NICKELGI) COMPLEXES CONTAINlNG' NON -CYCLIC AND MACROCYCUC UGANDS DERIVED FROM 2.3-BUTANEDIONE MONOHYDRAZONE AND 3,3-DIMETHYL- ‘ _ L2~BUTANEDIONE MONOHYDRAZONE Thesis for the Degreeof M. S'. - . MlCHlGAN STATE UNI-VERSH'Y _ ' DONGLAS BRMN BONFOEY - 1973 ‘ LIBRARY Michigan State Unxversity BINDING BY HOAB & SflNS' Ifi-g n-Aa —l--—-—- A..M ABSTRACT NICKEL(II) COMPLEXES CONTAINING NON-CYCLIC AND MACROCYCLIC LIGANDS DERIVED FROM 2,3-BUTANEDIONE MONOHYDRAZONE AND 3,5-DIMETHYL-1,2-BUTANEDIONE MONOHYDRAZONE By Douglas Brian Bonfoey Two new NiN202 complexes have been prepared and charac- terized. Reaction of 2,3-butanedione monohydrazone or 5,}- dimethy1-1,2-butanedione monohydrazone with acetone in the presence of nickel(II) ions results in a NiN202 complex. The complex contains a dinegatively charged tetradentate ligand in a square planar environment about the nickel(II) ion. ‘H nmr studies or these conplexes have established that the nickel(II) ion coordinates in an unsymmetrical mode result- ing in coordination to one five-membered aza ring, one six- membered diaza ring, and one five-membered triaza ring. Reaction of the complex based on 2,3-butanedione mono- hydrazone with 1,2—diaminoethane at 117‘0 resulted in a macrocyclic NiN4 complex, but no reaction was observed be- tween this complex and 1,3-diaminopropane. No reaction was observed between the NiN202 complex based on 3,3-dimethyl- 1,2-butanedione monohydrazone and 1,2-diaminoethane at117"C. The order of reactivity of the coordinated carbonyl Douglas Brian Bonroey groups of the NiN202 complexes is consistent with a mechan- ism involving nucleophilic attack at the carbon atom of the coordinated carbonyl group. NICKEL(II) COMPLEXES CONTAINING NON-CYCLIC AND MACROCYCLIC LIGANDS DERIVED FROM 2,5-BUTANEDIONE MONOHYDRAZONE AND 3,3-DIMETHYL-1,2—BUTANEDIONE MONOHYDRAZONE By Douglas Brian Bonfoey A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1975 ACKNOWLEDGMENT The author wishes to express his sincere appreciation to Dr. G. A. Nelson for his guidance during this work. D. B. B. ii TABLE OF CONTENTS LIST OF TABIIES O O O O O O O O O O O O O O O O O O 0 LIST OF FIGURE 0 O O O O O O O O O O O O O O I O O I. . II. III. INTRODUCTION . . . . . . . . . . . . . . . . . NOMENCLATURE . . . . . . . . . . . . . . . . . EXPERIMENTAL . . . . . . . . . . . . . . . . . Materials ................. Preparation of 2,3—Butanedione Mono- hydrazone . . . . . . . . . . . . . . . . Comments on Synthesis 0 e e e e e e e e e 0 Preparation of [3, 3'-Iso ro ylidenebis(azo)di- 2-but0n-2-OlatO] DiCkel 113, NiDMK o o o 0 Comments on Synthesis . . . . . . . . . . . Purification O O O O O O O O O O 0 O O O O 0 Preparation of (3, 3, 6 ,7, 12 ,13-hexamethyl-1, 2 ,4 ,5, 8 ,11-hexazacyclotr1deca-1, 4 ,6, 12- tetraenato-N', N‘, N' ,N”)nickel(II), Nimcy01°13:-’eo:-oeeoeeoeooe Reaction of NiDMK with 1,3-Diamino- propaneoeeeeeoeeeeeoeeeo Preparation of 3 ,3-Dimethyl-1, 2-butanedione Preparation of [13,1' -Isopropylidenebis- (azo)]bis(3, 3-dimethyl-1-buten-2-olat§] nickel(II),,N1TBK eooeeeeeooee Comments on Synthesis . . . . . . . . . . . iii Page vi oooooxA (I) 10 11 11 12 14 15 iv Reaction of NiTBK with 1,2-Diaminoethane IV. PHYSICAL MEASUREMENTS . . . . . . . . . . V. RESULTS AND DISCUSSION . . . . . . . . . . Characterization of NiDMK and NiTBK . . Reaction of Nickel Ketazines with Amines Reaction of NiDMK with 1,2-Diaminoethane Reaction of NiDMK with 1,3-Diamin0propane. Reaction of NiTBK with 1,2-Diaminopropane. Reaction of Coordinated Carbonyl Groups. Suggestions for Future Work . . . . . . VI 0 CONCLUSIONS 0 O O O O O O O O O O O O O 0 LIST OF REFERENCES . . . . . . . . . . . . . . . mmn O O O O O O O O I O O O O O O O O O O O Page 15 18 19 2o 30 31 36 57 38 40 41 42 43 LIST OF TABLES Mass Spectrum of NiDMK . . . . . . . . . . . Mass Spectrum of NiTBK . . . . . . . . . . . Chemical Shifts in ‘H nmr Spectrum of NiDMK. Mass Spectrum of NiDMcyclo 13 . . . . . . . Infrared Absorptions of NiDMK, NiTBK, and NiDMcyclo 13 . . . . . . . . . . . . . . . Page 22 25 28 32 ’43 Figure 1. 2. 3. 4. LIST OF FIGURES ‘H nmr Spectrum of NiDMK . .-. . . . . ‘H nmr Spectrum of NiTBK . . . . . . . ‘H nmr Spectrum of NiDMcyclo 13 . . . 100 MHz '3 nmr Spectrum of Dimethylene Bridge of NiDMcyclo 13 . . . . . . . vi Page 25 27 54 35 I. INTRODUCTION In recent years inorganic chemists have been investi- gating the effects of the metal ion on a ligand in coordi- nation compounds. Metal ions have been found to catalyze certain reactions-and to stabilize ligands that are unstable in the absence of a metal ion. Coordinated ligands have been found to undergo two‘ types of reactions that provide useful routes for organic synthesis. A "reaction of a coordinated ligand" denotes a chemical reaction in which the ligand undergoes a chemical reaction while it is coordinated to a metal ion. The "tem- plate effect'“' has been proposed to explain how a metal ion controls the steric course of a reaction. Acting as a tem- plate, the Ietal ion holds the ligands in a position that is favorable for a chemical reaction. The reactants then com- bine to form a product that would not be formed in the ab- sence of the metal ion. These reactions have been useful for the preparation of both cyclic and non-cyclic compounds. The cyclic compounds, called macrocycles, have 13-16 membered ring systems sur- rounding the metal ion. Macrocycles, particularly those con— taining four nitrogen donors coordinated in a plane about a metal ion, are