? WWI WNW — — — — — — ___. f-fl — —____ _____—._ — ___——_ 131 800 __THS THE CEEEE‘MSTRY OF ARSEND - I, 2 - DECARBA - CLCSO - DODECABGRANE (12) DEREVATWES Thesis for the Degree of M. S. MiCHtGAN STATE UNWERSHY ROBERT BERNARD ZABOROWSKI 1989 “‘7 ~- -- -~ uni-n. -a~ unamdm w] 4 L [1' R l R Y N11523:: ,.I"!t€ Univunity 2......m bTHES-‘ug ABSTRACT THE CHEMISTRY OF ARSINO-l,2-DICARBA- CLOSO-DODECABORANE(12) DERIVATIVES by Robert Bernard Zaborowski The compounds 1,2—bis(dimethylarsino)-l,2-dicarba—closo- dodecaborane(12), dimeric methylarsino(III)-l,2-dicarba-closo- dodecaborane(12) and the complexes formed by replacing two car- bonyl groups in nickel carbonyl, iron pentacarbonyl and molybdenum hexacarbonyl with the ligand, l,2—bis(dimethylarsino)-l,2-dicarba- closo-dodecaborane(12) were prepared. These new compounds were characterized by analysis, nmr and infrared data. Mossbauer data obtained on the iron complex suggests that the ligand is a potent electron donor, while infrared arguments suggest that the ligand may also function as an electron acceptor. Additional data on other attempted preparations is also discussed. THE CHEMISTRY OF ARSINO-1,2-DICARBA- CLOSO-DODECABORANE(12) DERIVATIVES By Robert Bernard Zaborowski A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1969 2"" , j. .5 1': .’ r;. ,A iv, u?» 4) '- (fi/ To my wife, Paula, Mom and Dad. ACKNOWLEDGEMENTS The author is deeply indebted to Dr. Kim Cohn for his guidance and personal interest throughout the course of this research. The author is grateful to Dr. P. G. Rasmussen of the University of Michigan, Ann Arbor, for obtaining the Mossbauer data, and to Mr. F. Parker and Dr. R. W. Parry of the University of Michigan, Ann Arbor, for the B11 nmr data. The author wishes to express his gratitude to the Department of Chemistry, Michigan State University, especially the Inorganic Division faculty members for their patient tutelage and effort and for the financial aid and experience gained as a Graduate Teaching Assistant. The author also wishes to express his deepest appreciation to his wife for her help in the final preparation of this manuscript and her devoted encouragement throughout. iii TABLE OF CONTENTS I I INTRODUCTION 0 O C O O O O O O O O O O O O O O O O O I I 0 EXPERIMENTAL O O I O O O O O O O O O O O O O O O A. Experimental Methods . . . . . . . . . . . . . B. Primary Starting Material . . . . . . . . . . . C. Synthesis . . . . . . . . . . . . . . . . . . 1) l, 2- -Bis(dimethy1arsino)- 1,2-dicarba-closo- dodecaborane(12) . . . . . . . . . . . . . 2) Dimeric Methylarsino(III)-l,2-dicarba- closo-dodecaborane(12) . . . . . . . . . . 3) l,2-Bis(dimethylarsino)-1,2-dicarba-closo- dodecaborane(12) nickel dicarbonyl . . . . 4) l,2—Bis(dimethylarsino)-l,2-dicarba-closo- dodecaborane(12) iron tricarbonyl . . . . . 5) 1,2—Bis(dimethylarsino)-l,2-dicarba-closo- ooum dodecaborane(12) molybdenum tetracarbonyl . D. Attempted Synthesis . l) Dimeric Methylarsino(III)-l, 2-dicarba-closo- 21 23 23 dodecaborane(12) nickel dicarbonyl . . 2) Dimeric Methylarsino(III)-l,2-dicarba-closo- dodecaborane(12) iron tricarbonyl . . . . . 3) Dichlordi[l,2-Bis(dimethylarsino)-l,2- dicarba-closo-dodecaborane] iron(III) Tetra- chloroferrate . III. DISCUSSION . IV. SUGGESTIONS FOR FUTURE RESEARCH BIBLIOGRAPHY . iv 25 28 36 37 LIST OF FIGURES FIGURE 1. 2. 3. 4. 5. 6. 8. 11 B nmr spectrum of B C H [As(CH ) ] in benzene 18$ $9 in 2 at an operating frequ 32.1 (CH3)33 was.used as an external reference . . . . . . . . Infrared spectrum of B C H [As(CH ) ] run as a nujol mull where x denégei ER. absogpéign due to nUJ 01 O 0 O O O O O O O O O O O O G O 0 O C 0 I 0 Infrared spectrum of [3 OC H oAsCH ]2 run as a nujol mull where x denotes thi agsérptiogs due to nujol. Infrared spectrum of (Bicars)Ni(CO) run as a nujol mull where x denotes the absorption; due to nujol. Infrared spectrum of (Bicars)Fe(CO) run as a nujol mull where x denotes the absorption due to nujol. Infrared spectrum of (Bicars)Mo(CO) run as a nuJol mull where x denotes the absorptions due to nujol. Infrared spectrum of (Bicars) FeCl FeCl run as a nujol mull where x denotes ghe agsorpéions due to HUJOIesseeeseesseeseessesss Schematic