 1.462 (ca.led) 1.43 (2) (obsd) 17.18 9334 0.065 complex. Our results once again indicate that caution must be exercised in drawing conclusions concerning the formation and nature of metal-oxygen complexes on the basis of a few observations and in the absence of firm structural evidence. The preparation of a stable coppcr-dioxygcn complex continues to present a challenge to the synthetic chemist. Our own efforts arc continuing in this area. • Experimental Sectioa Preparatioe of the FrM Upad L ud of Cu(L)OAc. The free ligand. I, and the Cu(II) acetate aimplcx Cu(l)OAc. 2. were prepared by the procedure of Sieg!. 11 (Cari,oayl){ l.>bis(l-( 4-metllylpyridyl)imino )i5oindolint )copper(I), C■(L)CO, 3. A solution of triethylamine (0.75 ml. 5.4 mmol) in ethanol (35 ml) was purged with CO and then added all al once to a mixture of the free ligand I (0.5 g, 1.5 mmol) and CuCI (0.15 g, 0.76 mmol) in a Schlenk apparatus. The red reaction mixture was stirred under CO. A finely divided yellow solid began to precipitate after 5 min and was collected by filtration under CO after 6 h. Tbe solid was washed with CO-saturated ethanol (2 x 5 ml) and dried in a slow stream of CO. The yield was 0.55 g (88%); •co 2068 r:m·•. Anal. Calcd for C 21 H 16 N 10Cu: C. 60.35; H. 3.85; N, 16.75; Cu. 15.2. Found: C, 60.05; H. 4.1; N, 16.9; Cu, 15.2. The carbonyl complex 3 was also prepared by reduction of the Cu(II) acetate complex 2 under nitrogen as follows. The Cu(II) complex Cu(l)OAc. 2 (0.5 g, I.I mmol). was wanned in toluene (70 ml. 60 °C). Sodium amalgam (5 g, 1%) was added with stirring to give a red solution and a gray solid. After I h the suspension was filtered. Tr<atmcnt of the red filtrate with CO led to gradual deposition of a finely divided yellow solid which was collected and washed as above. The yield was 0.3 g (65%) . M.....- of 0 2 Consumplioa and CO E>Oiulioa for !ht Reactioa ol Cu(l)CO, 3. witb o,. The carbonyl complex 3 in dry benzene was allowed to react with a premeasured amount of 0, at 25 °C. At the end of the r<action a Toepler pump was used to determine the amounts of CO evolved and of 0 2 remaining. In a typical experiment. dry benzene ( -15 ml) was distilled in a vacuum line onto a sample of the aimplex 3 (0. 144 g, 0.344 mmol) at -196 °C. The frozen mixture was wanned to 25 °C after a measured amount ofO 2 (1.01 mmol) had been added (panial pressure of 0 2 -200 torr). The suspension was stirred at 25 •c for 12 h. during which time the mixture turned from yellow to red. At the end of the reaction the total amount of 121 Table Ill. Poritional and Thermal Panmeten• fot (CuL), (ii<(),), 4 Cul Cu2 01 02 03 NIA N2A N3A N4A N5A CIA C2A C3A C4A CSA C6A .C7A CSA C9A C!0A CllA C!2A C!3A Cl4A C!SA C!6A Cl7A CIBA Cl9A C20A NIB N2B N3B N4B NSB CID C2B C3B C4B CSB C6B C7B C8B C9B CI0B CllB Cl2B Cl3B Cl4B Cl5B C!6B Cl7B Cl8B C19B C20B C21 "'" 5788" (7) 85542 (7) 6955 (3) 6990 (3) 5183 (4) 5804 (4) 3705 (4) 4351 (4) 4208 (4) 6217 (4) 6851 (5) 6962 (5) 5980 (6) 6065 (6) 4922 (5) 4830 (5) 3556 (5) 2387 (5) 1315 (6) 403 (6) 542 (6) 1607 (6) 2526 (5) 3792 (5) 5424 (5) 5796 (6) 6980 (6) 7377 (6) 7780 (6) 7384 (6) 7846 (4) 9285 (4) 9891 (4) 11088 (4) 9592 (4) 6853 (5) 6231 (6) 6585 (5) 5897 (6) 7610 (5) 8230 (5) 9975 (5) 11038 (5) 11544 (6) 12593 (6) 13079 (6) 12563 (5) II 549 (5) 10823 (5) 10545 (5) 11079 (6) 10748 (6) 11350 (7) 9835 (6) 9271 (6) 6322 (5) y/b z/c (I,, u,, Uu u., u., U,. 14707 (4) 37078 (4) 2438 (2) 3396 (2) 2389 (2) 2038 (2) 1213 (3) 627 (2) -374 (2) 697 (2) 2671 (3) 3106(3) 2925 (4) 3402 (4) 2289 (4) 1852 (3) 674 (3) -31 (3) -264 (4) -990 (4) -1444 (4) -1205 (3) -485 (3) -78 (3) -39 (3) -509 (3) -267 (3) -780 (4) 480 (4) 933 (3) 4606 (2) 4758 (3) 3806 (2) 2818 (2) 3120 (3) 4863 (3) 5428 (3) 5755 (3) 6354 (4) 5504 (3) 4943 (3) 4251 (3) 4073 (3) 4368 (4) 4066 (4) 3471 (4) 3182 (4) 3487 (3) 3322 (3) 2737 (3) 2243 (3) 2192 (4) 1692 (4) 2630 (4) 3057 (4) 2726 (3) 89121 (7) 90611 (7) 8768 (3) 7405 (3) 6895 (4) 10562 (4) 10390 (4) 8791 (4) 7072 (4) 7776 (4) 11301 (5) 12423 (5) 12885 (5) 14123 (6) 12151 (5) 11020 (5) 9440 (5) 8872 (5) 9190 (6) -8476 (7) 7516 (7) 7179 (6) 7906 (5) 7860 (5) 7065 (5) 6273 (5) 6202 (S) 5363 (6) 6953 (6) 7696 (6) 9633 (4) 12026 (4) 10816 (4) 10526 (4) 8374 (4) 8651 (6) 8936 (6) 10253 (6) 10581 (6) 11226 (5) 10901 (5) 11924 (5) 13135 (5) 14466 (6) 15358 (6) 14917 (6) 135-84 (6) 12703 (5) 11226 (5) 9131 (5) 8552 (5) 7193 (6) 6548 (7) 6432 (6) 7038 (5) 7642 (5) 450 (4) 480 (4) 47 (2) 53 (2) 55 (2) 42 (2) so (3) 44 (2) 53 (3) 55 (3) 48 (3) 56 (3) 67 (4) 73 (4) 63 (4) 47 (3) 45 (3) 50 (3) 56 (4) 47 (4) 60 (4) 63 (4) 41 (3) 48 (3) 59 (3) 74 (4) 77 (4) 79 (4) 61 (4) 66 (4) 52 (3) 68 (3) 46 (2) 49 (2) 43 (2) 58 (4) 58 (4) 55 (3) 63 (4) 59 (3) 53 (3) 51 (3) 46 (3) 64 (4) 85 (5) 68 (4) 51 (3) 41 (3) 46 (3) 52 (3) 65 (4) 72 (4) 115 (6) 79 (4) 68 (4) 44 (3) 456 (4) 518 (4) 46 (2) 53 (2) 64 (2) 49 (2) 59 (3) 46 (2) 46 (3) 47 (2) 523 (4) 520 (4) 44 (2) 52 (2) 78 (3) 49 (2) 58 (3) 55 (3) 67 (3) 61 (3) 57 (3) 48 (3) 46 (3) 54 (4) 58 (4) 55 (3) 62 (3) 63 (4) 80 (4) 105 (5) 109 (5) 93 (5) 69 (4) 92 (3) 102 (4) 8 (2) 9 (2) . 10 (2) 9 (2) 6 (2) 7 (2) 11 (2) 16 (2) 10 (3) 10 (3) 16 (3) 9 (4) 10 (3) 20 (3) 9 (2) 7 (3) 5 (3) 0 (3) -11 (3) 2 (3) 11 (2) IS (2) 18 (3) 17 (3) 32 (3) 35 (4) 14 (3) 6 (3) 11 (2) 15 (2) 9 (2) 10 (2) 13 (2) 20 (3) 20 (3) 11 (3) 25 (3) 13 (3) 10 (2) 9 (2) 6 (3) 22 (3) 25 (4) 30(4) 20 (3) 7 (2) 4 (2) 5 (2) 15 (3) 17 (3) 46 (5) 19 (4) 21 (3) 12 (2) 146 (4) 132 (4) 8 (2) 12 (2) -1 (2) 14 (2) 18 (2) 14 (2) 11 (2) 18 (2) 15 (3) 11 (3) 16 (3) 21 (3) 29 (3) 14 (3) 12 (3) 13 (3) 17 (3) 21 (4) 12 (4) 8 (4) 7 (3) 7 (3) 14 (3) 13 '(3) 24 (3) 24 (4) 33 (3) 35 (3) IS (2) 12 (2) 14 (2) 18 (2) 17 (2) 18 (3) 14 (3) 23 (3) 31 (4) 23 (3) 9 (3) 14 (3) 12 (3) 15 (3) 12 (3) 16 (3) 21 (3) 13 (2) 16 (3) 28 (3) 23 (3) 30 (3) 48 (5) 31 (3) 19 (3) 9 (2) 52 (3) 56 (4) 6 (2) 10 (2) 8 (2) 8 (2) 10 (2) 5 (2) 0(2) 4 (2) 6 (3) -2 (3) 3 (3) -14 (4) 10 (3) 13 (3) 18 (3) 19 (3) 18 (3) 30 (4) 16 (4) 10 (3) 18 (3) 9 (3) 7 (2) 2 (3) • 9 (3) -1 (3) 0 (3) -8 (3) 3 (2) 3 (2) 4 (2) 3 (2) 5 (2) 18 (3) 22 (3) 10 (3) 17 (4) 2 (3) 8 (3) 0(2) 3 (3) -3 (3) 7 (4) 21 (4) 12 (3) 9 (2) 8 (2) I (3) 4 (3) -4 (3) -3 (4) -1 (3) 8 (3) -s (2) 64(4) 66 (4) 77 (4) 121 (6) 74 (4) 56 (3) 50 (3) 57 (3) 78 (4) 95 (S) 66 (4) 51 (4) SI (3) 46 (3) 48 (3) . 53 (3) 62 (4) 78 °(4) 74 (4) 59 (4) 45 (2) 61 (3) 45 (2) 54 (3) 61 (3) 49 (3) 60 (4) 44 (3) 65 (4) 47 (3) 41 (3) 42 (3) 51 (3) 88 (4) 115 (6) 102 (5) 74 (4) 53 (3) 46 (3) so (3) 60 (4) 66 (4) 107 (6) 87 (4) 79 (4) so (3) 64 (4) SI (3) ss (3) 62 (4) 72 (4) 80 (4) 80 (4) 61 (3) 56 (3) 54 (3) 54 (3) 56 (3) 74 (4) 76 (4) 78 (4) 99 (5) 67 (4) 62 (3) 56 (3) 55 (3) 63 (4) 49 (4) 64 (4) 68 (4) 53 (3) 57 (3) 63 (3) 61 (4) 73 (4) 86 (5) 58 (4) 53 (4) 53 (3) • The fru:tional coordinates are multiplied by 10' for the copper atoms and by IO' for other atoms. The form of the temperature factor is cxp(-2,r'(U,,h'a*' + ... + 2U,,klb*c*)I . The U0 elements are multiplied by 10' for the copper atoms and by 10' for other atoms. Numbers in parentheses in this table azc estimated standard deviations in units of the last diizjt. gas remaining. X. was measured using a Toepler pump. The gas was then circulated aver CuO (240 °C) to oxidize CO to CO2, which was trapped at -196 °C. Measurement of the amount of CO 2 produced gave the amount of CO evolved (0.32 mmol) and, by difference (X -CO}, the amount of 0 2 remaining (0.85 mrnol). The amount of 0 2 at the stan (1.01 mrnol) less the amount of 0 2 remaining (0.85 mmol) gave the amount of 0 2 consumed. In this case, 95% of the total CO was recovered and the Cu/O 2 stoichiometry was 2.00/0.94. The infrared spectrum of the product isolated by suction filtration and air-drying was vinually identical with that of Cu(L)CO. 3. except for the absence of •co and slight changes near 1100 cm·•. No peaks attributable to •c...o or to carbonate were observable. The mass spectrum showed a large peak at 44 mass units (CO2), which docs not appear in the mass spectrum of Cu(L}CO, 3, as well as a significant peak at 456/458 mass units. (This latter peak appears very weakly in the spectrum of Cu(L)CO and has not been assigned.) Growtli al Sinp Cryslals from Cu(L)CO in 8enzme. The carbonyl complex Cu(L)CO, 3 (50 mg), was added to dry benzene (30 ml) in air and stirred for 5 min. giving a deep red-brown solution. The solution was filtered and then allowed to slowly evaporate at 25 °C. Under a microscope, dark parallelepipeds and dark amorphous mas.scs were observed (as well as some yellow crystals of the free ligand, which were separated by hand). The infrared spectrum of the crystalline material showed no obvious •c...o or any indication of carbonate, but elemental analysis was consistent with the formula C..,H 32 N 10Cu 2CO 3. Anal. Calcd: C, 58.65; H. 3.8; N, 16.7; Cu, 15.15. Found: C, 58.75; H. 3.95; N. 16.45; Cu. 14.8. Magnetic susceptibility measurements of the sample gave a"- of 1.5 i, 1 (25 °C, Faraday method). The mass spectrum of a parallelepiped crystal showed a strong peak at 44 mass units (CO 2), as well as significant 122 lkble IV. Positional and Thermal J>anmeten• for (CuL),(Cu\.OX,.