significant in their relation to coordination 2 compounds found in biological systems. The condensation of metal amine complexes with carbonyl compounds has been an area of interest in the reactions.of coordinated ligands. These reactions fall into two general classes. The first results in the formation of coordinated Schiff bases:2 “‘1 “33 fi CH —-N NH ——CH 2 2 2 CH M CH 2 2 \ __ /' CH2 NH2 fl CH2 C / \c Structure I The second is characterized by the linking of two coordinated amine groups by a three carbon bridge. Curtis5 prepared a macrocycle from the condensation of monocarbonyl compounds with coordinated diamines: 035\ ,932\ ’93} C C CHa/l I H I,CH3 - C CH3 032 033 Structure II Kerwin and Melsona’5 have synthesized and characterized the Nicke1(II) complexes shown in Structure III: F? o\ /N=N\ /R1 O/N¥\N N/9\R as: 2 ¢ ¢ NiMMK Structure III These complexes were prepared by refluxing benzilmonohydra- zone, nickel acetate, and a ketone (RqRZCO) in ethanol. Each complex contains two benzilmonohydrazone residues linked by a single carbon atom from the ketone. In the ab- sence of the metal ion benzilmonohydrazone condenses with acetone to give benzilacetone azine, 06115 can5 \c— / // (i \ thus demonstrating the ability of the "template effect" to influence the course of a reaction. Other authors have recently reported similar results. (Soedken and Peng6 have reported the synthesis of a macrocy- <3le by the reaction of butane-2,3-dione dihydrazone with formaldehyde in the presence of a metal ion. Alcock and 4 Tosker7 have reported ring closure of Structure IV by a re- action of the coordinated hydrazine with formaldehyde. R s/\s in . N32 “32 Structure IV Condensation reactions of coordinated carbonyl groups have been reported but the ability of the coordinated carbo- nyl group to react is dependent on both the type of amine and the nature of the ligand backbone. Referring to Struc- ture V, it has been reported8'10 that aliphatic diamines will react with the coordinated carbonyl only if R2 is -COOR or “COR 0 Structure V Kerwin4 also investigated the reactivity of the coordi- nated carbonyl with aliphatic amines. NiMMK reacted with ‘ 1,2-diaminoethane and 1,2-diaminopropane to produce a macro- cycle (Structure VI),whereas ethylamine and 5 1,3-diaminopr0pane reacted at only one of the coordinated carbonyl sites (Structure VII). These reactions are of in- terest because the ligand containing the coordinated carbo- nyl is different from those previously studied. ¢ ¢ ¢ ¢ \2 =0’ ‘9=<<’ “'me i/\N:N\/C\ /0 H3 O\Ni/N-—N\ (2/033 R \ / R ‘ /: C3 _ 19"Q\ ,——Q\ ¢ ¢ ¢ ¢ Ni-Hcyclo 13 (R=H) Ni—Ap So (R=(CH2)5NH2) Ni-Mcyclo 15 (R=CH3) Ni-E So (R=Cafis) Structure VI Structure VII The proposed mechanism for the condensation reaction involves nucleophilic attack by the amine at the carbon atom of the coordinated carbonyl. The carbon atom is believed to be electropositive due to the electron-withdrawing nature of the phenyl group and the positively charged metal ion. Since alkyl groups have less electron-withdrawing power than phe- nyl groups, or are electron-releasing, it was proposed to synthesize nickel ketazines with alkyl groups in place of the phenyl groups and to investigate the reactivity of these nickel ketazines towards amines. This thesis describes the preparation of two new nickel ketazines and the reactivity of these nickel ketazines to- wards amines. Nickel ketazines with methyl and t-butyl groups bonded to the coordinated carbonyl have been prepared and characterized. II. NOMENCLATURE The IUPAC nomenclature and structures of the compounds synthesized in the course of this project are shown in Struc- tures VIII and IX. Compounds of the type shown in Structure III were previously named nickel ketazines, and this general structure will be referred to as a nickel ketazine. NiDMK will refer to the nickel ketazine derived from 2,3-butane- dione monohydrazone. The "DM" indicates dimethyl substitu- tion on the hydrazone. NiTBK will refer to the nickel ke- tazine derived from t-butylglyoxal monohydrazone, where "TB" indicates t-butyl. The macrocycle, illustrated in Structure IX, derived from NiDMK will be abbreviated NiDMcyclo 13. The IUPAC name is derived from the fundamental ring system shown below: 12 1a m 11;}::CK1 z C—N N=N I ‘03 c—N N=N/ 'I s\ /s '9 C:::C 7 b 1,2,4,5,8,11-Hexazacyclotrideca-1,4,6,12-tetraene Rq _ /32 /c—c\ O N:: N CH >Ni< &/ 5 N N’ \033 a a Rq” §§C// 2 Structure VIII NiDMK (R1=R2=CH3) [3,3'-isopropylidenebis(azo)di-2-buten-2-olat§lnickel(II). NiTBK (R1=c(cna)3, R2=H) [13,1'-isop opylidenebis(azoi]bis(3,3'dimethyl-1-buten—2- olato nickel(II). CH5\ /QH3 0==C N/ \N‘~ CH’ \ / \N 0152 / N1\ /(|} / CH3 2‘N N '\\ I H CH3 CH /C N 5 \ c/ I CH} Structure IX NiDMcyclo 12 (3,3,6,7,12,13-hexamethyl-1,2,4,5,8,11~hexazacyclotrideca- 1,4,6,12-tetraenatoffi‘,Nfl,§},§fl)nickel(ll). III. EXPERIMENTAL Materials The following chemicals were used as supplied: 2 , B-Butanedione (Aldrich) 3,3-Dimethyl-2—butanone (Aldrich) Glyoxal - 40% aqueous solution (Aldrich) Hydrazine hydrate - 85% solution (Fisher Scientific Company) 1,2—Diaminoethane and 1,5-diaminopr0pane were distilled from sodium hydroxide under nitrogen and stored in a dry box under nitrogen. All other chemicals were of reagent grade or the equivalent. Preparation of 2,5-Butanedione Monohydrazone A single neck 250 ml round bottom flask was assembled with a reflux condenser, heating mantle,and magnetic stir- ring bar. 2,5-Butanedione (7.0 ml, 0.08 moles) was added