representation of predicted spectrum for B C H unit. Schematic representation of spectrum rieuitigg from selective overlapping . . . . . . . Page 10 12 15 17 19 22 26 34 LIST OF TABLES TABLE Page 1. Infrared carbonyl absorptions for various (L)Ni(C0)2 systems 0 O O O O O O O O O C O O I O O O O O O O O O O O 32 vi NOMENCLATURE The rapid development of research in the area of cage boron compounds was in part responsible for the recent development of a complete set of rules governing the nomenclature of such systems (1, 2). This nomenclature will be employed in naming all compounds in this manuscript with the exception of references to compounds which have been previously reported by other authors. These references will retain the nomenclature used in the original publication. In all work contained in the experimental section of this manuscript, the BlOCZHlO unit will be referred to as l,2-dicarba-closo-dodecaborane(12) rather than the common name carborane. l R. Adams, Inorg. Chem., 2, 1087 (1962). Inorg. Chem., Z, 1945 (1968). vii INTRODUCTION During the era from 1912 to 1936, Alfred Stock and his co- workers prepared and chemically characterized a number of the more common boron hydrides. Within the past decade the boron hydrides have received considerable attention as a result of the demands for new high energy fuels by the Defense Department (1). While all attempts to prepare useful fuels have proved futile, industrial chemists had uncovered a new class of stable polyhedral boranes. The polyhedral boranes encompass molecules or ions in which boron atoms alone or in combination with other atoms, describe a closed poly- hedron ranging from the tetrahedron to the icosahedron (2). Investi- gation of these polyhedral boranes revealed an aromatic character manifest not only in an extraordinary thermal stability, but also in a substitution chemistry centered on the exopolyhedral boron-hydrogen or carbon-hydrogen bonds. The low toxicity (3) of these polyhedral boranes and their high stability toward acids and bases is surprising when compared with the toxicity and stability of other boron hydrides (4). The icosahedral carboranes share this stability, but are less 2.. 12H12 ° The discovery of this new class of compounds, "aromatic" poly- stable to basic degradation than B hedral boranes, has opened up broad areas of research with a variety of specific synthetic and theoretical problems. Recent reports have suggested that the derivative chemistry of this class of compounds will rival that of aromatic hydrocarbons. Several reviews on the chemistry of some of these materials, particularly Ban2- and Bn-2C2Hn’ are available (5-11). One smaller area of interest developed when the synthesis of l,2-Dicarba-closo-dodecaboraue(12) (11) provided a new, presumably aromatic, structure. This synthesis, in turn, has led to extensive synthetic investigations of the derivative chemistry which is based on substitution at the two carbon atoms (12). In general, groups which are so substituted act as if they were attached to a very bulky, electron withdrawing moiety (7, 9). If appropriate groups are substituted at these two carbon atoms, it is possible to prepare a potential bifunctional ligand. Almost nothing is known of the steric and electronic properties of the carborane cage when present in such a ligand. In fact, only five reports have appeared which are concerned with complexes which contain ligands of this type. H. D. Smith (13) has reported that the reaction of nickel(II) chloride 6-hydrate with l,2-bis(diphenylphosphino)fgfcarborane, as well as with the corresponding derivatives containing one, two and three bromine atoms attached to the carborane nucleus, produced complexes containing two molecules of the bisphosphino ligand and one of nicke1(II) chloride. In addition, he obtained ligand-exchange data which suggested that the order of stability of the series described was [((C6H5)2P)2310H10C21N1C12<[((06H5)2P)2B10H1002]2NiC12<[((C6H5)2P)2 B10H9C2]2N1012<[((C6H5)2P)BIOH8BrZCZ]2NiC12<[((C6H5)2P)ZBIOH7Br3C2]N101é. Other complexes containing the ligand and Cu(II), Pt(II) and Pd(II) were prepared but not characterized. R. H. Holm and F. Rohrscheid (14) have