J . Phys. Chem. 1990, 94. 232-239
232
Methyl Cation Affinities of N, 0, and C Lone-Pair Donors Carol A. Deakyne**+ Geophysics Laboratory, Ionospheric Physics Dioision, Hanscom Air Force Base, Massachusetts 01 731-5000
and Michael Meot-Ner (Mautner) National Institute of Standards and Technology, Gaithersburg, Maryland 20899 (Received: June 19, 1989)
Methyl cation transfer equilibria were measured by use of pulsed high-pressure mass spectrometry. A ladder of A G O s o o values gives relative methyl cation affinities (MCAs) for several cyanides, ethers, and iodides, spanning a range of 15 kcal/mol. The ladder is anchored to an estimated MCA((CH,),O) = 93 kcal/mol, giving the following MCA values: CH3CN, 98; HCCCN, 95; i-C3H7CN,100; 1,4-dioxane, 96; c-C5Hlo0,98; c-C,H,O, 98; CHJ, 86; C2H51,90 kcal/mol. The MCAs of (CH3)*0, CH,CN, HCCCN, and several model compounds were calculated ab initio with basis sets ranging from MPn/6-31G* to MPn/6-31 I++G(2d,2p). The MP2/6-31G*(6-31G**) results agree best with the experimental results. MCAs show a good correlation with the amount of electron density transferred from the bases to the CH3+group, suggesting that the bonding is largely covalent. The bonding between a methyl cation and a base arises primary from the interaction between the (r-HOMO of the base and the u-LUMO of the methyl cation; *-effects are much less important.
Introduction Although there has been widespread theoretical and experimental interest in proton affinities in recent years,]-' relatively little work has been done on methyl cation affinities (MCAs). Reported experimental values have been determined via thermodynamic cycles utilizing enthalpies of formation or, to a much more limited extent, have been determined directly.2.8-10 The limited number of theoretical studies of MCAs include both ~emiempiricalll-l~ and ab initiol4J5 calculations. The latter were performed without complete geometry optimizations and without including electron correlation. Methyl cation affinities are useful quantities. They are indicative of the basicity of a carbon atom and can be utilized to garner information on hard vs soft bases. The methyl cation affinity of and a base correlates directly with its proton affinity ( PA)8-10916917 its intrinsic nucleophilicity.I6 Thus, we have carried out an investigation of the methyl cation affinities of several alcohols, nitriles, and isonitriles both experimentally and theoretically. The results have been used to examine (1) alkyl effects on the methyl cation affinity, (2) structural changes in the base brought about by the addition of the CH3+, and (3) the correlation between proton affinity and methyl cation affinity. The effect of diffuse functions, multiple sets of polarization functions, triple-{ basis sets, and electron correlation on the MCA have also been probed. Experimental and Theoretical Details The measurements were done by use of pulsed high-pressure mass spectrometry. As usual, reactions were initiated by a I-ms electron pulse, and ions were observed for an additional 1-4 ms. The reaction mixtures were composed of CH3C1or CH31as carrier gas and 1-20% of the reactants of interest. Rapid processes give (CH3)2CI+or (CH,)21+ which then react with the added nitriles or ethers to give RCNCH3+or R20CH3'. These ions react further to give the equilibria of interest. The methyl cation transfer reactions between the nitriles and cm3 ethers were very slow, with rate constants of 10-12-5 X s-I. Therefore, approach to equilibrium was slow and large concentrations of reactants were needed. The total source pressure was also high at 1-4 Torr. Checks were made to confirm that the equilibrium constant was independent of mixture composition and total source pressure. The calculations were carried out ab initio using the Gaussian 82 programI8 on a VAX 11/780 and a VAX 8650 computer. Fully optimized geometries were obtained by utilizing the 3-2 1 G and 6-3lC" (or 6-31G**) basis sets and the force relaxation 'On contract to GL from Wentworth Institute of Technology.
0022-3654/90/2094-0232$02.50/0
method.I9 The bond lengths were optimized to 0.001 8, and the bond angles to 0.1 O . The 6-31G* (or 6-31G**) equilibrium structures were used to compute single-point energies at several other basis set levels, the largest of which is denoted HF/6-311++G(2d,2p). This is a valence-triple-split basis with diffuse functions on all atoms, two sets of d functions on the non-hydrogen atoms, and two sets of p functions on the hydrogen atoms. Electron correlation was taken into account via second (MP2), third (MP3), and fourth (MP4SDTQ) order Mdler-Plesset perturbation theory.20 The frozen-core approximation was used.20 ( I ) For reviews see: (a) Dixon, D. A.; Lias, S. G. In Molecular Structure and Energetics; Liebman, J. F., Greenberg, A,, Eds.; VCH: Deerfield Beach, FL, 1987; Vol. 2, pp 233-314. (b) Bohme,D. K. In Interactions between Ions and Molecules; Auslws, P., Ed.; Plenum Press: New York, NY, 1975; pp 489-504. (c) Arnett, E. M. In Proton Trunsfer Reactions; Caldin, E. F., Gold, V., Eds.; Wiley: New York, NY, 1975; pp 79-101. (d) Kebarle, P. Annu. Rev. Phys. Chem. 1977,28,445. ( e ) Aue, D. H.; Bowers, M. T. In Gas Phase Ion Chemistry; Bowers, M. T., Ed.; Academic Press: New York, NY, 1979; Vol. 2, pp 1-51. (f) Taft, R. W. Prog. Phys. Org. Chem. 1983, 14, 248. (2) (a) Lias, S. G.; Liebman, J. F.; Levin, R. D. J . Phys. Chem. Ref Dara 1984, 13, 695. (b) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J . Phys. Chem. Ref. Data., Suppl. I 1988, 17, I . (3) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio Molecular Orbital Theory; Wiley: New York, NY, 1986. (4) Del Bene, J. E. J . Comput. Chem. 1985, 6, 296. (5) DeFrees, D. J.; McLean, A. D. J . Compur. Chem. 1986, 7, 321. (6) Meot-Ner (Mautner), M.; Karpas, Z.; Deakyne, C. A. J . Am. Chem. SOC.1986, 108, 3913. (7) Deakyne, C. A.; Meot-Ner (Mautner), M.; Buckley, T. J.; Metz, R. J. Chem. Phys. 1987, 86, 2334. (8) McMahon, T. B.; Heinis, T.; Nicol. G.; Hovey, J. K.; Kebarle, P. J . Am. Chem. SOC.1988, 110, 7591. (9) Huheey, J. E. Inorganic Chemistry: Principles of Structure and Reactiuity; Harper & Row: New York, NY, 1978; p 288. (10) Hine, J.; Weimar, R. D. J . Am. Chem. SOC.1965,87, 3387. (11) Ford, G. P.; Scribner, J. D. J . Comput. Chem. 1983, 4 , 594. (12) Sen Sharma, D. K.; DeHojer, S. M.; Kebarle, P. J . Am. Chem. Soc. 1985, 107, 3757. ( 1 3) McManus, S. P. J . Org. Chem. 1982, 47, 3070. (14) Vincent, M. A.; Radom, L. J . Am. Chem. SOC. 1978, 100, 3306. (15) Demontis, P.; Ercoli, R.; Gamba, A,; Suffritti, G. B.; Simonetta, M. J . Chem. SOC.,Perkins Trans. 2 1981, 488. ( I 6) Brauman, J. I.; Dodd, J. A.; Han, C.-C. In Nucleophilicity; Harris, J. M., McManus, S. P., Eds.; Advances in Chemistry Series 215; American Chemical Society: Washington, DC, 1987; pp 23-34. (17) Brauman, J. I.; Han, C.-C. J . Am. Chem. SOC.1988, 110, 5611. (18) Binkley, J. S.; Frisch, M.; Krishnan, R.; DeFrees, D. J.; Schlegel, H. B.; Whiteside, R. A.; Fluder, E. M.; Seeger, R.; Pople, J. A. Gaussian 82, release H, 1982, Carnegie-Mellon University. (19) (a) Pulay, P. Mol. Phys. 1969, 17, 197. (b) Schlegel, H. B.; Wolfe, S.; Bernardi, F. J . Chem. Phys. 1975, 63, 3632.
