Jordan, Norton
1
3CN M R of Methyl-Transition Metal Compounds
The existence of C M P glycosides of N,O-acylneuraminic acids has been demonstrated;2h for analogous reasons it has to be expected that these have the p configuration too. Taking into account the fact that in glycoconjugates only cy-NeuAc residues occur27 the /3-glycosidic configuration of C M P NeuAc is in accordance with the assumption that NeuAc residues are transferred to the acceptor molecule via a "single displacement mechanism" with inversion of configuration.6 So far, there is no evidence that a lipid intermediate is involved in this biosynthesis step since a twofold inversion of configuration would lead to sialic acid residues in fl linkages to glycoconjugates. It has been reported that in the biosynthesis of colominic acid (an 4 2 8) linked NeuAc polymer) by Escherichia coli28sialylundecaprenyl phosphate plays a role. In view of the foregoing this intermediate probably only acts as an acceptor molecule for additional NeuAc residues, thus initiating the polymer formation.
-
Acknowledgments. The authors are indebted to Dr. M . J. A. de Bie for helpful discussions. This investigation was supported by the Netherlands Foundation for Chemical Research (SON) with financial aid from the Netherlands Organization for the Advancement of Pure Research (ZWO) and by the Deutsche Forschungsgemeinschaft (Grant Scha 2 0 2 / 5 ) . References and Notes (1) Dedicated to Professor Dr. E. Havinga on the occasion of his 70th birthday. (2) Part of this work was presented at the lXth International Symposium on Carbohydrate Chemistry, London, April 10-14, 1978, Abstracts of Papers D7. (3) FOM-Institute for Atomic and Molecular Physics, Kruislaan 407, P.O. Box 41883, Amsterdam, The Netherlands. (4) Kean, E. L.: Roseman, S..J. Biol. Chem. 1966, 247, 5643-5650. (5) Van den Eijnden, D. H.; Van Dijk. W., Hoppe-Seyler's Z.Physiol. Chem.
4853
1972, 353, 1817-1820. (6) Sharon, N. "Complex Carbohydrates", Addison-Wesley: Reading, Mass., 1975: - - . DO r r 118-176. (7) McGuire, E. J. ''Biohgical Roles of Sialic Acid", Rosenberg. A,; Schengrund, C.-L., Eds.; Plenum Press: New York, 1976: pp 123-158. (8)Comb, D. G.; Watson, D. R.: Roseman, S.,J. Biol. Chem. 1966, 241. 5637-5642. (9) Stone, A. L.; Kolodny, E. H. Chem. Phys. Lipids 1971, 6, 274-279. (10) Delbaere, L. T. J.; James, M. N. G.; Lemieux, R. U. J. Am. Chern. Soc. 1973, 95,7866-7868. (11) Schwarcz, J. A.; Perlin, A. S. Can. J. Chem. 1972, 50, 3667-3676. (12) Wasylishen, R.; Schaefer, T. Can. J. Chem. 1973, 51, 961-973. (13) Spoormaker. T.; De Bie, M. J. A. Recl. Trav. Chim. Pays-Bas 1978, 97, 85-87. (14) Schauer, R.: Wember. M.; Ehrlich, K.: Haverkamp, J. Proceedings of the 4th International Symposium on Glycoconjugates, Woods Hole, Mass., 1977. (15) Kean. E. L.; Roseman, S. Methods Enzymol. 1966, 8, 208-215. (16) Yu. R. K.: Ledeen, R. J. 810l.Chem. 1969, 244, 1306-1313. (17) The intactness of CMP-NeuAc samples can easily be checked from the coupling constant 2Jc2_p (7.4 Hz) on the resonance of the NeuAc62 atom (6 101.1 ppm). (18) Dorman. D. E.: Roberts, J. D. Roc. Natl. Acad. Sci. U S A . 1970, 65, 19-26. (19) Karabatsos. G. J.; Orzech, Jr., C. E. J. Am. Chem. SOC.1965, 87, 560562. (20) Wasylishen, R. E.: Chum, K.; Bukata, J. Org. Magn. Reson. 1977, 9, 473-476. (21) Spoormaker, T.; De Bie, M. J. A. Red. Trav. Chim. Pays-Bas 1978, 97, 135-144. (22) (a) 360-MHz 'H NMR spectra of neutral D20 solutions of 1-4 and CMPNeuAc, unpublished results. (b) Brown, E. B.: Brey, W. S.;Weltner, W. Biochim. Biophys. Acta 1975, 399, 124-130. (c) Haverkamp, J.; Beau, J.-M.; Schauer, R., Hoppe-Seyler's Z.Physiol. Chem. 1979, 360. 159166. (23) Espersen, W. G.: Martin, R. 8 . J. Phys. Chem. 1976, 80, 741-745. (24) Hansen, P. E.; Feeney, J.; Roberts, G. C. K. J. Magn. Reson. 1975, 17, 249-261, (25) Czarniecki. M. F.; Thornton, E. R. J. Am. Chem. SOC. 1977, 99, 82738279. (26) (a) Buscher, H.-P.; Casals-Stenzel. J.; Schauer,'R.; Mestres-Ventura, P. Ew. J. Emchern. 1977, 77, 297-310. (b) Corfield, A. P: Ferreira &J Amral, c.: Wember, M.; Schauer, R. ibid. 1976, 68, 597-610. (27) Ledeen. R. W.; Yu, R. K. "Biological Roles of Sialic Acid", Rosenberg. A,; Schengrund, C.-L.. Eds.; Plenum Press: New York, 1976; pp 1-59. (28) Troy, F. A.: Vijay, I. K.; Tesche, N. J. Biol. Chem. 1975, 250, 156-163.