reported that similar bisphosphino ligands will replace CO from Ni(CO)4. On the basis of carbonyl absorptions they suggested that the bisphosphinofigf carborane group may accept electron density to a greater extent than conventional diphosphine ligands. Russian workers (15) have also reported that bisphosphino-g: carborane ligands will displace CO from Ni(CO)5 and Fe(CO)5 and will form complexes with NiI2 and PdClz. No spectral data were reported for these complexes. In a recent article Smith, Robinson and Papetti (16) reported that nicke1(II) and cobalt(II) salts react with 1,2-bis(mercapto)fgf carborane to yield square-planar MSZP2 and M842- complexes. On the basis of an empirical comparison of the electronic Spectrum with previously reported spectra, they report that the carborane nucleus does not provide an effective network for n-delocalization. To some extent, their conclusion is supported by an investigation on m7 and p-[a-carbonyl] benzoic acids by Hawthorne, Berry and Wegner (17). This study also suggested that the interaction of the gfcarborane group with the aryl group does 22£_allow electronic delocalization. In apparent contradiction to the conclusions of Smith and co— workers as well as Hawthorne and co-workers, Longuet-Higgins and Roberts (18) have proposed that back-bonding by n substituents on the isoelectronic B12H122- ion is possible. The controversy surrounding the electronic interaction of these bifunctional carborane ligands and their complexes prompted investi- gation into the synthetic possibilities of bisarsino-l,2-dicarba-closo- dodecaborane(12) derivatives and their subsequent use as complex- ing agents. Such ligands should be structurally analogous to o-phenylenediphosphines and -diarsines and should provide a source for detailed ligand-metal bonding studies. It was also hoped that such a ligand system, incorporating the carborane framework, would prove as versatile a complexing agent as o-phenylenebis- dimethylarsine (19). The only previously reported arsino compound incorporating the carborane framework is (o-PhCB C)3As (20). 10H10 EXPERIMENTAL Experimental Methods A Perkin-Elmer 237B grating spectrophotometer was employed for obtaining all infrared spectra. Solid spectra were run either as nujol or as fluorolube mulls. Solution spectra were run in either chloroform or methylene chloride. Proton nmr spectra were observed on a Varian Model A-6O nuclear magnetic resonance spectrOmeter operating at the ambient tempera- ture of the instrument. Tetramethylsilane was employed as an internal standard. Boron nmr absorptions were obtained on a Varian Model HA 100 spectrometer with an operating frequency of 32 MHz. Trimethylboron was used as an external reference, by the tube inter- change technique. Index of refraction measurements were obtained on a Bausch & Lomb Abbe Refractometer. A constant temperature bath was utilized for the water-cooling system. The Mfissbauer data were obtained in cooperation with the Chemistry Department of the University of Michigan. The instru- ment used employs an electromechanical tranducer for the drive operated in conjunction with a multichannel analyzer as previously described (21). The standard employed in the work was NazFe(CN)5 NO-ZH O. 2 All preparations were carried out under an atmosphere of nitrogen. Iron analyses were performed volumetrically employing potassium permanganate as titrant. The samples were dissolved in concentrated nitric acid and reduced to the ferrous state with stannous chloride solution (22). All other analyses were per- formed by Galbraith Laboratories, Knoxville, Tennessee. All melting points were recorded on a Thomas-Hoover Capillary Melting Point Apparatus. Primary Starting Material Primary starting material 1,2-dicarba-closo-dodecaborane(12), more commonly gfcarborane, was prepared directly from purified acetylene (Matheson) and decaborane (U.S. Department of the Air Force) (11,23, 24). The synthetic method of Hawthorne et a1. (23), was modified in the following manner. First, the number of sulfuric acid containing traps, as well as, the number of empty- safety traps in the purification train was increased. Second, the fluffy, cream colored