-OH),, 5 Cul Cul Cu3 Cu4 01 02 03 OS NIA N2A N3A N4A NSA 04A CIA C2A C3A 04A CSA C6A C7A CSA C9A CIOA CllA Cl2A Cl3A Cl4A CISA Cl6A Cl7A Cl8A Cl9A C20A NIB N2B N3B N4B NSB 04B CIB C2B C3B C4B CSB C6B C7B C8B C9B CIOB Cl!B Cl2B Cl3B Cl4B CISB Cl6B Cl7B Cl8B Cl9B C20B NIC N2C N3C N4C NSC 04C CIC C2C C3C C4C csc C6C C7C CBC C9C CIOC CllC z./11 y/b z/c u.. u,, u,, u., U,, u" 18219 (4) 2272 (4) 10238 (4) -6201 (S) 776 (2) 335 (2) 9 (2) 5009 (3) 2430 (2) 4051 (2) 3023 (2) 2579 (2) 1184 (2) 962 (4) 1855 (3) 2212 (3) 3161 (3) 3S44 (4) 3740 (3) 3384 (3) 3871 (3) 4614 (3) SS1S (3) 6096 (3) S692 (4) 4740 (3) 4200 (3) 3181 (3) 1614 (3) 1071 (3) 116 (3) -476 (4) -295 (3) 248 (3) -1098 (2) -1885 (2) -214 (2) 1233 (2) ISS8 (2) -473 (IS) -1211 (3) -2059 (3) -2865 (3) -3794 (3) - 2763 (3) -1891 (3) -1131 (3) -1181 (3) -1923 (3) -1721 (3) -826 (4) -76 (3) -281 (3) 333 (3) 1837 (3) 2780 (3) 3449 (3) 4455 (3) 3156 (3) 2233 (3) 1789 (2) 3244 (2) 2092 (2) !SIS (2) 263 (2) 421 (20) 1313 (3) 1662 (3) 2532 (3) 2901 (4) 3030 (3) 2663 (3) 2971 (3) 3628 (3) 4S72 (3) 4978 (3) 4460 (3) 21856 (5) 9950 (S) -12729 (S) -27299 (S) 1469 (2) -432 (2) -2397 (2) 8416 (3) 1687 (3) 1509 (3) 2465 (2) 3411 (3) 3161 (2) 1790 (3) 1568 (3) 1180 (3) 869 (3) 409 (4) IOOS (3) 1408 (3) 1984 (3) 2121 (3) 1790 (4) 2059 (4) 2643 (4) 2977 (4) 2704 (3) 2908 (3) 3585 (3) 4212 (3) 4448 (3) Sll4 (4) 4030 (3) 3400 (4) 1117 (2) 2134 (2) 1798 (2) 1806 (2) 888 (2) -184 (15) 645 (3) 6SS (3) 1178 (3) 1231 (4) 1648 (3) 1624 (3) 2191 (3) 273S (3) 3264 (3) 3684 (3) 3573 (4) 3049 (3) 2631 (3) 2033 (3) 1248 (3) 1063 (3) 522 (3) 315 (4) 168 (3) 363 (3) -2026 (2) -2426 (2) -1565 (2) -803 (2) -626 (2) -1756 (19) -2153 (4) -2744 (4) -3255 (3) -3916 (4) -3120 (3) -2508 (3) -2013 (3) -198S (3) -235S (3) -2229 (3) -1775(3) 13605 (3) 17345 (3) 20060 (3) 21148 (3) 1045 (I) 1428 (I) 1556 (1) 2057 (2) 558 (2) 1238 (2) 1953 (2) 2977 (2) 2033 (2) -175 (2) -83 (2) -667 (2) -611 (2) -1235 (3) 42 (2) 610 (2) 1811 (2) 24S9 (2) 2596 (2) 32SS (3) 37S6 (2) 3628 (2) 2962 (2) 2637 (2) 2686 (2) 3104 (2) 2877 (2) 3330 (3) 2204 (3) 1819 (2) 1143 (2) 2002 (2) 2511 (2) 3419 (2) 2341 (2) 245 (11) 481 (2) 3 (2) 186 (2) -336 (2) 863 (2) 1337 (2) 2507 (2) 3206 (2) 3747 (2) 4166 (2) 4S77 (2) 4311 (2) 3617 (2) 3172 (2) 3025 (2) 3375 (2) 3051 (2) 3442 (3) 2351 (2) 2025 (2) 1295 (2) 2081 (2) 2722 (2) 3746 (2) 2747 (2) 412 (14) 632 (2) 99 (2) 227 (2) -352 (3) 901 (2) 1428 (2) 2634 (2) 3314 (2) 3S09 (2) 4206 (2) 4682 (2) 333 (4) 301 (4) . 342 (4) 444 (4) 40 (2) 43 (2) ss (2) 84 (3) 34 (2) 32 (2) 29 (2) 40 (2) 37 (2) 5.21 (0.19)~ 33 (3) 62 (3) 56 (3) 74 (4) 43 (3) 41 (3) 33 (3) 36 (3) 41 (3) 37 (3) S6 (3) so (3) 39 (3) 42 (3) 45 (3) 53 (3) S3 (3) 66 (4) 41 (3) 47 (3) 31 (2) 30 (2) 28 (2) 34 (2) 29 (2) 4.~ 39 (3) ss (3) 36 (3) SI (3) 30 (2) 30 (2) 36 (3) 39 (3) 40 (3) ss (3) 70 (4) 46 (3) 42 (3) 4S (3) 34 (3) 40 (3) 27 (2) 36 (3) 32 (3) 37 (3) 39 (2) 30 (2) 27 (2) 37 (2) 33 (2) 4.Q()b 50 (3) 62 (3) 59 (3) 94 (4) 4S (3) 39 (3) 30 (3) 36 (3) 39 (3) 36 (3) so (3) 702 (5) 673 (S) 633 (5) 745 (6) 77 (2) 48 (2) 66 (2) 125 (4) 63 (3) 73 (3) ss (3) 61 (3) so (3) 419 (4) 375 (4) 394 (4) 519 (4) 31 (2) 40 (2) 45 (2) 115 (4) 31 (2) 42 (2) 40 (2) 44 (2) 45 (2) -34 (3) 67 (3) 53 (3) -44 (4) -20 (2) 0 (2) -18 (2) 9 (3) -3 (2) 2 (2) -5 (2) -9 (2) 3 (2) 4S (3) 38 (3) S6 (3) 116 (3) 6 (I) 18 (2) 28 (3) 3 (2) 8 (2) 2 (2) 3 (2) 6 (2) 98 (4) 21 .(3) 142 (3) 174 (4) 6 (2) 13 (2) S (2) 40 (3) 11 (2) 13 (2) 10 (2) S (2) 8 (2) 67 (4) 65 (4) S4 (4) 88 (S) 64 (4) S3 (3) S4 (3) ss (3) 72 (4) 87 (4) 99 (S) 81 (4j S9 (4) . .53 (3) 43 (3) so (3) 37 (3) 67 (4) 64 (4) 74 (4) 47 (2) 49 (3) 44 (2) SI (3) SI (3) ss (3) 34 (3) 46 (3) 59 (4) so (3) 4S (3) 50 (3) 46 (3) 52 (3) 68 (4) S2 (3) 48 (3) 49 (3) 4S (3) 54 (3) 62 (3) 78 (4) 103 (S) 74 (4) 52 (3) 38 (2) 47 (2) 34 (2) 36 (2) 38 (2) 5 (3) -12 (3) -s (3) -9 (3) -2 (3) -4 (2) -9 (2) -9 (2) I (3) -9 (3) -18 (3) -12 (3) -11 (2) -12 (2) -8 (2) -7 (2) -2 (2) I (3) 9 (3) I (3) 4 (2) S (2) I (2) 3 (2) 4 (2) 8 (2) 8 (2) 18 (3) 28 (3) IS (2) 12 (2) 6 (2) I (2) 6 (2) -7 (3) -14 (3) -1 (3) 2 (2) 4 (2) II (2) 17 (3) 25 (3) 33 (3) 14 (3) 4 (2) 6 (2) 12 (2) 7 (2) S (2) 6 (2) 22 (3) 13 (3) 11 (3) -4 (3) 14 (3) IS (3) 16 (3) 18 (3) 16 (3) 21 (3) 16 (3) I (3) 16 (3) 6 (3) 14 (3) S (3) 9 (3) 14 (3) 20 (3) 16 (3) 9 (2) 7 (2) S (2) 6 (2) 6 (2) S8 (3) 56 (3) S2 (3) 76 (4) 54 (3) 45 (3) 40 (3) 4S (3) 56 (3) 65 (4) 76 (4) 67 (4) 44 (3) 39 (3) 43 (3) 53 (3) 48 (3) 86 (4) SI (3) 56 (3) ss (3) 53 (3) 47 (3) 52 (3) 51 (3) 40 (3) 34 (3) 48 (3) 61 (3) so (3) 42 (3) 43 (3) 40 (3) SI (3) S3 (3) 36 (3) 41 (3) 39 (3) 39 (3) 40 (3) 36 (3) 53 (3) 70 (4) 51 (3) 38 (3) 36 (2) 41 (2) 34 (2) 36 (2) 37 (2) 6 (2) -6 (2) -4 (2) -6 (3) 5 (2) -1 (2) -2 (2) 3 (2) 6 (2) 12 (3) 3 (3) S (3) I (2) -4 (2) -6 (2) -3 (2) -2 (2) 4 (3) 6 (2) I (2) 4 (2) 0 (2) I (2) 3 (2) -1 (2) 6 (2) 2 (2) -3 (2) -2 (3) 4 (2) S (2) 11 (2) 14 (2) 14 (2) 27 (3) 18 (3) 7 (2) 17 (2) 8 (2) 4 (2) -2 (2) I (2) -1 (3) 13 (2) 9 (2) 8 (2) 7 (2) 4 (2) 3 ( 2) 6 (2) 4 (2) 7 (2) 19 (2) 21 (3) IS (2) 11 (2) 12 (2) 10(2) 10 (3) 7 (3) -2 (3) 6 (3) 7 (2) 10 (2) 8 (2) 8 (2) 14 (2) 19 (3) IS (3) 6 (2) 11 (2) 9 (2) 13 (2) 9 (2) 11 (2) 82 (4) 76 (4) 52 (3) 66 (4) 62 (4) so (3) 39 (3) 45 (3) 53 (3) 60(4) 67 (4) 49 (3) 43 (3) 48 (3) 73 (4) so (3) SI (3) 43 (3) 42 (3) 54 (3) 60(3) 4S (3) 9 (3) -0 (3) I (3) 7 (3) 8 (2) 3 (2) -s (2) 0(2) 6 (2) 4 (2) 4 (3) 12 (3) 10 (3) 17 (3) 41 (3) 15 (2) 17 (2) 3 (2) 6 (2) 8(2) -4 (2) -12 (2) 16 (3) 11 (3) 10 (3) S (3) 16 (3) 17 (3) IS (2) 14 (2) 13 (3) 16 (3) 9 (3) -s (1) 123 Table IV (Continued) Cl2C C13C Cl4C Cl5C Cl6C Cl7C CIBC C19C C20C NID N2D N3D N4D N5D 04D CID C2D C3D C4D CSD C6D C1D C8D C9D CI0D CllD Cl2D Cl3D Cl4D Cl5D Cl6D C17D C18D Cl9D C20D C21 C22 C23 C24 C25 C26 C27 C28 C29 C30 C31 C32 C33 C34 C35 x./.. y/b z/r u,. u., u., u,. u., u., 3522 (3) 3117 (3) 2146 (3) 590 (3) 4 (3) -916 (3) -1526 (3) -1252 (3) -655 (3) -1859 (2) -2890 (2) -1257 (2) 95 (3) 602 (3) -982 (31) -1801 (4) -2576 (4) -3484 (4) -4358 (4) -3548 (3) -2741 (3) -2210 (3) -2375 (3) -3179 (3) -3089 (4) -2226 (4) -1406 (3) -1500 (3) -801 (3) 771 (3) 1654 (3) 2357 (3) 3302 (4) 2169 (3) 1305 (4) 6173 (4) 6195 (4) 6826 (4) 7428 (4) 7438 (4) 6801 (5) 6417 (4) 6317 (4) 5463 (5) 4736 (4) 4852 (4) 5698 (5) 40 (10) 600 (6) 652 (7) -1412 (3) -1533 (3) -1257 (3) -509 (3) -65 (3) 286 (3) 800 (4) 142 (3) -287 (3) -2996 (3) -2289 (3) -2507 (3) -2486 (3) -3119 (3) -3826 (30) -3448 (4) -3642 (4) -3367 (4) -3570 (5) -2914 (4) -2739 (3) -2198 (3) -1751 (3) -1335 (4) -973 (4) -1037 (4) -1462 (3) -1811 (3) -2294 (3) -2961 (3) -3276 (4) -3776 (4) -4124 (4) -3944 (4) -3602 (4) 5389 (4) 5470 (4) 5963 (4) 6396 (4) 6319 (4) 5802 (4) 4259 (4) 5156 (4) 5701 (4) 5350 (4) 4444 (4) 3906 (4) 4134 (7) 4776 (I) S633 (7) 4495 (2) 51 (3) 36 (3) • 37 (3) 38(3) 48 (3) 40 (3) 6°1 (3) 30 (3) 36 (3) 47 (2) 38 (2) 42 (2) 54 (3) 47 (2) 62 (4) 47 (3) 44 (3) 40 (3) 51 (3) 40 (3) 76 (4) 65 (4) 66 (4) 65 (3) 68 (3) 59 (3) 59 (3) 61 (3) 42 (3) 39 (3) 7 (3) -2 (2) I (2) -4 (2) -1 (2) 4 (2) 12 (3) 5 (2) 3 (2) -8 (2) -1 (2) -1 (2) -6 (2) 3 (2) 0 (2) -1 (2) 3 (2) 5 (2) 13 (2) 18 (2) 25 (3) 8 (2) -1 (2) 12 (2) 11 (2) 10 (2) 7 (2) 16 (2) 7 (3) 14 (2) 16 (2) 11 (2) 7 (2) 15 (2) 25 (3) 17 (3) 18 (3) 8 (2) 18 (2) 21 (2) 17 (2) 22 (2) 72 (4) 82 (4) 69 (4) 119 (6) 74 (4) 61 (4) 54 (3) 55 (3) 68 (4) 73 (4) 72 (4) 60(4) 47 (3) 51 (3) 47 (3) 67 (4) 61 (4) 85 (5) 69 (4) 63 (4) 57 (4) 61 (4) 81 (5) 59 (3) 54 (3) 57 (3) 46 (3) 68 (4) 71 (4) 52 (3) 50 (3) 49 (3) 52 (3) 59 (3) 64 (3) 95 (4) 122 (5) 104 (5) 74 (4) 92 (5) 144 (6) 60 (4) 98 (5) 123 (6) -9 (3) -10 (3) -11 (3) -10(4) -8 (3) -3 (3) -5 (2) -8 (3) :-4 (3) -13 (3) -21 (3) -15 (3) -7 (2) -10 (3) -9 (3) -8 (3) -8 (3) -2 (3) 11 (3) I (3) 16 (3) -I (3) 18 (4) -7 (3) 14 (4) 53 (4) 36 (4) -1 (4) 22 (4) 22 (4) 16 (3) 10 (3) -5 (3) -9 (4) 3 (2) IS (2) -1 (3) -5 (3) 8 (3) -8 (4) 18 (3) 19 (3) 23 (3) 14 (3) 7 (3) -5 (3) -I (3) 10 (3) 15 (3) 24 (3) 18 (3) 29 (3) 36 (3) 50 (4) 36 (3) 22 (3) -10(4) 52 (4) 30 (4) -2 (4) 43 (4) 24 (4) 27 (3) 22 (3) 26 (3) 29 (3) 35 (4) 26 (3) -34 (8) 5 (7) -9 (7) 3798 (2) 3417 (2) 3431 (2) 3&83 (2) 3o50 (2) 4137(3) 2946 (2) 2531 (2) 1434 (2) 2207 (2) 2828 (2) 3801 (2) 2742 (2) 689 (21) 785 (2) 257 (2) 375 (2) -194 (3) 1030(2) 1562 (2) 2765 (2) 3450 (2) 3675 (3) 4371 (3) 4829 (2) 4611 (2) 3919 (2) 3514 (2) 3429 (2) 3812 (2) 3501 (3) 3898 (3) 2800 (3) 2446 (3) 3978 (3) 4643 (3) 5136 (3) 4975 (3) 4323 (3) 3814 (3) 1951 (3) 1892 (3) 1937 (3) 2041 (3) 2085 (3) 2047 (3) 202 (7) S79 (4) 462 (6) 39 (3) 42 (3) 37 (3) 60 (3) 69 (4) 53 (3) 45 (3) 49 (3) 52 (3) 47 (2) 50 (3) 64 (3) 4.QOb 59 (4) 63 (4) 59 (3) 80 (4) 44 (3) 48 (3) 38 (3) 49 (3) 44 (3) 66 (4) 85 (4) 64 (3) 51 (3) 50 (3) 56 (3) 49 (3) 46 (3) 53 (3) 44 (3) 57 (3) 83 (4) 46 (3) 69 (4) 61 (4) 104 (5) 137 (6) 115 (5) 114 (5) 12S (5) 90 (5) IOI (5) 158 (6) 227 (11) 138 (7) 171 (9) 84 (4) 75 (5) 94 (5) 135 (6) 88 (5) 88 (5) 109 (5) 87 (5) 84 (5) 61 (4) 92 (5) 103 (5) 55 (4) 133 (9) 132 (9) 100 (8) 48 (4) 57 (4) 70 (4) 78 (4) 71 (4) 87 (5) 73 (4) 149 (10) 11S (7) 151 (10) -IS (4) 3 (4) -97 (9) -47 (6) -58 (6) 11 (2) 11 (2) 18 (3) 32 (3) 24 (3) 4 (3) 14 (2) 2 (2) 10 (3) I (3) 9 (3) -2 (3) 29 (3) 23 (3) -29 (4) 15 (4) 23 (3) 5 (3) 63 (4) 33 (4) 33 (4) 40 (4) 35 (4) 25 (4) 17 (4) 34 (4) 93 (9) 41 (6) 79 (8) • See noteo, Table Ill. b Isotropic B. peaks at 326 (L), 327 (HL), and 389/391 (Cul) mass units. These do not appear in tbe spc:ctrum or Cu(L)CO, 3. The mass spc:ctrum of one or the amorphous mas,,:s showed a weak peak at 44 mass units, as well as a significant peak at 456/458 mus units. A subsequent crystal structure analysis or one or the parallelepipeds showed a carbonate-bridged