On the mechanism of ligand-assisted, copper-catalyzed benzylic amination by chloramine-T

Article abstract of DOI:10.1021/om100427s

The mechanism of hydrocarbon amination by chloramine-T derivatives catalyzed by (diimine)copper complexes has been investigated. The initial synthetic study of the reactions revealed ligand-accelerated catalysis, significant sensitivity to the electronic character of the substrates, and low to moderate enantioselectivities with homochiral ligands. Various mechanistic probes, both experimental and computational, have been focused on the C-H insertion process. A kinetic isotope effect of 4.6 was found in the amination of α-D(H)-cumenes catalyzed by [(diimine)Cu(solv)]Z. Amination of the isomeric substrates cis-and trans-4-tert-butyl-1-phenylcyclohexanes with 4-Me-C₆H₄SO₂NNaCl (chloramine-T) or 4-NO ₂-C₆H₄SO₂NNaCl (chloramine-N) catalyzed by [(diimine)Cu(CH₃CN)]PF₆ produced in all cases an approximately 1:1 mixture of the corresponding cis-and trans-4-tert-butyl-1- phenyl-1-sulfonaminocyclohexanes. Amination of the radical-clock substrate 1-phenyl-2-benzylcyclopropane with chloramine-T/(diimine)Cu(CH ₃CN)]PF₆ gave a mixture of ring-opened and cyclopropylmethylamino derivatives. Together, these results are most consistent with a stepwise insertion of an N-Ts(Ns) unit into the C-H bond, via carbon radicals, and a secondary contribution from a concerted insertion pathway. B3LYP and CASSCF computations suggest that the C-H insertion step involves the reaction of the hydrocarbon with a Cu-imido (nitrene) complex, [(diimine)Cu=NSO₂R]⁺. The ground-state triplet of the Cu-imido complex is calculated to be 3-13 kcal/mol more stable that the singlet complex, depending on the method and basis sets employed. The reaction of each complex with toluene is modeled to find that the C-H insertion transition state for the triplet (ΔG^? = 8.2 kcal/mol) is lower in energy than the singlet. The former reacts by a stepwise H-atom abstraction, while the latter reacts by a concerted C-H insertion. These results and kinetic isotope effect calculations for the singlet (2.9) and triplet (4.8) pathways, respectively, agree with the experimental observations (4.6) and point to a major role for the triplet complex in the stepwise, nonstereoselective insertion pathway.

Full text of DOI:10.1021/om100427s

3404 Organometallics 2010, 29, 3404–3412
DOI: 10.1021/om100427s
On the Mechanism of Ligand-Assisted, Copper-Catalyzed Benzylic
Amination by Chloramine-T
Dipti N. Barman,† Peng Liu,‡ Kendall N. Houk,‡ and Kenneth M. Nicholas*,†
†Department of Chemistry and Biochemistry, University of Oklahoma, Norman, Oklahoma 73019, and
‡Department of Chemistry and Biochemistry, University of California, Los Angeles, California 90095
Received May 5, 2010
The mechanism of hydrocarbon amination by chloramine-T derivatives catalyzed by (diimine)-
copper complexes has been investigated. The initial synthetic study of the reactions revealed ligand-
accelerated catalysis, significant sensitivity to the electronic character of the substrates, and low to
moderate enantioselectivities with homochiral ligands. Various mechanistic probes, both experi-
mental and computational, have been focused on the C-H insertion process. A kinetic isotope effect
of 4.6 was found in the amination of R-D(H)-cumenes catalyzed by [(diimine)Cu(solv)]Z. Amination
of the isomeric substrates cis- and trans-4-tert-butyl-1-phenylcyclohexanes with 4-Me-C6H4SO2-
NNaCl (chloramine-T) or 4-NO2-C6H4SO2NNaCl (chloramine-N) catalyzed by [(diimine)Cu-
(CH3CN)]PF6 produced in all cases an approximately 1:1 mixture of the corresponding cis- and
trans-4-tert-butyl-1-phenyl-1-sulfonaminocyclohexanes. Amination of the radical-clock substrate
1-phenyl-2-benzylcyclopropane with chloramine-T/(diimine)Cu(CH3CN)]PF6 gave a mixture of ring-
opened and cyclopropylmethylamino derivatives. Together, these results are most consistent with a
stepwise insertion of an N-Ts(Ns) unit into the C-H bond, via carbon radicals, and a secondary con-
tribution from a concerted insertion pathway. B3LYP and CASSCF computations suggest that the C-H
insertion step involves the reaction of the hydrocarbon with a Cu-imido (nitrene) complex, [(diimine)-
CudNSO2R]þ. The ground-state triplet of the Cu-imido complex is calculated to be 3-13 kcal/mol
more stable that the singlet complex, depending on the method and basis sets employed. The reaction
of each complex with toluene is modeled to find that the C-H insertion transition state for the triplet
(ΔG‡ =8.2 kcal/mol) is lower in energy than the singlet. The former reacts by a stepwise H-atom
abstraction, while the latter reacts by a concerted C-H insertion. These results and kinetic isotope effect
calculations for the singlet (2.9) and triplet (4.8) pathways, respectively, agree with the experimental
observations (4.6) and point to a major role for the triplet complex in the stepwise, nonstereoselective
insertion pathway.
Introduction
to develop new nitrogenation methods involving nitrenoid
transfer reactions, either via addition to C-C unsaturation,
i.e., aziridination,1 or by insertion into the C-H bonds of
hydrocarbons (eq 1).2 Complexes of many transition metals
have been found to catalyze such reactions using several N-
reagents, most commonly imidoiodinanes (ArIdNZ), azides
(Z-N3), or haloamine derivatives (ZNNaX).
Amines and their derivatives constitute the most impor-
tant class of synthetic and natural bioactive compounds, and
they find numerous important uses in medicine and agricul-
ture and as materials, dyes, and electronics. The importance
and ubiquity of the amino functionality has motivated chemists
*To whom correspondence should be addressed. E-mail: knicholas@
ou.edu.