to 100 ml absolute ethanol in the round bottom flask, and the solution was stirred vigorously while hydrazine monohydrate (4.8 ml, 0.08 moles) was injected rapidly from a syringe. The solution was refluxed for one hour and then immediately reduced to approximately 15 ml by rotary evaporation. The product crystallized as transparent needles at room temper- ature. The product was isolated by vacuum filtration, and 8 9 the crystals were washed three times with one-half milli- liter of ethanol. yield 2.4 g, 30% Comments on Synthesis 2,5-Butanedione monohydrazone was found to be unstable, forming a yellow powder upon standing at room temperature or in a warm concentrated ethanol solution. The yellow powder was believed to be cyclic or polymerized 2,5-butanedione monohydrazone. The polymer has no melting point but slowly decomposes up to a temperature of 250‘0. 2,5-Butanedione monohydrazone is very soluble in ethanol. The polymer is insoluble in ethanol and appears as a precipitate from ethanol. Isolation of 2,5-butanedione monohydrazone de- pends on rapid low temperature crystallization. Preparation of ggifi'-IsoEropylidenebis(azo)di-2-buten-2- o a an c e , 1 A 500 ml round bottom flask was assembled with a reflux condenser, heating mantle, and magnetic stirring bar. 2,}— Butanedione (1.75 ml, 0.02 moles) was added to 100 ml abso- lute ethanol in the round bottom flask, and the solution was heated to reflux. The solution was stirred vigorously while hydrazine monohydrate (1.2 ml, 0.02 moles) was injected rapidly from a syringe. The solution was refluxed for one hour, and then a hot solution of nickel acetate tetrahydrate (2.5 g, 0.01 mole) in 100 ml of 95% ethanol and 7.4 ml of acetone (0.1 mole) were added. An intense red color 10 appeared upon addition of the first few drops of the nickel acetate solution. After addition of all the nickel acetate, the solution was brown and appeared to be a suspension of very small particles. After six days the flask was removed, and the hot solu- tion was filtered through a medium porosity fritted funnel leaving a grey residue. The filtrate was allowed to evapo- rate from a beaker at room temperature until approximately 50 ml remained. Small golden crystals and an amorphous black solid were noted. Recrystallization of the golden crystals from ethanol and n-butanol at room temperature pro- duced larger red-brown crystals. Aggl;, Calculated for' 011318N402Ni: C, 44.44; H, 6.06; N, 18.85. Found: C, 44.64; H. 5.85; N. 19.03. yield 0.3 g, 18% (based on nickel) Comments on Synthesis 2,5-Butanedione has an intense yellow color, and it was noted that the drops from the condenser in the alcohol re- flux were also yellow. Two minutes after the addition of the hydrazine to the solution the drops were clear, suggest- ing that the 2,5—butanedione reacts rapidly with the hydra- zine. The one hour reflux at this point was arbitrary. The nickel acetate solution was prepared by slowly heating on a steam bath. Rapid heating produced insoluble nickel hydroxide. 11 Purification ' Recrystallization from ethanol and n-butanol must be done slowly or the crystals produced are extremely small making filtration difficult. The black material obtained in the first crystallization formed a sludge in the solution making filtration difficult and repeated recrystallization necessary to obtain a pure product. Experience indicates that the amount of black material increases with longer re- action times. A shorter reaction time may, therefore, give higher yields of product. It was found that the product could be sublimed at 110'-120'C under reduced pressure. The sublimed product was recovered as a yellow—brown powder that gave a cherry- red color in ethanol. Preoaration of 6 12 1 -hexamet.~1-1 2 4 8 11_ exazacyc o r '°°a" 0"” " O raena 0"... a“ ' " W31. 1; " I L" NiDMK (0.105 g, 0.45 mmoles) was weighed into a 25 ml flask. The flask was placed in a dry box under a nitrogen atmosphere. Approximately 5 ml of 1,2-diaminoethane was added, and the flask was heated and stirred. The NiDMK did not completely dissolve but after ten minutes red-orange crystals appeared. After thirty minutes the flask was al- lowed to cool, and the crystals were vacuum filtered using a medium porosity glass fritted funnel. The product was re- crystallized twice from acetone. Small red-orange needles were obtained. Anal. Calculated for 013322N6Ni: C, 48.59; 12 H, 6.85; N, 26.17. Found: c, 48.54; H, 6.82; N, 26.54. yield 0.060 g, 35% Reaction of NiDMK with 1,3-Diaminopropane 1‘_gThree days at room temperature. NiDMK (0.1 g, 0.45 mmoles) was stirred in 5 ml of 1,5—diaminopr0pane for three days at room temperature under a nitrogen atmosphere. The ketazine did not completely dissolve, and there was no color change during the reaction period. After three days the mixture was filtered through a medium porosity glass fritted funnel. Small orange-red crystals were recovered on the funnel and were recrystallized from acetone. The mass spectrum obtained on the crystals was identical to the mass spectrum of NiDMK. 2. Thirty_minutes in warm 1,5-diaminopropane. NiDMK (0.1 g) was weighed into a 25 ml flask. The flask was placed in a dry box under a nitrogen atmosphere, and 5 ml of 1,5-diaminopropane were added. The flask was heated on a hot plate with constant stirring. The temperature was con- trolled so that the solvent vapors slowly condensed on the sides of the flask. The NiDMK dissolved, giving a dark red solution. The color did not change during the reaction pe- riod. After thirty minutes the solution was allowed to cool,and crystals appeared. The crystals were removed by vacuum filtration through a medium porosity glass fritted funnel and recrystallized from acetone. The recrystalliza- tion produced orange-red crystals with a mass spectrum 13 identical to NiDMK. 