Photodissociation of Co−C Bond in Methyl- and Ethylcobalamin: An

Apr 17, 2009 - Fax: (502) 852-8149. E-mail: [email protected]., †. University of Silesia. , ‡. Wroclaw University of Technology. , §. University of...
0 downloads 0 Views 5MB Size
Download PDF
6898

J. Phys. Chem. B 2009, 113, 6898–6909

Photodissociation of Co-C Bond in Methyl- and Ethylcobalamin: An Insight from TD-DFT Calculations Piotr Lodowski,† Maria Jaworska,† Tadeusz Andrunio´w,‡ Manoj Kumar,§ and Pawel M. Kozlowski*,§ Department of Theoretical Chemistry, Institute of Chemistry, UniVersity of Silesia, Szkolna 9, PL-40 006 Katowice, Poland; Institute of Physical and Theoretical Chemistry, Department of Chemistry, Wroclaw UniVersity of Technology, 50-370 Wroclaw, Poland; and Department of Chemistry, UniVersity of LouisVille, LouisVille, Kentucky 40292 ReceiVed: NoVember 19, 2008; ReVised Manuscript ReceiVed: February 4, 2009

The mechanism of Co-C bond photodissociation in methylcobalamin (MeCbl) and ethylcobalamin (EtCbl) has been examined by means of time-dependent density functional theory (TD-DFT). The present contribution extends our recent study (J. Phys. Chem. B 2007, 111, 2419-2422) where relevant excited states involved in the photolysis of MeCbl have been identified. To obtain reliable structural models, the high-resolution crystal structure of MeCbl was used as the source of initial coordinates. The full MeCbl was simplified by replacing the corrin side chains by H atoms and the resulting geometry was optimized. The model of EtCbl was generated from the simplified structure of MeCbl by replacing methyl group with ethyl. For both models, the low-lying singlet and triplet excited states have been computed along the Co-C coordinate at TD-DFT/BP86/6-31G(d) level of theory. These calculations reveal that the photodissociation process is mediated by the repulsive 3 (σCo-C f σ*Co-C) triplet state. The overall mechanism of photodissociation for both systems is similar but energetic details are different, reflecting the difference in Co-C bond strength in MeCbl and EtCbl. In both cases the key intermediate involved in Co-C bond photodissociation is identified as first excited state (S1). The S1 intermediate has mixed character: it can be described as predominantly dCo f π*corrin metal-to-ligand charge transfer (MLCT) state with contribution from σ bond to corrin charge transfer (SBLCT) where upon electronic excitation the electron density shifts from the axial NIm-Co-C bonding to corrin ligand. The optimized geometry of the S1 indicates that the structure of the corrin remains essentially unchanged in comparison to ground state (S0). The major structural change occurs in the NIm-Co-C moiety, which becomes bent with elongated Co-C bond in S1 state. Finally, it is proposed that the photolysis of Co-C bond is in line with the mechanism of heme-CO photolysis, where participation of the dFe f π*porphyrin has been suggested. 1. Introduction The cleavage of the cobalt-carbon (Co-C) bond under nonenzymatic conditions has been extensively investigated to address mechanistic issues related to B12-dependent enzymatic catalysis.1 Various experimental techniques were applied to study isolated B12 cofactors, alkylcobalamins with upper axial ligand structurally modified, or model compounds having simplified corrin ring. In particular, photochemistry of B12 cofactors and their analogues have been studied using laserflash2 and continuous wave (CW)3 photolysis, kinetic magnetic field effect (MFE),4,5 chemically induced dynamic electron polarization (CIDEP),6,7 or chemically induced dynamic nuclear polarization (CIDNP).8 The time-resolved spectroscopic studies of Sension and co-workers9–18 provided the detailed analysis of photolysis for methyl- (MeCbl), ethyl- (EtCbl), n-propyl(PrCbl), and adenosylcobalamin (AdoCbl) in different solvents. The results obtained for MeCbl show that its photolysis depends on the excitation wavelength and the lifetime of metastable photoproduct is solvent dependent.18 In contrast, the photolysis * To whom correspondence should be addressed. Tel: (502) 852-6609. Fax: (502) 852-8149. E-mail: [email protected]. † University of Silesia. ‡ Wroclaw University of Technology. § University of Louisville.

of AdoCbl is wavelength independent and the resulting intermediate can have different type of spectrum, depending on solvent.18 Under nonenzymatic conditions, the scission of the Co-C bond can be induced thermally19–22 or photochemically9–18 to generate cob(II)alamin and alkyl radical, i.e., the same products (at least formally) that are formed in enzyme-catalyzed homolysis.23 While the thermally induced cleavage probes only the lowest electronic state, the bond scission generated by light involves manifold of low-lying excited states. Precisely which electronic excited states are involved and what is the mechanism of the Co-C bond photodissociation remains largely unexplained at molecular level. We have recently investigated, by means of time-dependent density functional theory (TD-DFT),24,25 the relevant excited states involved in the photolytic cleavage of Co-C bond in a structural model of MeCbl.26 It was found that photodissociation process is mediated by the repulsive 3 (σCo-C f σ*Co-C) triplet state. The key intermediate involved in photolysis event was identified as first excited state (S1) which has predominantly dCo f π*corrin metal-ligand charge transfer (MLCT) character. In the present theoretical account, the mechanism of photolysis is further explored for MeCbl and EtCbl complexes by means of TD-DFT. For both systems the detailed picture of low-lying excited states as well as their changes along the elongated Co-C coordinate is analyzed in

10.1021/jp810223h CCC: $40.75  2009 American Chemical Society Published on Web 04/17/2009

Co-C Bond in Methyl- and Ethylcobalamin

Figure 1. (Top) Molecular structure of B12 cofactors (R ) Me or Ado, R1 ) CH2CONH2, R2 ) CH2CH2CONH2, R3 ) (CH2)2CONHCH2CH(CH3)OPO3-). (Bottom) Structural models of MeCbl and EtCbl employed in present work and denoted as Im-[CoIII(corrin)]-R+ (R ) Me or Et).

order to simulate the dissociation process. Finally, the geometry and electronic properties of the S1 intermediate are characterized and changes are compared to the S0 state. 2. Computational Details All of the calculations reported in this work were carried out using nonlocal DFT with the nonhybrid Becke-Perdew (BP86)27 functional and the 6-31G(d) basis set (5d components). The use of the BP86 functional has been very successful in predicting structural properties of B12 cofactors,28,29 Co-C bond strength,30,31 and electronically excited states32 as well as reductive cleavage of the Co-C bond in complexes containing corrin33 or phthalocyanine equatorial ligands.34 Several types of time-dependent DFT (TD-DFT) calculations have been carried out to explore electronically excited states of model systems which were used to describe MeCbl and EtCbl (Figure 1). In the present case, TD-DFT calculations have been applied to assign electronic spectra of both complexes. To simulate the dissociation process, the Co-C bond was repeatedly stretched with the step size of 0.05 Å, and the groundstate geometry was reoptimized at every point. At each optimized point the manifold of singlet and triplet states were calculated using the TD-DFT/BP86/6-31G(d) level of theory. In both systems the key intermediate involved in Co-C bond photolysis was identified as first excited state (S1). Its structure was optimized and analyzed from electronic point of view. All calculations have been performed using the Gaussian 0335 and Turbomole36 suite of programs for electronic structure calculations. The Cartesian coordinates of all optimized structures used in current work can be found in the Supporting Information (Tabels S1-S4). 3. Results and Discussion 3.1. Structural Models of MeCbl and EtCbl. The simplified complexes of methyl- and ethylcobalamins were employed as

