2022
COMMUNICATIONS TO THE
thloride and isopropylation of dl-benzene, i t was obvious that o-deuteriocumene had a sharp and intense absorption at 15.8 p and m-deuteriocumene absorbed strongly at 15.0 p . The isopropylation of dl-benzene gave a product containing 86.0% dl-cumene (by mass spectra). This material was used on a n infrared standard with the assumption that twofifths of the deuterium was in the o-position. Products of the reaction of dimethyl sulfate with the cumyl anion were analyzed in the ethereal concentrate by gas-liquid Chromatography employing a Perkin-Elmer model 154B Vapor Fractometer operating at 150” and using helium as a carrier gas. tButylbenzene and p-cymene were well separated by a n “A”-column (purportedly di-n-decyl phthalate). P C y m e n e would have been detected at a molar concentration ‘ / b ~t h a t of the t-butylbenzene. Solubility Measurements.-Treatment of 0.1 mole of cumyl methyl ether with sodium-potassium alloy for 7 hours in 700 ml. of ethyl ether at room temperature followed by filtration gave 550 ml. of a n ether solution of the anion. Treatment of this solution with deuterium chloride gave a precipitate of alkali metal halide which was removed by filtration. The precipitate weighed 2.90 g. and was found t o contain less than 0.03% sodium by flame photometry. From the ethereal solution 4.75 g. of cumene was recovered. T h e 550 ml. of filtrate thus gave 0.038 mole of potassium chloride and 0.039 mole of cumene and the solubility of the anion in ether is indicated to he 0.07 M . Treatment of 0.1 mole of cuniyl methyl ether with the alloy in the presence of 100 ml. of the dimethyl ether of ethylene glycol for 7 hours a t room temperature did not give a homcgeneous solution. The solution was allowed t o settle and a sample of the homogeneous supernatant liquid was pipetted into methyl alcohol. The hydrolysate was analyzed for cumene, cumyl methyl ether and the dimethyl ether of ethylene glycol by gas-liquid chromatography a t 100”
EDITOR
VOI.
61
in the “X”-column. The procedure was calibrated by prepared samples and the results indicated that more than 95% of the cumyl methyl ether had been destroyed. The l ratio of cumene to diinethyl ether of ethylene g l y ~ after hydrolysis indicated a solubility of the anion in this solvent of 0.42 M . Reagents.-The cumyl methyl ether aiid p-brorr~ocunieiie used have been described p r e ~ i o u s l y . ’ ~Deuteriuni oxide (>99.5%/0) was obtained from the Stuart Oxygen C o . Deuterium chloride was prepared by the reaction of deuterium oxide with benzoyl chlorideI4 and should have possessed a purity of >9570. &Acetic acid was prepared by the hydrolysis of acetyl chloride with deuterium oxide. The d-acetic acid was isolated by distillation. Its isotopic purity was never measured. Phillips 99 mole Yc minimum n-pentane was used without purification as was hlallinckrodt analytical grade ether (containing a trace of alcohol as a preservative). The dimethyl ether of ethylene glycol (rlnsul Chem. Corp.) was refluxed over sodium until nu further rriction occurred. Constant boiling material was stored over sodium and under nitrogen before use. All other solvents were dried by calcium hydride.
Acknowledgment.-Mass spectra were obtained by RIr. G. P. Schacher. Infrared absorptions were obtained by RIr. D. Harmes and 1Iiss D. IIcClung. The i’apor Fractometer was operated by I I r . E. 11. Hadsell. The cumyl methyl cthcr was prepared by hlr. W. Reuter. (13) G. A. Russell, THIS J O U R N A L . 79, 38il (1%7); (1956). (14) H . C. Brown and C . Groat. i b i d . 64, 2 2 2 3 ( 1 9 1 2 ) .
78, 1047
SCHEXECTADY, N. Y .
COMMUNICATIONS T O T H E EDITOR INFRARED SPECTRA OF METAL CARBONYL HYDRIDES
Sir: Several years ago this laboratory located the first vibration involving a hydrogen atom in metal carbonyl hydrides - $03 cm. in HCo(C0)d. To take into account t h e range of possible consequences of t h e interaction expected between the hydrogen and t h e CO groups, a bonding model was proposed’ containing a degree of freedom. I n one limit the hydrogen was solely bonded t o the metal and in the other limit to CO groups-the demonstration of the exact state requiring further experimentation.lt2 Cotton has criticized this view and suggests t h a t the 703 cm.-l band arises from a Co-H stretching vibration.3,4 This laboratory has now studied in detail the infrared spectrum of DCo(C0)4, H2Fe(C0)4,HDFe(C0)4, HMn(C0)5, DMn(CO)S, and a number of related compounds 2 t o 33 p obtained with appropriate prisms and small gratings. The 703 and 330 cm.-’ bands of HCo(C0)d shift to 600 and 296 cm.-l in DCo(C0)d for frequency ratios of 1.17 and 1.12. These data establish t h a t C and/or 0 atoms participate along with the H or D atom in the motions (1) W. Edgell, C. hlagee and G . Gallup, THISJ O U R N A L , 78, 4185 (1950). (2) W. Edgell, A n n . Rev. P h y s . C h e m . , 8 , 353 (1957). (3) F. Cotton and G . Wilkinson, C h e m . and I n d . ( L o n d o n ) , 1305 (1956). (4) F. Cotton, TifIs J O U R N A I , 80. 4425 (195.8).