structure (CuL,)(l'•CO 3), 4. G'"1ti of Siaclt Crymk rn. Beu... ill dot Ab9alce of CMllaa Dioxide. The same crystal-growing process was carried out in a glovebox in which air near atmospheric pressure was continuously passed over Ascarite in order to remove any carbon dioxide. Again dark red parallelepipeds were formed, together with amorphous material. The infrared spc:ctrum or tbe sample was identical with tbat of the product isolated in the prevw cxpcrimcnL Several crystals were examined by photographic X-ray diffraction techniques. Some were found to have tbc cell dimensions of the bridging carbonate dima structure, 4, while otbers had a different cell. A structure determination oi these latter cryuals showed a high molecular weight tetramcric complex (CuL)i(CuLO)(J&-OHh, 5. X-r.y Data CollectiaL Preliminary precession and Weis.scnbcrg photographs showed triclinic symmetry for the crystals or (CuL)r (i,-CO 3), 4, aDd. (CuL)i(CuLO)(J&-OH)1, 5. F-.rthcr data were colJectal on a Datex-automatcd Gcncral Electric qllaflCT-circle diffractometer using Cu Ka radiation (X 1.5418 A). Lattice pa- rafflCICt'S (Table II) were demed from a least-5quares fit to the setting angles for 16 manually centered reflections for 4 and 17 for 5. Intensities were measured using 9-28 scans with a scanning rate of 1• /min, a takeoff angle of3°, and background a,unts of 20 sat each extremum; the scan range was 1.8° for 4 and varied from 1.8° (at 28 • 3°) to 3.0° (at 2S = 120°) for 5. For 4, 4578 reflections were measured out to 28 =II~. orwbich 4406 had positive net intensities and were used in the structure analysis; for 5, 93 34 reflections were measured out to 28 = 120°, or which 8709 were positive. Three check reflections measured periodically for each crystal showed no significant changes. The intensities for 5 were corrected for absorption,. (the crystal bad dimensions 0.10 X 0.10 X 0.37 mm), witb tbc transmission coefficients ranging from 0.79 to 0.86; intensities for 4 (0.08 x 0.15 x 0.38 mm) were not corrected. The observational variances a1(F0') were based on counting statistics plus an additional term (0.025S) 2, where S is tbe scan count. StnctJlre Sohrtioa aad RefinemHt The fwiction minimizal in the lcast-~uares refinements was 'f.w<F.7 - F/)1, with weights w equal to,,- (F.').'' The R index is DIFJ- IF, 11/'f.lFJ. Neutral-atom scattering factors for Cu were taken from Cromer and Wabcr,'° tb011e for C, N, and O from ref 42. and tbosc for H from Stcwan ct al." The real component or 124 62 anomalous dispenion was included for Cu. (c.L),(l'-<:0,), 4 Most of the atoms in the asymmetric: unit were found by direct methods." Differeoc:e Fouria' IIUIJ)I :ineldcd Ille positions of the ranaining nonhydrogen .atoms and. eYC11tuaUy, of the methyl hydrogen atoms; the nmaining hydrogen atoms positioned in idealized locations with C-H a 0.96 A. Final refinement was by blocked least squares, with the coordinates of 1he 56 heavier atoms in one ma1rix and lheir anisotropic temperature parameters, a scale factor, and an extinction paramc1e,- in a second. Tbe final R inde1 was 0.058 for 4578 observations and 506 paramcte,-s. A final difference map showed no fea1ures greater than 0.4 c/ A1. Final atomic parameters for lhc nonhydrogcn atoms arc given in Table III; those for lhe hydrogen atoms are available as supplementary material. (Od.h(Cal.O)(i,-OH)1, S. Positions for all a,pper, bridging oxygen, and nitrogen atoms were obtained by direa methods;" difTen:noo maps showed the positions of the ligand carbon atoms and of the 2.5 benzene molecules in lhe asymmetric unit. lsouopic refinement based on data out to 80° in 28 lowered R to 0.086. Hydrogen atoms were assigned ideali=I ooordinates (C-H • 0.96 A) except for those of the methyl groups and the bridging OH groupo. which were located on difference were maps. Afte,- initial block-diagonal anisotropic refinement, a difference map showed a number of features. The most prominent were two peaks of magnitude approximately 5 e/ A1, one being identified as a water molecule, 05, and the second located close to the ring carbon atom CI of ligand A; lhrce smaller peaks we,-e located close to the CI atoms of 1he other three ligands. These latte,- four peaks were . identified as oxygen atoms, 04, in panially occupied si1es. lo lbe final least-squares cycles, the matrices we,-c blocked in the following way. The coordinates of CIA and 04A, the anisotropic temperature parameters for CI A, and an isotropic temperature parameter and a population factor of 04A were relined in one matrix. Similar matrices were set up for CI Band 04B, for CIC .;,nd 04C, and for CI Band 04D, except that the isotropic temperature facton for these 04 atoms were no1 relined. In all four cases a panial hydrogen atom, with a population factor equal to the difference between 1.0 and tbal of the corresponding 04 alom, was placed in lhe idealized position (C-H 0.96 A); its paramctcn were 1101 relined. The ooordinates and anisouopic temperature parameters of the 1hrce carbon atoms C33-<:35 of lhe benzene ring lying on a cen1cr of symmetry were also included in a single mauix, since the thermal motion of this molecule was ve,-y large and the C--<: distances were irregular. (They remained irregular at the conclusion of the refinement, varying from 1.31 to 1.56 A. It is likely that this molecule is somewhat disonlcred.) The coordinates and anisotropic tcmpe,-ature parameters of the remaining Cu, 0, N, and C atoms were blocked into a 3 3 and a 6 X 6 matrix for each atom. . The final R index was 0.065 and the goodness of lit [:[w(F.' 2 1 2 F/) /(11 - p)] i was 1.87 for p 1125 parameters and 11 = 9334 reflections. The largest residual peaks on a difference map were 0.78 c/A' near Cu4 and 0.52 e/A' near Cul. Final atomic parameters for lhc nonbydrogcn atoms are given in Table IV; those for the hydrogen atoms are available as supplementary material. = X = Acknowledgment. We gratefully appreciate financial assistance from the National Science Foundation, the National Institutes of Health, and the International Copper Research Association. Repay No. 2, 62341-98-4; 3, 64811-68-3; 4. 68682-94-0; 5 (isomer A), 68715-72-0; 5 (isomer B), 68738-46-5. s.i,,ie.-wy Malffial Aftilallle: Positional and the,-mal pa• ramcters used for the hydrogen atoms in the refinement of 4 (Table V) and 5 (Table VD, individual bond distances (Table VU) and angles (Table VIII) for the tridcntate ligands. bond distances and angles for lhe benzene molecules in the unit ceU of 5 (Table IX), and listings of lhc structure factor amplitudes for 4 (Table X) and 5 (Table XI) (58 pages). Ordering information is given on any current masthead page. Referac:es ud Notes (I) J. Ptmch, P. Ailm,and W. E. BlumborJ, '"The Biocbani,uyo/Owa'", Academic Prea, New Yart. I 966. (2) W. H. Vamiate and A. Zubcrl,QhJcr in "Mola:ulat Mechanisml al Oxyga, Aaivatian", 0. Hayai,hi. F.d.. Academic Pr-. New Yon, 1974, p 371. (3) R. Malkin in "lnorpnic: Bioc:llcmiSl1y", G. I. Eichbom, Ed., Eberier. New York. 1973, p 689. <•I K. N. Raymond. Ed .. Ado. Chem. S,r.. Ne. 161. (l9'T7). (5) K. E. Van Holdund E F. J. Van Brogren.Biol. Al.oa""""' .. 5, 1 (1971). (6) R. I..Dfttie and R. Witten, "lnorpnic Biochemistry", G. I. Eichhorn, Ed .. Elsevier. New Y0<~ 1973, p ~ (7) J. Bonaventun, C. Bonaventun, and B. Sullivan. J. Exp. Zoo/ .. l'4, 155 (1975). (8) (a) R. L Jolley, Jr., L He Eva111, and H. S. Ma.son, Biochmi. Bioplty,. R,s. Com"'w•.. 46,878 (1972); (b) R. L Jolley, Jr .. L. H. Evans, N. Makino, and H. S. Mason. J. Biol. Clum .. U9. 335 ( 197•). (9) W. S. Cauabey, W. J. Wallace, J. A. Volpe. and S. Y01hikawa. "Tbo Enzymes", Vol. 13. 1975, p 299. (10) (al R. Aas&. R. S.-.J. Demmn, B. G. Malmstrom, B. Reinlwnmer, and T. Vam,gard. FESS utt.. 61. 115 (1976); (b) &oclum. BiophyJ. Rn. Comm""-, 79, 1204 (1976). (11) R. Brinden and J. Dcinum, FESS utt., 73, 14' (1977). (12) 0 . Farver, M. Goldberg, D. Lancet. and I. Pecht. Bioclum. BiopltyJ. R,s, Commwn., 73. •9• (I 976). (13) J. S. Locbr, T. B. Freedman. and T. M. Locbr, Bioclum. Biopl,ys. Ra. Commw11., 54,510 (197•). (t•) T. 8. Freedman, J. S. Locbr, and T. M. Locbr, J. Am. Clum. Soc .. '8, 2809 (1976). (15) T. J. Thamann. J. S. Loehr, and T . M. Loehr, J. Am. CMm. Soc., 99, 4187 (1977). (16) J.A. l..arnboe, T.G.Spro, N.S. Ferris, W. H. Woodruff. W. A. MaJte,e, and M. S. Ken, J. Am. Clum. Soc .. 99. 1979 (1977). (17) A. Zuberbultlcr in "Mew Ions in Biological Sysums". Vol 5, H. Sigel, Ed .. Marcel Dekker. New York. 1976, p 325. (IS) JC J. Oliver, T. N . Wal1er>, D. F. Coot, and C.-E. F. Richard, 1-,. C/t;M. Acta. 2-C. 85 (1977). (19) D. M. L. Goodgame. M. Goodgame. and G. W.R. Canham, Na,wr (London), ll2, 866 ( 1969). (20) C. E. Kramer, G. Davies, R. B. Davis, and R. W. Slaffll. J. Clwm. Soc., Clum. Commwn., 606 (1975). (21) C. S. Artus, J. L -Wilkinson, C. Mealli. T. J. Marks, and J. A. lbers, J. Am. Clum. Soc .. 96. 756,4 (197•). (ll) (a) S. J. Kim and T. Takizawa. J. Clttm. Soc., Clt,m. Commun .. 356 (1974): (b) Makromol. Chtm .. 176,891 (1975). (23) J. E. Buliawski. P. L Burk, M. F. Ludmann. and J. A. Osbam, J. Clwm. Soc .. Cli,m. Commun., •98 ( I977). (24) H. Sakurai, A. Yolt:oyama, and H. Tanaka, Clum. Pluum. Bull., II, 2373 (1970). (25) L. Graf and S. Fallab, Exp,ri,n1ia. ll, "6 (196-4). (26) L. J. Wilson. J. Clum. Soc., Clum . Commwn. . 634 (1978). (27) F0< eumple see: (a) R. D. Gray. J. Am. Clum. Soc., 91. 