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6899 structural models in the present theoretical quest. To obtain reliable models, the high-resolution crystal structure of MeCbl37 was used as the source of initial coordinates. The full MeCbl was simplified by replacing the corrin side chains by H atoms (Figure 1) and the resulting geometry was optimized. The model of EtCbl was generated from the simplified structure of MeCbl by replacing methyl group with ethyl. Previous theoretical studies have demonstrated that models based on naked corrin are capable of accurately describing electronic and spectroscopic properties of B12 cofactors. Tables 1 and 2 summarize some of the most relevant geometrical parameters of the optimized structures where selected bond lengths, bond angles, and dihedral angles are listed for MeCbl and EtCbl models. The structure of the corrin ring is very robust and changes very little with the nature of the ligands. The optimized Co-C bond lengths in Im-[CoIII(corrin)]R+ (R ) Me or Et, Figure 1) computed at BP86/6-31G(d) (5d components) level of theory are 1.968 and 1.999 Å for R ) Me and Et, respectively. The calculated bond lengths between cobalt and axial ligands are compared with the experimental values for MeCbl and CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide which has imidazole as the lower axial ligand. The optimized Co-C bond length of 1.968 Å is in good agreement with the corresponding distance of 1.972(8) Å found in the crystal structure of CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide38 and 1.979(4) Å in MeCbl.37 The calculated Co-NIm bond lengths are found to be 2.132 and 2.160 Å, for methyl and ethyl upper axial ligands, respectively. The value calculated for MeCbl model can be compared with experimental bond lengths of 2.093(6) Å for CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide38 and 2.163(4) Å for MeCbl.37 The calculated value is between these two experimental ones. BP86/6-31G(d)optimized bond distances for both complexes (Tables 1 and 2) are in line with well-established correlation for the octahedral Co σ-alkyl compounds known as inverse trans influence:39 in Im-[CoIII(corrin)]-Et+ both axial bonds are longer than those in Im-[CoIII(corrin)]-Me+. As expected, the optimized Co-C bond length is noticeably longer in ethyl-(EtCbl) than in analogous methyl-cobalt (MeCbl) complex. The Co-C bond dissociation energy (BDE) is predicted to be 37.0 kcal/mol for the MeCbl model system, upon inclusion of the zero-point vibrational energy (ZPE) correction, which is consistent with experimental values of 37 ( 3 kcal/mol (based on thermolysis40) and 36 ( 4 kcal/mol (employing calorimetric measurements41) respectively. For ethyl analogue the dissociation energy is lowered by ∼5 kcal/mol and its ZPE corrected value is found to be 32.0 kcal/mol. 3.2. Electronically Excited States of MeCbl and EtCbl. The simulated spectra were computed at the optimized groundstate equilibrium geometry employing the TD-DFT/BP86/631G(d) level of theory as shown in Figure 2. For both complexes the manifold of low-lying electronically excited states was computed to cover the spectral window up to 350 nm (∼5 eV). The lowest singlet and triplet electronic transition for MeCbl model important from photodissociation point of view are collected in Tables 3 and 4. The all calculated singlet and triplet electronic excitations can be found in Supporting Information (Tables S7-S10). In addition, molecular orbital analysis (MO diagram) together with the MO’s composition for MeCbl and EtCbl model systems are sketched in Figure S3 and Tables S5 and S6, respectively. By analyzing the occupied and unoccupied molecular orbitals, one can discern the important interactions between molecular fragments in MeCbl. The strongest cobalt-corrin interaction

6900

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Lodowski et al.

TABLE 1: Selected geometrical parameters and Mulliken charges on atoms or molecule fragments for Im-[CoIII(corrin)]-Me+ Geometrical parameters r (Å) Co-CMe Co-NIm Co-N21 Co-N24 Co-N22 Co-N23

Bond angles

S0

S1

1.968a 2.132b 1.871 1.867 1.933 1.932

1.983 2.034 1.876 1.870 1.970 1.975

CMe-Co-NIm N21-Co-N22 N22-Co-N23 N23-Co-N24 N24-Co-N21 Co-N21-C1 Co-N22-C9 Co-N23-C11 Co-N24-C19

Torsion angles

S0

S1

177.2 91.1 95.0 91.3 82.7 116.7 124.8 124.8 116.6

164.7 90.4 96.3 90.7 82.7 116.7 122.9 122.8 116.5

N21-N22-N23-Co N21-N22-N23-N24 Co-N22-C9-C10 Co-N22-C6-C5 Co-N23-C14-C15 Co-N21-C1-C19 NIm-Co-C10-CMe

S0

S1

-1.8 -4.2 1.1 -7.7 -1.5 33.2 179.1

-0.8 -4.0 -1.3 -6.5 -0.4 32.5 -178.5

Mulliken Charges COSMO Co CMe Me Im Corr

qS0

qS1

∆q (qS1 - qS0)

qS0

qS1

∆q (qS1 - qS0)

0.539 -0.576 -0.002 0.207 0.256

0.553 -0.569 0.098 0.293 0.056

0.014 0.007 0.100 0.086 -0.200

0.529 -0.589 -0.043 0.233 0.281

0.547 -0.582 0.049 0.323 0.081

0.019 0.007 0.092 0.090 -0.201

Excitation Energy (ES1 - E0(S1))c COSMO nm eV

795.0 1.56

779.9 1.59

a The experimental values: 1.972(8) Å for CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide from ref 38, 1.979(4) Å for methylcob(III)alamin from ref 37. b The experimental values: 2.093(6) Å for CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide from ref 38, 2.163(4) Å for methylcob(III)alamin from ref 37. c E0(S1) ) energy of the S0 state in the S1 geometry.

TABLE 2: Selected geometrical parameters and Mulliken charges on atoms or molecular fragments for Im-[CoIII(corrin)]-Et+ Geometrical Parameters r (Å) Co-CEt Co-NIm Co-N21 Co-N24 Co-N22 Co-N23

Bond angles

S0

S1

1.999 2.160 1.871 1.867 1.934 1.929

2.036 2.031 1.872 1.875 1.967 1.973

CEt-Co-NIm N21-Co-N22 N22-Co-N23 N23-Co-N24 N24-Co-N21 Co-N21-C1 Co-N22-C9 Co-N23-C11 Co-N24-C19

Torsion angles

S0

S1

175.7 90.9 95.0 91.5 82.7 116.7 124.8 124.8 117.0

164.1 90.5 96.2 90.9 82.5 116.8 123.1 123.0 116.7

N21-N22-N23-Co N21-N22-N23-N24 Co-N22-C9-C10 Co-N22-C6-C5 Co-N23-C14-C15 Co-N21-C1-C19 NIm-Co-C10-CEt

S0

S1

-2.4 -3.5 0.0 -8.0 -1.2 33.4 177.5

-1.2 -3.1 -2.0 -6.4 -0.3 33.3 179.5

Mulliken Charges COSMO Co CEt Et Im Corr

qS0

qS1

∆q (qS1 - qS0)

qS0

qS1

∆q (qS1 - qS0)

0.549 -0.371 0.031 0.203 0.218

0.571 -0.365 0.139 0.293 -0.003

0.023 0.006 0.108 0.090 -0.221

0.536 -0.372 -0.012 0.229 0.247

0.562 -0.366 0.088 0.322 0.028

0.026 0.006 0.100 0.093 -0.219

Excitation Energy (ES1 - E0(S1))a COSMO nm eV a

835.4 1.48

818.8 1.51

E0(S1) ) energy of the S0 state in the S1 geometry.

occurs between the free electron pairs of corrin nitrogens and the unoccupied dx2-y2 orbital (σ donation from corrin to cobalt).

