6598 Am. Chem. Soc., 96, 2109 (1974); (c) 6. V. DePamphilis, B. A. Averill, T. Herskovitz, L. Que, Jr.. and R. H. Holm, hid., 96, 4159 (1974); (d) M. Cerdonio, R.-H. Wang, J. Rawlings. and H. 6. Gray, hid., 96, 6534 (1974). (9) R. S. Gall, C. T.-W. Chu, and L. F. Dahl, J. Am. Chem. Soc.,96, 4019 (1974). (10) J. C. Slater and K. H. Johnson, Phys. Rev. 8,5, 844 (1972); K. H. Johnson, Adv. Ouantum Chem., 7 , 143 (1973). (11) K . H. Johnson, Annu. Rev. Phys. Chem., in press. (12) The cluster levels are labeled in terms of the irreducible representations of the Td Symmetry group and in terms of the principal atomic components of the orbital wavefunctions. The highest occupied orbital (lot*)is indicated by an arrow. (13) R . H. Holm, 6. A. Averill, T. Herskovitz, R. B. Frankel, H. 6. Gray. 0. Siiman, and F . J. Grunthaner. d Am. Chem. Soc.,96, 2644 (1974). (14) J. C. Slater and K. H. Johnson, Phys. Today, 27 (Oct), 34 (1974). (15) In contrast to oxidized ferredoxin, the highest occupied orbital in oxidized rubredoxin is predicted to be mainly sulfur-like in character: J. G. Norman, Jr.. and S. C. Jackels, J. Am. Chem. SOC.,97, 3833 (1975). (16) Calculations have not yet been carried out for [FeIS*4(SCH3)4]2- in the 4 6 symmetry because of computer funding limitations. (17) W. D. Phillips and M. Poe in "Iron-Sulfur Proteins", Vol. 11, W. Lovenberg. Ed., Academic Press, New York, N.Y., 1973, Chapter 7. (18) Address correspondence to this author at the Department of Chemistry, Stanford University, Stanford, California 94305.
SiMe3 > Ph, and that the low energy rotamers, i.e., those in which the bulky q5-C5H5 and R groups are mutually trans, are 550-700 cal mol-' more stable than the other two rotamers when L = PPh3 and R = Ph, 900-1500 cal mol-' when L = PPh3 and R = SiMe3. Armed with these data, we decided to study the diastereomeric relationships between a chiral metal entity, qS-C5H5Fe*COL-, and a chiral organic entity, -C*HPhSiMe3, both with ligands of quite different but clearly understood steric requirements. A compound of the type q5-C5H5FeCOLCHPhSiMe3should exist as two diastereomerically related pairs of enantiomers, each diastereomer existing as three staggered rotamers. 0
nCH&
C
RR(a)
RR(b)
L
C. Y. Yang, K. H. Johnson Department of Materials Science and Engineering Massachusetts Inst it u te of Technology Cambridge, Massachusetts 021 39
R. H. Holm* I n Department of Chemistry Massachusetts Institute of Technology Cambridge, Massachusetts 021 39
J. G . Norman, Jr. Department of Chemistry Unioersity of Washington Seattle, Washington 981 95 Received July 12, 1975
A Demonstration of Controlled Asymmetric Induction in Organoiron Chemistry. Suggestions Concerning the Specification of Chirality in Pseudotetrahedral Metal Complexes Containing Polyhapto Ligands Sir:
