2012
COMMUNICATIONS TO THE EDITOR
Vol. 84
is expected on the basis of the intensity of adsorbed light being directly proportional to the concentration for the short light paths used. To relate the CO yield to the oxalate decomposition, large scale experiments using 20 ml. of actinometer solution, 9.5 X M uranyl oxalate and 9.5 X M oxalic acid, were carried out. Oxalate decomposition was determined by titration with KMn04, and CO formation was determined by a combination of volume measurement and chromatography. Oxalate decomposed, yo 63 61 29
by a molecule of hydrogen halide, or by a completely concerted process in which hydrogen and halogen form bonds simultaneously from opposite sides of the olefin plane. The observation of predominant trans-addition does not necessarily rule out a two-step mechanism involving a classical carbonium ion since steric factors in the intermediate ion may favor transaddition. The reasoning involved6 requires that the second step of the addition proceed faster than rotation about the C-C bond so that transattack would lead to a product in a staggered configuration whereas cis-attack would give a product Oxalate dec./CO formed 3.7 3.3 3.4 in the unfavored eclipsed configuration. Such an Using a quantum yield of 0.60 for oxalate de- "eclipsing" effect is most easily visualized for composed6and the average mole ratio given above, aliphatic olefins, but the effects would still be a quantum yield of 0.17 + 0.02 is established for present in the cyclic systems studied by Hammond. CO formation in the actinometer solutions used. It would be desirable to study addition to an Comparisons of the sensitivity of our method with olefin in which steric effects do not favor one mode some other actinometers are given in Table I. of addition over the other. Addition of deuterium bromide to acenaphthylene meets this criterion. TABLE I In addition, this study provides an extreme test C O M P A R I S O N OF ACTINOMETER SENSITIVITIES for the n-complex mechanism since a classical ion Uranyl oxalate (titration) 6 x 10" would be resonance-stabilized while the n-comUranyl oxalate (colorimetric)* 3 x 1016 plex would be destabilized by the strain present in Malachite green leucocyanidea 6 X 1014 the five-membered ring. The fact that chlorine Uranyl oxalate (CO) 2 x 1014 adds cis to acenaphthylene6 in contrast to the The improved speed, simplicity, and accuracy of usually observed trans-addition may reflect such a method which does not relay on the consumption destabilization of n-complexes in acenaphthylene. We find that deuterium bromide adds to acenaphof a reactant, but on the formation of an easily detected substance such as carbon monoxide, may thylene to yield more than 85y0 of the cis-addition be useful for a variety of purposes. For our product. This result rules out a n-complex in the work the great sensitivity enabled us to work with product-determining step. Formation of an ion very small volumes, and i t may also be utilized to pair intermediate, which collapses before the broreduce otherwise prolonged exposure times or to mide ion can assume an orientation trans to the initially formed C-D bond, could account for the monitor the output of weak sources. We consider that, with suitable calibration, the predominant cis-addition. Alternatively, a oneuse of uranyl oxalate in the manner described above step concerted mechanism in which C-D and C-Br provides the most sensitive chemical actinometer bonds are formed simultaneously on the same side of the olefin plane would lead to a cis-addition presently available. product. (IO) Imperial Industries Limited, Fibres Division, Hookstone Road. Additions were carried out by passing deuterium Harrogate, England. bromide through a solution of acenaphthylene in DEPARTMENT OF CHEMISTRY The product was UNIVERSITY OF CALIFORXIA K. PORTER'O methylene chloride a t -GOo. isolated as a white crystalline solid, m.p. 68-O0, DAVIS,CALIFORNIA D. H. \'OLMAN after crystallization from petrol ether. When the RECEIVED APRIL 2, 10G2 bromide was treated with potassium tcrt-butoxide in tert-butyl alcohol, acenaphthylene was obtained THE STEREOCHEMISTRY OF THE POLAR ADDITION which was shown by n.m.r. analysis7 to contain OF HYDROGEN BROMIDE TO ACENAPHTHYLENE 95 f 10% of deuterium in the 1-position. It may Sir: be safely assumed that the elimination reaction is While the stereochemistry of the free radical stereospecifically trans since tyans-elimination has addition of hydrogen bromide to olefins has been been shown to be 740 times faster than cis-climinaIt folstudied extensively, 1 3 2 the polar addition of hydro- tion for the 1,2-di~hloroaccnaphthylenes.~ gen halides has received little attention. Ham( 5 ) An analogous argument has been given by Goering to explain rnond and co-workers have found that addition of trows free radical addition of hydrogen bromide. (6) b. J. Cristol, F. R. Stermitz and P. S. Ramey, ibid., 78, 4930 hydrogen bromide to 1,2-dimethylcy~lohexene~ and of hydrogen chloride to 1,2-dimethyl~yclopentene~(1956). (7) N.m.r. spectra were measured on a Varian high-resolution gives predominantly the trans-addition product. spectiorneter V-4300B operating to 40 Mc./sec. and equipped with a Hammond suggests that these results may be ac- Varian Field Stabilizer. Chemical shift values and coupling concommodated either by assuming that a n-complex stants were determined to an accuracy of f O . O 1 and 1 0 . 2 c.P.s., The amount of deuterium present in the acenaphthylintermediate intervenes which collapses on attack respectively. ene obtained in the elimination reaction was determined as the differ(1) H. L. Goering and D. W.Larson, J. Am. Chcm. Sac., 81, 5937 (1959). (2) P. S. Skell and R. G. Allen, ibid., 81, 5383 (1959). (3) G. S. Hammond and T. D. Nevitt, ibid., 76, 4121 (1954). (4) G. S. Hammond and C. H. Collins, ibid., 82, 6323 (1960).
ence between the number of protons found in authentic acenaphthylene (2.00) and the number of protons found in the unknown sample by integrated intensity measurement. A total of twenty-one determinations were made on the products of three different elimination reactions.