of this group of vibrations. Six vibrations of this character are observed between 300 and 800 cm.-’ in H*Fe(CO)d, the typical hydrogen contribution being less than for HCO(CO)~. Moreover, it appears that C-0 bending motions primarily couple with the hydrogen motion in these modes. The major components of the intense 5 p band of HJIn(C0)b are two overlapped bands with PQR structure apparently centering a t 2025,s and 2020 cm.-’. h much weaker band has its Q branch a t 1094.4 cm.-]. Among the weaker satellites is a band a t ITS2 crn.-l. The most intense bands a t longer wave lengths occur a t ‘735, 661, 615 and 45s cm.-’. T h e stronger bands of D l l n ( C O ) j are similar t o those of HMn(CO):, except t h a t the hydride bands a t 735, 664 and 613 ern.-' seem t o collapse into a deuteride band (doublet) a t 676 cm.-‘; the weaker band a t l7S2 cm.-l is replaced by one a t 1257 cm.-’. The 1752 and 1287 cni.-l bands are the hIn-H and Mn-D stretching vibrations 1 They correspond t o a Mn-H force constant of ca. 1.87md./*4. The following explanation is offered for the behavior of t h e bands near 1.5 p. The 735 and 615 cm.-’ bands result from t h e coupling of a weak, “pure” Rln-H bending mode with an intense C-0 bending mode, both of which have (uncoupled) frequencies near a second C-0 bending mode of different symmetry at 664 cm.-l. The deuteration shift of the (uncoupled) Mn-H frequency (ca. moves it
a)
COMMUNICATIONS TO THE EDITOR
April 20, 1959
beyond the range of effective coupling with t h e C-0 bending mode, which then appears together with the second C-0 mode a t ca. 676 cm.-'. The uncoupled &In-H bending frequency is thus fixed a t ca. (36.5 cm.-l! Thus both Mn-H stretching and bending frequencies are established. They are consistent with the metal-hydrogen bonding limit of t h e model of Edgell, Magee and Gallup,' and with those of Hieberb and Cotton and W i l k i n ~ o n . ~While not definitely proven, the above findings imply that the bonding of the hydrogen in HCo(CO), and H2Feis similar t o t h a t in Hhln(CO)5with greater hydrogen-CO coupling.
TABLE I Unpaired electrons (ohs.) b
(Cy),Ti (a~,)* n [(Cy)rMol+2 (ad* 0 (CYhV (a1,)Ye1g)* 3 [(CY)XrI (adz(elg)* 3 iCy),Cr (a~,)Yei,)? 2 [(CY)d+I (~I~)Y~I,)~ 1 (CyhMn (a1,)Ye1,)~ 5 (CyhFe (a~,)~(e~,)~ 0 KCY)ZCOI (a,)Ye~~)~ 0 0 [(CY?-2RhI+ [iCyhIrI (al,)Ye~,)~ n (CY)&U (al,)Ye~,)~ 0 (CY )2CO (a~,)*(ed~(ed~ 1 (Cy),Ni (a~,)~(e~)Ye~,)~ 2 Cy = C5H5. * Bibliography is given in reference 7. +
+
+
+
W. F. EDGELL G. ASATO
w. WILSOS
Assignment ( D m h field)
Species"
(5) W. Hieher and F. Leutert, Z. anorg. C h e m . , 204, 745 (1932); Dze C h r m i e , 66, 2 5 (1912).