56 (1969): (b) M. Giit.enspergor and A. D. Zubcrbllhlcr, H,ID. Chim. Acuz. 60. 2584 (1977); (cl A. L Crumbliss and L. J. GcstauL J. Caord. Clum. 5. 109 (1976); (d) J. A. Arce, E. Spodine. and W. Zamudio. J. 1-,. Nw:I. Clttm .. 37. I~ (1975); 31, 2029 (1976). (28) R. R. Gagnt. J. Am. Ch,m. Soc., 91, 6709 (1976). (29) R. R. Gagnt. J. L. Allison, R. S. Gall, and C. A. Koval. J. Am . Clwm. Soc., 99, 7170 ( I 977). (30) R. R. GagnE, J. L. AlliJon. and G. C. Lilemky, lnor,. Cli,m., 17. 3563 (1978). (31) R.R. GagnE, C. A. Koval, and T. J. Smith, J. Am. Clum. Soc., 99. 8367 (1977). (32) J. S. Tbompoon. T. J. Marks, and J. A. lbers, Proc. Natl. Acad. Sci. USA , 14. JI 14 (1977). (33) Y. Engdborghs, S. H. DcBruin, and R. IAatie. Biopltys. Clum., 4, 3•3 (1976). (34) 8. Salvato, A. Gbimti-Magaldi. and F. Ghin:tti, &ocltmustry, 13, •778 (197•). (35) W. 0. Sieg!. 1-,. Niu:/. Clum. utt .. 10, 825 (197•). (36) J. A. Elvidge and R. P. Linstc:ad. J. Clt,m. Soc., 5000 (1952). (37) M.A. Robiman. S. l. Trotz, and T. J. Hutley. 1-,. Cltmt.. 6,392 (1967). (38) M011 infran,d sp,aro obuin<d dwing this imestigation wen: uniformau... due to the Large number of ligand ablorptions and because the spectra were euc:ntially identical for all products in the region 600-1800 cm·•. (39) H. C. Fr,eman and G. Robinson, J. Ch,m. Soc .. 31,. ( 1965); R. L. Harlow and S. H. Simonsen, Acta Cry11allo,r. S<ct. B. 3l. 466 ( 1976). (.0) G. Barclay F. Hoskitu, J. Clum. Soc., 586 ( 1962); (bl R. J. Gene and M. R. Soow, J. C,_,_Soc, A. 2981 (1971); (c) K. Kaas and A. M. 5;n:mcn. Acta Crystallogr., Sttt . B, 29. 113 (1973): (d) P. D. Brotbertoo and A. H. White, J. Clum. Soc., Da/to,r Trans .. 2338 (1973); (e) R. L. Harlow and S. H. Simonxn, Acta Cry11allogr .. S«t. B, 31. 1313 (1975): (0 E. &taniok. H. C. Fr<eman. J. M. James, and C. E. Nocbolds, J. Clt,m. Soc. A, 2558 (1970). (•I) P. C. Healy and A.H. Wbite,J. CJ,a,,. Soc.. Daito,, Tra,a .. 1913 (1972); - abo ~fercnces ,l()d and 40e. (•2) G. A. Barclay and 8. F. Haslam, J. °""'- Soc, 2807 (1963): R. Ouabes, Y. Maoucbc, M. Perucaud. and P. Herpin. C.R. Htbd. Sta111:u Acad. Sci., S,r. C, 276, C28I (1973). (43) J. Chatt. M. Kubota, G. J. ~gh. F. C. March. R. Muon. and D. J. Yarrow. J. Clum. Soc., Chem. Commun., 1033 (1974); M. H. Meyer, P. Singh, W. E. Hatfield. and D. J. Hodssan. Acta Crystallogr., S«t. B, JO. 103 (197•): S. Krogsrud, S. Komiya, T. Ito, J. lber>, and A. yam&IDOIO, lflOrf. Cli,m., 15, 2798 (1976). (4') F. Zignan and H. D. Sdrusla, Z . Krisrallogr, !vistall-., Krisrallpltys., /vistallclwm., ll5, •16 (1972); P. Suooc. Ac1a C,ysta/Jogr, 22. 1"6 (1967). (45) "lnu:mational Tables fer X-R.ay Crysallograpby", Vol. Ill, Kynocb "'-, Birmingham. EnaJand, 1962. (a) A. and B. 125 (-46) Aver2ge of all phenol bond lengths for strvau,... with R las than 0.15 (47) (48) (49) (SO) (51) (52) (53) (54) in "Molecular Strvc;tures and Dimensions". Vol. Al. published for the Crystallographic Oat.a Centre. Cambridge. and the International Union of Crystallography by Oosthoc:k. Utrecht. Netherlands. 1972. M. I. Brua:: and A. P. P. Ost.auwski. J. Climr. Soc .. Da/10,t Tra,u .. 2433 ( 1973). C. Mcali. C. S. Arcus. J. L. Wilkinson. T. J. Marks. and J. A. Jbers. J. Am. Chtm. Soc .. '8. 71 I ( I 976). M. R. Chur<:hill. B. G. Dellocr. F. J. Rotella, 0. M. Abu Salah, and M. I. Brvce, lnor1. Co,m ., 14. 2051 (1975). J. 0. Alben, L. Yen, and N. J . Farrier, J. Am. Chtm. Soc., '2, 4-475 (1970). L. Y. Fager and J. 0 . Alben, Biocotmisrry, II . 4786 (1972). W. 0. Sieg!, S. J. lapportc, and J . P. Collman. I-,. Chtm ., 10, 2158 (1971). M. M. Taqui !ubn and A. E. Mancil, '"'"''· Chtm., 13, 2961 ( 1974). 8. Durham, T. J. Anderson, J. A. Switzer, J . F. Endicott, and M. D. Glick. ,,,,,,,_ 16. 271 { I 977). c,,m., (55) M. C. Weiss and V. L. Goedken, J. Am. Chtm . Soc.. '9, 3389 { 1976). {56) W.R . Busing and H. A. Lev). A<1a Crysrallorr., 10, 180 (1957). (57) Except for the programs o«n.r' and MULTAN ," the computer programs used were from the CRYM cryst.allographic computing system. (58) C. K. Johnson. "OR TEP. a Fonran Thermal Ellipsoid Plot Program", Repon ORNL-3794 Oak Ridge National Laboratory, Oak Ridge, Tenn., 1965. {59) (a) G. Germain, P. Main, and M. M. Woolfson, Acta Crysialloir .. S«t. A, 27,368 (1971); (b) P. Main. M. M. Woolfson, L. Lcssinger. G. Germain, and J. P. Declercq, "Multan 74, a System of Computing Programmes for the Automatic Solutions of Crystal Structures from X-Ray Diffraaion Data", University of York, York, England. and laboratoin: de Chimie Pllysiquc ct de Crist.allographie, louvain-la-Ncuvc, Belgium, Dec. 1974. (60) D. T. Cromer and J. T. Wabcr. Acra Crystallotr., II. 104 (1965). (61) R. F. Stewart. E. R. Davidson. and W. T. Simpson. J. CJ,~m. Pays., 42, 3175 (1965). (62) D. T. Cromer, Act~ Crystalloir .. II, I 7 ( 1965). 126 Appendix Two. Oxidation state of Copper in Reduced Copper(II) Macrocycles and Their Adducts The eighteen-electron CuLBF2 macrocycle was found to react with carbon monoxide and form a twenty-electron system. 1 The crystal structure 2 of CuLBF2 shows no unusual bond distances that would indicate macrocyle reduction rather than copper reduction upon synthesis from Cu(II)LBFiC1O4), and the macrocycle is slightly twisted toward tetrahedrality as expected for copper(!). The crystal structure 3 of CuLBF CO shows a very unusual five-coordinate 2 geometry, with the copper atom O. 96 A above the plane of the four nitrogen atoms and a 120° angle between a-diimine planes. Both of these structures are now listed in Cotton and Wilkinson's fourth edition text 4 as examples of ·new coordination types shown by copper(!). The crystal structures seem to indicate that CuLBF2 CO is the unusual state and CuLBF2 is the normal; examination of a series of infrared2 and carbon-13 nuclear magnetic resonance spectra by this author suggests the opposite. Cu(II)LBF2 (ClO4) has infrared bands assigned to the a-diimine unit at 1660 and 1590 cm- 1 • In four coordinate CuLBF2 there are no bands in this region, but two new bands are found at 1470 and 1320 cm- 1 • Loss of the infrared imine bands upon one-electron reduction sounds suspiciously like ligand reduction. In CuLBF2 CO the a -diimine bands are again found and are located at 1640 and 1560 cm- 1 • No ligand reduction is indicated by the infrared spectrum for CuLBF2CO. These infrared spectra are shown in Fig. 1. In the carbon-13 nuclear magnetic resonance spectra, the two types of 127 .,.. ..,.....:....;_ --~ -- n '.' ;. --- L.....-1-• ----! '----- 2200 2000 -,- . ....;..,.+C"" 2000 J900 l800 1700 1600 1500 WAVENUMBER CM-1 1400 1300 1 Fig. 1. Infrared spectra in the region 1200 to 2100 cm - of Cu(Il) LBF2 (ClO4 ) (top), CuLBF2 CO (center), and CuLBF2 (bottom). 1200 128 imine carbons can be expected to have different chemical shifts. For the Zn(II)LBF (ClO ) complex 2 the imine chemical shifts differ by 2 4 15.1 ppm. In one-electron reduced Cu(II)LBF2 (ClO4), the imine chemical shifts are only 4.4 ppm apart. The CuLBF2 CO complex again has the normal imine chemical shifts with a 14. 2 ppm difference. Data for these and other complexes are in Table 1. Table 1 13 C nmr imine chemical shifts (difference) in CD2 C~ Complex a-diimine infrared bands • CuLBF2 ClO4 1660, 1590 cm ZnLBF2 ClO4 1665, 1590 161. 4, 152.3 (15. l) CuLBF2 CO 1640, 1560 164.3, 150.1 (14. 2) CuLBF2PPh3 1630, 1550 161. 8, 150.3 (11. 5) 160. 3, 153.5 (6. 8) 160.1, 155.7 (4. 4) CuLBF2 -py CuLBF2 N-Meim 1620, 1540 CuLBF2 1470, 1320- -1 The stronger the binding constant of the fifth ligand3, 5 the more the imine spectra (both carbon-13 nuclear magnetic resonance and infrared) resemble those of the copper(II) and zino(II) complexes. The weaker the binding constant of the fifth ligand, the more the imine spectra resemble those of CuLBF2 with delocalization and spreading of the electron density. Only one similar study of nuclear magnetic resonance shifts of a macrocycle as a function of additional coordination to the metal has 129 been published. Busch and coworkers 6 found a correlation between the proton chemical shifts of the methyl groups in cooo.lt(III) macrocycles and the ligand field strength of the coordinated axial ligand. As the axial ligand field got stronger, the methyl group signals shifted to higher field, but the effect was very small. D. M. Ingle has examined the reduced NiLBF2 ClO4 compounds. 