0 1990 American Chemical Society
MCAs of N, 0, and C Lone-Pair Donors A G'(KCAL/MOL)
The Journal of Physical Chemistry, Vol. 94, No. 1 , 1990 233 AMCA
u-C,H,CN
MCA 101
I00
I-C,H,CN
I
99 98 98 98 98
t
96 95
93
2.9 S8.4
w,I
-2.8
6.5
90
4.1
86
Figure 1. Ladder of A G O m values from methyl cation transfer equilibria.
The methyl cation affinity is the negative of the enthalpy change for reaction 1 . The translational and rotational energies were
CH,+ + B
-+
AH298 = AEclec+ A&&
BCH,',
MCA = -AH298
+ AE%t8+ AE''$i's + A(PV)298
(1)
(2)
calculated classically.21 The pressure-volume work term was determined from the ideal gas law and is equal to -RT for this reaction. Normal-mode vibrational frequencies and standard statistical formulas were utilized to obtain AEY,&..zl The normal-mode vibrational frequencies were computed with the 3-21G equilibrium structures.22 AEC1=is the electronic energy change for the reaction and is given by eq 3, where ET is the total energy of the species. AECICC = ET(BCH3+)- ET(B) - ET(CH~+) (3)
Experimental Results The equilibria of interest were of type (4)or analogous reactions. R20CH3'
+ RCN
-
RCNCH,'
+ R20
(4)
Since many of the reactions are slow and involve positive activation energies, the measurements were done at a high temperature, 600
K. The slow reaction rates caused several experimental problems. First, in several cases it was difficult to be sure that the reaction had indeed reached equilibrium in the observable reaction time. In fact, with such slow reactions care must be taken since at low reactant concentrations a reaction may not proceed at observable rates. In this case ion ratios do not change with time, which may be misconstrued as equilibrium. In general, the relationship between k for ACH3+ B BCH3+ A and N(B) was such that 1 / ~ k N ( B ) 10, s-I; Le., the reaction half-life is comparable to the observation time. Our experience with faster reactions such as proton transfer is that equilibrium is achieved or closely approached under such conditions. For several reactions,
-
-+
-
+
(20) (a) M~ller,C.; Plesset, M. S . Phys. Reu. 1934, 46, 618. (b) Pople, J . A.; Binkley, J. S . ; Seeger, R. I n f . J . Quantum Chem., Symp. 1976, 10, 1.
only a lower limit for AGO could be obtained as is indicated in Figure 1 . Second, there was not a sufficient number of suitable compounds with closely spaced MCAs to create an interlocking ladder. For this reason, it was necessary to study reactions with fairly large AGO values, which compounded the difficulty in reaching equilibrium. As a result, we consider the error in the quoted AGOm values to be f 1 kcal/mol, and we assign the same error limit, conservatively, to all reactions. The interlocking ladder is given in Figure 1 . The assigned relative MCA values are a weighted average of several alternative pathways between pairs of compounds. Pathways with fewer steps and, therefore, less cumulative error were assigned higher weights. For all pairs of reactants, the weighted averages and direct measurements agree to within f 1 kcal/mol. Figure 1 shows MCAs relative to (CH3),0. Unfortunately, absolute MCAs are not known for any of the compounds investigated in this work. However, the MCA of (CH3),0 can be estimated with some confidence by using thermochemistry from ref 2. LiebmanZ3has extended Stanton's studies on bond energiestZ to derive interrelationship 1. (1) For an arbitrary RsM(z+l)CH,+, if the ionization potential (IP) of the fragment RsM(z+l) is significantly larger than the ionization potential of CH, (9.8 eV),2b then the bond strength in the ion RsM(z+l)-CH,+ is less than that of the isoelectronic neutral RsM(z)-CH,. If the IP of the fragment RsM(z+l) is either less than that of CH, or higher than but sufficiently close to that of CH,, then the bond strength of the isoelectronic neutral molecule RsM(z)-CH, is weaker than that of the ion RsM(z+l)-CH,+. In this expression BCH3+is represented by RsM(z+ 1)CH3+,where Rs represents the ligands bonded to the atom or group of atoms M with nuclear charge z + 1. For example, for (CH3),0+M(z+l) is 0 and Rs is (CH,),. Other interrelationships included in the article are the following.23 (2) Replacing a hydrogen with a methyl group increases the proton affinity of a base. (3) Methylation of a base decreases its ionization potential. (4)The magnitudes of the changes in molecular properties brought about by the substitution of hydrogens by other ligands decrease as the number of substitutions increases. The same reasoning used to explain the effect of a methyl group on the PA of a base, Le., a methyl group stabilizes positive charge more than a hydrogen, suggests another interrelationship. (5) Replacing a hydrogen with a methyl group increases the methyl cation affinity of a base. With the use of the above inequalities one can derive an upper and lower bound for the methyl cation affinity of dimethyl ether. The decrease in ionization potential that accompanies methyl substitution of CH,OH (0.83 eV) is larger than the decrease observed for CH3NH2(0.74 eV).zb This leads to the following inequality, namely, MCA((CH,),O) - MCA(CH,OH) > MCA((CH,),NH) - MCA(CH3NH2). The experimental methyl cation affinities of HzO, CH,OH, NH,, CH,NH,, and (CH,),NH are 68.5, 84, 105, 116.5, and 122.5 kcal/mol, respectively;2 thus, MCA(CH3),0 L 90 kcal/mol. An upper bound for the methyl cation affinity of dimethyl ether is obtained from the fourth interrelationship given above. The inequality derived from this interrelationship is MCA(CH,OH) - MCA(H20) L MCA((CH3),0) - MCA(CH,OH), yielding MCA((CH,),O) I99.5 kcal/ mol. The lower bound for MCA((CH,),O) calculated from interrelationship 1 is 80 kcal/mol, which is less useful than the lower bound given above. In applying interrelationship 1 , RsM(z) is (CH,),N, IP((CH,),O) = 10.025 eV, and D((CH&NCH,) = 80 k c a l / m ~ l . ~ A more exact estimate of MCA((CH,),O) can be determined by noting that in the sequences CH3NH3+ (CH,),NH2+ (CH,),NH+ and CH,PH,+ (CH3),PH2+ (CH,),PH+ each methyl substitution decreases the heat of formation ( M o rof) the ion by an equal amount., In the analogous sequence CH30H2+ (CHJ20H+ (CH3),0+, dAHoffor the first two ions is 5.5
-
-
-
--
-
(21) Del Bene, J. E.; Mettee, H. D.; Frisch, M. J.; Luke, B. T.; Pople, J.