The respective antibonding orbital is mixed with a π* orbital (LUMO+1, LUMO+2). The bonding orbital is low in energy,

Co-C Bond in Methyl- and Ethylcobalamin

Figure 2. Simulated electronic absorption spectrum of MeCbl (lower panel) and EtCbl (upper panel) based on the first few singlet excited states (required to complete the spectral window up to 350 nm) computed at the optimized ground-state equilibrium geometry of Im[CoIII(corrin)]-R+ (R ) Me or Et) employing the TD-DFT/BP86/631G(d) level of theory. Arrows indicate laser excitations at 520 and 400 nm, respectively.

and it is not shown in Figure S3. We can also find evidence of cobalt-corrin back-donation in HOMO and LUMO orbitals. This is an interaction between doubly occupied dyz and unoccupied π* orbitals. Orbital HOMO-6 is a σ bonding orbital between Co and the methyl group, which is also antibonding with respect to imidazole nitrogen. The relevant antibonding orbital is LUMO+3. Some admixture of σ bonding character can also be seen in HOMO-7 and HOMO-1 orbitals. The distinctive absorption features of the cobalt corrinoids can be attributed to a manifold of low-lying electronic excited states whose energies are sensitive to the nature of the axial ligands.42–48 The electronic excitations can be theoretically probed using TD-DFT as demonstrated in a recent theoretical analysis of the electronic spectra of MeCbl and its derivative in which the trans axial base was replaced by a water molecule (MeCbi-H2O).32 The electronic excitations strongly depend on the choice of the functional and for that reason the hybrid B3LYP and the gradient-corrected BP86 functionals have been examined in detail for MeCbl model systems. It was concluded that photochemistry of B12 cofactors (in which the Co-C bond is cleaved) is best described by BP86 functional. A more formal argument can be made from direct comparison of results obtained using different functional and using different level of theory for example CASSCF/CASPT2.49 For the sake of convenience the main results obtained using BP86 functional will only be briefly summarized. In both cobalamins the low-energy part of the electronic spectrum is commonly referred to as R/β band. In the experimental spectrum of MeCbl there are several peaks found in the R/β band, which have been interpreted as a vibrational progression of two distinct electronic transitions.15,50 The origin of this vibrational progression is still not fully understood, but most likely it is due to the change of geometry along normal modes of the corrin -CdCstretching vibrations that occurs upon the π f π* electronic excitation which contributes to the observed vibronic structure. TD-DFT calculations only partially support this assignment: the B3LYP-simulated spectrum has one intense transition in the R/β part, while there are three such transitions in the BP86 spectrum.32 Commonly used B3LYP functional describes the

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6901 lowest energy transitions as mainly of d/π f d character (only the first one is a π/d f π* transition). On the other hand in case of BP86 functional the first three transitions are of d/π f π* character and several subsequent ones are of d/π f π*/d type. The lowest energy singlet states S1-S4 are transitions from the highest occupied molecular orbitals (HOMO-3-HOMO) to the LUMO orbital. These transitions can be assigned to the R/β part of the spectrum. The relevant molecular orbitals involved in these transitions are presented in Figure 3. The lowest energy transitions (S1) is found at 566 nm in the calculated spectrum of MeCbl and it has a very small oscillator strength. It comes from HOMO-1 and HOMO-2 to LUMO excitations. The LUMO orbital is basically a π* orbital of corrin. HOMO-1 is a mixture of π orbital and d orbitals of dxz and dz2 character. It also extends to the methyl group. The HOMO-2 orbital is mainly dxz orbital of cobalt with some admixture of a π orbital of corrin. HOMO-1 and especially HOMO-2 have significant d contribution (see Table S5) and hence the S1 state can be classified as d f π* (MLCT) transition. Since HOMO-1 has the participation of dz2 and CH3 fragment, this transition can also be described as partly coming from the excitation of σ orbital. Sometimes these types of electronic transitions have been described in the literature as σ bond-ligand CT transition (SBLCT) (see, for example, studies of Daniel51 and Vlcek52). The next two transitions S2 and S3 calculated at 528 and 499 nm have significant oscillator strengths. The transition at 528 nm comes from HOMO to LUMO excitation and the transitions at 499 nm mainly from HOMO-2 to LUMO excitation. The HOMO orbital is the dyz orbital with admixture of π. Because HOMO and HOMO-2 have mixed character, these two transitions can be characterized as d/π f π* type. The S4 transition corresponds to the HOMO3fLUMO excitation and is of pure d f π* type. The higher energy part of the spectrum, up to 400 nm, belongs to so-called D/E region in cobalamins which has lower intensity than R/β band. Two electronic transitions observed at 427 and 400 nm in the MeCbl spectrum have been assigned to the transitions affiliated with D/E band. These transitions have dCo, πcorrin f dCo, and π*corrin character. The other transitions calculated in this region at 413, 388, and 382 nm with small oscillator strengths are of similar type, though excitations at 388 and 382 nm have transitions from dCo, πcorrin f σ*Co-C. Calculated electronic transitions can be compared with experimental data for CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide and MeCbl (Table 3 and Table S7). The imidazolylcobamide exhibited UV/VIS spectroscopic data typical for methylcobamides, and location of typical bands is very similar for both compounds. The similarity of experimental electronic spectra confirms the validity of the structural model used in the calculations. The agreement of the calculated spectrum with the experimental data shows that the BP86 functional properly describes the electronic transitions of MeCbl. In addition, the manifold of triplet excited states was also calculated. Since triplet excitations based on singlet groundstate wave function have zero transition dipole moment, they were not included in the simulated spectrum. Three of them lie energetically below first singlet excited-state S1 (566 nm). The lowest triplet state at 758 nm (T1) is a HOMO f LUMO transition, and the next two states, T2 and T3, are transitions from HOMO-1 and HOMO-2 to LUMO located at 655 and 594 nm, respectively. The molecular orbitals of EtCbl resemble those of MeCbl (Figure 3 and Figure S3, Table S6). Also, the relative position of MOs is very similar to that which occurs in the MeCbl. An

6902

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Lodowski et al.