B C H was dried in gag 2 instead of over 10 2 12 P205. Third, the final purification of the pure white solid, B10C2H12, was accomplished by filtration and dried in_vacuo. The identity of the product was confirmed by the recorded melting point of 297 i 20 (literature value 2980 (11)) and comparison of the infrared spectrum with the previously reported spectrum (25). The 1H nmr spectrum of the product dissolved in chloroform dis- played a broad CH resonance at -3.52 ppm from TMS (literature value 3.54 ppm (23)). The pure B was stored in an evacuated 10C2H12 dessicator over P205 prior to use. Synthesis J.emaimeamhmmammmmmwe The ligand B [As(CH 2, which we will call Bicars (26), 10C2H10 3)2] was prepared in a manner analogous to that employed in the preparation of B (25), but employing (CH AsBr (27, 28), instead 10C2H10[PR2]2 3)2 of (CH3)2PC1. Dimethylbromoarsine was prepared according to the method of Maier g£_al. (27). The method was modified slightly by elimination of the second trap which was to be cooled to -80° (27). The mixture of (CH3)2AsBr and CH3AsBr2 was effectively separated by distillation at 42 i 1 mm pressure to yield the pure (CH AsBr. The 1H nmr of the 3)2 (CH3)2AsBr fraction run neat exhibited a single resonance at -l.51 ppm which suggested complete separation. The compound was identified by its refractive index, 11])21 1.5684 (literature value nD20 1.5713 (28)) and boiling point of 50 i 20 at 42 mm (literature value 510 at 42 mm (28))- A slurry of dilithiocarborane (12), BlOCZHIOLiZ’ was freshly prepared by charging a nitrogen purged 100 ml. three-necked flask with a solution of n-butyllithium (0.055 mole) in 50 m1. dry ethyl ether. The flask was fitted with a magnetic stirrer, nitrogen inlet and an addition funnel with nitrogen outlet. This flask was then cooled to 0°. A solution of o-carborane (0.026 mole) in 20 ml. of dry ethyl ether was added and the mixture was stirred for ten minutes while it was maintained at 0°. After addition, the stirring was continued for another 45 minutes while the mixture was maintained at 0°. A solution of (CH AsBr (0.055 mole) in 20 ml. of dry ethyl 3)2 ether was then added to the cooled BlDCZHIOLiZ slurry over a period of 25 minutes. Reaction appeared to be immediate as indicated by diappearance of the white BlOHlOCZLiZ slurry. The slight yellow 'mixture was allowed to warm to 200 and stirred for an additional 30 minutes. It was then refluxed for 1 hour. The solvents were removed by evaporation which was accomplished by the continued pass- age of nitrogen throught the system for 24 hours at 23°. Separationcnfthe pure white crystalline product was effected by immediate precipitation upon the addition of 50 ml. of water to a saturated solution of the reaction products in ethyl ether. This procedure hydrolyzed the LiBr formed in the reaction and effected a quantitative separation of the desired product. The product was then collected by filtration. The product was purified by dissolu- tion in ethyl ether and recrystallization upon addition of water. It was then dried in zagug_to remove all traces of water. The pro- duct may also be purified by recrystallization from a minimum amount of hot n—hexane. The formula, B10C2H10[As(CH3)2]2, data. The 1H nmr spectrum of the ligand dissolved in carbon tetra- is supported by all spectral chloride and carbon disulfide exhibits a single resonance absorption at -l.22 ppm from TMS due to the methyl groups attached to the arsenic atoms. A chemical shift of -Ou86 ppm from TMS is recorded when the ligand, Bicars, is dissolved in benzene. The B11 nmr spectra of the ligand (29), as shown in Figure 1, consists of a set of four distinct absorptions of relative intensity 1:2:4:3 at 84.7, 90.1, 94.4 and 99.3 ppm from (CH3)3B respectively. 10 84.7 90.1 94.4 99.3 ‘7 H 0 Figure l. B11 nmr spectrum of B [As(CH in benzene at 10C2H10 3)2]2 an operating frequency of 32.1 MHz. (CH3)BB was used as an external reference. 