7' 8 NiLBF2 also lacks the normal a -diimine infrared oo.nds while NiLBF2CO has them. While the CuLBF2 complexes can be studied by nuclear magnetic resonance as above, the NiLBF2 complexes can be studied by electron paramagnetic resonance and fhe location of the unpaired electron determined. The epr spectrum of reduced NiLBF2 (ClO4 ) indicates that the electron goes into the ligand, but the electron is on the metal in NiLBF2 CO. A simple bonding description9 suggests the orbital scheme shown in Fig. 2. Figure 2A applies to the M(II) complexes [Cu(II) shown, (A) (B) cr*=•~~D CJ ~ o• • o •oa~ w*=~~·~ M N M N Fig. 2. Interaction of d orbitals with a- diimine orbitalso Ni(II) has one less electron]. Addition of an electron somehow lowers the ,r* orbital and the result is Fig. 2B (CuLBF2 shown, NiLBF2 has one 130 less electron). Note that in this description "Cu(I)" is really Cu(III) ligand dianion and "Ni(I)" is really Ni(II) ligand radical anion. Population of the 1r* orbital both changes the vibrational frequencies and equalizes the chemical shifts· of the imines. Coordination of a fifth ligand changes the geometry and results in an orbital diagram like Fig. 2A (NiLBF2 CO shown, CuLBF2 CO has one more electron). Thus the true metal(I) species are the five coordinate complexes in this bonding description. The designation of the four coordinate species as reduced ligand complexes must be modified by the reminder that they are actually quite covalent in nature. ·Copper(III) is one extreme and copper(!) is the ether. The variation of the infrared and nuclear magnetic resonance parameters with strength of axial ligand binding demonstrates this. Investigation of additional Ni(I) complexes by D. M. Ingle 8 reveals that the above dicussion is only applicable to ligands containing a-diimine groups. For. nonconjugated imines like Me 6 -diene or Me 6 - tetraene, reduction of the metal(II) complex does yield a true fourcoordinate metal(!) complex. Copper(!) and nickel(!) complexes with nonconjugated imines also form five-coordinate species. The CO binding constants of a number of these complexes have-been reported. Table 2 summarizes these, as well as gives the CO binding constants for all new complexes reported in this thesis. This work verifies that the tetraene complexes bind CO much more strongly than the diene complexes. An interesting comparison is the effect on the reduction pctential of adding the two extra methyl groups going from Me 6 -diene to Me8 -diene 131 Table 2 Complex Ef * Keo ** NiMe 6 -diene( Cl04)2 -1.241 4. 7 X 10 8 NiMe8 -diene{Cl04) 2 -1. 216 4. 8 X 102 ·this work NiMe8 -tetraene{Cl04) 2 -1.101 NiMe 6 -tetraene( ClO4) 2 -1.081 2. 0 X 10 4 4. 5 X 10 Ni(DMGBF2) 2 -0.786 2. 8 X 10 5 8 NiLBFlCl04) -0.617 2. 7 X 10 3 8 Ni-butyl-LBF2 (Cl04) -0.798 1. 3 X 10 !Q, CuMe8tetraene( ClO4)2 -0.762 6.1 X 10 !!, CuMe 6 -tetraene{Cl04)2 .1, CuMe 6 -diene(Cl04) 2 -0.683 2. 5 X 10 this work -0.656 4. 7 X 10 5 ~' CuMe8 -diene(ClOJ 2 -0.594 < 10 this work !, CuLBF2 {Cl04) -0.456 1. 2 X 105 5 §_, Cu(DMGBF2 ) 2 -0.438 8. 8 X 10 5 5 ~' Cu-butyl-LBF2 {Cl04 ) -0.270 1. 2 X 10 3 5 4 4 4 4 4 Reference this work .~. 8 . 8 this work *On pt or Hg electrode (no significant differences noted) in dimethylformamide. Potentials converted to values versus SHE using ferrocene as an internal standard. 1 O **By the method in reference 5, using the concentration of CO in dimethylformamide 11 under one atmosphere of CO as 4. 64 x 10~ M. 3 132 and from Me 6 -tetraene to Me8 -tetraene. 1) The reduction potentials of the copper complexes are affected more than those of the nickel complexes. 2) The reduction potentials of the diene complexes are shifted positive. 3) The reduction potentials of the tetraene complexes • are shifted negative. 133 References 1. Ro R. Gagne, J. Amer. Chem. Soc., ll, 6709-6710 (1976). 2o R. R. Gagne, J. L. Allison, and G. C. Lisensky, Inorg. Chem. !1, 3563-3571 (1978). 3. R. R. Gagne, J. L. Allison, R. S. Gall, and C. A. Koval, J. Amer. Chem. Soc., ---. 99, 7170-7178 (1977). 4. F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry," Wiley, N. Y., 4th ed., 1980, p. 799. 5. R. R. Gagne, J. L. Allison, and D. M. Ingle, Inorg. Chem., ll, 2767-2774 (1979). 6. E. S. Gore, J. C. Dabrowiak, and D. H. Busch, J.C. S. Chem. Comm., 923-924 (1972). 7. R. R. Gagne and D. M. Ingle, J. Amer. Chem. Soc., 102, 1444""""" 1446 (1980). 8. R. R. Gagne and D. M. Ingle, Inorg. Chem., accepted for publication. 9. R. R. Gagne, J. K. Burdett, D. M. Ingle and P. D. Williams, private communication. 10. R. R. Gagne, C. A. Koval, and G. C. Lisensky, Inorg. Chem., 19 (1980) in press. ,...,.. 11. "Properties and Uses of Dimethylformamide," DuPont Industrial Chemicals Dept., Wilmington, Del., 1976. 134 Appendix Three. 13 C nmr chemical shifts.* N N e N N a b i C f h C a b,d,e f, g,h i Me 6 -diene(HC1O4) 2 174.3 40.8 43.1,46.7,57.7 20.8,23.9,23.9 Me 8 -diene(H ClO4) 2 173.0 42.1 46.0,51.7,57.6 19.6,24.1,24.4 18.2 NiMe 6 -diene(ClO4) 2 182.1 43.4 51.2,52.2,54.3 22.3,22.7,24.4 NiMe 8 -diene(ClO4) 2 .183. 0 49.0 51.0,52.9,58.7 21.7,23.4,24.1 17.9 NiMe 6 -tetraene(ClO4) 2 186.5 174.1 50.7,57.7,64.3 24.5,25.5,25.5 NiMe 8 -tetraene(ClO4) 2 185.9 176.3 50.2,57.4,69.5 23.9,24.6,25.6 16. 3 168.8 164.7 55.6,57.8,59.4 23.0,26.3,26.5 17.2 CuMe8 -tetraene( C104) *ppm vs. internal tetramethylsilane. Dimethylsulfoxide-d6 used for Ni compounds and acetone-d6 used for the Cu compound. Spectra of other Cu compounds were net obtainable due to solubility and disproportionation. 135 Proposals 136 Proposal I. Synthetic Phosphorus Chemistry A synthetic organic phosphorus chemistry scheme is suggested whereby the syntheses of the first, a~. phosphorus donor, binucleating macrocycles are outlined. Mononucleating all phosphorus donor macrocycles have only been prepared for the first time in the last few years, and similar synthetic reactions will be used. 137 Synthetic Phosphorus Chemistry Much of the interest in macrocyclic ligand systems has been in the field of cyclic polyethers. These oxygen donor compounds are used for the ligation of the relatively hard alkali and alkaline earth metal cations. Although some softer nitrogen and sulfur containing cyclic ligands have been used for transition metals, only recently have the first tetrakis (tert. -phosphino) macrocycles, shown in Fig. 1, been reported. These macrocycles are much more appropriate for ligation of the softer transition metal~ M(O), Pd(II), pt(II), pt(IV), Cu(I), Ag(I), Au(I), Cd(II), Hg(I) and Hg(II). It is proposed that similar Fig. 1. Known tetrakis (tert. -phosphine) macrocycles. l- 4 138 synthetic techniques be used to obtain the first all phosphorus donor macrocyclic complexes capable of complexing two metal ions. While this proposal concerns synthetic phosphorus chemistry, such binuclear metal complexes should exhibit interesting properties based on the greater covalency and extension of the phosphorus ligand orbitals. Several problems have limited previous work with phosphorus ligands in general. Phosphorus compounds are susceptible to air oxidation and all reactions must be carried out under O2 -free conditions. Tervalent phosphorus compounds with attached alkyl groups are especially sensitive since 1h e electron donating alkyl groups increase the already high electron density on the phosphorus atoms. As Crofts 5 has pointed out, the nauseating smells, tendency to spontaneous inflammability, and toxicity, have not helped to make phosphine chemistry popular. Moreover, few starting materials are commercially available. Synthetic routes must be planned from simple compounds as is done here. The reaction of chloranil with the commercially available phosphonite (iso-Pr0) 2 PPh is the beginning of the syntheses proposed here and should give the product shown in Eq. 1. (Powers and Graham (1) PlO-uoR-) P\\ 8 have previously noted that when chloranil reacts with the tertiary phosphite (iso-Pr0) 3 P, all four chlorine atoms are replaced by phosphonate groups). Tetravalent RgP = 0 compounds can in general be 6 139 reduced to trivalent phosphorus by treatment with LiA1H4 • 7 ' 8 In addition to removal of the = 0 by LiA1H4 , halogens or -OR groups are exchanged with H. LiA1H4 treatment will also probably result in reduction of the ring to give an aromatic product (Eq. 2). Since phosphines with P-H bonds can be further reacted with alkali metals ~ P(o-tsoP.-)Ph (2) P.. Uso-Pto) P D 0 LtA\Ha, Plo-uoP..> P\\ 8 H 1 "' or vinyl phosphines, the product, !, in Eq. 2 is the common key starting material for several new macrocyclic compounds. 3 Another needed intermediate has been used by Kyba for synthesis of a mononucleating macrocy.cle and is prepared according to Eq. 3. Ph (3) ~PH ~PH ~ 2 " The Arbusov reaction (Eq. 4), 9 where A and B may be alkyl, aryl, alkoxy, or aryloxy groups, is the best studied phosphorus-carbon bond forming reaction. Since the reaction is favored for A and B 140 A OR ,ef (4) 8 +RX + r..• 1· electron donating, the order of reactivity is aryoxy < alkoxy < aryl < alkyl. The usual I > Br > Cl reactivity is found and primary halides work best. Of significant interest in macrocycle synthesis is when 0 ll 0 II .. 10 XR'X is used to form ABP-R-PAB, followed by LiAlH4 reduction (Eq. 5). P-H bonds react with BuLi or PhLi yielding P-Li, which upon reaction with alkyl halide RX gives P-R and net phosphorus carbon bond formation. 3, ll, 12 Intermediates 2 and 3 would be treated in " " this way to give 4 and 5 (Scheme I). " ....... 141 Scheme I Ph @r 2 "' Ph PH Sult ®:Li 0 PH---. ~ PLt Ph Ph Ph Ph Ph Ph ~ i, HPCH2C.Hl.PH + t! Bult ~ Li.Pclia.C.H,PLi. +Z.BuH LtPCHi_CHa.P L, + exc.ess ClC.H1 CH&Cl-4z.Cl ~ P\,. ---+ Cl C.Hz.C.H:LC.Ha.P C.H,_ C~ PCH&.CHi, C.t-1~C.l . 5 "' . Lastly, ! could be similarly treated with BuLi {Eq. 6) and then (6) reacted with 4 and 5 to give the desired new macrocyclic ligands {Scheme II), "' "' Macrocycles related to .1 and ~ could also be prepared by using different legnth alkyl halides in Eq. 5 and Scheme I. In summary, the syntheses outlined here, starting from (iso-Pr0) 2 PPh and (PhP) 5 , should yield the first all phosphorus donor binucleating macrocycles. Scheme II Lt Ph 6 " 5 " Lt Ph • 6 " 4 " 143 Appendix. Phosphorous compound names. Trivalent Tetravalent Phosphines Phosphine oxides RsP, I¼PH, RPH2 .RsP=O, l¼PO(H) Phosphinites Phosphinates I¼P(OR) I¼PO(OR) I¼POH phosphinous acid l¼PO(OH) 2 ° phosphonic acid Phosphonites Phosphonates RP(OR) 2 RPO(OR) 2 RPO~ 2 phosphonous acid RPO(OH) 2 1 ° phosphonic acid Phosphites Phosphates 3° P(OR) 3 3° PO(OR) 3 2° (RO) 2POH 2° (RO) 2PO(OH) 1 ° (RO)PO2H 2 1 ° (RO)PO(OH) 2 if RO = X, halophosphite if RO = X, halophosphate Table after M. Porthault, in "Chimie Organique du Phosphore," Colloques Internationaux du Centre National de la Recherche Scientifique, No. 182 (1970). 