3C N M R Relaxation Mechanisms in Methyl-Transition Metal Compounds Richard F. Jordan and Jack R. Norton*' Contributionf r o m the Department of Chemistry, Princeton University, Princeton, New Jersey 08540. Received February 5, I979
Abstract: The 13C N M R relaxation mechanisms in transition metal-methyl compounds have been investigated. The methyl carbons in cis-Os(CO),(CH3)2, ( T - C ~ H ~ ) M O ( C O ) ~ C H ( T, ,- C ~ H ~ ) F ~ ( C O ) ~(~-C5H5)2Zr(CH3)2, CH~, and CH3AuPPh3 relax by the dipolar and spin-rotation mechanisms. The methyl carbon in CH3Re(C0)5 shows an additional contribution due to scalar relaxation of the second kind. The relaxation of the methyl and methylene carbons in the tautomeric clusters Os3(CO)lo(CH3)(H) F! Os3(CO)l,,(CH,)(H)2 is strictly dipolar. Estimates of the methyl rotation barriers from the spin-rotation relaxation times are reported.
Introduction The use of I3C N M R relaxation times as probes of the structures and dynamics of organic and main-group organometallic compounds is well e ~ t a b l i s h e dHowever, .~ only a few studies involving transition-metal organometallic compounds have a ~ p e a r e d ,and ~ , ~to our knowledge only one involving an alkyl carbon c bonded to a transition metaL5 We have thus undertaken studies to determine the I 3C(methyl) relaxation mechanisms of a number of representative transition metal* To z horn correspondence should be addressed a t the Department of Chemistry. Colorado State University. Fort Collins, Colo. 80523.
0002-78631791 1501-4853$01.OO/O
methyl complexes. In a preliminary communication6 we reported the 13C T I ' Sand ~ C H ' Sfor the methyl carbons in ( T C ~ H S ) F ~ ( C O ) ~ C( TH-~C,S H ~ ) M O ( C O ) ~ C H Os(Co)4~, (CH3)2, and ( T - C ~ H S ) ~ Z ~ ( CHerein H ~ ) ~ .we report variable-temperature (and in several instances variable-field) relaxation studies of the above compounds as well as of CH3AuPPh3, CH3Re(C0)5, and Os3(CO) ,o(CH3)(H). Potential Relaxation Mechanisms for Methyl Carbons. A major contribution to the spin-lattice relaxation rate R I ( I / T I ) of the methyl carbons in transition metal-methyl complexes is expected to be dipole-dipole relaxation Rl DD, due to the methyl h y d r c ~ g e n sIn . ~ the ~ ~ extreme narrowing limit this 0 1979 American Chemical Society
4854
Journal of the American Chemical Society
rate is given by eq I , i n which 7, is the carbon-hydrogen reorientation correlation time.2bRIDDand the non-dipole-dipole contribution Rlothercan be determined from R I and the n u clear Overhauser enhancement ~ C by H eq 2.8 However, when ~ C is H near the theoretical maximum of 1.99 (Le., the nondipole-dipole contribution is small), there is a large uncertainty in the calculated value of Rlother.