TABLE 3: Ten Lowest Singlet States for Im-[CoIII(corrin)]-Me+ expt λ (nm) (E (eV)) lower axial ligand E (eV)

λ (nm)

f

%

character

S1

2.18

566.4

0.0005

S2 S3

2.34 2.48

528.8 499.6

0.0251 0.0513

S4 S5 S6

2.58 2.90 3.00

479.4 427.4 413.0

0.0124 0.0208 0.0033

S7

3.09

400.9

0.0131

S8

3.18

388.9

0.0049

S9

3.24

382.1

0.0038

S10

3.27

378.1

0.0039

22 72 72 12 54 16 84 68 46 12 12 12 8 58 22 26 24 20 16 16 38 84

HOMO-2(d,π) f LUMO(π*) HOMO-1(π,d) f LUMO(π*) HOMO(d,π) f LUMO(π*) HOMO-3(d) f LUMO(π*) HOMO-2(d,π) f LUMO(π*) HOMO-1(π,d) f LUMO(π*) HOMO-3(d) f LUMO(π*) HOMO(d,π) f LUMO+1(d, π*) HOMO-1(π,d) f LUMO+1(d, π*) HOMO-1(π,d) f LUMO+2(π*,d) HOMO(d,π) f LUMO+1(d, π*) HOMO(d,π) f LUMO+2(π*,d) HOMO-1(π,d) f LUMO+1(d, π*) HOMO(d,π) f LUMO+2(π*,d) HOMO-4(π,d) f LUMO(π*) HOMO-2(d,π) f LUMO+1(d, π*) HOMO-1(π,d) f LUMO+2(π*,d) HOMO(d,π) f LUMO+3(σ*) HOMO-4(π,d) f LUMO(π*) HOMO-2(d,π) f LUMO+2 (π*,d) HOMO(d,π) f LUMO+3(σ*) HOMO-5(πIm) f LUMO(π*)

a

DBIa

imidazoleb

527 (2.35) 486 (2.55)

528 (2.35) 480 (2.58)

429 (2.89)

428 (2.90)

401 (3.09)

The experimental values for MeCbl from ref 42. b The experimental values for CoR-(1H-imidazol-1-yl)-Coβ-methylcob(III)amide from ref

38.

TABLE 4: Ten Lowest Triplet States for Im-[CoIII(corrin)]-Me+ E (eV)

λ (nm)

%

character

T1 T2 T3 T4

1.63 1.89 2.08 2.24

757.9 654.5 594.0 551.8

T5

2.36

523.2

T6 T7

2.44 2.56

507.0 483.0

T8

2.65

467.5

T9

2.67

462.9

T10

2.80

441.7

58 53 52 39 19 22 10 15 43 36 23 12 13 7 13 8 11 8 20 30 8 8

HOMO(d,π) f LUMO(π*) HOMO-1(π,d) f LUMO(π*) HOMO-2(d,π) f LUMO(π*) HOMO(d,π) f LUMO+1(d, π*) HOMO(d,π) f LUMO+2(π*,d) HOMO-2(d,π) f LUMO+1(d, π*) HOMO-2(d,π) f LUMO+2(π*,d) HOMO-1(π,d) f LUMO+1(d, π*) HOMO-3(d) f LUMO(π*) HOMO-3(d) f LUMO+1(d, π*) HOMO-3(d) f LUMO+2(π*,d) HOMO-1(π,d) f LUMO+1(d, π*) HOMO-1(π,d) f LUMO+2(π*,d) HOMO(d,π) f LUMO+1(d, π*) HOMO(d,π) f LUMO+2(π*,d) HOMO-1(π,d) f LUMO+1(d, π*) HOMO-1(π,d) f LUMO+2(π*,d) HOMO(d,π) f LUMO+1(d, π*) HOMO(d,π) f LUMO+2(π*,d) HOMO-2(d,π) f LUMO+1(d, π*) HOMO-2(d,π) f LUMO+2(π*,d) HOMO-1(π,d) f LUMO+1(d, π*)

important difference can be found in the case of a π orbital of cobalt-ethyl bond (HOMO-6). As can be seen from Figure S3, this orbital energy is much higher as compared to the corresponding orbital in the MeCbl. The singlet and triplet excited states of EtCbl are shown in Tables S9 and S10. The excited states of EtCbl are slightly lower in energy than those of MeCbl, but they are identical in character. 3.3. Mechanism of Photolysis. To simulate the dissociation process of MeCbl and EtCbl, the Co-C bond was subjected to repeated elongation with an increment of 0.05 Å at each point. The ground-state geometry was reoptimized at each point, and the vertical excitation energies were calculated using BP86/631G(d) level of theory. For each optimized geometry we

TABLE 5: Energy Differences between Minimum of Singlet Excited States and the Intersection Point with Dissociative Triplet State of 3(σCo-C f σ*Co-C) (Interpolated from Figure 4) ∆E (kcal/mol) S1 S2 S3 S4 S5

Im-[CoIII(corrin)]-Me+

Im-[CoIII(corrin)]-Et+

4.11 3.32 1.40 1.80 0.15

2.18 1.76 0.25 0.74 0.02

calculated the energies of relevant singlet and triplet excited states. The resulting potential energy profiles are displayed in Figure 4. The triplet excited state with a large participation of 3 (σCo-C f σ*Co-C) excitation is marked with a black line. The manifold of low-lying excited states qualitatively looks very similar for both systems but is different quantitatively. The energy curves corresponding to EtCbl are shallower and energetically lower than those representing MeCbl (Figure 4). Table 5 provides more detailed comparison between these two systems. The energy difference between the minimum on the calculated energy curve of the excited state and the intersection point of this curve with the triplet 3(σCo-C f σ*Co-C) state was estimated for several states of MeCbl and EtCbl. On average the energy difference is lowered by a factor of 2 for first few lowest singlet states. It may be noted that these barriers are estimated on the basis of excited-state minima and crossing points which were not optimized. However, their relative values are meaningful. The mechanism of Co-C bond photolysis in MeCbl and EtCbl can be elaborated based on energy curves sketched in Figure 4. The photolysis process is initiated with excitation by light of either 520 nm near maximum of R/β band or 400 nm near upper limit of D/E just below γ band (Figure 2).13 According to TD-DFT calculations, the S1-S4 singlet excited states belong to the R/β band while S5-S10 singlet states to the

Co-C Bond in Methyl- and Ethylcobalamin

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6903

Figure 3. Isosurface plots of selected frontier MOs for (a) MeCbl and (b) EtCbl, based on the BP86/6-31G calculations. In addition, Figure S3 in the Supporting Information provides detailed MO analysis for both the model compounds under investigation.

Figure 4. Potential energy curves of the lowest-excited singlet (red) and triplet (blue) states of the Im-[CoIII(corrin)]-R+ (R ) Me or Et) complexes along the Co-C bond stretch computed at TD-DFT/BP86/6-31G(d). The black line corresponds to the triplet excited state with a large participation of 3(σCo-C f σ*Co-C) excitation.

D/E band (Figure 4; see also Table S7, Supporting Information), respectively. The low-lying S1-S4 states have non-repulsive character and S2-S3 cross at ∼2.5 Å while S4 and S5 states display an avoided crossing at ∼2.3 Å. For EtCbl S2-S3 crossing occurs at ∼2.6 Å, while S4 and S5 exhibit avoided crossing at ∼2.35 Å. The higher states S5-S10 display multiple crossings, even at small Co-C distances. In both systems the

low-lying electronically excited states possess bound character and therefore the mechanism of photolysis which directly involves singlet excited states leading to the methyl or ethyl radical and excited cob(II)alamin cannot be considered as valid. TD-DFT calculations clearly indicate that the singlet state with 1 (σCo-C f σ*Co-C) character does not participate in the photolysis process. There are two reasons for that; first is that the

6904

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Lodowski et al.

Figure 5. Relevant excited states involved in Co-C bond scission corresponding to 520 nm (left panel) and 400 nm (right panel) laser excitations.