144 References 1. L. Horner, Phosphorus, §_, 63 (1975). 2. T. A. DelDonno and W. Rosen, J. Amer. Chem. Soc., 99, 8051-8052 (1977); Inorg. Chem., ll, 3714-3716 (1978). 3. ""' E. P. Kym, C. W. Hudson, M. J. McPhaul, and A. M. John, J. Amer. Chem. Soc., ~, 8053-8054 (1977). 4. R. E. Davis, C. W. Hudson, and E. P. Kyba, J. Amer. Chem. Soc., .!QQ, 3642-3643 (1978). 5. . P. C. Crofts, Quart. Rev., ,!!, 341-366 (1958). 6. J. F. Powers and G. R. Graham, Abstract 134th American Chemical Society Meeting, 87P (1958); U.S. Patent 2,935,578 (1960); Chem. Abstr., 54, 19598 (1960). ""' 7. L. Horner, H. Hoffmann and P. Beck, Chem. Ber., 91, 1583""' 1589 (1958). 8. L. Maier in "Organic Phosphorus Compounds," G. M. Kosolapoff and L. Maier, Eds., Wiley-Interscience, New York, 1972, Vol. 1, pp. 4-7. 9. R. G. Harvey and E. R. DeSombre, Topics in Phosphorus Chemistry, !, 57-111 (1964). 10. A. H. Ford-Moore and J. H. WflJiarns, J. Chem. Soc., 1465. (1947). 11. K. Issleib aJ?-d F. Krech, Chem. Ber., ~' 2656-2669 (1961). 12. R. Uriarte, T. J. Mazanec, K. 0~ Tau, and D. W. Meek, Inorg. Chem., .!i, 79-85 (1980). 145 Proposal n. Polypyrrole Coated Electrodes As Diodes It is proposed to constrct a DC power supply by conversion of an AC source using a polypyrrole coated electrode solution cell in the circuit. 146 Polypyrrole Coated Electrodes As Diodes A recent communication1 concerning electrochemical properties of polypyrrole coated platinum electrodes deserves further inquiry. The published observation is that for coated electrodes, a cyclic volta;m ogram . of ferrocene (E O positive of that of polypyrrole) could be obtained while a cyclic voltammogram of nitrobenzene (E O negative of that of polypyrrole) could not. First it is proposed to verify these results. The observed differences may have been due to the differing approach of ferrocene and nitrobenzene to the polypyrrole surface. The published conclusion was that current is carried by the electrode only at potentials positive of the polypyrrole oxidation and no current is carried at pc:tentials negative of the polypyrrole reduction. This conclusion is better verified by using one compound with two reversible waves, one positive of polypyrrole and one negative. Ru(acac) 3 is one possibility. It is further proposed to make a solution analogue of a half wave rectifier, by replacing the diode in a normal circuit 2 with a cell containing a polypyrrole coated electrode (Fig. 1). Application of an AC voltage alternating about the polypyrrole E O should result in a DC current output. Observation on an oscilloscope of such a waveform would be conclusive proof of the published hypc:thesis. A solution analogue of a full wave rectifier is also possible by .adding a second polypyrrole electrode (Fig. 2). The output from this DC power supply might even be filtered somewhat due to the faradaic capacitance of the cell. 147 solution cell with polypyrrole electrode v· lT\ I Fig. L Fig. 2. Half-wave rectifier. 148 Another proposed area of further inquiry is the amount of current capable of being carried by the oxidized polypyrrole layer. Presumably this can be controlled by changing the size of the electrode area. For large currents (~ 10 Amp or greater) vacuum tubes are still used rather than solid state diodes. There are two distinct disadvantages to use of the vacuum tubes. First, the vacuum tube's filament has high power requirements (~4 Amp at 6. 3 V). Second, the vacuum tube requires a vacuum and the glass tube used to maintain that vacuum is fragile. Thus replacement of vacuum tubes in high current power supplies by a polypyrrole solution cell has possible commercial potential. To adequately evaluate such potential, further experiments are needed to investigate the breakdown voltage of the reduced polypyrrole and the lifetime of such an electrode surface. 149 References 1. A. F. Diaz and J. I. Castillo, J.C.S. Chem. Commun., 397-398 {1980). 2. A. J. Diefendorfer, "Principles of Electronic Instrumentation," Saunders, Philadelphia, 1972, pp. 152-173. 150 Proposal m. Gas Chromatographic Resolution of Enantiomers It is proposed that gas-liquid chromatographic separation of enantiomers will be enhanced by use of optically active solid supports loaded with an optically active liquid phase. 151 Gas Chromatographic Resolution of Enantiomers Optical isomers occur widely in.nature and are utilized for pharmaceutical purposes, for experimental biochemical studies, and as intermediates in the syntheses of biologically active products. Methods which distinguish or resolve enantiomers are very important, especially in pharmaceutical applications where the two enantiomers can have different biological effects. The use of gas chromatography to resolve enantiomers has mainly been achieved by conversion of the racemic mixture to a mixture of diastereomers by a suitable chemical reaction with an optically active reagent. l, 2 The diastereomers are then separated by normal gas chromatographic techniques. Previous studies of gas-liquid chromatography for the direct separation of enantiomers have centered on the use of optically active liquid phases, most notably N-trifluoroacetyl amino acid esters, and carbonyl-bis (amino acid esters) (Fig. 1). 3 Several observations suggest that more ordered optically active phases increase the 0 R II *I Fig. 1. CF3 -C-N-C-C-OR' H I II N-trifluroacetylaminoacidester H 0 0 R R II *I H*I CF3 -C-N-C-C-N-C-C-OR' N-trifluoroacetyldipeptide ester H I II f II H O R' 0 H 0 R' I 11 *I ROC-*c-N-C-N-C-C-OR II I I I i 11 OHH HBO Carbonyl-bis(amino acid ester) 152 enantiomeric separation. (1) Corbin and Rogers 4 have discovered that use of a carbonyl-bis(L-valine isopropyl ester) liquid phase in an open column operated below the liquid phase melting point gave greatly enhanced separations. They suggested the enhanced activity resulted from a more fixed interaction geometry. (2) Lochmuller and Souter 5 ' 6 studied the same carbonyl-bis(L-valine isopropyl ester) as a 5% loading on 100-120 mesh AWDMCS Chromosorb G in a 6 ft x ½" O.D. column. The operation temperature was chosen such that the "liquid" phase was a smectic liquid crystal mesa-phase. The largest relative retention ever observed for any gas chromatographic enantiomer separation was obtained. While use of liquid crystal mesophases generally seem to give greater separation than the corresponding isotropic liquids, the operating temperatures are somewhat limited. For the carbonyl-bis(L-valine isopropyl ester) the smectic liquid crystal phase exists in the region 99 to 109 ° C. Consequently only materials volatile in this narrow range can be examined and much of the utility of gas chromatographic analysis is lost. It is proposed that optically active solid supports in addition to continued use of optically active liquid phases may enhance enantiomeric separation under a wider range of operating conditions. The underlying theory is that if good separation has been observed for the more ordered chiral "liquid" phases, where presumably the separation is due to a diastereomeric complex formed by solute-solvent interactions, the ordering of the liquid phase could also be accomplished by diastereomeric interactions with the optically active solid support. 153 Gas liquid chromatography accomplishes a separation by partitioning a sample between a mobile gas phase and a thin layer of nonvolatile liquid held on a solid support. While the solid support is usually considered inert, such is not actually the case. The commonly used diatomaceous supports tend to absorb sample molecules particularly when lightly loaded with liquid phase or when nonpolar liquid phases are used. This is the reason that the silanol groups are often treated with dimethyldichlorosilane to convert them to silyl ethers and reduce surface activity. Since the nonoptically active solid supports generally used are silicates, use of the optically active silicate quartz (SiO ) 7 seems a 2 good place to begin this investigation. Quartz is the most important optically active crystal, both because of its widespread occurrence and because of its high optical activity. In quartz crystals the silica groups are joined to each other in left-handed screw arrangements in • some crystals and right-handed screw arrangements in others. These two types of crystals would be separated and ground into the usual mesh size for loading into a column. Both optical isomers will be tested with the carbonyl-bis(L-valence isopropyl ester) liquid phase and various degrees of dimeth~dichlorosilane pretreatment. The same isomers to be separated as used by Corbin, Rogers, Lochmiiller and Souter would be chromatographed to check for enhanced separation. Further inquiry would investigate other optically active solid supports. 154 References 1. E. Gil-AvandD. Nurok, Adv. Chromatography, !.Q, 99-172 (1974). 2. P. Newman, "Optical Resolution Procedures for Chemical Compounds," Manhattan College, N. Y., 1978, Vol. 1, pp. 565582. 