5. One hour at the boiling_point of 1,5-diaminopro- pane (155'9). NiDMK (0.1 g) was weighed into a 25 ml flask. The flask was placed in a dry box under a nitrogen atmos-' phere, and 5 ml of 1,3-diaminopr0pane were added. The mix- ture was heated with constant stirring on a hot plate for one hour. The temperature was maintained at just below the boiling point of 1,5-diaminopropane. The NiDMK dissolved at this temperature and did not recrystallize upon standing overnight at room temperature. The solution was trans- ferred to a round bottom flask, and the 1,5-diaminopropane was removed by rotary evaporation. An attempt to recrys- tallize the product from acetone produced a dark red oil. 4. Six days at~50°0. A Schlenk tube containing 0.1 g NiDMK and 5 ml of 1,5-diaminopr0pane (handled under a nitro- gen atmosphere) was connected to a nitrogen source and par- tially immersed in an oil bath maintained at 50’C. The NiDMK did not completely dissolve but the solution turned to a very dark red color. After six days the NiDMK appeared to be completely dissolved and did not recrystallize on cooling. The solution was transferred to a round bottom flask, and the liquid was removed by rotary evaporation. A dark red oil remained in the flask. The oil was insoluble in acetone but was soluble in ethanol. Attempts to recrystallize from ethanol and t-amylalcohol produced only the red oil. 14 Prgparation of 5.5-Dimethyl-1,2-butanedione Following the method of Taylor11 3,5-dimethyl-2-buta- none (65 g, 0.65 moles) was refluxed with selenium dioxide (47 g, 0.42 moles) at 110‘-120'C for 20 hours. The appa- ratus consisted of a 250 ml single neck round bottom flask assembled with a reflux condenser, oil bath, and magnetic stirring bar. After 20 hours the reflux column was replaced with a condenser, and a distillation was carried out at the reaction temperature. Below 80’0 a yellow liquid and an im- miscible clear liquid were distilled. At 85’C a homogeneous yellow oil was distilled. This liquid was allowed to stand in a stoppered flask, and after two days it solidified to a waxy white solid which had a melting range of 84.5’-85.5°C. Taylorqq identified the white solid as 5,5-dimethyl-1,2-bu- tanedione hemihydrate with a melting point of 85°C. yield 15 g, 51% (based on selenium dioxide) Preparation of'lP,1’-Isoa o lidenebis azo bis(5,5-dimethyl- — u en- -o_a o)]nicke.< , l A 500 ml round bottom flask was assembled with a reflux condenser, heating mantle, and magnetic stirring bar. 3,}— Dimethyl-1,2-butanedione (2.5 89 0.02 moles) was added to 100 ml absolute ethanol, and the solution was heated. Hydrazine monohydrate (1.2 ml, 0.02 moles) was added, and the warm so- lution was stirred for thirty minutes. During the stirring period the solution slowly turned yellow. After thirty mi- nutes a hot solution of nickel acetate tetrahydrate (2.5 g, 0.01 moles) in 100 ml of 95% ethanol and 7.4 ml (0.1 moles) 15 of acetone were added. The initial drops of nickel acetate produced a red precipitate and a cherry-red solution. After addition of the nickel acetate solution, the mixture was a very intense dark red. The solution was then heated to re- flux. During six days of refluxing the solution became red- brown, and a precipitate was noted on the sides of the flask. After six days the flask was removed, and the hot solution was vacuum filtered through a medium porosity glass fritted funnel. A light green solid was removed. The filtrate was allowed to stand in a beaker at room tempera- ture. Brown crystals formed and were removed by vacuum fil- tration. Recrystallization from ethanol gave dark red crys- tals. Anal. Calculated for 015H26N402Ni: C, 50.99; H, 7.56; N, 15.86. Found: C, 51.25; H, 7.40; N, 16.02. yield 0.4 g, 11.4% (based on nickel) Comments on Synthesis The nickel acetate solution was prepared by slowly heating on a steam bath. Rapid heating produces insoluble nickel hydroxide. Reaction of NiTBK with 1,2-Diaminoethane 1. Thirty minutes near the boiling point. NiTBK (0.1 g) was weighed into a 25 ml flask. The flask was placed in a dry box under a nitrogen atmosphere. 1,2-Diami- noethane (5 ml) was added, and the mixture was heated, with 16 stirring, to Just below the boiling point on a hot plate- After thirty minutes the solution was allowed to cool, and the solids were removed by filtration through a medium po- rosity glass fritted funnel. Recrystallization from acetone produced orange crystals which were identified by their mass spectrum as NiTBK. 2. One hour at the boiling point (117‘C). The proce- dure described above was followed except that the solution was allowed to boil gently for one hour. The NiTBK dis- solved under these conditions, and upon cooling only a small amount crystallized out of solution. The mixture was transferred to a round bottom flask, and the 1,2-diaminoethane was removed by rotary evaporation. The acetone-soluble residue was removed and recrystallized from acetone. Orange crystals, identical to those recovered in procedure 1, were recovered. 