Figure 6. Potential energy surfaces representing the lowest triplet states plotted as a function of axial bond lengths (expressed in Å).

state possessing such a character is energetically very high and none of the 30 calculated singlet states have dominant contribution from 1(σCo-C f σ*Co-C) excitation. Moreover, such a state would not dissociate to cob(II)alamin and •CH3, but rather to the ionic fragments. The resulting conclusion can be drawn from a more accurate picture obtained for pentaamminemethylcobalt complex, CoCH3(NH3)52+, employing CASSCF/CASPT2 level of theory.53 On the other hand, the triplet state, which has 3(σCo-C f σ*Co-C) character, can lead to a proper cleavage. This state has a repulsive character and in dissociation limit gives proper products, i.e., cob(II)alamin and •CH3. Such state, denoted with a black line and labeled as 3(σCo-C f σ*Co-C) in Figure 4, was obtained by connecting triplet transitions with significant contributions of excitations from occupied orbitals of σCo-C character to unoccupied orbitals of σ*Co-C nature. For both models at the equilibrium geometry, the 3(σCo-C f σ*Co-C) state is quite high in energy, but its energy drops significantly above ∼2.4 Å. This might be due to the reason that at longer Co-CMe or Co-CEt bond distances this state is not properly described

by wave function which has one-determinant character. This distance is actually shorter than 2.65 Å at which the instability in the ground-state wave function occurs and above which the UHF type description is required. Fortunately, the photochemical events occur at much shorter distances, where the 3(σCo-C f σ*Co-C) state behaves properly. The dissociative 3(σCo-C f σ*Co-C) state becomes the lowest energy state above Co-C bond distance of 2.35 Å. At this point it displays an avoided crossing with the T1 state (Figure 4) which arises from the HOMOfLUMO (dCo, πcorrin f π*corrin) excitation. The presence of the avoided crossing between T1 and 3 (σCo-C f σ*Co-C) does not preclude the conical intersection between these two states that may exist along some other coordinate, e.g., like change in geometry of the corrin ring. The solution to this problem requires a multireference wave function and analysis of excited states potential energy surfaces using energy derivatives. By employing TD-DFT framework we were able to further characterize this point by optimizing the geometry of the lowest triplet state. Then we computed low-lying excited triplet states employing the BP86/6-31G(d) level of theory. It was found that the third triplet state (designated as T3) here represents the dissociative state (σCo-C f σ*Co-C). When the Co-C bond is short (around 2.36 Å), this dissociative state undergoes avoided crossing with T2 state and upon further elongation to 2.43 Å, T3 state becomes the lowest lying triplet state. The analysis of spin densities affiliated with corrin and cobalt moieties provides further insight into the mechanism of photolytic cleavage of Co-C bond (Figure 6). As long as the Co-C bond length remains up to 2.43 Å, the whole electronic density stays on corrin macrocycle; only at very large Co-NIm bond distances (around 2.50 Å) there is electronic density transfer from equatorial ligand to the Co-C bond. At this point, T3 and T1 states are widely separated. But as the Co-C bond is lengthened beyond 2.43 Å, T3 and T1 states exhibit avoided crossing and the electronic density swiftly moves from corrin moiety into Co-C axial plane which can be seen from the spin density profiles of corrin and cobalt shown in Figure 7. The present TD-DFT calculation can provide a rationale why the photolysis of MeCbl is wavelength dependent. It has been

Co-C Bond in Methyl- and Ethylcobalamin

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6905

Figure 8. Superposition of S0 and S1 states corresponding to (a) MeCbl and (b) EtCbl, respectively. Figure 7. Spin density (shown along z-axis) of corrin (a) and cobalt (b) entities in the lowest T1 state plotted as a function of axial bond lengths (expressed in Å).

demonstrated by Sension and co-workers13 that the excitation of MeCbl results either in direct bond homolysis or in rapid internal conversion and population of the S1 excited state. At short-wavelength excitation (400 nm), only 25% of MeCbl results in bond homolysis while at 520 nm (R/β band) the whole MeCbl results in formation of the S1 excited state. For MeCbl this state has a lifetime of 1 and 2.4 ns in water and ethylene glycol, respectively. From the S1 state only 14% of MeCbl undergoes photolysis, while the rest of the MeCbl reverts back to the ground state. For EtCbl the S1 state has a lifetime of 48 ps and the complete photolysis occurs from this state. The electronic spectrum of S1 state resembles spectra of cob(III)alamins with non-alkyl axial ligands.13 Based on energy curves plotted along the Co-C bond (see Figure 5), much deeper insight can be obtained into the wavelength dependency of the photolysis of MeCbl. At excitation of 520 nm no prompt photolysis is observed, instead a longlived (∼1 ns) metastable product is formed, with the electronic spectrum similar to cob(III)alamin species, with a weakly bound axial ligand.13 Subsequently, this product in part (∼14%) undergoes photolysis, and the remaining part reverts to the ground state. According to TD-DFT calculations the laser excitation at 520 nm leads to the excited state in the R/β band, presumably S2 or S3, because of large oscillator strengths (Figure

5 left panel). The S1-S3 states are very close in energy and fast transition to the lowest one, i.e., S1, is most likely. From this state the photodissociation may occur through inter system crossing (ISC) with the dissociative 3(σCo-C f σ*Co-C) state. The relaxation to the ground state can take place by a radiationless transition from S1 (most probably through a conical intersection involving some energy barrier). Crossing of the two states, S1 and dissociative 3(σCo-C f σ*Co-C), involves a small energy barrier (Table 5), which may provide an explanation for the long lifetime of the metastable state. From S1 state only 14% of MeCbl forms photolysis product, the remaining part, through internal conversion, goes back to S0. The internal conversion barrier in water is measured to be 10.5 kcal/mol.18 It means that intersystem crossing barrier is larger than 10.5 kcal/mol, hence the estimated barrier of 4.1 kcal/mol is too small. On the other hand, spin-orbit coupling may also play a role in this process. From this perspective, S1 state would be a likely candidate for a long-lived transient species detected in the photolysis of MeCbl. The triplet state has not been noticed during the experimental study concerning the photochemical internal conversions. One may surmise that the internal conversion from the dissociative 3(σCo-C f σ*Co-C) state to T1 or T2 may occur. If such a triplet state was formed in small amounts, then considering the fact that T1 state is quite low in energy (1.63 eV, see Table 4) there should be a low-barrier deactivation path to S0 (most probably with a much smaller barrier than that for S1). The excitation at 400 nm leads in part (about 25%) to a prompt photolysis and in remaining part to the formation of

6906

J. Phys. Chem. B, Vol. 113, No. 19, 2009

Lodowski et al.

Figure 9. Molecular orbital diagram for MeCbl in the S0 and S1 optimized geometry.

the long-lived metastable state. The energy diagram (Figure 4) shows that S5-S10 crosses with 3(σCo-C f σ*Co-C) state practically without any energy barrier. This explains the prompt photolysis at excitation to the high part of the D/E spectrum. Alternatively, the consecutive transitions from an excited state in this region to the lowest energy singlet S1 state may take place, which again leads to long-lived transient product. According to magnetic measurements on a model complex reported by Grissom and co-workers,8 the lowest triplet state from which photolysis occurs should be excluded as possible metastable product. Since the singlet states are energetically close, such transitions are very likely and they can be very fast. It is possible that there might be intersections between these states along some coordinate other than the Co-C distance, e.g., corrin distortion, which would make such internal conversion process very efficient. The comparison of the energy curves for S1 state of MeCbl and EtCbl (Figure 4) shows that this state for EtCbl has shallower minimum than MeCbl. The photolysis from S1 state takes place by intersystem crossing with the triplet state of dissociative character 3(σCo-C f σ*Co-C) and this intersystem crossing occurs at Co-C bond length of 2.2 Å for both systems under investigation (Figure 4). The estimated energy difference between the minimum of S1 state and the crossing point of S1 with 3(σCo-C f σ*Co-C) state is 4.1 kcal for MeCbl and 2.2 kcal for EtCbl (Table 5). It shows that energy barrier for intersystem crossing is smaller for EtCbl, which explains the complete photolysis from this state for EtCbl, and only partial for MeCbl. In addition, the spin-orbit interaction can also influence the intersystem crossing, but such effect was not probed in the present account.