(1) (a) Evans, D. A.; Faul, M. M.; Bilodeau, M. T.; Anderson, B. A.;
Barnes, D. M. J. Am. Chem. Soc. 1993, 115, 5328. (b) Mueller, P.; Fruit, C.
Chem. Rev. 2003, 103, 2905. (c) Liang, J.-L.; Huang, J.-S.; Yu, X.-Q.; Zhu,
N.; Che, C.-M. Chem.;Eur. J. 2002, 8, 1563. (d) Li, Z.; Quan, R. W.;
Jacobsen, E. N. J. Am. Chem. Soc. 1995, 117, 5889. (e) Mairena, M. A.;
Diaz-Requejo, M. M.; Belderrain, T. R.; Nicasio, M. C.; Trofimenko, S.;
Perez, P. J. Organometallics 2004, 23, 233.
(2) (a) Li, Z.; Capretto, D. A.; Rahaman, R.; He, C. Angew. Chem.,
Int. Ed. 2007, 46, 5184. (b) Thu, H.-Y.; Yu, W.-Y.; Che, C.-M. J. Am. Chem.
Soc. 2006, 128, 9048. (c) Reed, S. A.; Mazzotti, A. R.; White, M. C. J. Am.
Chem. Soc. 2009, 131, 11701. (d) Wang, Z.; Zhang, Y.; Fu, H.; Jiang, Y.;
Zhao, Y. Org. Lett. 2008, 10, 1863. (e) Collet, F.; Dodd, R. H.; Dauban, P.
Chem. Commun. 2009, 5061–5074. (f) Davies, H. M. L.; Long, M. S.
Angew. Chem., Int. Ed. 2005, 44, 3518–3520. (g) Kalita, B.; Lamar, A. A.;
Nicholas, K. M. Chem. Commun. 2008, 4291.
The chemo-, regio-, and stereoselectivities of these emerg-
ing nitrogenation processes are of paramount synthetic import-
ance. Several systems studied to date effect aziridinations with
moderate to excellent enantioselectivities employing chiral metal
complex catalysts.3 Interestingly, some of these same reagent/
catalyst systems do not react stereospecifically; for example,
(3) Gillespie, K. M.; Sanders, C. J.; O’Shaughnessy, P.; Westmoreland,
I.; Thickitt, C. P.; Scott, P. J. Org. Chem. 2002, 67, 3450–3458.
r
pubs.acs.org/Organometallics
Published on Web 07/12/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 15, 2010 3405
cis-olefins, unless constrained within a ring, often produce cis/
trans aziridine product mixtures,4 suggestive of a stepwise
formation of the two C-N bonds. The stereoselectivity of
C-H aminations has not been investigated extensively, but
intramolecular L*2Rh2-5 and Ru(pybox*)6-catalyzed reac-
tions, reagent-based [ArS*(O)(NSO2Ar)NH2]/Rh-catalyzed
intermolecular reactions,7 and (salen)Mn-catalyzed intermole-
cular aminations of benzylic substrates8 have been reported
with moderate to high enantioselectivities, consistent with a
concerted C-H insertion process.
Scheme 1
These selectivity features are determined primarily by the
nature of the reactive N-species that is involved in the step(s)
in which the C-N bond(s) is(are) formed. For the most part
the reactive intermediates in metal-catalyzed aziridinations
and C-H aminations are unknown. It has been generally
assumed, with some supporting evidence, that metal-imido
(nitrene) complexes, LMdNZ, are the active N-transfer
agents, rather than free nitrenes. Only in a few instances
have structurally characterized imido complexes been iso-
lated and shown to effect aziridination or C-H insertions.
Che and co-workers have investigated the reactions of iso-
lable (porphyrin)Ru(NSO2R)2 complexes, which both azir-
idinate olefins and aminate hydrocarbon C-H bonds.9
Kinetics and redox potentials have been interpreted in terms
of stepwise radical pathways with both classes of substrates.
Recently, Warren and co-workers reported on the C-H inser-
tion reactivity of dinuclear [(ketoiminato)Cu-N(adamantyl)]2.10
Kinetic isotope effects, correlations with C-H bond dissociation
energies, and DFT computational studies by the Cundari
group11 suggest that a monometallic LCu(NR) singlet diradical
species is responsible for the C-H insertion.
Aside from the systems involving isolable metal-imido
complexes, some insights have been gained from mechanistic
probes of other metal-catalyzed aziridinations and amina-
tions presumed to involve imido-metal intermediates. The
C-H(D) kinetic isotope effects, Hammett studies, radical-
clock probes, and high intramolecular stereoselectivities
observed in the Rh2X4-catalyzed C-H aminations point to
the intervention of an electrophilic imido-Rh2 complex, which
effects concerted C-H insertion.12 Among the mechanistic
studies of copper-catalyzed aziridination, the general picture
is less clear. Some Cu-catalyzed aziridinations employing sui-
table chiral ligands achieve high enantioselectivities,13 although
olefin stereochemistry is not always retained as expected for a
concerted process. Kinetic analysis, Hammett substituent
studies, and the common stereoselectivity found independent
of the aminating agent have been interpreted in terms of the
intervention of an electrophilic LCudNZ(þ) as the active
nitrogenating species.14 Computational studies of various
imido-Cu complexes with different N-donor auxiliary ligands
have mostly supported a ground-state triplet state for such
species15 (the Warren/Cundari system notwithstanding11),
which reacts stepwise as a N-centered radical with substrate-
dependent competing C-C rotation versus C-N ring closure.