Catalysis by chiral transition metal complexes of the addition of, for instance, hydrogen and silanes to prochiral olefins and ketones often results in induction of asymmetry into the reaction products.'V2 Although the reasons for such stereoselectivity are not, in general, understood, analogy with a number of well-studied, stoichiometric, stereoselective organic reaction^^-^ suggests that steric control may be very important in determining the relative free energies of formation of diastereomeric products, intermediates, and/ or activated complexes. We have recently shown,6 using ' H N M R spectroscopy, that steric effects strongly influence the relative energies of the staggered conformations of the compounds sS-C5H5FeCOLCH2R ( L = tertiary phosphine donors; R = Ph, SiMe3). It was demonstrated that the order of decreasing ligand steric requirements is q5-C5H5 > L > CO and
RR(c)
The enantiomeric R S and SS isomers would also exist, of course. (Suggestions concerning a convention for specifying the chirality of these stereoisomers will be discussed below,) By analogy with similar primary alkyl complexes,6 it is reasonable to expect that the most stable rotamer of each diastereomer would be that in which bulky q5-C5H5 and SiMe3 groups are mutually trans, i.e., S R ( a ) and RR(a) and their enantiomers. Furthermore, it is possible to predict that S R ( a ) should be somewhat more stable than RR(a), as the small CO is gauche to both the Ph and the SiMe3 groups in the former, while L is gauche to Ph and SiMe3 in the latter. Thus the RS-SR stereoisomers may be expected to be somewhat more stable than the RR-SS stereoisomers. Identification of RS-SR and R R - S S diastereomers should be possible using 'H N M R spectroscopy, as the ahydrogen of the coordinated alkyl group is gauche to L in S R ( a ) but trans to L in RR(a). Thus, if the barriers to rotation about the iron-carbon bonds are sufficiently high that individual rotamers can be distinguished in the N M R spectrum, then 3 J p ~for S R ( a ) should be less than 3 J p ~for RR(a).6 If on the other hand, barriers to rotations are sufficiently low that time-averaged spectra are obtained, then the observed 3 J p of ~ S R should decrease with decreasing temperature, while that of R R should increase.6 Treatment of racemic q5-C5H5FeCOLI (L = PPh3, P(OPh)3) with racemic Me3SiPhCHMgBr a t 35' in 2:l benzene-ethyl ether solution gave stable products of the formula q5-C5H5FeCOLCHPhSiMe3' (L = PPh3 ( l ) , P(OPh)3 (2)). Surprisingly, the ' H N M R spectra (Table I )
Table 1. Chemical Shift and Coupling Constant Data Compound
a-CH
1
2
'Ha
3 2
i3cb
3 a In
~
rl*-C,H 5
7.7 HZ 10.1 HZ 10.5 HZ 18 H z C 18 HzC
6 4 . 3 7 , J p ~= 1.5 HZ 6 4.07,Jp~ = 1.0 HZ 6 4.30,Jp~ = 1.0 Hz 6 82.84,Jpc = 1.5 HZ 6 82.70,Jpc= 1.4 HZ
6 1.40, 3JpH = 6 2.24, 3JpH = 6 2.13, ?IpH =
6 2.9,Jpc 6 1.6,Jpc
%
Sine 6
0.30
6 0.33 6
0.35
6 2.21 6 2.20
Ph 6
-7.15
fi -7.14 6 -7.14 6 121-157 6
121-157
C,D, solution. b In CDCl, solution, in ppm downfield from internal TMS. C Doublet partially obscured by stronger MeSi resonance.