3. C. H. Lochmiiller and R. W. Souter, J. Chromatography, ill, 283-302 (1975). 4. J. A. Corbin and L. B. Rogers, Anal. Chem., ,..,._ 42, 974-979 (1970). 5. C. H. Lochmiiller and R. W. Souter, J. Chromatography, ~' 243 (1973) 6. 0 C. H. Lochmuller and R. W. Souter, J. Chromatography, ~' 41 (1974). 7. E. A. Wood, "Crystals and Light," Van Nostrand, Princeton, N. J., 1964, pp. 136-144. 155 Proposal IV. Synthesis of Linked Macrocyclic Ligands Synthesis of a tetramine from malonic ester for the purpose of deriving linked macrocyclic ligands is proposed. The ultimate aim is to study metal-metal interactions where the interaction, either direct or through bridging ligands, is along the z axis. 156 Synthesis of Linked Macrocyclic Ligands Metal atoms or ions held in close proximity at selected distances are of interest because of possible cooperative chemical and electronic interactions and multiple electron redox processeso Indeed, many proteins and enzymes contain two or more metal atoms in sophisticated systems. Studies of binuclear complexes involve measurement of the electrochemical properties, 1 magnetic interactions, 2 electron paramagnetic spectra and electronic absorption spectroscopy. Most binuclear complexes prepared to date contain metal atoms side-by-side. 3 Fewer binuclear complexes have been examined where the metal-metal interaction is along the z axis, either directly or through bridging ligands, although metal(acetate) dimers, 4 some solid state d8 metal 2 complexes, and the face-to-face porphyrin complexes 5 are nctable exceptions (see Fig. 1). L ._I .._ . ., c(,:/R~ y-c'e-:,~ . Fig. 1. Metal-metal interaction along the z axis. A series of ligands that will hold two metal atoms in such a configuration could be derived from the tetramine, (H 2 NCH 2 ) 2 CHRCH(CH2 NH2 ) 2 , synthesized as shown in Scheme I. 157 Scheme I EtOH> Malonic ester is used as the starting active methylene compound since the mono- rather than di-alkylated product can in general be obtained. 6 With a, w-dihaloalkanes cyclization could be a problem but even for Br(CH2 ) 3 Br, whose cyclization is normally rapid, 7 yields of greater than fifty percent of the tetraester are obtained when an excess of one equivalent of malonic ester is used. 8 The dihaloxylenes should present no such problem in any case. Formation of the amide and reduction to the amine appears to be ·straightforward, 9 but alternate routes from the tetraester to the tetramine are also possible. 10 Once the tetramine has been synthesized, a variety of linked macrocycles are available by Schiff base condensation with aldehydes or ketones. One macrocycle, whose mononuclear version is quite readily synthesized, involves the condensation with 2, 3-butanedionemonaxime (Scheme 2). 11 After addition of the metal ion, the oxime bridging hydrogen atom can be replaced by a BF2 group by treatment with BF3 • Et2 O. ll 158 Scheme 2 NH~ 0 II II • . . . n~ 0 "' I ~ ' H Any metal-metal interaction will be dependent on the length of the link and the conformation of the m~crocycles. Molecular models show that all the isomers are interconvertible. Addition of a bridging ligand will tie the metals together in the conformation illustrated above, with one macrocycle lined up vertically over the other. Use of imidazole as a bridging ligand is of interest from the viewpoint of the complex as a model for multi-metal protein centers. Imidazole has been found as a briqging ligand in superaxide dismutase (Cu-Im-Zn) 12 and proposed as a bridging ligand for cytochrome axidase {Fe-Im-Cu), 13 but few non-polymeric model complexes have been prepared. 14 - 18 Other bridging nitrogen heterocycles, 19 such as pyrazine, 20 would also be examined. Diaxygen as a bridging ligand is possible when the macrocycle link is short; for face-to-face cobalt porphyrins, Chang has observed5c both products in Fig. 2, depending on the link length. Fig. 2. Face to face cobalt porphyrins 159 A subject of great interest is when the two metals are different. The first step in such a synthesis of mixed metal linked macrocycles involves condensing an N2 O2 donor tetradentate ligand with the tetramine, similar to the preparation of a known mononuclear macrocycle {Scheme 3). 21 After isolation of the mononuclear complex with the Scheme 3 linked diamine, condensation with a second tetradentate metal complex, where the metal is different, would be accomplished. The bridging ligand would again be added as before. Alternatively, two Jager-type ligands {Fig. 3) 22 , 23 could be similarly condensed to form additional Fig. 3. Jager ligands mixed-metal linked macrocycles. In the Jager ligand the extra carbonyl functionality is necessary to activate the ligand toward 160 further condensation. It would be highly desirable to synthesize doubly linked macrocycles with the macrocycles face-to-face. The Jager-type ligands could be synthesized using the tetramine for the first condensation, as shown in Scheme 4. • Addition of the metal, and use of high-dilution Scheme 4 He ( 0 Cz.Hs)., + + t. -+ techniques for the addition of a second tetramine, could give the desired product. Synthesis of the doubly linked macrocycle may, however, be a low-yield process with predominantly polymers produced. 161 References 1. T. J. Meyer, Advances in Chemistry, 1§.Q, 73-84 (1976). 2. D. J. Hodgson, Prog. Inorg. Chem., ~' 173-241 (1974). 3. . S. E. Groh, Israel J. Chem., _!&, 277-307 (1976/77)0 4. R.J.Doedens, Progr. Inorg. Chem., ll, 209-31 (1976). 5. J. Po Collman, C. M. Elliot, T. R. Halbert, and B. So Tavrog, Proc. Natl. Acad. Sci. USA, 74, 18-22 (1977); """ N. E. Kagan, D. Mauzerall and R. B. Merrifield, J. Amer. Chem. Soc., 99, 5484-6 (1977); and C. K. Chang, Advances """ in Chemistry, ill, 162-77 (1978). 6. A. C. Cope, H. Lo Holmes and H. 0. House, Organic Reactions, .,.._ 9, 107-331 (1957) . 7. A. C. Knipe and C. J. M. St:irling, J. Chem. Soc. (B), 67-71 (1968). 8. J. Cason and H. Rapoport, "Laboratory Text in Organic Chemistry," Preritice Hall, Englewood Cliffs, N. J., 1955, pp. 295-308. 9. N. G. Gaylord, "Reduction with Complex Metal Hydrides," Interscience, N. Y., 1956, pp. 544-551. 10. M. s. Newman, D. H. Buscg, G. E. Cheney, and C. R. Gustafson, Inorg. Chem., ll, 2890-3 (1972). 11. ~- R. Gagne, J. L. AlliSJn, R. s. Gall, and c. A. Koval, J. Amer. Chem. Soc.;~' 7170-8 (1977). 12. J. S. Richardson, K. A. Thomas, B. H. Rubin, and D. C. Richardson, Proc. Natl. Acad. Sci. USA, ll, 1349-53 (1975). 162 References (continued) 130 Go Palmer, G. To Babcock, and L. E. Vickery, Proc. Natl. Acad. Sci. USA, 73, 2206-10 (1976). """ 14. G. Kolks, C. R. Frihart, H. N. Rabinowitz and S. J. Lippard, J. Amer. Chem. Soc., .,.,-. 98, 5720-1 (1976) . 150 D. Kovacs and R. E. Shepherd, J. Inorg. Biochem., !Q, 67-88 (1979). 16. Go Kolks, S. J. Lippard, J. Amer. Chem. Soc., .,..,,.., 99, 5804-5806 (1977). .. 17. M. S. Haddad and D. N. Hendrickson, Inorg. Chem., ll, 2622-2630 (1978). 18. C. L. O'Young, J. C. Dewan, H. R. Lilienthal, and S. J. Lippard, J. Amer. Chemo Soc., .,...,...,.. 100, 7291-7300 (1978) . 19. 20. M. Inoue and M. Kubo, Coord. Chem. Rev., 21, 1-27 (1976). """ J. H. Fuhrhop, M. Baccouche, and M. Btinzel, Angew. Chem. Int. Ed. Engl., )&_, 322-323 (1980). 21. M. G:reen, J. Smith, and Po Ao Tasker, Inorg. Chim. Acta, .§_, 17-24 and 65-69 (1971). 22. L. Wolf and E. G. Jager, Z. Anorg. Allg. Chem., ~ ' 76-91 (1966); E. Jager, Z. Chem., .,... 8, 30 and 392-3 and 470-1 (1968) . 23. D. P. Riley and D. H. Busch, Inorg. Syn., ll, 36-40 (1978). 163 Proposal V. Substitution Kinetics of Five-Coordinate Complexes Many kinetic studies have been done on the ligand substituVon reactions of four- and six-coordinate metal complexes. Much less studied kinetically are the ligand substitution reactions of fivecoordinate complexes. The proposal is made to study the monodentate ligand substitution reactions of M(terpyridine)X2 and M[(N,N'-di-(3-aminopropyl)piperazine)X]Y. The object is to determine if an associative or dissociative pathway is fallowed, and if the sixteen and eighteen electron rule applies to ligand substitution in five-coordinate complexes. 164 Substitution Kinetics of Five-Coordinate Complexes A large number of studies have been done on the mechanisms of substitution of four- and six-coordinate metal complexes. Squareplanar rf metal complexes are found to undergo substitution reactions with associative mechanisms involving a five-coordinate intermediate. Six-coordinate complexes substitute ligands by predominantly dissociative pathways, again involving a five-coordinate intermediate. 