As part of an ongoing project to discover and elucidate
new metal-catalyzed nitrogenation reactions of hydro-
carbons, we recently have focused on the amination of benzylic
substrates. An efficient system for benzylic amination was
first developed employing readily available anhydrous chlor-
amine-T (MeC6H4SO2NNaCl) with commercial [Cu(CH3-
CN)4]PF6 (1) as catalyst.16 We then sought to identify sui-
table ligand partners for copper that could provide tailored
catalytic activity and selectivity. Several ligand types, includ-
ing phosphines, amines, R-amino acids, and diimines (Schiff
bases), were indeed found to enhance catalytic amination
activity (ligand-accelerated catalysis)17 relative to weakly
ligated 1.18 Within a series of para-substituted aryldiimine
ligands and aryl-substituted hydrocarbons the catalytic acti-
vity was found to increase with more electron-deficient
ligands and more electron-rich substrates, indicative of an
electrophilic aminating species, possibly [L2CuNSO2Ar]þ
(Scheme 1). A survey of the amination of prochiral benzylic
substrates by chloramine-T catalyzed by complexes of homo-
chiral amino acids and diimines, however, achieved only low
to modest enantioselectivities (0-35%). These results caused
us to consider further the identity of the aminating species in
these reactions and the origin of its limited enantioselectivity.
To address these issues, we have conducted a mechanistic
study, including both experimental and computational
probes, focusing on the crucial C-H insertion process. The
results of this investigation are reported herein.
(4) Caine, D.; O’Brien, P.; Rosser, C. M. Org. Lett. 2002, 4, 1923–
1926.
(5) (a) Fleming, J. J.; Fiori, K. W.; Du Bois, J. J. Am. Chem. Soc.
2003, 125, 2028. (b) Fiori, K. W.; Fleming, J. J.; Du Bois, J. Angew. Chem.,
Int. Ed. 2004, 43, 4349.
(6) Milczek, E.; Boudet, N.; Blakey, S. Angew. Chem., Int. Ed. 2008,
47, 6825–6828.
(7) Shou, W. G.; Li, J.; Guo, T.; Lin, Z.; Jia, G. Organometallics 2009,
28, 6847–6854.
Results and Discussion
Experimental Studies. The primary focus of our experi-
mental mechanistic investigation is the C-H/C-N substitu-
tion process itself. The key questions are, (1) is the substitution
(8) Ohta, C.; Katsuki, T. Tetrahedron Lett. 2001, 42, 3885.
(9) (a) Leung, S. K.-Y.; Huang, J.-S.; Liang, J.-L.; Che, C.-M.; Zhou,
Z.-Y. Angew. Chem., Int. Ed. 2003, 42, 340. (b) Liang, J.-L.; Yuan, S.-X.;
Chan, P. W. H.; Che, C.-M. Org. Lett. 2002, 4, 4507. (c) Liang, J.-L.; Huang,
J.-S.; Yu, X.-Q.; Zhu, N.; Che, C.-M. Chem.;Eur. J. 2002, 8, 1563.
(d) Liang, J.-L.; Yuan, S.-X.; Huang, J.-S.; Yu, W.-Y.; Che, C.-M. Angew.
Chem., Int. Ed. 2002, 41, 3465.
(14) Li, Z.; Conser, K. R.; Jacobsen, E. N. J. Am. Chem. Soc. 1993,
115, 5326.
(15) (a) Brandt, P.; So1dergren, M. J.; Andersson, P. G.; Norrby,
P. O. J. Am. Chem. Soc. 2000, 122, 8013–8020. (b) Conradie, J.; Ghosh, A.
J. Chem. Theory Comput. 2007, 3, 689. (c) Meng, Q.; Wang, F.; Qu, X.;
Zhou, J.; Li, M. THEOCHEM 2007, 815, 111. (d) Comba, P.; Lang, C.;
Lopez de Laorden, C.; Muruganantham, A.; Rajaraman, G.; Wadepohl, H.;
Zajaczkowski, M. Chem.;Eur. J. 2008, 14, 5313. (e) Cundari, T. R.;
Dinescu, A.; Kazi, A. B. Inorg. Chem. 2008, 47, 10067.
(10) Badiei, Y. M.; Krishnaswamy, A.; Melzer, M. M.; Warren, T. H.
J. Am. Chem. Soc. 2006, 128, 15056.
(11) Badiei, Y. M.; Dinescu, A.; Dai, X.; Palomino, R. M.; Heinemann,
F. W.; Cundari, T. R.; Warren, T. H. Angew. Chem., Int. Ed. 2008, 47, 9961.
(12) Fiori, K. W.; Espino, C. G.; Brodsky, B. H.; Du Bois, J.
Tetrahedron 2009, 65, 3042.
(16) Bhuyan, R.; Nicholas, K. M. Org. Lett. 2007, 9, 3957.
ꢀ
€
(17) Jacobsen, E. N.; Marko, I. E.; Mungall, W. S.; Schroder, G. W.;
Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 1968.
(18) Barman, D. N.; Nicholas, K. M. Tetrahedron Lett. 2010, 51,
1815.
€
(13) Muller, P.; Fruit, C. Chem. Rev. 2003, 103, 2905–2920.

3406 Organometallics, Vol. 29, No. 15, 2010
Barman et al.
Scheme 2
cation was generated at C-1, ring-opened amination pro-
ducts, e.g., N-(1,4-diphenylbut-3-enyl)-4-methylbenzene-
sulfonamide (12), would be produced via rapid ring-opening
(ca. 2 ꢀ109 s-1) of the intermediate.19 Amination of trans-10
under the standard conditions with chloramine-T catalyzed by
[(3)Cu(CH3CN)]PF6 produced a mixture of cyclic sulfonamide
11 (6%) and ring-opened products 12-C/12-T(18%). Hencewe
conclude that the nitrenoid C-H insertion occurs stepwise and
proceeds via H-atom (or hydride) abstraction to form the C-1
radical (or cation). Although the formation of ring-opened
products cannot unambiguously differentiate radical from
carbocation intermediates, the relatively modest substrate sub-
stituent effects that were observed in our synthetic studies20 and
the computational modeling (vide infra) favor the intervention
of radicals.
reaction rate-limiting? (2) is the process concerted or stepwise?
(3) is it stereoselective and, if so, in what sense, i.e., retention/
inversion? and (4) what is the active aminating agent?