3.4. Electronic and Structural Properties of S1 Intermediate State. The key metastable photoproduct involved in the Co-C bond photolysis has been identified as an S1 state and is described in the literature as having predominantly the dCo f π*corrin metal-to-ligand charge transfer (MLCT) character (see, for example, recent study of Sension and co-workers18 and references therein). To obtain a more realistic picture of this intermediate, the S1 state was optimized for MeCbl and EtCbl model compounds. Figure 8 depicts results for both cobalamins where the resulting optimized geometry of metastable photoproduct (S1) was superimposed upon that of the ground state (S0). More detailed description in terms of optimized structural parameters is given in Tables 1 and 2, respectively. The S1 state geometry of MeCbl is characterized by a slightly longer Co-CMe bond (lengthened by 0.015 Å) and shorter Co-NIm bond (contracted by about 0.1 Å). The axial NIm-Co-CMe arrangement, which is almost linear in S0 (177°), is bent in S1 (164°). The geometry of corrin ring does not change much and the dihedral angles are close to those in S0 state (Table 1). The energy difference between S1 and S0 is 1.56 eV. From S1 state internal conversion occurs to the ground state and the barrier for this process was measured at 10.5 kcal/mol (44 kJ/mol) in water and 7.2 kcal/mol (30 kJ/mol) in ethylene glycol.18 The larger energy barrier in more polar solvent indicates that the S1 state has a charge-transfer character. In order to understand to what extent solvation effects may influence properties of S1 state, additional calculations were performed in presence of solvent (i.e., water) modeled via COSMO (Tables 1 and 2). All predicted changes were negligible, pointing out that solvent effects do not change the character of the S1 state. A similar analysis for EtCbl model shows that the changes in axial bond distances are more pronounced in the S1 state; the Co-CEt bond is longer

Co-C Bond in Methyl- and Ethylcobalamin

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6907

Figure 10. Electron density difference between the S0 and S1 states for MeCbl and EtCbl: (a) isosurface plot value of 0.003 and (b) cross-section contour plot along the axial bonding (the section plane is shown below).

by 0.037 Å relative to the ground state while the Co-NIm bond is 0.129 Å shorter (Table 2). The NIm-Co-CEt bending angle is similar to that in MeCbl. The relatively elongated Co-CEt bond in S1 of EtCbl indicates that this state is more weakly bound than the S1 state of MeCbl, and the dissociation from this state will be more facile. The energy difference between S1 and S0 is 1.48 eV for EtCbl, which is slightly lower than that for MeCbl. From the S1 state the complete photodissociation of Co-CEt bond occurs. The difference in the properties of S1 state of MeCbl and EtCbl and in their photolytical reactivity can be attributed to the different Co-C bond strengths in MeCbl and EtCbl complexes, respectively. In case of MeCbl system, the optimized Co-C bond length is 1.968 Å while its BDE (including its ZPE correction term) is found to be 37.0 kcal/ mol. For EtCbl model compound, the optimized Co-C bond distance is 1.999 Å and its ZPE-corrected BDE is 32.2 kcal/ mol. In Figure 9 the molecular orbital diagram for MeCbl at S0 and S1 geometries is presented. The respective diagram for EtCbl is very similar (Figure S4). In the S0 geometry HOMO and HOMO-1 are very close in energy. The HOMO-1 is of π/d character and HOMO is of d/π type. The lowest vertical transition is mainly a HOMO-1 f LUMO transition. S1 state at its lowest energy geometry arises from HOMO f LUMO excitation. Coming from S0 to S1 geometry the two highest occupied orbitals change order while the character of the lowest electronic transition basically remains the same. However, the

participation of dCo orbital in HOMO increases from 23% to 41%, while at the same time the participation of πcorrin orbital diminishes from 61% to 40% (Figure 9). It can be seen that the MLCT character of S1 state is more pronounced in the S1 geometry than in the ground-state geometry. Figure 10 gives the pictorial description of electron density difference between the ground-state S0 and first excited singlet state S1. Clearly, the density difference reflects the nature of the frontier HOMO and LUMO orbitals of the ground state at the S1 geometry for MeCbl and EtCbl, respectively. The plot of electronic density difference between S0 and S1 in the geometry of S1 reveals that the electron density decreases on metal atom (Co) as well as along NIm-Co-R axis (R ) Me or Et) while it increases on corrin. The Mulliken charges of S0 and S1 state (Tables 1 and 2) are in agreement with such a picture. The positive charge on Co increases slightly, while there is substantial enhancement in positive charge density on methyl or ethyl and imidazole, and significant positive charge reduction on corrin. Density difference plots and Mulliken population analysis confirm partial charge transfer from axial NIm-Co-CR bond to corrin. In addition, further analysis also shows that partial charge transfer from axial bonding to corrin is somewhat larger in EtCbl than in MeCbl. Such description of the S1 excited state is in agreement with the experimentally observed sensitivity with respect to the polarity of the solvent. However, present TDDFT calculations only partially support previous assignment that the S1 intermediate state is of MLCT character. Closer inspection

6908

J. Phys. Chem. B, Vol. 113, No. 19, 2009

of electron density changes points out that upon electronic excitation density is not solely shifted from the cobalt but from the axial NIm-Co-CR bonding to corrin ligand. These types of electronic transitions are described as σ bond-ligand CT transition (SBLCT).51,52 Taking into account that MLCT component is important in this electronic transition, the most appropriate way is to describe this state as having mixed MLCT/ SBLCT character. Clearly, this state cannot be described as Co(III)-CH3- (d f CH3) as has been suggested in ref 18. The latter one is of very high energy as shown for simple cobalt-methyl complex.50 4. Summary and Conclusions This work is among the first few studies where DFT and TDDFT have been applied to investigate photolysis of Co-C bond in cobalamins. Based on these calculations, a mechanism of MeCbl and EtCbl photolysis is proposed according to which after excitation to R/β band (520 nm excitation wavelength) the internal conversion leads to the lowest excited-state S1. From this state, two paths are possible: one is photolysis, and the other is internal conversion from S1 to the ground state. The photolysis occurs by an intersystem crossing with a dissociative 3(σCo-C f σ*Co-C) triplet state. There is a barrier in this process which is smaller for EtCbl than for MeCbl. This difference explains why MeCbl undergoes partial photolysis (most of MeCbl reverting back to the ground state) while in case of EtCbl, photolysis is the main process. Excitation to the D/E band with 400 nm wavelength gives 25% photolysis products, and the rest forms S1 state. The prompt photolysis is a result of a practically barrierless crossing of high lying singlet states with the dissociative 3(σCo-C f σ*Co-C) triplet state. The reversion to the long-living S1 state results from a path of consecutive internal conversion between singlet states. The geometry of the S1 state was optimized for both systems and its comparison with the ground-state points out that the changes in corrin ring are rather small. The largest changes are related to bending angle NIm-Co-CR (R ) Me or Et) while the Co-CR bond length is elongated by ∼0.02 and 0.04 Å for MeCbl and EtCbl, respectively. While the Co-C coordinate is a driving force for the photolysis process, the similar Co-CR bond lengths in the S0 and S1 states indicate that other coordinates, for example alkyl bending, may play a role in the S1-S0 internal conversion. The S1 intermediate has mixed character: it can be described as predominantly dCo f π*corrin MLCT state with contribution from SBLCT where upon electronic excitation the electron density shifts from the axial NIm-Co-C bonding to corrin ligand. This is in agreement with experimental findings which reveal a Co(III) character in the transient electronic spectrum and stabilization of S1 state in the polar solvent. This state may be compared to carboxyhemoglobin54 where experimentally it was shown that the photolysis of CO occurs from d f π* excited state. Recently, this finding was also confirmed by theoretical calculations.55,56 Acknowledgment. The TURBOMOLE calculations were carried out in the Academic Computer Centre CYFRONET of the University of Science and Technology in Cracow, ACC CYFRONET AGH, Krako´w, Poland, http://www.cyfronet.pl, under grant No. MNiSW/SGI3700/US´laski/111/200 and MNiSW/ IBM_BC_HS21/US´laski /111/2007 and in the Wroclaw Centre for Networking and Supercomputing, WCSS, Wrocław, Poland, http://www.wcss.wroc.pl, under grant No. 51/96. Supporting Information Available: The Cartesian coordinates generated by the study, the relevant MOs, and TD-DFT