Jourual of the American Chemical Society
/
97:22
/
October 29. I975
6599 of each in C6D6 contained only one set of resonances, q7-C7H7, q6-C6H6, $-C5H5, and q3-C3H5 would be considsuggesting that, in each case, either only a single diaered pseudo-atoms of atomic weights 84, 72, 60, and 36, restereomer was formed or that the diastereomers of each spectively. The convention thus preserves the advantages of the topological approach of Cahn et a1.I8 and, by retaining have completely coincident spectra. Either interpretation their sequence-rule procedure, at the same time allows discontrasts with the behavior of the compounds q5-C5H5FeCOPPh3R (R = CH(OEt)Me,8 ~ ~ ~ ~ O - C H D C H D C tinction M ~ ~ , ~between, for instance, q 6 - C & 5 and ?7'-C6H7, threo-CHDCHDPh,lO CH2CHMePh1'), for which the diaq5-C6H7, $-CsH4Me, and 775-CsH5.'9 stereomers are readily distinguishable by N M R spectroscoChirality of the complexes reported here has been aspy and are formed in approximately equimolar quantities. signed in accord with the above procedure, the iron designaIn an effort to settle the question of chemical shift coincition preceding that of carbon as suggested by Reich-Rohrdence, an attempt was made to insert SO2 into the iron-carwig and Wojcicki.' I We invite comment and suggestions bon bonds of the complexes; it seemed very unlikely that the concerning the pseudo-atom convention. IH N M R spectra of the diastereomeric sulfinates would Acknowledgments. Financial assistance from Queen's also be coincident.I2 Surprisingly, treatment of 2 with SO2 University and the National Research Council of Canada under a variety of conditions did not result in the formation made this research possible. of a stable sulfinate but rather in epimerization of 2 to give mixtures of 2 and the R R - S S diastereomer 3. As expected References and Notes the IH and the I3C N M R spectra of 2 and 3 are different (1) W. Knowies, Ann. N.Y. Acad. Sci., 214, 119 (1973) (Table I). Identification of 2 and 3 as the RS-SR and (2)L. Marko and B. Heil, Catal. Rev., 8,269 (1973). R R - S S diastereomers, respectively, was tentatively made (3)J. D. Morrison and H. S. Mosher, "Asymmetric Organic Reactions", Prentice-Hall. Englewood Cliffs, N.J., 1971, by studying the variation of 3 J with ~ ~temperature (see (4)E. Ruch and I. Ugi, Top. Stereochem.. 4, 99 (1969). above); as expected, 3 J p of ~ 2 (and 1) decreased, that of 3 (5) L. Salem, J. Am. Chem. SOC.,95,94 (1973). increased smoothly with decreasing temperature. FormaJ. (6)K. Stanley and M. C. Baird, horg. Nucl. Chem. Left., 10, 1 1 1 1 (1974): Am. Chem. Soc., 97,4292(1975). tion of 3 appeared to be kinetically controlled, as only 2 was (7)Attempts to prepare the compounds by photolysis of solutions of obtained in significant quantities under conditions in which q5-C5HsFe(C0)&HPhSiMe3 and L resulted in considerable decomposition. equilibrium between 2 and 3 appeared to have been reached (8)A. Davison and D. L. Reger. J. Am. Chem. SOC.,94,9237 (1972). in the presence of SO2.I4 Thus, dramatic steric effects on (9)P. L. Bock, D. J. Boschetto, J. R. Rasmussen. J. P. Demers, and G. M. the thermodynamically controlled stereoselective formation Whitesides, J. Am. Chem. Soc., 96,2814 (1974). (10)D. Slack and M.C. Baird, unpublished results. of these iron alkyl complexes has been demonstrated. Al(11) P. Reich-Rohrwig and A. Wojcicki, Inorg. Chem., 13, 2457 (1974). though it is not known whether the other, similar8-" diaste(12)We thank Professor A. Wojcicki for this suggestion, based on the premise that the SO2 insertion reaction should be stereospecific. Thus reacreomeric iron complexes have been prepared under conditions of SO2 with q5-C5H5Fe(C0)2R(R = threoCHDCHDCMe3,' threotions in which the diastereomeric pairs had equilibrated CHDCHDPh'O) proceed with inversion of configuration at the a-carbon with each other, it is probably significant that none of them atom, while reaction of SO?with q5-C5H5FeCOPPh3CH2COmentholate'3 proceeds with