Our preliminary substrate and ligand-based reactivity survey
(op. cit.) showed significant electronic effects, suggesting that
the C-H insertion step could be rate-limiting. To address this
issue further, we carried out a competitive amination reaction
between cumene and R-d1-cumene to determine the kinetic
isotope effect (Scheme 2). A 1:1 mixture of the isotopomers
2-H(D) was added to the preformed complex from diimine 3
and [Cu(CH3CN)4]PF6 in CH3CN (20 °C), followed by addi-
tion of anhydrous chloramine-T. After a minimal workup to
avoid H/D exchange GC-MS analysis of the reaction mixture
found the ratio of sulfonamides 4-H/4-D and the isotope effect
to be 4.6 at 40% conversion; analysis of the unreacted cumene
gave a kinetic isotope effect (KIE) of 3.5. This sizable primary
KIE indicates that C-H bond-breaking is at least partially
rate-limiting.
To further probe the nature of the C-H insertion step,
especially its concertedness, we examined the stereoselectivity
of the Cu-catalyzed amination with a pair of diastereomeric
hydrocarbon substrates, cis- and trans-1-tert-butyl-4-phenyl-
cyclohexane (5-C, 5-T). A concerted insertion pathway is
expected to be stereospecific; that is, the respective isomeric
substrates should produce opposite isomeric products with
high selectivity. Each substrate was separately treated with
chloramine-T in the presence of the preformed catalyst from
biphenyldiimine ligand 7 and [Cu(CH3CN)]4PF6 under typical
conditions (Scheme 3).17 Chromatographic purification of the
mixture from the reaction with cis-1-tert-butyl-4-phenylcyclo-
hexane (5-C) provided an inseparable mixture of the isomeric
1-tert-butyl-4-phenyl-tosylamide-cyclohexanes (8-C/8-T, 33%
yield) in a 1:1 ratio. trans-1-tert-Butyl-4-phenylcyclohexane
(5-T) was similarly aminated to give a nearly identical isomeric
mixture of amination products 8-C/8-T in 28% combined
yield. These results strongly suggest the primary operation of
a nonconcerted process for the C-H insertion, likely proceed-
ing through a common intermediate.
We wondered whether the electronic character of the ami-
nating species could be sufficiently altered to affect the con-
certedness of the insertion step and hence its stereoselectivity.
To test this idea, the nosylamination of trans-1-tert-butyl-4-
phenylcyclohexane (5-T) was examined using 4-O2N-C6H4-
SO2NNaCl (6b), “chloramine-N” (Scheme 3). In this case a
separable mixture of isomeric 4-tert-butyl-1-phenyl-1-nosyl-
cyclohexanes (9-C/9-T) was obtained (38% combined yield),
still in a nearly 1:1 ratio. Clearly, the nosylamino transfer agent
still effects C-H insertion primarily by a stepwise, nonstereo-
selective process.
The nonstereoselective feature of the C-H/N substitution
suggested the possibility of a stereorandomizing carbon
radical or carbocation intermediate. To test the validity of
this hypothesis, we chose the benzylic cyclopropyl substrate
10 for amination, anticipating that if either a radical or a
Computational Studies. To understand the mechanism and
concertedness of the Cu-catalyzed amination reactions of
chloramine-T, we have performed DFT and CASSCF calcu-
lations on both singlet and triplet pathways of this reaction.
We focused on the C-H insertion reaction of the Cu-imido
complex since this process is stereoselectivity-determining
and, based on the KIE results, at least partially rate-limiting.
The mechanisms of a related reaction thought to involve
Cu-imido intermediates, Cu-catalyzed alkene aziridinations
with PhIdNTs and TsN3, have been investigated previously
by both experiment and theory. Kinetic and theoretical
studies by Jacobson,1d Norrby,15a and Comba15d suggest
that the aziridination reactions occur via a Cu(I)/Cu(III)
cycle involving a Cu-imido complex LCudNTs as the reac-
tive intermediate. DFT calculations also suggest that Cu(II)
precatalysts can enter the catalytic cycle through reduction
by PhIdNTs to Cu(I). On the other hand, alternative
mechanisms involving Cu(II) active species were proposed
by Perez21 and Vedernikov.22
ꢀ
Due to their high reactivities, no terminal imido-copper
complexes have been structurally characterized. However,
their structure and spin state have been investigated using
several levels of theory. Density functional theory calcula-
tions suggest a ground triplet state for cationic LCu-imido
complexes.15 However, recent CASSCF11,15e and ccCA23
(correlation consistent composite approach) calculations
by Cundari and co-workers suggest that a biradical singlet is
the ground state for neutral (β-diketiminate)Cu(NPh) and
(β-diketiminate)Cu(NH) complexes. A recent DFT study on
Ni-nitrene complexes by the Cundari group suggested that
the singlet-triplet gap is affected by the nitrene substituent.
The singlet ground states are favored when aryl/alkyl groups
are on the nitrene,24 while the triplet is favored with electron-
withdrawing heteroatom groups (like -SO2R).25
In this study, we investigated the structures and spin state
of Cu-imido complexes and the C-H insertion transition
(19) Davis, A. F.; Song, M. M.; Alexander, A. J. Org. Chem. 2006,
71, 2779.
(20) Johnson, C. C.; Lippard, S. J.; Liu, K. E.; Newcomb, M. J. Am.
Chem. Soc. 1993, 115, 939.
ꢀ
(21) Dıaz-Requejo, M. M.; Perez, P. J.; Brookhart, M.; Templeton,
´
J. L. Organometallics 1997, 16, 4399.
(22) Mohr, F.; Binfield, S. A.; Fettinger, J. C.; Vedernikov, A. N.
J. Org. Chem. 2005, 70, 4833.
(23) Tekarli, S. M.; Williams, T. G.; Cundari, T. R. J. Chem. Theory
Comput. 2009, 5, 2959.
(24) (a) Cundari, T. R.; Vaddadi, S. THEOCHEM 2006, 801, 47.
(b) Cundari, T. R.; Pierpont, A. W.; Vaddadi, S. J. Organomet. Chem. 2007,
692, 4551. (c) Cundari, T. R.; Jiminez-Halla, J. O. C.; Morello, G. R.;
Vaddadi, S. J. Am. Chem. Soc. 2008, 130, 13051.