11 The infrared spectrum is shown in Figure 2. The spectrum exhibits characteristic absorptions at 724(3) and 2570(3) cmfl. The absorption at 724(3) cm.1 is attributed to the B-H cage structure and the one at 2570(3) cm"1 ascribed to the B-H stretching mode. Other absorptions appear at 862, 902(m), 982(w), 1080(m) and 1262(w) cm-l. Analysis: Calculated for C6H22AszBlo: C, 20.46; H, 6.30; As, 42.02; B, 30.72. Found: C, 20.66; H, 6.46; As, 41.79; B, 31.09. Molecular weight determined cryoscopically in benzene is 344 g/mole (theoretical 353 g/mole). The melting point is 111 i 20 which com- pares favorably with the melting point of 1110 obtained in separate work by H. D. Smith (30). The compound was found to be readily soluble in acetone, benzene and ethyl ether. It is extremely insoluble in water. The compound appears to be quite stable in air. 12 omo .Hofisc ou map sowuauomom ecu mouoaov x mumsa Hana Hohsa m we can NHNAmmUVm The preparation was attempted analogous to the preparation of (Bicars)Ni(C0)2 as described on page 16 of this manuscript. A 1.65 mmole sample of the ligand, Bicars, about 15 ml. of dry n-hexane and an excess of Ni(CO)4 (6.45 mmole) were employed. The reaction was allowed to proceed for some 12 hours until the Ni(CO)4 began to decompose in the reaction system as evidenced by formation of a nicke1(0) deposit. After removing all volatile materials by distillation in ygggg, a white product remained. The product was purified by recrystalliza- tion from benzene. A11 spectral evidence suggests no reaction occurred between the [B10C2H10ASCH3]2 and Ni(C0)4. The infrared spectrum of the white product failed to exhibit any absorptions in the carbonyl stretching region. The 1H nmr of the material in benzene was identical with the observed spectrum of the starting material [B ASCH3]2. 10C2H10 24 Bissais.¥a£hxlstsiseéééizzixéasiesakaz:isssaseéssaketansiiéi izsstttisatkesxi The preparation was attempted in a manner analogous to the preparation of (Bicars)Fe(C0)3 as described on page 18 of this manuscript. An excess of freshly distilled Fe(CO)5 (4.6 mmole) and a 1.2 mmole sample of [B10C2H10ASCH3]2 were employed in the reaction mixture. The mixture was heated to 150 i 100 for 5 hours in the combustion furnace. After removal of the unreacted Fe(CO)5 by distillation in gaggg_a dark residue remained. All spectral data suggest no reaction occurred between the The 1H nmr spectrum of the dark Fe(CO)5 and [B A3CH 1002H10 312' residue dissolved in benzene was identical with the observed spectrum of the starting material [B1002H10A3(CH3)]2. 25 Whtmwzéiemamssemseashetassilzhficmma EEHQEMBIREERBEE The product was prepared using Bicars and FeFl3 in a manner similar to that employed in the preparation of (Diars)2FeC12FeCl4 (35). A 3.05 mmole sample of freshly prepared anhydrous FeCl (36) 3 was transferred under nitrogen atmosphere to a 50 m1. flask. About 40 m1. of dry ethyl ether was added to the flask and the solution was filtered to remove insoluble impurities. The filtered solution was then transferred to an addition funnel. A 3.02 mmole sample of the ligand, was introduced into a 250 ml. single-neck flask equipped with a side-arm for nitrogen inlet. About 40 m1. of dry ethyl ether was added to the flask and the solution stirred with a magnetic stirrer. The addition funnel containing the FeCl solution was attached to the flask and the 3 entire system purged with nitrogen. The FeCl3 solution was added over a period of 10 minutes. The dark mixture was stirred for an additional 15 minutes at 200. The reaction was then allowed to reflux for an additional 30 minutes. As the solvent evaporated during refluxing a deep red crust formed on the side of the reaction vessel. The remaining solvent was then removed by distillation in_g§ggg_to yield the deep blood- red product. The product was purified by thoroughly washing it with benzene to remove any unreacted starting materials. Final purifica- tion consisted of recrystallization from a small amount of warm benzene. The infrared spectrum of the product as shown in Figure 7 suggests formation of a complex containing the ligand Bicars. Present in the 26 mmuoooo x oumnz Hana aches m no any ooqa 2 «Home N oooa P mooH .HOnso ou moo maofiuouomnm onu HommNAmumonv mo annuoomm omumumoH .n ouswwm oomH ooom ooom oooq b P - b AH aov hososvoum 27 spectrum are the characteristic B-H stretching and cage absorp- tions at 1265(br) and 726(3) cm.1 respectively. Analysis: Calculated for C12H44A34B20Fe2: C, 14.01; H, 4.31; Fe, 10.86. Found: C, 12.88, H, 3.95; Fe, 10.94. The deep blood-red product appears to be destroyed by allow- ing it to react with water. A white residue is obtained after this treatment. The infrared spectrum of the white residue is identical to the observed spectrum of the ligand, Bicars. DISCUSSION The possibility of preparing bisarsinocarborane derivatives was established with the following examples. 