5., Five hours at the boiling point. The procedure described above was followed except that the solution was boiled gently for five hours. The NiTBK dissolved and did not recrystallize upon cooling. The 1,2-diaminoethane was removed by rotary evaporation. The acetone soluble mate- rial was removed leaving a dark red residue which was found to be soluble in ethanol. Attempts to recrystallize this material from ethanol and n-butanol failed, leaving a dark red oil. 17 4. Twenty_hours at 80°C. NiTBK (0.1 g) and 1,2-diami- noethane (5 ml) were placed in a Schlenk tube. The 1,2-dia- minoethane was handled in a dry box under a nitrogen atmos- phere. The Schlenk tube was connected to a drying tube packed with Aquasorb and placed in an oil bath maintained at 80'C. After twenty hours the NiTBK had dissolved, and the solution was a very dark red color. Recrystallization did not occur upon cooling. The 1,2-diaminoethane was re- moved by rotary evaporation, leaving a dark red material. This material was insoluble in acetone but soluble in al- cohol. Attempts to recrystallize this material from ethanol, n-butanol, isopropyl alcohol and t-amyl alcohol produced only a red oil or an amorphous solid. IV. PHYSICAL MEASUREMENTS Infrared spectra were obtained using Nujol and Fluoro- lube mull techniques with a Perkin-Elmer Model 457 spectrom- eter. The ‘H nmr spectra were obtained with a Varian HA-1OO in CD015 using TMS as an internal reference. Mass spectra were determined with a Hitachi-Perkin-Elmer EMU-60. Elemen- tal analyses were performed by Chemalytics, Tempe, Arizona. Molecular weight determinations were performed by Galbraith Laboratories, Inc., Knoxville, Tennessee. 18 V. RESULTS AND DISCUSSION The starting point of this project was the preparation of 2,3-butanedione monohydrazone and 3,5-dimethyl-1,2-bu- tanedione monohydrazone from a-diketones. It was soon dis- covered that the monohydrazones were unstable, tending to polymerize, and that handling would be difficult. It was also found that the reactivity of the carbonyl group of the a-diketone or monohydrazone was dependent upon the nature of the alkyl substituent. Glyoxal reacted with hydrazine so rapidly that isolation of a monohydrazone was not possible. 2,5-Butanedione reacts rapidly with hydrazine to form the monohydrazOne and polymerizes in the solid state at room temperature. Newman and Kahle12 reported that preparation of dipivaloyl monohydrazone from dipivaloyl took two days with an acid catalyst. The reactivity of the carbonyl there- fore depends upon the alkyl substituent and decreases in the order H>CH5>C(CH3)5. This series proved useful in the preparation of monohydrazones from unsymmetrical diketones. 5,5-Dimethyl-1,2-butanedione monohydrazone with hydrazine condensation at the carbonyl group with the hydrogen sub- stituent was prepared using mild reaction conditions. In order to avoid the handling difficulties associated with the monohydrazones these compounds were prepared "in 19 2O situ" by reaction of equimolar amounts of a-diketone and hydrazine monohydrate in alcohol. A short reaction period was observed, and then nickel acetate and acetone were ad- ded to the solution. The color of the solution changed from yellow to red upon addition of the nickel acetate, indica- ting that the nickel ions had been complexed in a square planar environment. The solution was filtered. The nickel ketazines were recrystallized from the filtrate at room temperature. This procedure is very similar to the proce- dure described by Kerwin and Melson4’5 for the preparation of nickel ketazines based on benzilmonohydrazone. NiDMK and NiTBK were prepared in this manner,and yields in the neighborhood of 10-20 percent were obtained. It was believed that the yields could be improved by optimizing the reaction conditions for the preparation of the monohydrazones and the reaction period of the nickel ketazines. Characterization of NiDMK and NiTBK The elemental analyses for NiDMK and NiTBK, given in Section III, are in close agreement with the theoretical values calculated from the proposed structures. NiDMK was found to be monomeric in carbon tetrachloride by vapor phase osmometry. In the infrared spectra of NiDMK and NiTBK (Appendix, Table V, p. 45), the absence of N32 stretching vibrations indicates that a chemical reaction has taken place with 21 elimination of the hydrogens from the primary amine groups of the monohydrazone. The absence of a carbonyl absorption is consistent with the formation of a C-O-Ni bond. The strongest absorptions in the spectrum of NiDMK were at 1522 and 1150 cm'q. The spectrum of NiTBK contained three strong absorptions at 1450, 1540, and 1180 cm‘q. Two of these ab- sorptions were common to the spectra of NiDMK and NiTBK. The common bands at 1522,1540, 1150, and 1180 cm'1 are in the C-0 and C-N stretching region of the infrared. A study of the fragmentation patterns of their mass spectra provided evidence of the structures of the nickel ketazines. The mass spectra of NiDMK and NiTBK are listed in Tables 1 and 2. In each case the highest mass observed corresponded to the parent ion for the pr0posed structure of the nickel ketazines. Nickel containing fragments were i— dentified by the relative abundance of the nickel isotopes. The tentative assignment of the fragments supported the pro- posed structure, however, the fragmentation pattern did not provide evidence that would establish which mode of coordi- nation is preferred by the nickel ion. In the symmetrical mode (Structure X) the nickel ion coordinates to two five- membered aza rings and a six-membered tetraaza ring. In the unsymmetrical mode (Structure XI) the nickel ion coordinates to a five-membered aza ring, a six-membered diaza ring, and a five-membered triaza ring. The nickel ion is associated with chelate rings in several of the fragments. These frag- ments could be attributed to either a five-membered aza ring 22 Table I. Mass Spectrum of NiDMK M7e' Assignment 296 Enema" P-15 ES - 035* P-28 5 - 001*, E: - 113* __ 2 ._ + _ _+ 198 ,N1\ ,0“; ,Ni\ ,an o ,N—N-C\ o _c\ /t.