The reorientation correlation time 7, is temperature dependent (eq 3)9 and thus the dipolar contribution decreases with increasing temperature (eq 4 ) . 7,
= 7 0 exp(EA/RT)
In (RIDD)= E A / R T
(3)
+ constant
(4)
Possible non-dipole-dipole mechanism^^^,^ for a methyl carbon include chemical shift anisotropy, spin rotation, and scalar relaxation. The first contribution is field dependent and is described by eq 5 , in which 611and u l are the shielding tensor components and the extreme narrowing limit is assumed. It has been shown to be negligible for organic methyl carbonsI0 and is small for first-row transition-metal carbonyl carbons. I
For the molecules studied in this work, the spin-rotation contribution RISR to the methyl carbon relaxation will be predominantly due to internal rotation of the methyl group about the M-C In the extreme narrowing limit the magnitude of this contribution is given to reasonable accuracy by
where Cil is the axial component of the spin-rotation coupling tensor, and 711is the angular momentum correlation time for rotation about the symmetry axis, in this case the M-C bond.I2 Substitution of an approximate expression for 711,eq 7 (in which Ill is the moment of inertia about the M-C bond, k B is the Boltzmann constant, and /3 is an empirical into eq 6 gives the temperature dependence of RISR, eq 8. I n
-
711
=P
d&
(7)
contrast with other relaxation mechanisms this contribution increases with increasing temperature. For organic methyl carbons RISR approaches the limiting value of s'-I (at 38 "C) when the barrier to methyl rotation is very low (i.e., the methyl group is a free rotor) and increases approximately linearly with the barrier to rotation V0.l2 Scalar relaxation of the second kind is possible for methyl carbons that are bonded to metals with quadrupole moments, but is likely to be important only in CH3Re(C0)5.'4%'5 In the extreme narrowing limit this relaxation mechanism is described by 1 = RISC = 8 r 2 J 2 1 ( I + 1){ T IWQT ~ Q ) * (9) Tisc 3 1 (A in which T I Qand I are the relaxation time and spin of the quadrupolar nucleus, J is the coupling constant between it and the relaxing nucleus, and Au is the difference in their Larmor freq~encies.~ TIQ(Re) is not known, but it can be estimated from the
+
]
/ lOl:l7
August IS, 1979
TIQ(jSMn) in CH3Mn(CO)s. The quadrupolar relaxation rate
(in which q is the asymmetry parameter and Q' is the quadrupole coupling constant9) is proportional to (Q')* when 11 is small and 7, is constant.I6 The R e isotopes and j5Mn all have I = 5/2 and for corresponding compounds Q'(Re)/Q'(Mn) = IO,l5 so TIQ(Re) and TIQ(Mn) are approximately related by eq 1 1 .
(11)
TIQ(Mn) in CH3Mn(CO)j can be estimated from the "Mn line width (1.69 G a t 14.86 MHz)I7 to be 180 p s , which gives TlQ(Re) 1.8 ps. For C H 3 R e ( C 0 ) j in a 23.5-kG field Aw = 15.6 Mrad/s (weighted average for Ig7Reand '*'Re) and (AwTlQ)* = 790 >> I , so that eq 9 reduces to
Thus the scalar contribution to the 13C relaxation in CH3R e ( C 0 ) s should be both temperature ( 7 c , m o l c c u l a r 0: a Ha2) dependent. eXp(EA.molccular/RT))ahd field
Results and Discussion Dipole-Dipole and Spin-Rotation Relaxation. The variable-temperature data for all the M-I3CH3 compounds studied are summarized in Table I. It is clear from the less than maximum values for QCH that there is a significant nondipolar contribution to the relaxation. For O S ( C O ) ~ ( C H ~()T~- , CjHdFe(C0)2CH3, ( T - C S H ~ ) M ~ ( C O ) ~ C( KH-~C, ~ H ~ ) Z Zr(CH3)2, and CH3AuPPh3, Tlother decreases markedly with increasing temperature, indicating that the spin-rotation contribution is substantial. The high-field measurements (Table 11) show that for Os(C0)4(CH3)2 Ti is field independent and thus that the chemical shift anisotropy contribution is negligible.'