1. Li\ ___C,Li + 2(CH3)2AsBr (CH3)2 :SH3)2 (3(7/ 3 \\A8 A8 + 2 LiBr BlOHlO V\\c--c// (I) \o BlOH10 Li Li B H 2. \C—C/ + CH3AsBr2 ————) / 10 10 \(V / 0 + LiBr 10 10 C C CH: A! ‘31 CH 3' S 3‘ 3 \c- c/ 0 (II) BlOHlO Both species (I and 11) display stability toward air oxida- tion and hydrolysis in contrast to the behavior of Diarsine (o- phenylenebisdimethylarsine). The possibility of employing the new bisarsino-1,2-dicarba- closo-dodecaborane(12) derivatives as ligands was then investi- gated. A series of reactions was conducted which established that 28 29 B10C2H10[AS(CH carbonyl group in metal carbonyls and, in so doing, to form new 3)2]2, or Bicars, was capable of replacing the 4, 5 and 6 coordinate complexes according to the following equation: 0 \\ // + M(C0)n _______9 + 2 CO where: M=Ni n=4 M=Fe n=5 M=Mo n=6 All attempts to prepare the analogous Ni (34) and Fe complexes with dimeric methylarsino(III)-l,2—dicarba-closo-dodecaborane(12) (species II) failed. This may be a result of the steric hinderance due to the large B cage units. Such a hypothesis is supported 10C2H10 by the preliminary x-ray data on the ligand Bicars (37) which shows extremely short As-C bond distances. As short and possibly shorter As-C bond distances in the system: 30 //////C C\\\\\ BB 0 0BH 10 10 ////’ 10 10 \\\\\c c \AIS/ CH3 would bring the two large B cage units even closer together 10C2H10 making it nearly impossible to get a metal atom close enough for bonding through the arsenic atoms. In addition, the position of the two methyl groups would pose further steric problems. The difficulties encountered in the attempted preparation of the complex (Bicars)2FeCleeCl4 may be a reflection of the ease with which the compound hydrolyzes. Both iron analysis and infrared data suggest the presence of the desired product. The carbon-hydrogen data may be explained by possible hydrolysis of the product during shipping and handling. The Mfissbauer spectrum of the iron complex, (Bicars)Fe(CO)3, has provided some information about the electronic properties of this new ligand. The isomer shift of +0.18 mm/sec. is typical of low valent iron species. The strongly electron donating properties of the ligand is reflected in the negative shift compared to Fe(CO)5 (literature value of -.282 mm/sec. (38) corrected for difference in standard and temperature to +.201 mm/sec.). If one agrees with the idea of dn-pn back—bonding, it may be concluded from this 31 evidence that forward coordination dominates over back donation. This idea is supported by Collins and Pettit (39) in a M633bauer study of the (©3P)2Fe(CO)3 system. The isomer shift for the complex of +0.16 mm/sec. (39) (cited value corrected for difference in standard employed) compares favorably with that of the (Bicars) Fe(CO)3 system, suggesting a similarity in bonding properties of the two ligands. Table 1 shows the infrared absorptions attributed to carbonyl stretching modes for various (L)Ni(CO)2 systems where L is a variety of bisphosphino and arsino ligands. The complex (Bicars)Ni(CO)2 exhibits two strong infrared absorptions which may be easily attribut- ed to the carbonyl stretching modes at 2027 and 1968 cm—1. The lower frequencies compared with Ni(CO)4 (literature value 2057 cm“-1 (40)) suggest that the ligand Bicars, is a weaker n-acceptor than the carbonyl group in line with other phosphino and arsino derivatives. Further, these frequencies are about 20 to 30 cm”1 higher than those for the comparable complex containing Diarsine as a ligand, Ni(CO)2[ng6H4(A3Me This suggests that the ligand, Bicars 2),]. incorporating the carborane framework, may back accept electron density to a greater extent than the more conventional arsine ligand. A similar