—_c\c CR3 (1“ R on \ / SHE 35 , on; (.3 ”a L .4 156 ,Ni\ + I ,Ni\— + ‘1: 4"“ 9 i / = C‘ N on; we . on; or “5 L. _J 140 on on fi” _ Ni -+_ Ni\ §b=cl 3 9135 7"“ o \N=N—q =0! ca , CH ‘03 C 5 . 3 5 .CH ‘ _ _ _ 5 (:115 128 '— ,Ni _+ f3; 68: 99 ? *For nickel-containing species the value forwb i is given. 23 Table 2. Mass Spectrum of NiTBK M/e“ Assignment 552 Earsnfl+ 13-15 E - 0331+ + + P-28 E-Na'], E-cd‘] P-85 E: - OCCWHBBT + + P-115 E) - OCC(CHB)BCHCH‘;J ’ E, - mc(m3)2m083:| 226 V Ni 03'” i\ on o’ \N-N=C< 3 T o’N N—c’ \c=c' on; c N \CH ’ ‘H (on )’o I?” ““95 .. 3 3 H 170 r ,Ni\1 + N 3H1 bo14). ore? omov #N.v mmer M «I c _ ml 1.1.3; W23 I I 28 mm.m no.m mm.r mm.r om.r mm.m obfinoanoonpoa nopnwu mm.m mo.m sm.r mm.r m¢.v w.m odfimasmwn nonnmo mm.m ae.m ee.m rm.e om.v mm.a anouosoem we.m oo.m mm.e em.e mm.e om.me assesses mass 6 v 26.3mm onononchnosoz on» Assn a V 93on 959230 pnobaom sons masons Hesse: Harps: assasoo oaseoeaoan sang «a asses-omens m. on.” seesaw Rosanne .m oases 29 resonance structures of the five- and six-membered rings. The six-membered diaza ring contains a pair of conjugated double bonds. A resonance structure involving the double bonds can be written: ,Nj... ..Ni\ 0 N- 0 N- I II II I 0350\‘9/N CHgC‘ ('g’xN CH5 CH3 The resonance structure results in a delocalization of charge within the chelate ring and would be expected to in- crease the stability of the six-membered chelate ring. ‘The five-membered aza ring does not contain a pair of conjugated double bonds. The resonance structure involves a separation of charge between two adjacent nitrogen atoms and involves atoms located outside of the chelate ring: /Nl% 3N1\9 GD 0 NZN- 0 N—N- E__é N: n 03/ — CH CH/ CH 3 3 3 Further work is necessary to establish the electron distri- bution within the chelate rings; however, if the sixrmembered ring is favored by resonance stabilization over the five- membered ring, then the unsymmetrical structure would be preferred. To summarize, NiDMK and NiTBK have been characterized by their infrared, ‘H nmr, and mass spectra as complexes 50 containing dinegativsly charged ligands in which two mono- hydrazone residues are linked by a single carbon atom from the acetone. The tetradentate ligand is coordinated about the nickel(II) ion in a square planar environment. In ad- dition, the 'H nmr spectra of the nickel ketazines are con- sistent with a structure in which the nickel(II) ion is co- ordinated in an unsymmetrical mode. The unsymmetrical structure was not detected in the nickel ketazines prepared by Kerwin and Melson‘“5 because of the complexity of the 'H nmr absorptions of the phenyl groups. Reaction of Nickel Ketazines with Amines The reactivity of the coordinated carbonyl groups of NiDMK and NiTBK toward amines was investigated using reac- tion conditions similar to those described by Kerwin“ for the preparation of NiN4 and NiN30 complexes from NiMMK. The reactions of the nickel ketazines with neat amines were found to be temperature dependent. Many of the reactions failed at high temperatures producing a red oil. The struc- tures of the oily products could not be deduced from their mass spectra, and attempts to recrystallize the oil from various solvents were unsuccessful. Low temperatures re- sulted in recovery of the nickel ketazine from the reaction mixture. Failure to obtain a product from the reaction con- ditions investigated was not taken as evidence that the pro- duct could not be prepared but only that the reaction did not take place under the specified conditions. 51 Reaction of NiDMK with 1,2-Diaminoethane A boiling mixture of NiDMK and 1,2-diaminoethane pro- duced orange-red crystals. The crystals were isolated and characterized as a macrocyclic NiN4 complex formed by con- densation of both coordinated carbonyl groups of the nickel ketazine with the amine groups of 1,2-diaminoethane. The elemental analysis of NiDMcyclo 15, given in Sec- tion III, is in close agreement with the theoretical values calculated from the proposed structure. The infrared spectrum of NiDMcyclo 15 (Appendix, Table 5, p. 45) differs significantly from the spectrum of NiDMK. The spectrum of NiDMK contained absorptions in the C-0 and C-N stretching region at 1522 and 1150 cm'q. The absorption at 1522 cm’1 in the spectrum of NiDMK is absent from the spectrum of NiDMcyclo 15. However, the absorption at 1150 cm'1 is present, and a strong absorption appears at 1170 cm”. A reasonable explanation of this data would be to assign the absorption at 1522 cm’1 to s C-0 stretching vibration, which is lost upon reaction of NiDMK with 1,2-diaminoethane; the band at 1150 cm'1 to the C-N stretching vibration of the ni- ckel ketazine; and the band at 1170 our1 to the C-N stretché ing vibration of the ethylenediamine bridge. Numerous ab- sorptions in the spectrum of NiDMcyclo 15 could not be as- signed to specific functional group vibrations and were be- lieved to be group vibrations of the chelate rings. The mass spectrum of NiDMcyclo 15 (Table 4) provided evidence to support the proposed reaction. The highest 52 Table 4. Mass Spectrumof NiDMcyclo 15 M/e' Assignment 520 fisarentfl + 19-15 I} - CH + P-28 E: - 112]”, P-56 E) - Ncaecnzlfl +, [p - NC(CI13)é| + P-69 E) - CBBCNCHZCHQ + P-84 E - NNC(CHB)2] + N 19.97 E) - CHacNNC(CH5)£]+ P-111 Ea - 033CNNC(CH3)2N:I+ P-125 E) - CHacNNC(Cfia)2'mfl + 168 FIR -+ r n\ W“ Ni ,Ni\ N’ \ =N CH, \CH CE I a 3 , CH3 ‘- " L. _ I‘ + I + 154 /Ni\ '1 NINE; N N—N \__I | N :35} 035 d , 3 CH3 ._ ..J 140 I“ Ni ‘1 + / \ N N 5:6 C/ \ -33 CH}— 126 I’Ni-N-—/0=C\ + __ 033 CH3 ‘For nickel-containing species the value for 58Ni is given. 