(25) Cundari, T. R.; Morello, G. R. J. Org. Chem. 2009, 74, 5711.

Article
Organometallics, Vol. 29, No. 15, 2010 3407
Scheme 3
Scheme 4
Table 1. Singlet/Triplet Splitting of the [(diimine)Cud
NSO2Me]+ Complex 13 by Various Levels of Calculations
triplet ground state.15a Single-point CASSCF calculations on
B3LYP/LANL2DZ-6-31G(d) geometries using a 10-electron-
10-orbital active space and a 12-electron-12-orbital active
space predicted very similar singlet/triplet splitting, suggesting
a (10,10) active space is adequate for CASSCF calculations.26
CASSCF geometry optimizations using CEP-31G(d) and
LANL2DZ basis sets predict the triplet is 7.5 and 11.1 kcal/
mol more stable than the singlet, respectively. This is in con-
trast to the CASSCF study by Cundari and co-workers on the
(β-diketiminate)Cu-nitrene complexes, which predicts a sing-
let ground state. Since the Cu-N covalent bond order is
greater27 and thus more polarized in the singlet complex, Cu is
more positively charged in the singlet complex than in the
triplet. The Mulliken charges on the Cu of the singlet and the
triplet complexes are 0.600 and 0.554, respectively. Electron-
rich ligands such as β-diketiminate may stabilize the positive
charge on the Cu atom and thus stabilize the singlet. At a
reviewer’s suggestion the effects of the N-substituent and the
denticity of the sulfonimido ligand on the singlet/triplet gap
were also evaluated. B3LYP calculations on the singlet/triplet
splitting of [(diimine)CudN-Ph]+ and the κ1 isomer of
[(diimine)CudNSO2Me]+ found that both complexes have
triplet ground states with larger S/T splittings than the [κ2-
(diimine)CudNSO2Me]+ complex 13.28 This suggests that
the oxygen atom bound to the Cu in the [κ2-(diimine)Cud
NSO2Me]+ complex stabilizes the singlet by donating elec-
trons to the Cu which is more positively charged in the singlet
complex.
method
ΔE (kcal/mol)a ΔG (kcal/mol)a
B3LYP/LANL2DZ-6-31G(d)
B3LYP/SDD-6-311+G(d)
B3LYP/6-311+G(d)
CASSCF(10,10)/CEP-
31G(d)//B3LYP/LANL2DZ-6-31G(d)
CASSCF(12,12)/CEP-
-3.3
-1.8
-5.8
-4.3
-0.2
-2.8
-10.1
-12.6b
-10.7
-13.2b
31G(d)//B3LYP/LANL2DZ-6-31G(d)
CASSCF(10,10)/CEP-31G(d)
CASSCF(10,10)/LANL2DZ
-7.5
-11.6
-7.5b
-11.1b
a Electronic energies and Gibbs free energies are in kcal/mol, defined
as E(triplet) - E(singlet). Negative number means triplet ground state.
b Thermal and entropic corrections are calculated at the B3LYP/
LANL2DZ-6-31G(d) level.
state at several theoretical levels. N,N0-Dimethylene-1,2-
ethylenediamine was used as a model ligand. Calculations
using a real ligand that has been used experimentally suggest
the singlet/triplet splitting is similar to the model system (see
Supporting Information for details). N-SO2Me was used as a
surrogate for the N-SO2Tol in the calculations.
We first calculated the singlet/triplet splitting of [(diimine)
CudNSO2Me]+ 13 using B3LYP and CASSCF with several
basis sets. The results are shown in Table 1. Both the B3LYP
and CASSCF calculations suggest a triplet ground state for
the Cu-imido complex, although the singlet/triplet splitting
varies from -2.8 to -13.2 kcal/mol by different methods.
B3LYP calculations with a mixed basis set of LANL2DZ
effective core potential basis set for Cu and 6-31G(d) for
other atoms predict that the triplet complex is 5.8 kcal/mol
more stable than the singlet complex. Using a smaller-core
pseudopotential SDD for Cu and 6-311+G(d) for other
atoms, a slightly smaller S/T splitting of -4.3 kcal/mol was
predicted. The full-electron basis set 6-311+G(d) on all
atoms including Cu predicts an even smaller splitting of
-2.8 kcal/mol. This agrees with the previous B3LYP study
by Norrby and co-workers, whose calculations with a full-
electron basis set predict small singlet/triplet splitting, while
effective core potential basis sets predict a much more stable
The CASSCF frontier molecular orbitals of the singlet and
triplet Cu-imido complexes are shown in Figure 1. The two
singly occupied orbitals in the triplet complex are the in-
plane Cu dxy orbital and the out-of-plane p orbital of the
imido N, respectively (Figure 1b). The spin density of the
triplet complex is mainly located at the imido N (1.180) and
(26) Cundari et al. also used a (10,10) active space with the CEP-31G
.
(d) basis set in their CASSCF study on Cu-nitrene complexes. See ref 15e
(27) NBO calculations predict that the Cu-N Wiberg bond indices
are 0.454 and 0.771 for the triplet and singlet complex, respectively.
(28) See Supporting Information for more details about these
calculations.

3408 Organometallics, Vol. 29, No. 15, 2010
Barman et al.
Figure 1. Frontier molecular orbitals of singlet- and triplet-13 from CASSCF(10,10)/CEP-3G(d) calculations.
Figure 2. Transition-state structures of singlet- and triplet-TS14 from B3LYP/LANL2DZ-6-31G(d) calculations.
Cu (0.437) atoms, in agreement with the FMO analysis. The
Cu-imido-N bond length in the triplet complex is longer
than in the singlet complex (2.00 and 1.87 A, respectively, for
B3LYP/LANL2DZ-6-31G(d)-optimized geometries; 2.04
and 1.96 A, respectively, for CASSCF(10,10)/CEP-31G(d)-
optimized geometries). These results suggest that the un-
paired electrons of the triplet Cu-imido complex are mainly
located on the imido N and Cu atoms and that subsequent
C-H insertion is most likely to occur at the imido N atom.