A. J . Phys. Chem. 1983.87, 3279. (22) Pople, J. A.; Krishnan, R.; Schlegel, H. B.; Binkley, J. S . I n f . J . Quantum Chem., Symp. 1979, 13, 325.
(23) Liebman, J. F. In Molecular Structure and Energetics; Liebman, J. F., Greenberg, A., Eds.;VCH: Deerfield Beach, FL, 1987; Vol. 4, pp 49-70. (24) Stanton, R. E. J . Chem. Phys. 1963, 39, 2368.
Deakyne and Meot-Ner ,CH3NC
.HNC
120r
CiHjNY2 CH3llV2
'
I
,,.-'
/
110r
H I
,I--.o-*
H
078
\
I459
I119
'070
C - N =C,-H H
7
-0-'.0~9
0.7.70
~.,S(I
1072
C3r
H
U
i
I
.
70t
H2O
2 ? 11
200
180
160
P A (kcal mol)
Figure 3. Methyl cation affinity vs proton affinity for neutral bases. The correlation coefficient is 0.737 and the slope is 0.71 .__ -
cs
C3"
Figure 2. Equilibrium geometries calculated from 6-31G* or 6-31G** (HC3NCH3+)optimizations. Bond lengths are in angstroms, and bond angles are in degrees. Atomic charges from population analysis are given in italics. kcal/moL2 Assuming the same value holds for the last two ions gives AHof((CH,),O+) = 124.5 kcal/mol and MCA((CH,),O) = 93.5 kcal/mol. If the trends in the MCAs of the series N H 3 CH3NH2 (CH,),NH are extrapolated to the series H 2 0 CH30H (CH3)20, the value obtained for the MCA of ( C H J 2 0 is 92.5 kcal/mol. The average estimated value of MCA((CH,),O) is 93 f 1 kcal/mol. As a matter of interest we note that AH"f((CH,),N+) = 129.5 kcal/mol and MCA((CH,),N) = 126 f 1 kcal/mol from these methods. This MCA is, of course, far above the range of the present measurements. The MCA of CH3CN can be estimated from the proton affinities of CH$N and H C N as follows.2 PA(CH,CN) - PA(HCN) = 17.0 kcal/mol, and the slope of the plot of MCA vs PA in Figure 3 is 0.71. From this information, MCA(CH,CN) - MCA(HCN) = 12 kcal/mol. Using 87.1 kcal/mo12 for the MCA of HCN yields 99 f 3 kcal/mol for the MCA of CH3CN. The large uncertainty is due in part to the scatter in the MCA vs PA correlations. From the above estimates the difference in the MCA of CH3CN and (CH,),O is 6 kcal/mol, in reasonable agreement with the measured difference of 4.5 kcal/mol (Figure 1). This supports both the measured and estimated values. The relative MCAs in Figure 1 were measured at 600 K. However, since the number of vibrations and rotations are equal for both sides of eq 4, the C, values are also equal to a good approximation. Therefore, the relative measured MCAs have been combined with the estimated MCA((CH,),O) at 298 K to obtain the absolute MCAs at 298 K reported in Figure I . We note that HCN with a MCA of 87.1 kcal/mo12 falls within the range of the present measurements and would serve as a suitable reference standard. However, attempts to measure a methyl transfer equilibrium with C H J led to the formation of HCNCH3+ irreversibly, corresponding to an apparent MCA(CH,I) - MCA(HCN) > 5 kcal/mol. Experiments with CH3C H O (MCA = 88 kcal/mo12) as a reference standard also produced no satisfactory equilibria. Another possible reference reaction is (CH3)21++ H2S CH3SH2+ CHJ. However, the MCAs of H2S and CH,I are comparable (84 and 86 kcal/mol, respectively2) and no competing proton-transfer reactions occur. It is of interest to compare trends in methyl cation affinities with proton affinities. For the ethers the variations in the two
--
-
.
~
/
root UP?
/
--
-
+
I 160
170
180 PA (kcalimol)
_-_ 190
-
200
..
Figure 4. Methyl cation affinity vs proton affinity for water, methanol, and dimethyl ether. Experimental data, 0; MP3/6-31+G*//HF/631G* data, +; MP2/6-3l+G*//HF/6-31G* data, A; HF/6-31+G*/ /HF/6-31G* data, See text for correlation coefficients and slopes. follow quantitatively (ether, PA, MCA): ( ~ ~ 3 1 0.0, ~ 0 0.0; , 1,4-dioxane, 1.7, 3.2; THF, 6.7, 5.4; THP, 7.6, 5.3. In contrast, the correlation for the nitriles is poor: CH,CN, 0.0, 0.0; HCN, -17.0, -1 1; CHICICN, -8.9,0.9; HCCCN, -7.7, -2.7; C,H,CN, 4.2, 1.3; n-C3H7CN, 5.3, 3.4; i-C3H7CN, 5.9, 2.4. A further discussion of MCA vs PA is given below.
Computational Results and Analysis The 6-3 lG*-optimized structures of CH,NCH+, CH3CNCH3+, (CH3)20H+,(CH&O+, and HCCCNCH3+ (6-31G**) are depicted in Figure 2. The angle between the LCOC bisector and the 0-H bond is 38.7' in protonated dimethyl ether. The equilibrium geometries of the other molecules can be found in the l i t e r a t ~ r e . ~Atomic ~ ~ ~ . charges ~~ from Mulliken population analysis are included in the figure (in italics).27 Figures 3 and 4 are graphs of methyl cation affinity vs proton affinity for a series of bases. Table I gives the total electronic energies of the molecules at various basis set levels. Other ET data utilized to calculate MCAs and PAS are from ref 4, 7, and 25. The data for CH3CNH+, HCCCN, and HCCCNH+ are from ref 7. The 6-31 +G** and (25) Whiteside, R. A.; Frisch, M. J.; Binkley, J. S.; DeFrees, D. J.; Schlegel, H. B.; Raghavachari, K.; Pople, J. A. 'Carnegie-Mellon Quantum Chemistry Archive"; Department of Chemistry, Carnegie-Mellon University: Pittsburgh, PA. ( 2 6 ) Boyd, R. J.; Jones, W. E.; Ling, K. W. Chem. Phys. 1981, 58, 203. (27) Mulliken, R. S. J . Chem. Phys. 1955, 23, 1833.