1 The sixteen and eighteen electron rule, as stated by Tolman, 2 says that diamagnetic organometallic complexes of transition metals may exist in spectroscopically o:r kinetically detectable concentrations only if the metal's valence shell contains sixteen or eighteen electrons. Thus sixteen electron square planar a8 systems go through an associa"'. tive intermediate with eighteen electrons, and eighteen electron octahedral complexes go through a dissociative intermediate with sixteen electrons. Considerably less work has been done on the substitution reactions of five-coordinate complexes. At first glance, steric considerations would suggest these complexes could undergo substitution with an associative six-coordinate intermediate, or a dissociative four-coordinate intermediate. The sixteen and eighteen electron rule for diamagnetic complexes would require that the eighteen electron five-coordinate complexes substitute with a sixteen electron dissociation intermediate. Monodentate diagmagnetic eighteen electron five-coordinate complexes are found to undergo the substitution reaction MDJ{ + Y - MD4 Y + X 165 with characteristically dissociative kinetics~-lO [Co(2, 6-lutidine-1axide) 5 ][Cl04 ] is a seventeen electron species which substitutes dissociatively o ll The only sixteen electron five-coordinate series of complexes that has been studied13 is Mn(NO)(C0) L. For substitution of CO by 3 phosphine Y, the mechanism depends on L and Y. For L = CO and Y = PPhg or P(0Ph) 3 the mechanism is associative with an eighteen electron intermediate as expected. For L = Y = P(OPh) 3 the reaction is dissociative with a fourteen electron intermediate. For L =Y = PPh3 , the associative and dissociative pathways are both used to the same extent. For L = Y = PBUs the dissociative pathway is favored, but beth mechanisms occur. The dependence on L could be due to the stabilization of the four-coordinate intermediate., or the lack of room for the entering ligand, and the sixteen and eighteen electron rule is not always followed. Five coordinate compounds with multidentate ligands are even less predictable. The (Ni(o-phenylenebisdimethylarsine)2 Br] SCN complex exchanges Br- for SCN- dissociatively as expected for an eighteen electron complex. 12 The eighteen electron pt(Il) tris(o-diphenylarsinophenyl)arsine halide mono-cation and pt(Il) tris(o-dimethylarsinophenyl)arsine halide moo.a-cation react associatively through nineteen or twenty electron intermediates. 14 , 15 Another troublesome system has been studied kinetically by Sweigert and coworkers (Eq. 1), 16 - 20 where X and Y are different phosphine ligands. The complexes themselves, as well as their substitution reaction intermediates, violate the sixteen and eighteen 166 . (1) X Z RJ(s,u~cR +Y➔ R s,, 's_) R electron rule. They react associatively for (R = Ph, CF3 ) fourteen electron (Fe, z=O) and fifteen electron (Co, z=O) systems and by both paths for (R = CF3 , CN) sixteen electron (Co, z=-1) systems. The substitution kinetics of a five-coordinate high spin seventeen electron iron(m) (p-CH3 -tetraphenylporphyrin) complex have been elegantly studied by LaMar and coworkers and found to be associative with an eighteen electron intermediate. 27 - 3o This is the only macrocycle whose five-coordinate substitution kinetics have been examined. The reaction is found to proceed with inversion and is studied by observing the averaging of the inequivalent m-H proton nuclear magnetic resonance signals as a function of added x- concentration (see Fig. 1). Fig. 1. Inversion of a macrocycle upon fifth-ligand substitution. 167 The above review of five-coordinate substitution reactions (five-coordinate reactions with bidentate ligands have been ignored) shows several main themes. Eighteen eiectron diamagnetic monodentate-ligand five-coordinate complexes substitute dissociatively, as expected from the sixteen and eighteen electron rule. The one sixteen electron system that has been studied substitutes by both pathways, which is not expected from the sixteen and eighteen electron rule. With multidentate ligands, more information is needed concerning the kinetics of five-coordinate substitution reactions before further systematization is possible. I propose to study the substitution kinetics of five-coordinate M(terpyridyl)X2 and M [ (NN' -di-(3-aminopropyl)piperazine)X] Y complexes (Fig. 2). The purpose is threefold: (1) to gain more Fi"'. 2. Five-coordinate metal complexes M(terpy)~ (left) and fM(dapp)xt (right). information about five-coordinate substitutions in general, (2) to learn more about fifth ligand substitution pathways as a function of the metal ion, and (3) to learn more about ligand substitution in high spin complexes. 168 The crystallographically characterized geometry of Zn(terpy)C¼ is distorted trigonal bipyramidal, 31 , 32 while that of [Ni(dapp)Cl] Cl is square pyramidal. 33 ' 34 • In both cases a series of complexes with other Mand X have been prepared whose X-ray powder patterns indicate isostructural five-coordinate complexes. 31 , 35 , 36 , 38 These complexes are M(terpy)X2 (for M =Mn, Co, Ni, Cu, Zn, Cd, Hg and X =Cl, Br, and I), 35 - 37 and [M(dapp)X]Y (for M =Co, Ni, Cu and X = Cl, Br, I, NCS and Y = X, ClO4 , PF 6) • 33,38 Ligand !ability is suff~cient to achieve substitution reactions, as noted by previous workers as a tendency toward disproportionation in some solvents (notably water). 39 Non-aqueous organic solvents should permit kinetic studies. Given suitable behavior of these complexes in organic solvents, a variety of Mand X will be examined for typical associative and dissociative kinetics (see Appendix) o 169 Appendix Kinetics of the 8Nl Dissociative Mechanism MD4 +Y Applying the steady state approximation to the first step yields d[MD ] •• 4 dt = 0 = K 1 [MD4X] -k_1 [MD4] [X] -k:a [MD4] [Y] , and rearranging, gives [MD4 ] = k 1 [MD4 X] k:a [Y] + k_l[X] • Substituting this in the rate expression gives k1 k:a [Y] rate = k:a [MDJ [Y] = - - - - [MD4X] . k1 [X] + k:a [Y] When k:a [Y] is small compared to k_ 1 [X] , the rate will be [Y] . kobs -[MD4X]. Excess X will retard the rate. [X] When k:a [Y] is large compared to k_ 1 [X], the rate will be k 1 [ MD4X] and a first-order reaction is observed. For a dissociative mechanism a plot of kobs vs. (Y] will be 1 1 curved, while a plct of vs. - - will be linear with a non-zero kobs [Y] 1 intercept = . 1ti[MD4X] 170 Since a dissociative mechanism involves the production two units from one, .6.Sf is usually positive. In general the rate will depend on electronic stabilization of the four-coordinate intermediate. Racemization is possible. Kinetics of the 8N2 associative mechanism MDJ{Y ~ k_l Applying the steady state approximation to the first step gives .. d[MD4XY] • dt = 0 = k 1 [MD4X] [Y] -k_1 [MD4 XY] -~ [MD4XY] , . k1 [MD4X] [Y] and rearranging, gives [MD4XY] = - - - - - . k_l +~ Substituting into the rate expression results in k1 1½ [MD4X] [Y] k_l +~ • Thus the rate = kobs [Y] [MD4X] and a second-order reaction is observed. Excess X should have no effect on the rate. A plot of kobs vs. [Y] will be linear. Since an associative mechanism involves the production of one unit from two, ~sf is usually negative. In general the rate will depend on the steric crowding of the complex. I The reaction usually proceeds by inversion. 171 References 1. K. F o Purcell and J. C. Kotz, "Inorganic Chemistry," W. B. Saunders Co., Philadelphia, 1977. 2. C. A. Tolman, Chem. Soc. Rev., .,..., 1, 337 (1972) . 3. R. F. Heck, J. Amer. Chem. Soc., .,...,.,..., 85, 651 (1963) . 4. G. Cardaci and V. Narcisco, J.C. s. Dalton, 2292 (1972). 5. E. E. Siefert, R. J. Angelici, J. Organomet. Chem., ~' 374 (1967). 6. B. F. G. Johnson, J. Lewis and M. V. Twigg, J.C.S. Dalton, 1876 (1975). 7. A. Poe and M. V. Twigg, Inorg. Chem., ]1, 2982 (1974). 8. C. G. Grimes and R. G. Pearson, Inorg. Chem., ll, 970 (1974). 9. P. Meakin and J. P. Jesson, J. Amer. Chem. Soc., ,_ 96, 5751 (1974). 10. P. M. Enriquez and S. S. Zumdahl, Inorg. Chem., g, 2475 (1973). 11. P. ~- Enriquez, S. s. Zumdahl, and L. Forshey, J.C.S. Chem. Comm., 1527 (1970). E, 113 (1977). 12. D. Ao Sweigart, Inorg. Chem. Acta. 13. Ho Wawersik and F. Basolo, J. Amer. Chem. Soc., .,..,,.._ 89, 4626 (1967). 14. R. G. Pearson, M. M. Muir and L. M. Venanzi, J. Chem. Soc., 5521 (1965). 15. T. D. B. Morgan and M. L. Tobe, Inorg. Chim. Acta, §_, 563 16. D. A. Sweigart, D. E. Cooper and J. M. Millican, Inorg. Chem., ,,.._,.._ 13, 1272 (1974). 172 References (continued) 17. D. A. Sweigart, J.C.S. Chem. Comm., 688 (1975). 18. D. A. Sweigart, Inorg. Chim. Acta, JJ!, 179 (1976) o 19. D. A. Sweigart and D. G. DeWit, Inorg. Chem., ~' 1582 (1970). 20. D. A. Sweigart, Inorg. Chim. Acta, _!!, 317 (1974). 21. H. B. Gray and E. Billig, J. Amer. Chem. Soc., 85, 2019 """ (1963). 22. C. H. Langford, E. Billig, S. I. Shupack and H. B. Gray, J. Amer. Chem. Soc., -... 86, 2958 (1964). 23. A. L. Balch, Inorg. Chem., §_, 2158 (1967). 24. J. A. McCleverty, Progr. Inorg. Chem., !Q, 49 (1968). 25. J. A. McCleverty, N. M. Atherton, N. G. Connelly and C. J. Winscom, J. Chem. Soc. {A), 2249 (1969)0 26. J. A. McCleverty and D. G. Orchard, J. Chem. Soc. {A), 626 (1971). 27. -- F. A. Walker and G. N. LaMar, Ann. N. Y. Acad. Sci. , 206, 328 (1973). 28. G. N. LaMar, J. Amer. Chem. Soc., -... 95, 1662 (1973). 29. G. N. LaMar, Pure Appl. Chem., ~' 13 (1974). 30. R. V. Snyder and G. N. LaMar, J. Amer. Chem. Soc., .,..,,.. 98, .4419 (1976). 31. D. E. C. Corbridge and E. G. Cox, J. Chem. Soc., 594-603 (1956). 32. F. W. B. Einstein and B. R. Penfold, Acta Cryst., !Q, 924-926 (1966). 33. N. A. Bailey, J. G. GibsonandE. D. McKenzie, J.C.S. Chem. Comm., 741-742 (1969). 173 References (continued) 34. J. G. Gibson and E. D. McKenzie, J. Chem. Soco Dalton, 989-992 (1974). 35. J. S. Judge, W. M. Reiff, G. M. Intille, P. Ballway, and W. Ao Baker, Jr., J. Inorg. Nucl. Chem., ~' 1711-1716 (1967). 36. C. M. Harris, T. N. Lockyer, and N. C. stephenson, Aust. J. Chem., 19, 1741-3 (1966). 37. ---- G. Morgan and F. H. Burstall, J. Chem. Soc., 1649-1655 (1937). 38. J. G. Gibson and E. D. McKenzie, J. Chem. Soc. Ser. A., 1029-1038 (1971). 39. R. Hogg and R. G. Wilkins, J. Chem. Soc., 341-350 (1962).