The C-H insertion occurs via hydrogen atom transfer
from the toluene to the amido nitrogen atom of the LCu-
imido complex. In the triplet C-H insertion transition state
triplet-TS14 (Figure 2), the toluene hydrogen approaches the
perpendicular position of the imido-N, since the radical
character is mainly located at the out-of-plane p orbital of
the nitrogen. The angle of the attack is 119° as predicted by
B3LYP calculations and 141° by CASSCF calculations.
In contrast, in the singlet transition state, singlet-TS14
(Figure 2), the toluene attacks the in-plane position of the
Cu-imido complex. The HOMO and LUMO of the singlet
Cu-imido complex involve the σ(Cu-N) bonding and the
σ
(Cu-N)* antibonding orbitals, respectively; both are in the
same plane of the Cu-imido complex (Figure 1a). B3LYP
calculations predict the angle of the attack via the singlet
pathway is 170°, and CASSCF predicts 165°.
The triplet C-H insertion transition state is 2.0 and
1.3 kcal/mol more stable than the singlet transition state,
as predicted by B3LYP/LANL2DZ-6-31G(d) and CASSCF
(10,10)/CEP-31G(d) calculations, respectively. IRC calcula-
tions on both singlet and triplet transition states were per-
formed to determine the concertedness of both pathways.
The IRC calculations on the singlet transition state singlet-
TS14 lead directly to the amination product and the Cu
being bound to the oxygen atom in the product complex

Article
Organometallics, Vol. 29, No. 15, 2010 3409
Figure 3. Potential energy profile for the singlet pathway at the B3LYP/LANL2DZ-6-31G(d) level.
Figure 4. Potential energy profile for the triplet pathway at the B3LYP/LANL2DZ-6-31G(d) level.
singlet-16. In contrast, the triplet transition state triplet-
TS14 leads to a radical pair intermediate triplet-15, a weakly
bound complex of benzyl and [LCu-NH-SO2Me]+ radicals.
The positive charge is mainly located on the Cu-imido
complex, and the benzyl radical is almost neutral. This
suggests the intermediate in the triplet C-H insertion is a
carbon radical rather than a carbocation. Recombination of
the radical pair may occur via multiple mechanisms: spin
crossover to form a singlet radical pair followed by rapid
radical combination or intermolecular radical coupling. In
either case, carbon configuration may be lost during the
formation of the singlet LCu-product complex singlet-16
from the radical pair intermediate triplet-15. The triplet LCu-
product complex triplet-16 is much less stable than the singlet.
The Gibbs free energy and enthalpy profiles for C-H
insertion of toluene via singlet and triplet pathways were
calculated at the B3LYP/LANL2DZ-6-31G(d) level and are
shown in Figures 3 and 4, respectively. Transition states and
intermediates in both profiles are with respect to the most
stable triplet LCu-imido complex triplet-13. Steps involving
spin crossover or radical termination are marked with
dashed lines.

3410 Organometallics, Vol. 29, No. 15, 2010
Scheme 5. Calculated and Experimental Kinetic Isotope Effects
Barman et al.
Scheme 6. C-H Activation Barriers at the B3LYP/LANL2DZ-6-31G(d)
Scheme 7. Mechanistic Pathways for (diimine)Cu-Catalyzed Amination by Chloramine-T
Kinetic isotope effects on the C-H insertion of toluene
were calculated at the B3LYP/LANL2DZ-6-31G(d) level.
The resulting KIEs for the singlet and triplet pathways are
2.91 and 4.83, respectively. The experimentally observed KIE
for the reaction with toluene is 4.6 (Scheme 5), in agreement
with the calculated KIE for the triplet pathway.
We have also investigated the activation barriers for C-H
insertion of representative electron-rich and electron-poor
toluene derivatives (Scheme 6 (a)-(c)). The triplet C-H
insertion transition state is favored in all cases. The more
electron-rich methylanisole requires a lower barrier in the
C-H insertion relative to toluene, and nitrotoluene is found

Article
Organometallics, Vol. 29, No. 15, 2010 3411
to have the largest barrier. These results agree with the
qualitative experimental dependence of the rate and amina-
tion efficiency on the substrate structure.18 This indicates
that the Cu-imido complex 13 is electrophilic and that triplet-
13 is more electrophilic than singlet-13 since the singlet/
triplet splitting of the transition state is larger for electron-
rich substrates.
meter. GC-mass spectra (electron impact) were obtained on an
Agilent benchtop HP6890/MSD5973. The following compounds
were prepared by literature methods: diimine ligands 3 and 7,18
N-chloro-N-sodio-4-nitrobenzenesulfonamide,29 cyclohexane deri-
vatives 5-C and 5-T,30 and cyclopropyl substrate 10.31
Preparation of d1-Cumene. The procedure was modified from
the previously reported preparation.32 Commercially available
cumene (2.0 mL, 19 mmol) was mixed with 10 wt % of Pd-C
and 8 mL of D2O, and the reaction flask was fitted with two
balloons full of hydrogen. The mixture was stirred at room
temperature overnight and then filtered through filter paper to
remove the Pd-C. The reaction mixture was partitioned be-
tween ethyl acetate and water, and the organic phase was dried
over MgSO4 and concentrated by rotary evaporation. The
deuterium content of the recovered cumene was determined by
1H NMR integration of the methyl groups to the tertiary
hydrogen.
D-Isotope Effect Determination. Commercial Cu(CH3CN)4-
PF6 (19 mg, 0.049 mmol), diimine 3 (19 mg, 0.049 mmol), and
5 mL of dry CH3CN were added to a round-bottom flask con-
taining dry molecular sieves (4 A, ca. 200 mg) under argon. To
the well-stirred suspension was added the 1:1 mixture of cumene
and deuterated cumene (68 μL, 0.49 mmol). Anhydrous chlor-
amine-T (89 mg, 0.39 mmol) was added after one hour, and the
mixture was stirred at room temperature overnight. The mixture
was filtered through Celite to remove the precipitated copper
salts and then analyzed by gas chromatography/mass spectro-
metry (GC-MS). The ratio of the d0-product to the d1-product
(4-H/4-D) was 4.6:1.0. GC-MS (EI) analysis of the unreacted
cumene-H/D in two separate runs gave kH/kD values of 3.3 and
3.7 (av=3.5).