CHEMISTRY DEP.4RTMENT PLJRDLT USIVERSITY LAFAYBTTE, IXDIASA RECEIVED MARCH14, 1959
2023
a
C. ?LSGELL
cene analogs assuming V, > V , and a formal charge of -1 on each ring. Also listed are the observed paramagnetic moments expressed as number of unpaired electrons. D m h LIGAND FIELDS AND "SANDWICH" COMPLEXES The predictions agree with experiment except for (Cy),V, (Cy)&r f, and (Cy)&fn which suggest Sir: Numerous molecular orbital treatments of ferro- t h a t a weak field treatment may be required. The cene and its analogs have been p u b l i ~ h e d . ~ - ~considerations given here should apply also to the There is presented here a strong-field, ligand-field benzene metal complexes. The author is indebted model for sandwich-type complexes with the t o Professor G. W. Watt for the suggestion of a field general formula M p ( R U ) z . Here M' is a transition theory approach t o the problem and for a prelimimetal atom in its pth valence state and RY is a nary correlation of the data. planar ring carrying a charge v. The field on Mp DEPARTMENTS OF CHEMISTRY ASD PHYSICS F. A. MATSEN OF TEXAS from rings RY is due to the formal charges on the THEUNIVERSITY AUSTIN12, TEXAS ring and/or virtual charges arising from,mutual RECEIVED FEBRUARY 18, 1959 polarization of M and R. I t is assumed t h a t the actual field is adequately reproduced by one of Dah symmetry and t h a t the field uncouples the SYNTHETIC STUDIES ON SPHINGOLIPIDS. V. THE d electrons on the transition metal atom.s Since SYNTHESIS OF DIHYDROCEREBROSIDES t h e field is axially symmetric, m remains a good Sir : quantum number, the d functions of different [ m ( The structure of the cerebrosides has been estabare not mixed, and the five d orbitals have a t most ~ Va, in which three different energies. Let a = r sin @ be the lished by Carter and c o - ~ o r k e r s l -as coordinate perpendicular, and z = r cos e the co- R is a long-chain fatty acid residue. We wish to ordinate parallel to the symmetry axis. The per- report the synthesis of palmitoyl- and stearoylturbing field contains no odd powers in z because dihydrocerebrosides (Vb and Vc). As key intermediate we employed the substituted of the symmetry plane through X I p . The field is ) assumed t o be continuous a t t h e origin so t h a t it cis-oxazoline I (R1 = C H ~ ( C H Z.4) ~ ~Hydrolysis contains no first power in a. To the second power with diluted hydrochloric acid gave erythro-30-benzoyldihydrosphinogosine (II), which was not v = vo + v*u2 vzz2 isolated, but was acylated directly in the presence = O[' L + v,(az + 2") - (V. - V.)9 of sodium acetate t o give the erythro form of the The portion of the perturbation in braces is spheri- amidoester 111. (IIIb: m.p. 74-75.5'; found: cally symmetric, produces no d splitting, and is C, 76.8; H, 11.8; N, 2.2; IIIc: m.p. 73-75': neglected. With hydrogenic orbitals of effective found: C, 76.9; H, 11.35; N, 2.0.) charge 2, E(esn) = -3Dd, E(e1,) = -9Dd, and When a benzene solution of I11 was shaken with E(al,) = -11Dd, where Dd = 6 ( V a - V,) ( Z / a , J 2 tetraacetyl-a-D-galactosyl, bromide5 in the presence is t h e D m h ligand field strength. In Table I are of freshly-prepared silver carbonate,6 a 60y0 yield listed the D m h electron configurations of some ferro- of IV was obtained. (IVh: m.p. 43-45'; found: C, 67.9; H , 9.7; N, 1.8; IVc: m.p. 43-45'; (1) This research was supported by a grant from the Petroleum Research Fund administered by the American Chemical Society. found: C, 67.9; H, 9.3; N, 1.6). Saponification (2) H. H. Jaff6. J . Chcm. P h y s . , 21, 156 (1953). (1) H. E. Carter, 0. Nalbandov and P. A. Tavormina, J . B i d . Chcm., (3) W. M o 5 t t , T H I S JOURNAL,76, 3386 (1954).
+
(4) J. D. Dunitz and L. E. Orgel, J. Chcm. P h y s . , 23,954 (1355). (5) E. Ruch, Rcc. Trow. C h i m . , 76, 638 (1956). (6) M. Yamazaki, J . Chcm. P h y s . , 24, 1200 (1956). (7) A. D. Liehr and C. J. Ballhausen. Acta Chem. Scond., 11, 207
(1957). ( 8 ) See, for example, W. hloffitt and C. J. Ballhausen. Ann. Rev. P h y s . Chcm., 7,107 (1956).
194, 197 (1951).
(2) H. E. Carter and F. L. Greenwood, ibid.. 199, 283 (1952). (3) H. E. Carter and Y . Fujino, ibid.. 221, 879 (1956). (4) D. Shapiro, H. M. Flowers and S. Spector-Shefer, THISJOURNAL, in press. (5) A. Robertson, J . Chcm. SOC.,1820 (1929). (6) W. Koenigs and E. Knorr, Be?., 84, 957 (1901).