MCAs of N, 0, and C Lone-Pair Donors
The Journal of Physical Chemistry, Vol. 94, No. I , 1990 235
TABLE I: Calculated Total Energies ( E T )in hartrees
molecule CH3+
H20 H30+ CH30H CH3OH2' (CHd2O (CHd@H+
(CHp)@+ HCN HCNH+ HNC
CH3CN CH3CNH+
CH3NC C H ,NC H+
C H 3CN C H 3+ HCCCNe H CCCN H HCCCNCH3tc
basis 6-31+G* 6-31+G** 6-31 l++G(2d,2p) 6-31+G* 6-31+G* 6-31+G* 6-3 1 +G* 6-31 I++G(2d,2p) 6-31G* 6-31G** 6-3 1 +G* 6-31G* 6-31G** 6-31+G* 6-3 I+G** 6-311+G** 6-31G* 6-31G** 6-31+G* 6-3 1+G* 6-31 I++G(2d,2p) 6-31+G* 6-31+G* 6-3 l+G** 6-31 I+G** 6-31 I++G(2d,2p) 6-31G' 6-3 1 +G* 6-31G* 6-31+G* 6-311+G** 6-31 I++G(2d,2p) 6-31G* 6-3 1G** 6-31+G* 6-31G* 6-31G** 6-31+G* 6-31+G** 6-3 I 1+G** 6-31 I++G(2d,2p) 6-31G* 6-31G** 6-31+G* 6-31G* 6-31+G* 6-31G' 6-31+G* 6-31G* 6-31G** 6-31+G*
HF -39.230 89 -39.236 70 -39.246 41 -76.017 7 1 -76.290 60 -1 15.04092 -1 15.340 65 -1 15.391 23 -1 54.064 74' -1 54.074 00 -154.069 33 -154.382 38 -154.39881 -154.38066 -1 54.400 32 -154.43099 -193.421 48 -193.43621 -193.42288 -92.878 63 -92.904 99 -93.15949 -92.861 08 -92.865 44 -92.884 62 -92.892 73 -131.927 53' -131.931 17 -132.23675 -132.237 28 -132.273 54 -1 32.280 41 -131.894 36b -13 1.899 02 -1 3 1.900 82 -132.221 99 -132.229 12 -1 32.222 87 -1 32.230 06 -132.25651 -132.26276 -171.294 02 -171.304 3 1 -17 1.294 90 -168.54943 -168.55504 -168.84686 -168.84879 -207.905 47 -207.91248 -207.907 19
MP2 -39.325 54 -39.347 04 -39.365 14 -76.208 70 -76.476 42 -115.35645 -1 15.648 56 -1 15.772 62 -1 54.502 06 -1 54.550 76 -154.513 03 -154.813 14 -1 54.870 43 -154.814 64
MP3 -39.342 04 -39.365 04
MP4'
-76.21 2 99 -76.483 56 -115.371 00 -1 15.666 36 -1 54.526 80 -1 54.578 95
-154.536 85 -1 54.840 97 -154,901 63 -154.842 17
-193.981 71
-194.018 28
-193.988 54 -93.16032 -93.22007 -93.432 39 -93.132 05 -93.14070 -93.17 1 80 -93.225 41 -132.333 72 -132.34065 -132.635 35 -1 32.637 43 -132.711 08 -132.741 40 -132.29008 -1 32.3 13 64 -1 32.300 91 -132.621 32 -132.65051 -132.623 82 -132.652 90 - 13 2.694 69 -132.725 27 -171.82035 -171.865 83 -171.82364 -169.071 03 -1 69.079 87 -169.362 84 -169.366 57 -208.549 11 -208.578 67 -208.55402
-194.025 IO -93. I6203
-1 54.546 32
-1 54.855 15
-1 94.044 80
-93.438 39 -93.1 39 03 -93.147 91 -132.345 38 -132.351 74 -132.651 43 -132.653 56
-132.371 19
-1 32.305 67 -1 32.330 80
-132.331 19
-132.31573 -132.635 58 -132.66650 -132.638 13 -132.668 92 -171.84448 -171.847 87 -169.07263 -169.08074 -169.367 22 -169.370 87 -208.561 50
-1 32.674 94
-1 32.660 66
-1 7 1.875 66
-169.1 11 92 -169.405 84 -208.60761
-208.566 44
'MP4SDTQ. bReference 25. CThetotal energies were obtained with HF/6-31G** optimized geometries. For all the other molecules, the total energies were obtained with HF/6-31G* optimized geometries.
6-31 I+G** data for C H 3 0 H , CH30H2+,and HCN are from ref 4. All other total electronic energies from the literature are from ref 25. Table I1 presents the zero-point energies (ZPE) and internal energies (E298)of the molecules and ions. The internal energies were computed employing the methods described above.
The methyl cation affinities are given in Table 111. The experimental MCAs reported for H,O, CH30H, HNC, and HCN are obtained by utilizing eq 1 and heats of formation from ref 2 and 6. The measured methyl cation affinity of CH3CN (98 kcal/mol) yields 18 1 kcal/mol for AHor(CH3CNCH3+). This result combined with AHor(CH3NC) = 42 kcal/mo12 gives a MCA of 122 kcal/mol for CH3NC. The normal-mode vibrational frequencies obtained with the 3-21G basis set are listed in Table IV. The values in the table have been scaled by 0.89, the recommended scale factor for frequencies obtained at this level of c a l ~ u l a t i o n . ~ ~ ~ ~
Table V tabulates Rxc, AqCT, the X-C Mulliken overlap pop~lation,~' AqR(tot), and cu. Rxc is the length of the X - C bond in the cations, where X = C, N , or 0. AqCT is the amount of electron density transferred from the base to the CH3+moiety. AqR(tot) is the amount of electron density transferred to the rest of the base from the hydrogen(s) bonded directly to the oxygen or CN group and/or from the hydrogens of the methyl group(s) bonded to the oxygen or C N group. cu is the energy in hartrees of the highest occupied u molecular orbital (u-HOMO). These terms provide information on the delocalization, Le., charge transfer and polarization, component of the methyl cation affin-
it^.',^^ A . Energetics. Examination of the results tabulated in Table 111 leads to several general conclusions. First, the Hartree-Fock
MCAs are too small compared to the experimental MCAs at every (28) Pople, J. A,; Schlegel, H. B.; Krishnan, R.; DeFrees, D. J.; Binkley, J. S.; Frisch, M. J.; Whiteside, R. A,; Hout, R. F., Jr.; Hehre, W. J. I n t . J . Quantum Chem., Symp. 1981, 15, 269. (29) Kollman, P.; Rothenberg, S. J . Am. Chem. SOC.1977, 99, 1333.