Aminosulfonation of cis-1-tert-Butyl-4-phenylcyclohexane.
Commercial Cu(CH3CN)4PF6 (28 mg, 0.073 mmol), diimine 7
(35 mg, 0.073 mmol), and 5 mL of dry CH3CN were added to a
round-bottom flask containing dry molecular sieves (4 A, ca.
250 mg) under argon. To the well-stirred suspension was added
cis-1-tert-butyl-4-phenylcyclohexane19 (5-C, 0.16 g, 0.73 mmol).
Anhydrous chloramine-T (0.25 g, 1.1 mmol) was added to the
reaction vessel after 30 min and the mixture stirred at room
temperature overnight. The mixture was then filtered through
Celite powder to remove insoluble metal-containing materials,
and the solvent was evaporated under reduced pressure. The
crude residue was purified by preparative TLC (3:1 hexane/ethyl
acetate). The ratio of the inseparable isomeric products 8-T and
8-C was determined by integration of the tertiary butyl group
of each isomer (found 1.0:1.0). Combined yield: 9.3 mg (33%).
1H NMR (CDCl3, 300 MHz): δ 8.96 (s, 1H), 7.88 (d, 2H, J =
8.1), 7.33-7.25 (m, 14H), 7.13 (d, 2H, J = 8.1), 7.04 (d, 2H, J =
8.4), 5.57 (s, 1H), 2.56 (d, 2H, J = 12.9), 2.47 (s, 3H), 2.45
(s, 3H), 2.22 (d, 2H, J = 12), 1.93-1.62 (m, 8H), 1.3-1.09 (m,
6H), 0.95 (s, 9H), 0.90 (s, 9H). 13C NMR (CDCl3, 75 MHz):
δ 167.0, 163.4, 146.9, 146.2, 142.9, 141.9, 140.6, 139.9, 129.5,
129.2, 128.8, 128.5, 127.7, 126.7, 126.2, 125.3, 59.6, 59.9, 47.5,
47.3, 38.63, 36.1, 32.6, 27.7, 23.2, 23.0, 22.8, 22.2, 21.7, 21.6. ESI-
MS: m/z 408.21 (M+Na) 449.21 (M+CH3CN+Na).
To study the electronic effects of substituents on the
sulfonamide reagent, activation energies of the C-H insertion
with two Cu-imido complexes, LCu-N-SO2CH3, and LCu-N-
SO2CF3, were calculated using B3LYP. The results are given in
Scheme 6 (b) and (d). Both Cu-imido complexes have a triplet
ground state. An electron-withdrawing group on the sulfon-
amide decreases the singlet/triplet gap of the Cu-imido com-
plex from 5.8 kcal/mol for LCu-N-SO2CH3 to 2.3 kcal/mol for
LCu-N-SO2CF3. Both complexes still prefer the triplet path-
way in C-H insertion. The singlet/triplet gaps of the C-H
insertion transition states are similar: 2.0 kcal/mol for LCu-N-
SO2CH3 and 1.6 kcal/mol for LCu-N-SO2CF3. An electron-
withdrawing group on the sulfonamide decreases the C-H
insertion activation energies for both the singlet and triplet
pathways. These results suggest that sulfonamides with elec-
tron-withdrawing groups are more reactive, while the C-H
insertion via the triplet pathway is slightly more favorable.
On the basis of these experimental and computational
studies, we propose the mechanistic pathways outlined in
Scheme 7. Considering all the evidence, the triplet pathway likely
dominates. However, the modest (but nonzero) enantioselec-
tivities observed in reactions catalyzed by chiral diimine com-
plexes18 and the significant amount of ring-retained product
from the radical-clock substrate suggest that the singlet
pathway contributes to a minor extent.
Conclusions
The combined experimental observations of low enantio-
selectivity, a substantial primary kinetic isotope effect, neg-
ligible diastereoselectivity, and the formation of ring-opened
products with the cyclopropyl substrate together are indicative
of the predominant operation of a stepwise (nonconcerted)
C-H insertion process, most likely through the intermediacy of
C-centered radicals. Computational results support this picture
with the primary reactive N-species being the triplet (diimine)
Cu(NZ) complex, which acts as an N-centered radical, stepwise
abstracting a H-atom from the substrate followed by radical
rebound to form the C-N bond. A minor pathway involving
concerted, stereoselective insertion via the singlet complex
requires a somewhat higher activation energy but may also
contribute. A primary kinetic isotope effect was observed, and
calculations suggest that the KIE corresponds to the triplet
C-H insertion transition state.
Aminosulfonation of trans-1-tert-Butyl-4-phenylcyclohexane.
Commercial Cu(CH3CN)4PF6 (0.028 g, 0.073 mmol), diimine
ligand 7 (35 mg, 0.073 mmol), and 5 mL of dry CH3CN were
added to a round-bottom flask with dry molecular sieves (4 A,
ca. 200 mg) under an argon atmosphere. trans-1-tert-Butyl-4-
phenylcyclohexane (5-T, 0.158 g, 0.73 mmol) was then added to
the mixture. Anhydrous chloramine-T (251 mg, 1.1 mmol) was
added after 30 min, and the reaction mixture was stirred at room
Experimental Section
General Procedures. Acetonitrile was refluxed over and dis-
tilled from calcium hydride. Some of the hydrocarbon substrates
were obtained commercially and distilled before use. Commercial
chloramine-T trihydrate was dried in a drying pistol over refluxing
toluene under vacuum (0.1 mm) for 4-5 h; no thermal instability
or decomposition was noted in any samples dried in this way. Cu-
(CH3CN)4PF6 was obtained commercially. NMR spectra were
recorded on a Varian Mercury NMR spectrometer at 300 MHz
for 1H and 75 MHz for 13C NMR spectra. All chemical shifts (ppm)
reported are relative to the NMR solvent peak. Electrospray (ESI)
mass spectra were recorded on a Finnigan TSQ 700 mass spectro-
(29) Heintzelman, R. W.; Swern, D. Synthesis 1976, 11, 731.