May 20, 1962
COMMUNICATIONS TO THE EDITOR
2013
Wave Lenglh. Millimicrons lows that the addition must have been more than 85% cis. Since the reaction was rapid in the absence of light at --BOo, and since styrene was found to add hydrogen bromide in the presence of acenaphthylene under identical conditions to give exclusively (2-bromoethyl)-benzene, the reaction cannot involve a radical mechanism. The identity of the styrene hydrobromide was established by n.m.r. analysis. N.m.r. studies' confirm the above results. The bromide obtained from addition of deuterium bromide to acenaphthylene gives an n.m.r. spectrum with a complex multiplet extending from 2 r to 3 7, a doublet a t 4.25 r with a splitting of 7.4 c.P.s., and an unresolved broad signal a t 6.15 r. When this bromide is treated with lithium bromide in Frequency, cm:' rIO-' acetone, a significant change in the spectrum takes place. The signal a t 4.25 T becomes four lines of Fig. 1.-The visible and near infrared spectrum of nearly equal intensity. The outer lines corre- [ N(CH8)4]2[Co(ONO~)~], 0.01 molar, in acetonitrile containspond with the original doublet while the inner ing 7.6 X lo-* mole liter-I of [(CeHs)a(n-C,He)P]NOa. The lines are separated by 1.9 c.P.s.. Assignment mean energies of the absorption bands have been taken as of the 7.4 C.P.S. and 1.9 C.P.S.coupling to the cis- 18,700 and 8,100 an.-'. and trans-orientation of protons is consistent with the theoretical8 and experimentalg couplings found The infrared spectrum (mull) shows all of the for dihedral angles of about 0" and 120°, respec- bands characteristic of coordinated nitrate ions' tively. Further verification is obtained when (i.e., M-ON02) and no others which cannot be hydrogen bromide is added to the acenaphthylene assigned to the [N(CH&]+ ions, thus supporting containing >85% deuterium in the 1-position. the view that all nitrate ions are coordinated to give The n.m.r. spectrum of the bromide in this case [Co(ON02)4I2-in the crystalline compound. There shows two strong lines a t 4.25 r separated by 1.9 is also evidence that the same complex anions are C.P.S. Weak signals could also be detected cor- present in acetonitrile solutions of the compound. responding to the 7.4 C.P.S. splitting. Thus the molar conductance of a 0.003 molar soluIt was possible to estimate from integrated peak tion in acetonitrile a t 25' was 285 ohm-' mole-' intensities that no less than 85% of the cis-addition which may be compared with a value of 266 ohm-' product was obtained. This is consistent with the mole-' for a 0.003 molar solution of [N(CH3)4]2results obtained from the elimination reaction. [CO(NCS)~] in the same solvent. The infrared Similar results were obtained for addition reactions spectrum of a 0.1 molar solution in acetonitrile carried out in petroleum ether and pentane. We showed bands characteristic of coordinated nitrate are currently investigating additions to other olefins ions. There were also very weak bands attributaand with other hydrogen halides in order to evalu- ble to uncoordinated nitrate ions, indicating that ate the various factors which determine the stereo- slight dissociation of the complex anion occurs in chemistry of the process. this solvent. (8) M. Karplus. J . Chcm. P h y s . , SO, 11 (1959). The electronic absorption spectrum was recorded (9) F. A. L. Anet, Can. J. Chem., 39, 789 (1961). with successively increased additions of [ (C6H5)3(10) National Science Foundation Predoctoral Fellow. (n-CdHg)P]N03 in order to repress dissociation of GEORGE HERBERT JONES LABORATORY OF CHEMISTRY DEPARTMENT MICHAEL J. S. DEWAR the complex anion. It was found that the spectra THEUNIVERSITY OF CHICAGO ROBERTC. FA HEY'^ of 0.01 molar solutions in acetonitrile were identical CHICAGO 37, ILLINOIS, USA. for all phosphonium nitrate concentrations equal RECEIVED MARCH30, 1962 to or greater than 0.01 molar although the spectra of such solutions differed slightly, mainly in band THE PREPARATION AND CHARACTERIZATION OF A intensities, from that of a solution containing only COMPOUND CONTAINING the cobalt compound. The spectrum is shown in TETRANITRATOCOBALTATE(11) Fig. 1. Sir: The magnetic moment of [N(CH3)4]2[Co(ONO2)41 We wish to report the preparation and characterization of a compound which is novel and interesting is 4.50 h0.05 B.M. (calculated from susceptibility in two respects. The compound is tetramethyl- data with corrections for diamagnetism and temammonium tetranitratocobaltate(II), [N(CH3)4]2- perature - independent paramagnetism2) which [Co(ON02)4]. This hygroscopic, violet compound strongly implies2that the Co(I1) is in a tetrahedral was prepared by dissolving [N(CH3)4]N03 and environment. This compound is of interest since, to the best of Co(N03)~ 2H20 in 2:l mole ratio in nitromethane. our knowledge, i t constitutes the first reasonably On addition of chloroform, the product separated in a crystalline state, and was filtered, washed with well-authenticated example of a compound con(1) B. M. Gatehouse, $. E. Livingstone and R. s. Nyholm, J . Chcm. chloroform and dried in vacuum. Soc., 4222 (1957). Anal. Calcd. for C B H 2 4 C ~ N 6 0 C, ~ 2 :21.1; H, (2) F. A. Cotton, D. M. L. Goodgame and M. Goodgame. J . A m . 5.3; N, 18.4. Found: C, 20.9; H, 5.5; N, 17.9. Chem. SOC.,8 3 , 4690 (1961).