236 The Journal of Physical Chemistry, Vol. 94, No. I , 1990 TABLE 11: Calculated Zero-Point Energies (ZPE)O and Internal Energies ( E d b in kcal/mol point molecule group ZPE E298 CH1+ D i..h 18.5 1 20.30 13.95 H ,0 c2, 12.17 H30+ c3, 19.58 21.50 32.53 CH,OH c, 30.43 cs 37.29 39.88 CH,0H2+ 50.59 (CH,)*O c2, 47.79 c2, 54.74 58.15 (CH,)@H+ 16.35 (cH 3)@+ c3,: 72.27 C,, 10.14 11.70 HCN HCNH' C,, 17.36 18.99 C," 9.8 1 11.46 HNC CH3CN c 3 c 27.59 29.79 CH3CNH+ c3, 34.30 36.62 C3" 27.19 29.57 CH3NC 36.88 C H ,N C H+ c3ti 34.46 C H ,CN C H 3+ c3, 5 1.40 54.95 HCCCN C,, 17.44 19.78 26.87 HCCCNH+ C ,, 24.39 44.90 HCCCNCH3+ c3, 41.47 a Calculated from 3-2 I G optimum geometries; corrected by 0.89. bSee text for formula.
basis set level. Second, accounting for electron correlation at the MP2 level enlarges the magnitudes of the MCAs by 14-24 kcal/mol. The net effect of the MP3 and MP4 corrections is to reduce the MP2 correction by 1-4 kcal/mol. The increases in the MCAs of the oxygen and nitrogen bases are similar in magnitude while those for the carbon bases are larger. Third, the size of the correlation correction for a given molecule is quite constant regardless of basis set. It changes by no more than 2 kcal/mol. Thus, the electron correlation contribution for the smallest basis set (6-31G*) can be used to approximate the contribution for the larger basis sets. In addition, the contribution TABLE 111: Methyl Cation Affinities (MCA) in kcal/mol" molecule basis
H2
0
CH30H
(CH3)2O HCN
HNC
CHjCN CHjNC HCCCN
6-31G* 6-31G** 6-3 1+G* 6-3I+G** 6-311+G** 6-31 I++G(2d,2p) 6-31G* 6-3 IC** 6-31+G* 6-31+G** 6-31G* 6-31G** 6-3 1 +G* 6-31G* 6-31G** 6-31+G* 6-3 l+G** 6-31 l+G** 6-31 I++G(2d,2p) 6-31G' 6-31G** 6-3 1+G* 6-31+G** 6-311+G** 6-31 I++G(2d,2p) 6-31G* 6-3 1 G** 6-31+G* 6-31G* 6-31G** 6-3 1+G* 6-31G* 6-3 IC** 6-3 1 +G*
HF 56.2 56.2 52.7 52.0 51.2 50.3 68.3 68.0 63.6 65.1 74.2 74.1 72.1 68.6 68.3 66.8 66.4 65.7 65.6 90.3 90.1 86.9 86.5 86.9 87.4 80.9 80.7 79.1 101.6 101.5 97.9 74.5 74. I 72.2
Deakyne and Meot-Ner SCHEME I c-'> C ,O-0 N 0-0 C2"
-
O C I O - O C I 0 - o NO - o C 2 0 M09
MOI I
for the unsubstituted molecule of a series can be used to approximate the contribution for all the molecules of the series. However, since the latter conclusion is based on such a limited number of systems, it requires further testing. Fourth, the MCAs obtained with and without polarization functions on the hydrogens are very similar. They vary by no more than 1.5 kcal/mol and usually by less than 0.5 kcal/mol, which is substantially less than what is generally observed for proton affinities3 The differences tend to be greater when diffuse functions are included in the basis set than when they are not. Most of the calculations for the larger molecules were carried out without polarization functions on the hydrogens. Fifth, the addition of diffuse functions to the basis set lowers the calculated MCAs by 2-6 kcal/mol. Further expansion of the basis set to 6-31 1+G** and 6-31 1++G(2d,2p) changes the MCAs by about 1 kcal/mol. Similar results were obtained in comparable studies of the effect of basis set size on proton a f f i n i t i e ~ . ~ , ~ The net outcome is that even the best calculations carried out in this study yield MCAs that are too small by 4-7 kcal/mol compared to the experimental MCAs. In fact, the MP2/6-3iG* results are closest in magnitude to the experimental values. Since the changes in the MCAs are much smaller for the MP3 and MP4 calculations than they are for the MP2 calculations, this suggests that the discrepancies in the experimental and theoretical MCAs will not be decreased significantly by higher order Mdler-Plesset calculations. Rather, larger basis sets are needed to reduce the error to 1-2 kcal/mol. It should be noted, however, that MCAs are difficult to obtain experimentally (this work and ref 8) and that many of the experimental values reported herein depend on the accuracy of the dimethyl ether MCA. In addition, there is some question about whether the proton affinity and, therefore, the heat of formation of H N C should be l ~ w e r . ~If. it~ is~ lower,
MP2 72.3 72. I 66.7 65.8 66.2 64.9 85.0 84.8 78.5
MP3 69.5 69.4 64.8 64. I
MP4 71.1
exDtb
82.2 81.9 76.3
81.2
84.0
92.1
89.2
91.0
93
89.2 84.6 85.0 82.3 82.5 82.8 83.6 113.2 113.3 108.6 108.6 110.7 1 10.9 97.1 97.4 94.5 124.2 124.6 119.2 91.7 92.4 89.0
86.9 81.7 82.0 79.8 80.0
82.7
87.1
68.5
108.1 108.0 104.0 103.8
109.0
118
94.6
95.3
98
92.4 119.3
120.2
122
114.8 88.2
89.8
95
85.9
MCAs calculated utilizing total electronic energies from this work and ref 4, 7, and 25. See text for details. bThis work or ref 2.
MCAs of N, 0, and C Lone-Pair Donors
The Journal of Physical Chemistry, Vol. 94, No. I , 1990 237
TABLE IV: Vibrational Frequencies in cm-I'
point group
molecule
sYm
frequency
A' A" A' A" A1 A2
390, 745, 1107, 1419, 1446, 1616, 2913, 3032, 3359 91, 833, 1194, 1454, 3028, 3457 109, 255, 331, 736, 1137, 1206, 1447, 1460, 1473, 2905, 3008, 3014, 3436 99, 872, 1061, 1108, 1375, 1425, 1451, 1461, 2904, 3008, 3013 212, 670, 1163, 1473, 1477, 2902, 2997 133, 1086, 1454, 3002 133, 357, 902, 1111, 1228, 1427, 1458, 1475, 2897, 2994, 3008 675, 1413, 2241, 2886, 3182 335, 956, 1133, 1447, 2984 587, 911, 1415, 1421, 2393, 2853, 2889 2 199, 519, 1089, 1135, 1434, 1456, 2927, 2982 571,985, 1419, 2127, 2411, 2883, 3195 148. 331, 738, 936, 1130, 1452, 2977
CH30H2'
c,
(CHd20H'
c,
(CHAP
C3"
CH3NCHt
C3"
CH3CNCH3+
C3"
AI E AI A2
HCCCNCH,'
C3"
AI
E
E E
"Obtained with the 3-21G basis set; corrected by 0.89. TABLE V: Parameters Related to the Methyl Cation Affinity
molecule AH,,,'
x-C' RXCb 0.p.