33 observed mass corresponded to the singly charged parent ion. The fragmentation pattern contained residues that could be assigned to either a five-membered aza ring with a pendant nitrogen atom or a sixrmembered diaza ring. Thus, the fragmentation pattern was consistent with either a sym- metrical or an unsymmetrical structure. The 1H nmr spectrum of NiDMcyclo 15 (Figure 3) Pro- vided evidence that the nickel ion is coordinated in an un- symmetrical mode. The spectrum consisted of five singlets and two sets of triplets with relative areas of 6:5:5:5:5:2: 2. Absorption assignments were based on the relative areas. The geminal methyl groups appear as a singlet at 8 1.57 Ppm. Four singlets at 8 1.95, 1.99, 2.10, and 2.58 ppm are as- signed to the monohydrazone methyl groups. The two methy- lene groups appear as two sets of "triplets" centered at 85.65 and 4.10 ppm. A computer simulation (Figure 4) of ‘ this pattern was achieved by assuming an AA'BB' system with JAB = -15 Hz, JAA' a 4.7 Hz, and JAB' s 7.1 Hz. This pat- tern is consistent with an unsymmetrical structure since the 032 groups are bound to nitrogen atoms in chelate rings of different size. Furthermore, the hydrogens of each 032 group must be in different chemical environments (axial and equitorial). The axial hydrogens will be deshielded by the electrons in the dz. orbital of the nickel(II) ion while the equitorial hydrogens almost in the plane of the ligand will not be influenced by these electrons.45’1‘ Several confor- mations of the dimethylene bridge were considered, but based 54 me oaohozmfiz mo ssnpooam as: m. eEeQeQ Q m.m a.m m.m m.a h j .m osdwfim 35 JVUL JWL J A I l i I I L l 4.10 5.90 5.70 5.50 5 p.p.m. Figure 4. 100 MHz ‘H nmr Spectrum of Dimethylene Bridge of NiDMcyclo 15. (Experimental: above. Computer simulation: 56 on the computer simulation the proposed conformation was ac- cepted as the best explanation of the ‘H nmr spectrum. Resonance considerations very similar to those proposed for the nickel ketazines can be proposed for the chelate rings of the macrocycle. The six-membered chelate ring with conjugated double bonds provides a resonance structure re- sulting in delocalization of charge: ,Ni., ,.N:I\ -N N- "F N- I II I I 3 a 5 as He 5 The resonance structure of the five-membered chelate ring results in a charge separation and involves atoms outside the chelate ring: ‘/Niu .Ni\€> G) N\ lNZN- N; IN— N CHP::QCH /C— QCH 3 5 w} 3 If resonance stabilization favors the sixrmembered ring the unsymmetrical structure would be preferred. Molecular framework models indicate that unfavorable strain is not produced in either the symmetrical or unsym- metrical structures. A To briefly summarize, the infrared, ‘H nmr and mass spectra are consistent with the formatiOn of a dinegatively charged 15-membered macrocyclic ligand coordinated in a square planar environment about the nickel ion. The ‘H nmr 37 spectrum establishes an unsymmetrical mode of coordination. Molecular framework models indicate that unfavorable strain is not produced in the unsymmetrical structure, however, it is probable that a slight tetrahedral twisting of the ligand system takes place to reduce some apparent strain inherent in the purely planar configuration;15 Reaction of NiDMK with 1,3-Diaminopropane Kerwin4 obtained an NiN5O complex from the reaction of NiMMK with neat 1,3-diaminopropane at ambient temperature for three days under a nitrogen atmosphere. These conditions failed to give a reaction between NiDMK and 1,3-diaminopro- pane. Several attempts, using different reaction condi- tions, failed to produce a reaction. The conditions and results are summarized below: 1. 3 days L N. R. Ambient temperature? 2 30 minutes g) N R warm I 5. 1 hour of; red oil 135‘0 4. six days .; red oil 50°C Reaction of NiTBK with 1,2—Diaminopropane NiTBK failed to react with 1,2—diaminoethane under con- ditions which produced macrocyclic products with NiMMK4 and NiDMK. Several attempts were made to obtain a product from heating NiTBK in neat 1,2-diaminoethane under a nitrogen 58 atmosphere. The conditions and results are listed below: 1 30 minutes \ N. R. 117'0 ’ 2 # 1 hour g!) N. R. 117'0 ’ 3. 5 hours __. red oil 117:0 ’ 4. 