observation regarding the phosphinocarborane ligands shown in Table 1, whose carbonyl absorptions compare favorably with those of the ligand Bicars, has also been mention- ed by Rohrscheid and Holm (14). It is interesting to note that the Mfissbauer and IR data confirm the fact that o-carborane possesses the ability to serve 32 Table 1. Infrared carbonyl absorptions for various (L)Ni(CO) 2 systems. All spectra run as solution spectra in dichloromethane. Ligand Absorptions in cm- 06H4[A3(CH3)2]2 2001 (41) 1934 310C2H10[A3(CH3)2]2 2027 1968 B10C2H10[P(CH3)2]2 2013 (14) 1955 B1002H10[P(C6H5)2]2 2021 (14) 1966 33 as both electron donor and acceptor, already chemically established (42). ll . . . The B nmr spectra of the ligand, Bicars, as displayed in Figure 1 may be explained in the following manner. Assuming a nearly icosahedral structure for the B unit as in o- 10C2H10 carborane (42); one readily recognizes the existance of four different types of boron atoms, namely, 3(6), 8(10), 9(12) and 4(5, 7, 11). This would give rise to four singlets. The absorp- tions, however are further split by spin-spin coupling due to the protons attached to the borons to give a predicted spectrum of four doublets of relative integral intensities, 2:2:2:4. Such a spectrum has been reported for the B11 of o-carborane at 60 MHz (43, 44). It has been found that at a lower frequency the spectrum obtained may be explained in terms of selective overlapping of the expected doublets (43). The predicted spectrum consisting of four doublets is shown schematicly in Figure 8a. With a reduction in resolution, doublets a & b coalesce so that the high-field member of a and the low-field member of b are nearly superimposed. At the lower frequency, the doublet b is nearly centered on the low-field member of c and the doublets c and d are so closely overlapped as to be indistinguish- able. This behavior is represented in Figure 8b. The resultant Spectrum consists of four singlets of relative intensity 1:2:4:3 as shown. This corresponds exactly to the observed Bll nmr spectrum 11 of Bicars, Figure l. The higher degree of resolution in the B of Bicars over that of o-carborane (45) suggests an increased 34 Figure 8a. Schematic representation of predicted spectrum for BIOCZHIO unit. Figure 8b. Schematic representation of spectrum resulting from selective overlapping. 35 differentiation in the types of borons in the molecule. This change in environment about the carbon atoms may be due to the presence of the As(CH3)2 groups. SUGGESTIONS FOR FUTURE RESEARCH Further research should be directed toward the preparation of a compound suitable for epr studies. Such studies should provide informa- tion regarding spin density transmission in the Bicars-metal system. For example, the complex (Bicars)Fe(CO)ZCl may be prepared by reaction of (Bicars)Fe(C0)3 in HCl solution followed by air oxidation and the complex (Bicars)2FeC12-B4 by replacement of the FeCl4 of (Bicars)2FeC12-FeC14 with B¢4-. Such systems would be useful for epr studies because they should contain Fe d1 systems and in addition they possess a high degree of symmetry which simplifies interpretation of the epr measurements results. M6ssbauer studies of these systems would provide additional information concerning the nature of the Fe-ligand bonding. The chemical shift is a direct indication of the Fe s-electron density. Interpretation of this information in terms of the Fe d-electrons would provide an indication of the donor-acceptor properties of the ligand, Bicars. Single crystal x—ray diffraction studies of the compound, dimeric methylarsino(III)-l,2-dicarba-closo-dodecaborane(12) should be initiated. Structural information for this compound should help clarify the reasons underlying its inability to displace CO from either Fe(CO)5 or Ni(C0)4. This inability may be a consequence of steric hinderance arising from extremely short C-As bond distances. Such short bond distances could result from resonance stabilization requiring the BlOCZHlO unit to accept a partial negative charge. Such behavior is consistant with the established donor-acceptor properties of the B nucleus 10C2H10 (17, 26). 36 2. 5. 6. 7. 10. 11. 12. 13. 