Aqm
AqR(tot)
e., d
H20 CH3OH (CH3)20 HCN CH,CN HNC CH3NC
1.400 1.480 1.465 1.459 1.454 1.461 1.468
0.396 0.445 0.474 0.484 0.514 0.626 0.657
0.254 0.454 0.610 0.137 0.207 0.075 0.210
-0.57080 -0.49738 -0.47552 -0.57380 -0.54948 -0.47852 -0.46035
56.2 68.3 74.2 68.6 80.9 90.3 101.6
0.138 0.192 0.233 0.217 0.246 0.459 0.465
HF/6-31G*//HF/6-31G* methyl cation affinity in kcal/mol. bX-C bond length in A, where X is 0, N , or C. cX-C Mulliken
overlap population.
Energy in hartrees.
this would also decrease the value of the experimental methyl cation affinity for HNC. Thus, some of the deviation in the experimental and theoretical data may be due to inaccuracies in the experimental data. In contrast, the conclusions regarding relative MCAs are much more encouraging. Compare the computed and experimental MCAs of the other bases with those of (CH3)20. Expanding the size of the basis set and incorporating electron correlation in the calculations does improve the relative values of the MCAs. However, the improvement is generally minor. This conclusion is reinforced by doing a further comparison among HCN, HNC, H 2 0 , and CH30H, for which a greater number of basis sets were employed. Consequently, if qualitative rather than quantitative results are desired, these data indicate that small basis sets and Hartree-Fock theory can be utilized to compute MCAs. This conclusion is not unexpected since a similar conclusion has been made for proton affinities.Ia Interrelationships and Bounds. The interrelationships given above yield upper or lower bounds for the MCAs of several of the molecules other than (CH3)20.23Consider interrelationship 1. The ionization potentials of H C N (14.0 eV) and H 2 0 (12.6 eV) are both considerably larger than that of CH, (9.8 eV).zb Interrelationship I suggests that the MCAs of HCN and HzO are smaller than the bond dissociation energies of HCC-CH3 (1 17 kcal/mol) and H2N-CH, (84 kcal/mol), respectively.z The ionization potential of CH,OH (1 1.O eV) is within about 1 eV of that of CH3.2b In this case interrelationship 1 suggests that the bond dissociation energy of CH3N(H)CH3 (84 kcal/mol)2 is a lower bound for the MCA of C H 3 0 H . Examination of the data in Table 111 shows that all of the above inequalities are true but that interrelationship 1 leads to a good estimate of only MCA(CH3OH). The data are not available to employ interrelationship 1 for the other molecules in Table 111. However, an upper bound for the MCA of CH3CN can be obtained by considering the other interrelationships. The difference in the ionization potentials of NH3 and CH3NHzis 1.2 eV while the difference for HCN and CH3CN is 0.9 eV.2b This yields the inequality MCA(CH3CN) - MCA(30) Deakyne, C. A. Manuscript in preparation.
(HCN) MCA(CH3NH2) - MCA(NH3); thus, MCA(CH3CN) I 99 kcal/mol, in very good agreement with the experimental value (Figure 1). It should be noted, however, that applying this reasoning to the proton affinity of CH3CN leads to the incorrect inequality PA(CH3CN) I 181.4 kcal/moL2 The proton affinity of CH3CN is 188 kcal/mol.2 B. RNC us RCN. The bonding between a methyl cation and a base primarily arises from the interaction between the a-HOMO of the base and the u-LUMO of the methyl cation. Since H N C and CH,NC are stronger u-donors than H C N and CH,CN,,' respectively, it is expected that the R N C MCAs will be larger than the R C N MCAs and this is what is observed (Table 111). In fact, AMCA(CH,NC-CH,CN) and AMCA(HNC-HCN) are greater than 20 kcal/mol, with the former difference slightly smaller than the latter. Of course, the MCA of CH3NC must be larger than the MCA of CH,CN since the resultant cation, CH3CNCH3+,is identical for both bases and CH3NC is less stable than CH3CN (Table 111). The relative values of the R N C and RCN methyl cation affinities are improved with respect to the experimental values by including electron correlation in the calculations (Table 111). The correlation contribution to the MCA is larger for the isonitriles than for the nitriles, so part of the computed AMCA(RNC-RCN) is the result of the differences in the magnitudes of the correlation contributions. However, electron correlation accounts for only 3-5 kcal/mol of AMCA(RNC-RCN); the remainder is obtained at the Hartree-Fock level. Geometries. Adding a methyl cation to a cyano or an isocyano base affects the geometry of the base. For CH3NC and H N C the addition of the CH3+ shortens the C=N bond and lengthens the N-X (X = H, C) bond. Both C=N bond shortenings are about 0.03 A, but the N-C bond lengthens by 0.03 A while the N-H bond lengthens by only about 0.01 A. The same trends are observed for the nitriles, but in general the changes are smaller. The addition of the CH3+ shortens the e N bond by 0.01 A for both nitriles, lengthens the C-H bond of H C N by 0.01 A, and leaves the C-C bond of CH,CN essentially unchanged. Similar bond length changes are observed when these bases are protonated.6 However, the changes brought about by protonation are generally larger, as expected from the much stronger interaction between the proton and base than between the methyl cation and base. The changes in the C=N and N-C bond lengths when CH3+ is added to CH3NC can be rationalized by comparing the orbital composition of the u-HOMO of CH3NC with the orbital composition of the molecular orbital (MO) to which it corresponds in CH3CNCH3+. This analysis will not be applied to the observed variations in the CH,CN bond lengths, since these variations are too small for their origins to be elucidated by this method. The pair of orbitals for CH3NC (MO 11 and MO 9) are shown schematically in Scheme I. (31) Howell, J. A. S.; Saillard, J.-Y.; Le Beuze, A,; Jaouen, G. J . Chem. SOC.,Dalton Trans. 1982, 2533.