20 hours ; red oil 80‘0 Reactivity of Coordinated Carbonyl Groups This preliminary investigation of the reactions of the coordinated carbonyl groups with amines suggests that the order of reactivity of the nickel ketazines is consistent with the order expected for a mechanism involving nucleo- phic attack by the amine group at the carbon atom of the coordinated carbonyl group. The electron density at the carbon atom is influenced by the positively charged nickel ion and the alkyl or aryl group bonded to the carbon atom. Phenyl groups are electron-withdrawing; methyl groups are weakly electron-releasing; and t—butyl groups are elec- tron-releasing. The electropositive nature of the carbon atom depends upon the nature of the substituents and de- creases in the order phenyl>methyl>t-buty1. If the me- chanism involves nucleophilic attack at the carbon atom, the order of reactivity would be NiMMK>NiDHK>NiTBK. Kerwin“ found that NiMMK reacted with 1,2-diaminoethane and 1,5-di- aminopropane at room temperature to form NiN4 and NiNao 39 complexes. NiDHK reacted with 1,2-diaminoethane at 117’0 to form a NiN4 macrocycle but did not react with 1,5-diamino- propane at room temperature. NiTBK did not react with 1,2- diaminoethane at 117’0. Various mechanisms involving nu- cleophilic attack at the carbon atom may be proposed to ac- count for the observed order of reactivity. The discovery that nickel ketazines are coordinated in an unsymmetrical mode is significant to the study of the mechanism of amine condensation reactions with the coordi- nated carbonyl. It is now apparent that the coordinated carbonyl groups are not equivalent. If the resonance are gument is valid then the carbonyl group in the five-membered ring is essentially a negatively charged oxygen atom bonded to a carbon atom while the carbonyl group in the siXbmem- bored ring does not carry a full charge and has more double bond character. The reactivity of these groups towards amines would be expected to be different. Kerwin4 reported that the reaction of NiMMK with ethylamine or 1,3-diamino- propane at 100‘0 resulted in amine condensation at only one of the coordinated carbonyl groups. This suggests a large difference in the reactivity of the carbonyl groups. Reac- tion at both sites is achieved only with 1,2-diamines with the formation of a macrocycle. The second condensation must either take place by a different mechanism or involve a stronger nucleophile, and the chain length of the diamine must be the determining factor in the second condensation reaction. #0 An interesting similarity is noted between the six-mem— bered ring of the nickel ketazines and the six-membered ring of the bis(acetylacetone)ethylenediamine metal complex. Both ring systems contain conjugated double bonds, and reso- nance structures resulting in electron delocalization can be written for both ring structures. The failure of the coor- dinated carbonyl group of the bis(acetylacetone)ethylenedi- amine metal complex to react with aliphatic amines suggests that the coordinated carbonyl group of the sixbmembered ring of the nickel ketazines may be less susceptable to nucleo- philic attack by primary amines. §gggestions for Future Work A crystal structure determination of the complex pre- pared in the course of this project would confirm the un- symmetrical structure of the complexes providing rearrange- ment does not occur upon crystallization. The conformation of the dimethylene bridge of NiDMcyclo 13 could also be con- firmed by this method. A study of the bond lengths of the five- and six-membered rings would establish the degree of electron delocalization associated with each ring. In addi- tion, a crystal structure study of one of the NiN30 complexes prepared by Kerwin4 would establish the site of the first condensation reaction. VI. CONCLUSIONS In summary, two new nickel ketazines have been prepared and characterized. The nickel ketazines were found to be coordinated in an unsymmetrical mode. A preliminary inves- tigation of the reactivity of the nickel ketazines towards amines resulted in a macrocyclic Niflk complex. The order of the reactivity of the nickel ketazines is consistent with a mechanism involving nucleophilic attack by the amine at the carbon atom of the coordinated carbonyl group. 41 LIST OF REFERENCES 10. 11. 12. 13. ’14. 15. H. D. N. C. C. V. N. LIST OF REFERENCES D. Busch, Helv. Chim. Acta, Special Issue, 174 (1967). A. House.and N. F. Curtis, J. Amer. Chem. Soc., §§, 225 (1964). F. Curtis, Coord. Chem. Rev., 5, 3 (1968). M. Kerwin, "Nickel (II) Complexes Containing Non- cyclic and Macrocyclic Ligands Derived from Benzil Monohydrazone" (unpublished Doctor's dessertation, Michigan State University, 1972). M. Kerwin and G. A. Nelson, Inor Chem., 11, 726 (1972). L. Goedken and S. M. Peng, J. Chem. Soc. Chem. Comm., 62 (1975). W. Alcock and P. A. Tosker, J. Chem. Soc. Chem. Comm., 1259 (1972). G. Jager, Z. Chem., 8, 30, 392, 470 (1968). J. Truex and R. H. Holm, J. Amer. Chem. Soc., 25, 4529 (1972)- St. C. Black and M. J. Lane, Aust. J. Chem., g1, 2039 (1970). w. J. Taylor, N. H. Callow and C. R. W. Frances, J. Chem. Soc., 257 (19393- S. Newman and G. R. Kahle J. Org. Chem. 2 666 (1958). . 9 .2. D. Buckingham and P. J. Stevens, J. Chem. Soc., 4583 (1964). G. Warner, N. J. Rose, and D. H. Busch, J. Amer. Chem. Soc., 29, 6938 (1968). F. Richardson and R. E. Sievers, J. Amer. Chem. Soc., 95:. 4131* (1972). 42 APPENDIX 43 Table 5. Infrared Absorptions of NiDMK, NiTBK, and NiDMcyclo 15 (Technique - Nujol Mulls) NiDMK 2920(8). 2875(H). 1540(W). 1450(M). 1425(M). 1402(M). 1362(M). 1322(3). 127501). 1210(M). 1150(M), 1035(W), 980(M), 795(W). 702(W). 590(V) NiTBK 2910(3). 2850(3). 1500(Sh9M). 1450(5). 1408(3). 1375(W). 1540(3), 1262(w), 1218(M), 1184(3), 1150—1120(3 bands M), 1045(W). 1120C”), 962(8). 930C“). 895(5). 815(W). 792(W)o 748(W). 710(W). 670(W). 625(W). 685(W). 670(W). 650(W) NiDMcyclo 13 2970(Sh), 2955(Sh), 2910(3), 2360(3), 1520(3), 1460(3), 1430(W), 1397(3). 1370(8), 1345(W). 1328(W). 1315(3). 1268(M), 1218(M), 1170(3), 1150(Sh,M), 1050(w), 1030(w), 990(W). 975(“). 928C“). 380(W). 320C“), 700C"), 525(W) ”'IIIIIIHII 1111M IIUEIEIIIMITLIIIII“