Lodowski et al. results. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) (a) B12; Dolphin, D., Ed.; Wiley-Interscience: New York, 1982. (b) Banerjee, R. Chem. Biol. 1997, 4, 175–186. (c) Ludwig, M. L.; Matthews, R. G. Annu. ReV. Biochem. 1997, 66, 269–313. (d) Kra¨utler, B., Arigoni, D., Golding, B. T., Eds.; Vitamin B12 and B12 Proteins; WileyVCH: New York, 1998. (e) Marzilli, L. G. In Bioinorganic Catalysis; Reedijk, J., Bouwman, E., Eds.; Marcel Dekker: New York, 1999; pp 423468. (f) Banerjee, R. Chemistry and Biochemistry of B12; John Wiley & Sons: New York, 1999. (g) Toraya, T. Cell. Mol. Life Sci. 2000, 57, 106– 127. (h) Banerjee, R. Biochemistry 2001, 40, 6191–6198. (i) Matthews, R. G. Acc. Chem. Res. 2001, 34, 681–689. (j) Banerjee, R.; Ragsdale, S. W. Annu. ReV. Biochem. 2003, 72, 209–247. (k) Banerjee, R. Chem. ReV. 2003, 103, 2083–2094. (l) Toraya, T. Chem. ReV. 2003, 103, 2095–2127. (m) Brown, K. L. Chem. ReV. 2005, 105, 2075–2149. (n) Randaccio, L.; Geremia, S.; Nardin, G.; Wuerges, J. Coord. Chem. ReV. 2006, 250, 1332– 1350. (o) Randaccio, L.; Geremia, S.; Wuerges, J. J. Organomet. Chem. 2007, 692, 1198–1215. (2) (a) Endicott, J. F.; Netzel, T. L. J. Am. Chem. Soc. 1979, 101, 4000– 4002. (b) Chagovetz, A. M.; Grissom, C. B. J. Am. Chem. Soc. 1993, 115, 12152–12157. (c) Lott, W. B.; Chagovetz, A. M.; Grissom, C. B. J. Am. Chem. Soc. 1995, 117, 12194–12201. (d) Chen, E.; Chance, M. R. J. Biol. Chem. 1990, 265, 12987–12994. (e) Endicott, J. F.; Ferraudi, G. J. J. Am. Chem. Soc. 1977, 99, 243–245. (3) Chen, E.; Chance, M. R. Biochemistry 1993, 32, 1480–1487. (4) Grissom, C. B.; Chagovetz, A. M. Z. Phys. Chem. 1993, 182, 181– 188. (5) Natarajan, E.; Grissom, C. B. Photochem. Photobiol. 1996, 64, 286– 295. (6) Sakaguchi, Y.; Hayashi, H.; I’Haya, Y. J. J. Phys. Chem. 1990, 94, 291–291. (7) Rao, D. N. R.; Symons, M. C. R. J. Chem. Soc., Chem. Commun. 1982, 954–955. (8) Kruppa, A. I.; Taraban, M. B.; Leshina, T. V.; Natarajan, E.; Grissom, C. B. Inorg. Chem. 1997, 36, 758–759. (9) Walker, L. A., II.; Jarrett, J. T.; Anderson, N. A.; Pullen, S. H.; Matthews, R. G.; Sension, R. J. J. Am. Chem. Soc. 1998, 120, 3597–3603. (10) Walker, L. A., II.; Shiang, J. J.; Anderson, N. A.; Pullen, S. H.; Sension, R. J. J. Am. Chem. Soc. 1998, 120, 7286–7292. (11) Shiang, J. J.; Walker, L. A., II.; Anderson, N. A.; Cole, A. G.; Sension, R. J. J. Phys. Chem. B 1999, 103, 10532–10539. (12) Yoder, L. M.; Cole, A. G.; Walker, L. A., II.; Sension, R. J. J. Phys. Chem. B 2001, 105, 12180–12188. (13) Cole, A. G.; Yoder, L. M.; Shiang, J. J.; Anderson, N. A.; Walker, L. A., II.; Banaszak Holl, M. M.; Sension, R. J. J. Am. Chem. Soc. 2002, 124, 434–441. (14) Sension, R. J.; Cole, A. G.; Harris, A. D.; Fox, C. C.; Woodbury, N. W.; Lin, S.; Marsh, E. N. G. J. Am. Chem. Soc. 2004, 126, 1598–1599. (15) Sension, R. J.; Harris, A. D.; Stickrath, A.; Cole, A. G.; Fox, C. C.; Marsh, E. N. G. J. Phys. Chem. B 2005, 109, 18146–18152. (16) Sension, R. J.; Harris, A. D.; Cole, A. G. J. Phys. Chem. B 2005, 109, 21954–21962. (17) Shiang, J. J.; Cole, A. G.; Sension, R. J.; Hang, K.; Weng, Y.; Trommel, J. S.; Marzilli, L. G.; Lian, T.; Anderson, J. Am. Chem. Soc. 2006, 128, 801–808. (18) Harris, D. A.; Stickrath, A. B.; Carroll, E. C.; Sension, R. J. J. Am. Chem. Soc. 2007, 129, 7578–7585. (19) Halpern, J.; Kim, S.-H.; Leung, T. W. J. Am. Chem. Soc. 1984, 106, 8317–8319. (20) Garr, C. D.; Sirovatka, J. M.; Finke, R. G. J. Am. Chem. Soc. 1996, 118, 11124–11154. (21) (a) Hay, B. P.; Finke, R. G. J. Am. Chem. Soc. 1986, 108, 4820– 4829. (b) Hay, B. P.; Finke, R. G. J. Am. Chem. Soc. 1987, 109, 8012– 8018. (22) Finke, R. Vitamin B12 and B12 Proteins; Wiley-VCH: New York, 1998; Chapter 25. (23) Halpern, J. Science 1985, 227, 869–875. (24) Runge, E.; Gross, E. K. U. Phys. ReV. Lett. 1984, 52, 997–1000. (25) (a) Baerends, E. J.; Ricciardi, G.; Rosa, A.; van Gisergen, S. J. A. Coor. Chem. ReV. 2002, 230, 5–27. (b) Dreuw, A.; Head-Gordon, M. Chem. ReV. 2005, 105, 4009–4037. (c) Dreuw, A. ChemPhysChem 2006, 7, 2259– 2274. (26) Jaworska, M.; Lodowski, P.; Andrunio´w, T.; Kozlowski, P. M. J. Phys. Chem. B 2007, 111, 2419–2422. (27) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Perdew, J. P. Phys. ReV. B 1986, 33, 8822–8824. (28) Rovira, C.; Biarnes, X.; Kunc, K. Inorg. Chem. 2004, 43, 6628– 6632. (29) Rovira, C.; Kozlowski, P. M. J. Phys. Chem. 2007, 111, 3251– 3257.