Deakyne and Meot-Ner
238 The Journal of Physical Chemistry, Vol. 94, No. 1. 1990 Scheme I shows that the addition of CH3+to CH3NCgenerates charge density redistributions in M O 1 I that (1) greatly reduce the electron density in the Cl(2s) atomic orbital (AO) and (2) reduce the electron density in the N(2s) and Cl(2p,) AOs (i.e., the p orbital along the C-N-C internuclear axis). Furthermore, adding the CH3+group causes the sign of the N(2s) and C2(2s) orbitals to reverse. These modifications convert the C1(2s)-N(2s) and C,( 2pX)-N(2s) overlaps from antibonding to bonding, convert the Cl(2p,)-N(2s) and C2(2s)-N(2p,) overlaps from bonding to antibonding, and greatly diminish the Cl(2s)-N(2p,) antibonding overlap. Another relevant interaction is that between the C2(2p,) and N(2p,) orbitals. This interaction is either bonding or nonbonding in all of the CH3NC occupied orbitals. Addition of the CH3+significantly decreases the electron density in both of these 2p, orbitals which reduces the bonding overlap between them. In net, the C,=N bond is shortened and the N-C, bond is lengthened, although the competing effects make the changes smaller than might be expected from the large charge shifts involved. Examination of the CH,+-induced charge redistributions in the degenerate set of *-HOMOS of CH,NC shows that, in contrast to what is actually observed, the effect of the charge shifts is to lengthen the C=N bond and to shorten the N-C bond. When all of the a-orbitals are considered, the overall result is that the bond length changes essentially cancel. As noted above, protonating CH,NC produces bond length changes similar to those obtained when a CH3+ moiety is added to CH,NC6 However, the electron density shifts responsible for the changes are not entirely the same in both cases. The pertinent interactions when CH3NC is protonated are the C1(2s)-N(2s), C I(2s)-N( 2p,), C I(2p,)-N( 2p,), and C2(2s)-N( 2s,2pX).6Charge redistributions in the *-orbitals are not important here either.6 C. Alkyl Effects. Substituting a methyl group for a hydrogen increases the methyl cation affinity of a base (Table HI), in agreement with interrelationship 5 (see above). For the cyano and isocyano bases the magnitude of the alkyl effect on the MCA is independent of basis set size, inclusion of electron correlation, and level of correlation. The observation is less true for the oxygen bases where the variation with respect to the first factor can be as large as 4 kcal/mol. With the exception of C H 3 0 H the computed methyl effects are overestimated compared to the experimental results. Since the electron correlation contribution to the MCA difference is less than 2 kcal/mol when a hydrogen is replaced by a methyl group, the increase in MCA produced by this substitution can be explained by the Hartree-Fock calculations. Delocalization effects are known to be important in the enhancement of the proton affinity when a methyl group replaces a h y d r ~ g e n . ~ Similarly, .~~,~~ these effects are important in the enhancement of the MCA when a methyl group replaces a hydrogen. Alkyl groups increase the polarizability of a base, since they supply more electron density to the proton- or CH,+-accepting atom than a hydrogen does.6*29,32 For the MCA this is exemplified by the rise in the total amount of electron density transferred from the base to the CH3+ moiety Aqcr, by the rise in the amount of electron density transferred from the hydrogens of the base to the CH,+ moiety AqR(tot), and by the destabilization of the a-HOMO when methyl substitution occurs (Table V). Furthermore, although the X-C bond lengths in t h e methyl-substituted bases are often longer than in the unsubstituted bases, the X-C overlap populations are larger for the methyl-substituted bases, again indicating a larger delocalization component of the bond energy when a hydrogen is replaced by a methyl group. While redistribution of electron density through the u-orbitals makes the major contribution to the alkyl effect, redistribution of electron density through the *-orbitals also contributes. Consider a simple PMO treatment of the interaction between a methyl group and a C N or a CH3CN+fragment. Charge shifts through the *-orbitals stabilize the cations if ( I ) the *-HOMO of the CH, fragment interacts with the T-LUMO of the C N or CH,CN+ fragment and (2) the interaction is stronger for the (32) Umeyama, H.; Morokuma, K . J . Am. Chem. SOC.1976, 98, 4400.
CH3CN+ fragment than for the C N fragment. The interaction between the CH,CN+ x-LUMO and CH, *-HOMO is stronger than that between the C N T-LUMO and CH, *-HOMO. The difference in energy between the two former orbitals is -0.4 au; the difference in energy between the two latter orbitals is -0.7 au. However, the above more favorable interaction between CH3CN+and CH, is partially offset by the interaction between the T-HOMO of the base fragment and the T-LUMO of the CH3 fragment. In this case the C N *-HOMO - CH, T-LUMO energy difference is -0.8 au and the CH3CN+ *-HOMO - CH3 T LUMO difference is 1.I au. The argument holds for the other bases as well. D. Plots of MCA vs PA. From both an experimental and calculational point of view, it is difficult to obtain accurate MCAs. However, as long as the PA of a base is known, its MCA can be estimated from a plot of methyl cation affinity vs proton affinity for a series of related bases.8,9~16*'7 Moreover, as is shown below, plots of experimental data are reproduced quite well by plots of theoretical data. Graphs of methyl cation affinity vs. proton affinity have been shown to be linear for a wide range of Such a plot is given in Figure 3 for all of the bases considered in this study plus some amines, iodides, and thiols. The methyl cation affinities of the additional bases were determined via eq 1 using heats of formation from ref 2. The proton affinities were also taken from ref 2. The correlation coefficient for this plot is 0.737 and the slope is 0.71. If the slope and the correlation coefficient had been unity, then the differences in heats of formation of BH+ and BCH3+ would have been identical for every B (see eq 1). Alternatively, as shown by the cycles in eq 6 and 7, a slope and correlation coefficient of unity would have meant that the difference in the B+-H and B+-CH3 bond strengths is the same for every B (see eq 8). Note that eq 8 indicates that a large part
-
BCH,+(g)
- + + - + CH,(g)
CH,+(g) + e
CH,(g)
B+(g) net:
BCH3+(g) BH+(g)
D(B+-CHJ
B+(g)
e
B(g)
CH,+(g)
H(g)
H(g)
B+(g)
B(g)
+ B+(g)
+ -
+
H+(g) + e e
IP(CH3) -IP(B) MCA(B)
(6)
D(B+-H) IPW) -IP(B)
B(g)
net: BH+(g) H+(g) + B(g) PA(B) (7) PA(B) - MCA(B) = [D(B+-H) - D(B+-CH,)] [IP(H) - IP(CH,)] (8)
+
of the difference in the magnitudes of PA(B) and MCA(B) is the result of the difference in the magnitudes of IP(H) and IP(CH3), 88 kcal/mol.2b The graphs show less scatter when a series of related bases are pl0tted8,~~J~ This has been done for water, methanol, and dimethyl ether in Figure 4. Now the correlation coefficient is 1 .OOO and the slope is 1.0. If all the alcohols and ethers included in Figure 3 had been included in Figure 4, the slope would have been 0.92 and the correlation coefficient would have been 0.984. When MP3/6-31 +G*//HF/6-31G* proton affinities (this work and ref 25) and methyl cation affinities are plotted, the slope is 0.96 and the correlation coefficient is 0.969. Slightly better results are obtained when the MP2/6-3l+G*//HF/6-31G* data are utilized (slope, 0.99; correlation coefficient, 0.977). The HF/6-3 1 G*//HF/6-3lG* data yield a line translated downward with a slope of 0.77 and a correlation coefficient of 0.989. The analogous graphs for the nitrile bases show similar patterns. While the data are limited and a more extensive study is required, these results again indicate that relatively small basis sets yield reliable trends in MCAs and PAs.6*29,32 The results also suggest that plots of theoretical data can be used to obtain a good estimate of the MCA of a base as long as an accurate experimental or theoretical value of its PA is known.