retention of configuration at iron. is as crowded as are 1, 2, and 3, and thus none of them (13)T. C. Flood and D. L. Miles, J. Am. Chern. Soc., 95,6460 (1973). would experience the same degree of steric control. (14)K. Stanley, D. Groves, and M. C. Baird, J. Am. Chem. Soc., following paper in this issue. Because of a dearth of information concerning the abso(15)Pure AppliedChem., 28, l(1971). lute configurations of pseudo-four-coordinate complexes (16)A. Davison and N. Martinez, J. Organomet. Chem., 74, C17 (1974). containing polyhapto ligands, such as those discussed here, (17)H. Alt, M. Herberhold. C. G. Kreiter, and H. Strack, J. Organornet. Chem., 77,353 (1974). there has been little need for a systematic convention for the (18)R . S. Cahn, C. Ingold. and V . Prelog. Angew. Chem., Int. Ed. €rig/., 5, specification of chirality. The accepted IUPAC Rules for 385 11966) ..(19)Attempts to apply the sequence rules rigorously would fail because unNomenclature of Inorganic ChemistryI5 are not particularambiguous and widely-accepted valence bond structures for these deloly applicable, and recourse has been made'6,17to the widely calized structures do not exist. Cyclohexadienyi (q5-C5H7) and methylcyaccepted R and S conventions for tetrahedral compounds, clopentadienyi (q5-C5H4Me)both contain "C5" groups, the former with two C-C bonds to CH2. the latter one C-C bond to methyl. Thus the foras developed by Cahn, Ingold, and Prelog.I8 Unfortunately mer, containing in essence a "C5" = CHz group, would have priority the latter conventions, while clearly specifying sequence over the latter, a "Cg"-CH3 group.'8 Ambiguities may arise in the case of polyhapto ligands containing heteroatoms, as a q6-Cs ligand would priorities for monohapto and polydentate ligands, do not have a higher pseudo-atomic weight than a q5-C40 ligand, although the apply to polyhapto ligands, with the result that different exsequence-rule procedures would probably assign the latter higher priortensions of the R - S conventions have already appeared in ity. The former approach would appear to be simpler. the organometallic literature.'6-17Thus while both Davison K. Stanley, M. C . Baird* and MartinezI6 and Alt et aI.I7 arbitrarily choose to assign Department of Chemistry, Queen's University higher priorities to polyhapto than to monohapto ligands, Kingston, Ontario, Canada K 7 L 3N6 i.e., q5-C5H5 > q2-C2H4 > v'-CH3, the former appear to Receiued July 3, I975 prefer, in the case of monohapto ligands containing a donor atom of higher atomic weight than carbon, to assign such a ligand higher priority than ?-C5H5. Alt et al., on the other hand, would appear to assign such a ligand lower priority. Besides the lack of consistency in the two approaches, both Novel Sulfur Dioxide-Induced Epimerization of are also inconsistent with accepted priority rules,I8 which Complexes of the Type q5-C5H5Fe*COLC*HRR' would assign lower priority to a hydrocarbon ligand than to Sir: a carbon-bonded ligand containing a @-oxygen or -nitrogen atom. We have recently shown' that (RS-SR)-v5-C5H5FeCOWhile any priority convention would be purely arbitrary P(OPh)3CHPhSiMe3 ( l ) , obtained by treating racemic and would undoubtedly evolve as inconsistencies appear, the q5-CsHsFeCOP(OPh)31 with racemic Me3SiPhCHMgBr, present time is perhaps appropriate to suggest a more conreacts with SO2 to form mixtures of 1 and its thermodysistent approach. In keeping with the s ~ g g e s t i o n ' ~that . ~ ~ namically less stable diastereomer, (RR-SS)polyhapto ligands be assigned high priorities, we suggest v5-C5HsFeCOP(OPh)3CHPhSiMe3(2). The result was that such ligands be considered pseudo-atoms of atomic surprising because compounds such as 1 and 2 normally weight equal to the sum of the atomic weights of all the react very readily with SO2 to form stable S-sulfinate comatoms bonded to the metal atom. Thus the ligands p l e ~ e s . Furthermore, ~.~ SO2 insertion reactions into metal_ _ . I
Communications to the Editor