Co-C Bond in Methyl- and Ethylcobalamin (30) Jensen, K. P.; Ryde, U. J. Phys. Chem. A 2003, 107, 7539–7545. (31) Kuta, J.; Patchkovskii, S.; Zgierski, M. Z.; Kozlowski, P. M. J. Comput. Chem. 2006, 27, 1429–1437. (32) Andrunio´w, T.; Jaworska, M.; Lodowski, P.; Zgierski, M. Z.; Dreos, R.; Randaccio, L.; Kozlowski, P. M. J. Chem. Phys. 2008, 129, 085101. (33) Kozlowski, P. M.; Kuta, J.; Galezowski, W. J. Phys. Chem. B 2007, 111, 7638–7645. (34) Galezowski, W.; Kuta, J.; Kozlowski, P. M. J. Phys. Chem. B 2008, 112, 3177–3183. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, reVision C.02; Gaussian, Inc.: Wallingford, CT, 2004. (36) (a) Ahlrichs, R.; Ba¨r, M.; Ha¨ser, M.; Horn, H.; Ko¨lmel, C. Chem. Phys. Lett. 1989, 162, 165–169. (b) Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995, 102, 346. (c) Furche, F.; Ahlrichs, R. J. Chem. Phys. 2002, 117, 7433. (d) Furche, F.; Rappoport, D. Density functional methods for excited states: equilibrium structure and electronic spectra. Computational Photochemistry; Olivucci, M., Ed.; Elsevier: Amsterdam, 2005; Theoretical and Computational Chemistry, Vol. 16, Chapter III. (e) TURBOMOLE has been designed by the Quantum Chemistry Group, University of Karlsruhe, Germany, since 1988. The following members of the group have made contributions: Ahlrichs, R.; Bär, M.; Baron, H.; Bauernschmitt, R.; Böcker, S.; Crawford, N.; Deglmann, P.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche, F.; Haase, F.; Hãser, M.; Hättig, C.; Hellweg, A.; Horn, H.; Huber, C.; Huniar, U.; Kattannek, M.; Köhn, A.; Kölmel, C.; Kollwitz, M.; May, K.; Nava, P.; Ochsenfeld, C.; öhm, H.; Patzelt, H.; Rappoport, D.; Rubner, O.; Schäfer, A.; Schneider, U.; Sierka, M.; Treutler, O.; Unterreiner, B.; Arnim, M.; Weigend, F.; Weis, P.; Weiss, H.; http://www.turbomole.com. (37) Randaccio, L.; Furlan, M.; Geremia, S.; Slouf, M.; Srnova, I.; Toffoli, D. Inorg. Chem. 2000, 39, 3403–3413. (38) Fasching, M.; Schmidt, W.; Kra¨utler, B.; Stupperich, E.; Schmidt, A.; Kratky, C. HelV. Chim. Acta 2000, 83, 2295–2316.

J. Phys. Chem. B, Vol. 113, No. 19, 2009 6909 (39) (a) Bresciani-Pahor, N.; Forcolin, M.; Marzilli, L. G.; Randaccio, L.; Summers, J. S.; Toscano, P. J. Coord. Chem. ReV. 1985, 63, 1–125. (b) Randaccio, L.; Bresciani-Pahor, N.; Zangrando, E. Chem. Soc. ReV. 1989, 18, 225–250. (c) DeRidder, D. J. A.; Zangrando, E.; Burgi, H.-B. J. Mol. Struct. 1996, 374, 63–83. (d) Zou, X.; Brown, K. L. Inorg. Chim. Acta 1998, 267, 305–308. (e) Randaccio, L.; Geremia, S.; Stener, M.; Toffoli, D.; Zangrando, E. Eur. J. Inorg. Chem. 2002, 2002, 93–103. (f) Andruniow, T.; Kuta, J.; Zgierski, M. Z.; Kozlowski, P. M. Chem. Phys. Lett. 2005, 410, 410–416. (40) Martin, B. D.; Finke, R. G. J. Am. Chem. Soc. 1992, 114, 585– 592. (41) Hung, R. R.; Grabowski, J. J. J. Am. Chem. Soc. 1999, 121, 1359– 1364. (42) Giannotti, C. In B12; Dolphin, D., Ed.; John Wiley & Sons: New York, 1982; Vol. 1, Chapter 11, pp 393-430. (43) Pratt, J. M. In Chemistry and Biochemistry of B12; Banerjee, R., Ed.; John Wiley & Sons: New York, 1999; Chapter 5, pp 113-164. (44) Hill, J. A.; Pratt, J. M.; Williams, R. J. P. J. Chem. Soc. 1964, 5149–5153. (45) Pratt, J. M.; Thorp, R. G. J. Chem. Soc. 1966, 187–191. (46) Firth, R. A.; Hill, H. A. O.; Mann, B. E.; Pratt, J. M.; Thorp, R. G.; Williams, R. J. P. J. Chem. Soc. 1968, 2419–2428. (47) Dong, S.; Padmakumar, R.; Banerjee, R.; Spiro, T. G. Inorg. Chim. Acta 1998, 270, 392–398. (48) Toohey, J. I. Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 934–942. (49) CASSCF/CASPT2 calculations for a simplified structural model of MeCbl with an additional symmetry imposed with respect to the plane bisecting the corrin ring indicate that the lowest exited state computed at the TD-DFT/BP86 agrees better with multi-comfigurational analysis rather than the one based on TD-DFT/B3LYP. (50) Brooks, A. J.; Vlasie, M.; Banerjee, R.; Brunold, T. C. J. Am. Chem. Soc. 2004, 126, 8167. (51) Daniel, C. Coord. Chem. ReV. 2003, 238-239, 143–166. (52) Vlc`ek, A., Jr.; Za´lis`, S. Coord. Chem. ReV. 2007, 251, 258–287. (53) In case of pentaamminemethylcobalt complex, CoCH3(NH3)52+, it can be demonstrated, employing CASSCF/CASPT2 level of theory, that the cleavage of the Co-C bond gives radical fragments, rather than ionic. (54) Franzen, S.; Kiger, L.; Popart, C.; Martin, J. L. Biophys. J. 2001, 80, 2372–2385. (55) Dunietz, B. D.; Dreuw, A.; Head-Gordon, M. J. Phys. Chem. B 2003, 107, 5623–5629. (56) Dreuw, A.; Dunietz, B. D.; Head-Gordon, M. J. Am. Chem. Soc. 2002, 124, 12070–12071.

JP810223H