* As this is the case where such a contribution is most likely to be detected (Os(C0)4(CH3)2 has the largest TI of the compounds studied, and RICSA is probably largest for carbons bonded to heavy metal^'^,^^), and, as such effects are rarely important (see Introduction and ref 1 l ) , we conclude that relaxation by chemical shift anisotropy is negligible in the Fe, Mo, Zr, and Au compounds as well. We thus conclude that for the compounds above Tlother= TISR(internal).A plot (Figure 1) of R l o t h evs. r fl (eq 8) for Os(C0)4(CH3)2 is not perfectly linear but, in view of the many approximations built into eq 6-8 and the large uncertainties in the calculated values of R l o t h e r , it offers satisfactory confirmation of our conclusion. The dipole-dipole contributions show the expected temperature dependence (eq 4) and thus plots (Figure 2 ) of In (1 / T IDD) vs. 1/ T are linear. The slopes give reasonable values for the reorientation activation energies (kcal/mol): 2.25 for O S ( C O ) ~ ( C H ~ )3.45 ~ , for CH3Re(CO)j, 2.48 for ( T C5Hj)Fe(C0)2CH3, 2.90 for ( T - C ~ H ~ ) M O ( C O ) ~3.21 CH~, for (~-CgHj)2Zr(CH3)2,and 2.84 for CH3AuPPh3. (The last two figures are based on only two experimental points.) Relaxation Mechanisms in I3CH3Re(CO)5. The very low values of ~ C (Tables H I and 11) for CHjRe(C0)' indicate that relaxation is predominantly nondipolar. By the argument of the previous section the chemical shift anisotropy contribution is probably small, so that spin-rotation and scalar relaxation of the second kind are the only nondipolar mechanisms possible: R,other = R I S R + R 1 S c (13)
Jordan, Norton
/ I3CN M R of Methyl-Transition Metal Compounds
4855
Table I. Variable Temperature I3C(Methyl) Relaxation Data a t 25.1 M H z (23.5 k G ) compd
0
-30
Ti VCH
TiDDc Tlothcr ' Ti
8.47 1.77 9.57 76.6 f 20
17c H TlDD
T iolhcr
12.9 1.46 11.7 48.4 f 7 4.24 I .99 4.24 >206 li
temp, "C 38
29 13.7 1.18 23.2 33.7 f 4
4.41 I .99 4.41 >440d
TI VCH
TiDD T,Olhcr
TI W H
TiDD T Iother Ti
3.22 1.99 3.22
VC H
TiDD Tlother
>160d
2.33 0.46 10.1 3.03
TI VCH
TiDD Tlolhcr a
All T I ' Sare f5%. NOES are f0.05 or as indicated.
14.8 I . I2 f 0.05 26.3 f 3 33.8 f 4 7.33 I .89 f 0.02 7.72 f 0.5 150 f 37 5.13 I .99 f 0.02 5.13 f 0.3 >510d 10.0 1.82 f 0.02 10.9 f 0.6 120 f 20 4.6 1.5 f 0.05 6.1 f 0 . 5 19. f 3
50
63
17.5 0.97 36.1 34.0 f 3
17.3 0.87 39.8 30.6 f 3 8.80 I .77 9.89 79
70
7.22 1.81 7.94 80.0 14.2 I .60 17.7 72.5
4.73 0.19 49.5 5.24
3.92 0.24 32.7 4.45
3.03 0.32 18.9 3.61
Calculated from eq 2.
90 14.0 0.68 41.2 21.2 f 2
Minimum limit estimated from eq 2 using minimum
WH.
Table II. 13C(Methyl) Relaxation Data a t 67.8 M H z " (63.5 kG) compd T I ,s 1)CH TlDD T,other OS~COMCH~)? C H3Re(COh
14.3 8.06
1.1 I 0.79
25.6 20.3
32.4 13.4
Temperature 34 "C.
eq 12 in the Introduction) is important. The field dependence of the scalar contribution can be used to formulate a pair of simultaneous equations (eq 14 and 15) which describes the nondipolar relaxation in two different fields H I and H r . R l o t h e r ( H i ) = R I S C ( H ~+ ) R~SR (14) Rlother(H2) = (HI/H2)2 RISC(Hi)
/" I5
IeT'
I8
19
2,mllr,
Figure 1. Plot of the nondipolar relaxation rate of the methyl carbon of Os(CO)d(CH3)2 vs. +The line is the least-squares fit to the data, and the bars represent estimated uncertainties.
As indicated by the fact that Rlotheris much greater than s-I a t 38 "C (the maximum spin-rotation contribution found for organic methyls' 2 , and decreases with increasing temperature, the spin-rotation contribution is minor. On the other hand, the fact that Rlother decreases markedly with increasing field suggests that scalar relaxation (as described by l/25
+ RISR
(1 5)
Solving these equations using R lother values obtained at 23.5 (Table I*') and 63.5 kG (Table 11) yields the values of T i S R = 22 s and TISC = 4.6 s in the former field at 34 OC. Scalar relaxation of the second kind is thus the dominant relaxation mechanism for the methyl carbon in CH3Re(CO)s.22.23The observation of field dependence of this type (i.e., Rother0: 1 / H o 2 ) ,is, in the extreme narrowing limit, diagnostic for such a scalar contribution. J R ~ -can I ~be c estimated as 640 Hz from eq 12 using the estimated value of TIQ(Re) and the experimental value of RlSc.This value is reasonable in light of known M-C couplings (e.g., J19spt-i3c is 594 Hz in cis-(CH3)2Pt(P(CH3)2Ph)rr4) and is consistent with the fact that no Re-I3C coupling is observed (Le., 2 r J T l Q = 7.2 X