+-
J . Phys. Chem. 1990, 94, 239-243 Acknowledgment. We thank Drs. M. J. Henchman and J. F. Liebman for helpful discussions. The support of the Air Force Geophysics Laboratory Information Resources Center is gratefully acknowledged. Registry No. CHJ,74-88-4; CH,CI,74-87-3; (CHJ2C1+, 24400-15-5; (CH3)21', 24400-13-3; H20,7732-18-5;CHSOH,67-56-1;(CH,),O, 1 1 5-10-6;HCN,74-90-8; HNC,75-13-8;CHSCN,75-05-8;CHSNC,
239
593-75-9; HCCCN,1070-71-9; C2H5NH2, 75-04-7; CHSNH2, 74-89-5; NH3,7664-41-7;C4H@, 109-99-9;C5H100, 142-68-7;C4H802,12391-1; C ~ H S O108-95-2; H~ C2H50H364-17-5;r-C4H90Ht 75-65-0; H2S9 7783-06-4; C2H51, 75-03-6; CHSSH, 74-93-1; i-CSH7, 78-82-0; C2H5CN, 107-12-0; CH30H2+, 17836-08-7;(CHS)20H+, 17009-82-4; (CH3)30+, 43625-65-6;CH3NCH+,64709-60-0;CH3CNCH3+, 21 963-22-4; HCCCNCH,', 123701-12-2;CH,', 14531-53-4;HSO', 13968-08-6; HCNH', 21 107-92-6;HCCCNH', 76092-42-7.
Metal-to-Ligand Charge-Transfer (MLCT) Photochemistry. Experimental Evidence for the Participation of a Higher Lying MLCT State in Polypyridyl Complexes of Ruthenium(I I ) and Osmium( I I ) Richard S. Lumpkin,t Edward M. Kober,t Laura A. Worl,+ Zakir Murtaza,t and Thomas J. Meyer*qt Department of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27599, and Division of Isotope and Nuclear Chemistry, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (Received: July 25, 1988; In Final Form: May 22, 1989)
Evidence is presented which supports the existence of a thermally accessible, higher lying metal-to-ligand charge-transfer (MLCT) excited state for polypyridyl complexes of Ru(I1) and Os(I1). This state is shown to exist for complexes both in solution and in the solid state. It is typically found to occur at 300-800 cm-l above the lowest lying MLCT state and to have a decay rate constant of lo6-lo8 s-l. This decay rate constant is about 10 times faster than the decay rate constant for the next lower lying MLCT state. The decay process appears to be almost completely nonradiative with the radiative decay rate constant estimated to be lo6 s-l or slower. The higher lying MLCT states can make a significant contribution to nonradiative decay at room temperature.
Introduction The results of extensive studies on the photochemical and photophysical properties of polypyridyl complexes of Ru( 11) and Os(I1) are now a~ailable.I-~At temperatures below 77 K Crosby and co-workers found some time ago that the temperature dependences of excited-state lifetimes (7)and emission quantum yields (&) for [ R ~ ( b p y ) ~(bpy ] ~ + is 2,2'-bipyridine) and related tris-chelates in poly(methy1 methacrylate) matrices could be reasonably well explained by the presence of three low-lying metal-to-ligand charge-transfer (MLCT) excited states whose populations are in thermal equilibrium." At temperatures above 77 K, evidence has been obtained for an additional decay channel or channels either in the solid state or in solution.s8 These latter processes play an important role in determining excited-state properties under ambient conditions, and their origins are the subject of this investigation. One of the processes that appear at higher temperatures has been identified as a thermally activated population of a metalcentered, dd state (or states). It has been suggested that these states lead to the photoinduced ligand loss chemistry that is observed for many polypyridyl complexes of R U ( I I ) ~ ~and ' V ~for certain ~ such carbonyl-containing polypyridyl complexes of O S ( I I ) . ~For complexes, lifetimes in fluid solution are generally highly temperature dependent and over wide temperature ranges can be fit to the kinetic expression shown in eq la. More generally, for decay from two nondegenerate states which are in thermal equilibrium, the temperature dependence is given by eq 1b. Results derived by using eq l a will be distinguishable from those obtained by using eq 1 b only if AE, 2 3kBT.
'University of North Carolina.
Los Alamos National Laboratory.
0022-3654/90/2094-0239$02.50/0
7
For complexes where dd states appear to play a significant role in dictating the temperature-dependent properties of the observed (1) Meyer, T. J. Pure Appl. Chem. 1986,50, 1293. (2) Ferguson, J.; Herren, F.; Krausz, E. R.;Maeder, M.; Vrbancich, J. Coord. Chem. Rev. 1985,64,21. (3) (a) Seddon, K. R.Coord. Chem. Rev. 1982,41,79.(b) Kalyanasundaram, K. Coord. Chem. Rev. 1982,41, 159. (4) (a) Hager, C. D.; Crosby, G. A. J . A m . Chem. SOC.1975,97,7031. (b) Hager, G. D.; Watts, R.J.; Crosby, G. A. J . Am. Chem. SOC.1975,97, 7037. (c) Lacky, D. E.; Pankuch, B. J.; Crosby, G. A. J . Phys. Chem. 1980, 84,2068. (5)(a) Allsopp, S.R.;Cox, A.; Jenkins, S.H.; Kemp, T. J.; Tunstal, S. M. Chem. Phys. Lett. 1976,43,135. (b) Allsopp, S.R.; Cox, A.; Kemp, T. J.; Reed,W. J. J. Chem. SOC.,Faraday Trans. I 1978,74,1275. (c) Allsopp, S . R.; Cox, A.; Kemp, T. J.; Reed,W. J.; Carassiti, V.; Traverso, 0. J . Chem. SOC.,Faraday Trans. 1 1979,75,353. (6) (a) Van Houten, J.; Watts, R. J. J . A m . Chem. SOC.1976,98,4853. (b) Van Houten, J.; Watts, R. J. Inorg. Chem. 1978,17,3381. (7)(a) Wallace, W.M.; Hoggard, P. E.Inorg. Chem. 1980,19,2141.(b) Porter, G.B.; Sparks, R. H. J . Phofochem. 1980,13, 123. (8) (a) Juris, A.; Barigelletti, F.; Balzani, V.; Belser, P.; Von Zelewsky, A. Inorg. Chem. 1985,24, 202. (b) Barigelletti, F.; Juris, A.; Balzani, V.; Belser, P.; Von Zelewsky, A. Inorg. Chem. 1983,22, 3335. (c) Barigelletti, F.;Belser, P.; Von Zelewsky, A.; Juris, A.; Balzani, V. J . Phys. Chem. 1985, 89,3680. (9) (a) Durham, B.; Caspar, J. V.; Nagle, J. K.; Meyer, T. J. J . Am. Chem. SOC.1982,104.4803.(b) Casoar. J. V.: M e w . T. J. J . Am. Chem. Soc. 1983. 105,5583. (c) Caspar; J. V.;Meyer, T. J. inorg. Chem. 1983,22,2444. (d) Pinnick, D. V.; Durham, B. Inorg. Chem. 1984,23,1440. (e) Wacholtz, W. M.; Auerbach, R. A.; Schmehl, R.H.; Ollino, M.; Cherry, W. R. Inorg. Chem. 1985,24, 1758.
0 1990 American Chemical Society