14. BIBLIOGRAPHY D. R. Martin, J. Chem. Ed., 36, 208, (1959). E. L. Muetterties, and F. Klanberg, Inorg. Chem., 5, 1955, (1966). W. H. Sweet, A. N. Soloway, and R. L. Wright, J. Pharmacol. Expil. Therap., 137, 263, (1962). F. A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry", 2nd. ed. rev., Interscience, New York, N.Y., 1966, pp. 272-275. M. F. Hawthorne, Endeavor, 25, 146 (1966). W. N. Lipscomb, Science, 153, 373 (1966). M. F. Hawthorne, "The Chemistry of Boron and Its Compounds", E. L. Muetterties, Ed. John Wiley and Sons, New York, 1967, pp. 286-323. R. L. Hughes, I. C. Smith, and E. W. Lawless, "Production of the Boranes and Related Research", R. T. Holzmann, Ed., Academic Press, New York, 1967, pp. 163-184. T. Onak, "Advances in Organometallic Chemistry", F. G. A. Stone and R. West, Ed., Academic Press, New York, 1965, pp. 263-363. E. L. Muetterties, and W. H. Knoth, Chem. Eng. News, 44, 88, (1966). 3” T. Heying, J. W. Ayer, Jr., 8. L. Clark, D. J. Mangold, H. L. Goldstein, M. Hillman, R. J. Pollak, and J. W. Syzmanski, Inorg. Chem., 2, 1089 (1963). T. Heying, J. W. Ager, Jr., S. L. Clark, R. P. Alexander, S. Papetti, J. A. Reid, and S. I. Trotz, Inorg. Chem., 2, 1097 (1963). “ H. D. Smith, J. Amer. Chem. Soc., 81, 1817 (1965). R. H. Holm and F. Rohrscheid, J. Organometal. Chem., 4, 338 (1965). 37 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 38 L. T. Zakharkin and G. G. Zhigareva, Izv. Akad. Nauk SSSR, Ser. Khim., 932 (1965). H. D. Smith, M. A. Robinson and S. Papetti, Inogg. Chem., 5, 1014 (1967). M. F. Hawthorne, T. F. Berry and P. A. Wegner, J. Amer. Chem. Soc., 87, 4746 (1965). H. F. Longuet-Higgins and M. D. Roberts, Proc. Roy. Soc., (London), A230, 110 (1965). For a summary of many of these complexes see, "Chelating Agents and Metal Chelates,: F. P. Dwyer and D. P. Mellor, Ed.,Academic Press Inc., New York, N.Y., 1964, Chapter 3. L. J. Zukharkin, V. I. Bregardze and 0. Yu. Okhlobystin, J. Organometal. Chem., 4, 211 (1965). J. Ulrich, Ph. D. Dissertation, University of Michigan, Ann Arbor, Michigan, 1967. A. I. Vogel, "A Textbook of Quantitative Inorganic Analysis including Elementary Instrumental Analysis", 3rd ed., John Wiley & Sons, New York, N.Y., 1963 , pp. 287-289. M. F. Hawthorne, T. D. Andrews, P. M. Garrett, F. P. Olsen, M. Reintjes, F. N. Tebbe, L. F. Warren, P. A. wegner and D. C. Young, Inorg. Syn. lg, 91 (1967). S. Papetti and T. L. Heying, J. Amer. Chem. Soc., 86, 2295 (1964). R. P. Alexander and H. A. Schroeder, Inorg. Chem., 2, 1107 (1963). R. B. Zaborowski and K. Cohn, Insig. Chem., in press, (1969). L. Maier, E. G. Rochow and W. C. Fernelius, J. Inorg. and Nucl. Chem., 16, 213 (1961). L. Maier, Inorg. Syn. Z, 82 (1964). B11 spectrum was obtained by Mr. F. Parker and Dr. R. W. Parry of the University of Michigan, Ann Arbor, Michigan. H. D. Smith, private communication, U. S. Borax Research Co., Anaheim, California. H. Nigam, R. Nyholm and D. V. Rao Ramana, J. Chem. Soc., 1397, (1959). 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 39 G. Brauer, Ed., "Handbook of Inorganic Preparative Chemistry”, Academic Press, New York, N.Y., 1963, pg. 1743. Spectrum obtained by Dr. P. G. Rasmussen of the Chemistry Department of the University of Michigan, Ann Arbor, Michigan. C. M. Kerwin, private communication, Chemistry Department, Michigan State University, East Lansing, Michigan. R. S. Nyholm, J. Chem. Soc., 851 (1950). A. R. Pray, InorgLHSyn., 5, 153 (1957). J. J. Stezowski and H. A. Eick, private communication, Chemistry Department, Michigan State University, East Lansing, Michigan. L. M. Epstein, J. Chem. Phys. 36, 2731 (1962). R. L. Collins and R. Pettit, J. Amer. Chem. Soc., 85, 2332, (1963). W. E. Edgell, J. Huff, J. Thomas, H. Lehman, C. Angell and G. Aato, J. Amer. Chem. Soc., 82, 1254 (1960). C. C. Barraclough, J. Lewis and R. S. Nyholm, J. Chem. Soc., 2852, (1961). J. A. Potonza and W. N. Lipscomb, J. Amer. Chem. Soc., 86, 1874, (1964). R. L. Piling, F. N. Tebbe, M. F. Hawthorne and E. A. Pier, Proc. Chem. Soc., 402, (1964). G. D. Vickers, H. Agahigian, E. A. Pier and H. Schroeder, Inorg, Chem. 5, 693 (1966). H. Schroeder, T. L. Heying and J. B. Reines, Inorg. Chem., 2, 1092 (1965).