(30) Garbisch, E. W., Jr.; Patterson, D. B. J. Am. Chem. Soc. 1963, 85,
3228.
(31) Bumgardner, C. L.; Iwerks, H. J. Am. Chem. Soc. 1966, 88, 5518.
(32) Kurita, T.; Hattori, K.; Seki, S.; Mizumoto, T.; Aoki, F.;
Yamada, Y.; Ikawa, K.; Maegawa, T.; Monguchi, Y.; Sajiki, H.
Chem.;Eur. J. 2008, 14, 664.

3412 Organometallics, Vol. 29, No. 15, 2010
Barman et al.
temperature overnight. The mixture was then filtered through
Celite powder, and the solvent was evaporated under reduced
pressure. The crude residue was purified by preparative TLC
(3:1 hexane/ethyl acetate). Combined yield of 8-C/8-T: 8.0 mg,
yield 28%, white solid. The ratio of the isomers was determined
by integration of the tertiary butyl group of each (found 1.0:1.0).
1H NMR (CDCl3, 300 MHz): δ 8.96 (s, 1H), 7.88 (d, 2H, J =
8.1), 7.33-7.25 (m, 14H), 7.13 (d, 2H, J = 8.1), 7.04 (d, 2H, J =
8.4), 5.57 (s, 1H), 2.56 (d, 2H, J = 12.9), 2.47 (s, 3H), 2.45
(s, 3H), 2.22 (d, 2H, J = 12), 1.93-1.62 (m, 8H), 1.3-1.09 (m,
6H), 0.95 (s, 9H), 0.90 (s, 9H). 13C NMR (CDCl3, 75 MHz): δ
167.0, 163.4, 146.9, 146.2, 142.9, 141.9, 140.6, 139.9, 129.5,
129.2, 128.8, 128.5, 127.7, 126.7, 126.2, 125.3, 59.6, 59.9, 47.5,
47.3, 38.6, 36.1, 32.6, 27.7, 23.2, 23.0, 22.8, 22.2, 21.7, 21.6. ESI-
MS: m/z 408.21 (M+Na), 449.21 (M+CH3CN+Na).
Aminonosylation of trans-1-tert-Butyl-4-phenylcyclohexane.
Commercial Cu(CH3CN)4PF6 (10 mg, 0.025 mmol), diimine 7
(13 mg, 0.025 mmol), and 5 mL of dry CH3CN were added to a
round-bottom flask containing dry molecular sieves (4 A, ca. 200
mg) under argon. To the well-stirred suspension was added trans-1-
tert-butyl-4-phenylcyclohexane (5-T, 55 mg, 0.25 mmol). An-
hydrous N-chloro-N-sodio-4-nitrobenzenesulfonamide (6b, 80 mg,
0.31 mmol) was added into the reaction vessel after one hour, and
the mixture was stirred at room temperature overnight. The mix-
ture was then filtered through Celite to remove the precipitated
copper salts, and the solvent was evaporated under reduced pres-
sure. NMR analysis of the crude product showed a 1:1 ratio of
isomers from integration of the tert-butyl protons. The isomers 9-C
and 9-T were separated by preparative TLC (3:1 hexane/ethyl
acetate) but not stereochemically assigned. Rf(isomer 1) = 0.38;
yield 12 mg (12%), white solid, mp 207-208 °C. 1H NMR(CDCl3,
300 MHz): δ 7.92(d, 2H, J = 8.7), 7.33 (d, 2H, J = 8.7), 7.11 (d,
2H, J = 8.4), 7.04-6.97 (m, 3H), 4.80 (s, 1H), 2.76 (d, 2H, J =
11.1), 1.75 (t, 2H, J = 12.9, J = 12.9), 1.58 (d, 2H, J = 13.2),
0.91-0.78 (m, 3H), 0.634 (s, 9H). 13C NMR (CDCl3, 75 MHz):
δ 147.7, 137.5, 135.2, 128.2, 128.1, 127.8, 127.6, 123.4, 60.9,
47.5, 38.3, 32.2, 27.3, 23.6. ESI-MS (m/z): 416.21 (M + Na).
Rf(isomer 2)=0.15; yield 10 mg (10%); mp >220 °C. 1H NMR
(CDCl3, 300 MHz): δ 8.05 (d, 2H, J = 8.7), 7.34-7.24 (m, 10H),
5.63 (s, 1H), 2.39 (d, 2H, J = 12.9), 1.75-1.62 (m, 6H), 1.18-1.031
(m, 2H), 0.9 (s, 9H).
to afford the cyclopropylmethyl amine derivative 11; yield 5 mg
(6%), white solid. 1H NMR (CDCl3, 300 MHz): δ 7.66 (d, 2H,
J = 8.1), 7.26-7.17 (m, 5H), 7.12-6.96 (m, 5H), 6.57 (d, 2H,
J = 7.2), 4.96 (d, 1H, J = 10.5), 4.87 (d, 1H, J = 3), 4.61 (dd,
1H, J = 4.5, J = 5.1), 3.23-3.17 (m, 1H), 2.38 (s, 3H), 2.24-2.1
(m, 2H). 13C NMR (CDCl3, 75 MHz): δ 144.1, 139.0, 137.6,
136.2, 130.0, 128.6, 128.5, 128.4, 128.3, 127.1, 127, 126.7, 67.0,
58.6, 57.4, 38.4, 21.6. ESI-MS (m/z): 408.21 (M+Na). The
NMR spectra for the isolated ring-opened products 12-C and
12-T (18 mg combined, 18% yield) were identical to those reported
previously.20
Computational Methods. Density functional calculations were
performed using the B3LYP functional in Gaussian 03.33
CASSCF Calculations. CASSCF calculations were carried
out using GAMESS (ver. 12 Jan 2009 R1).34 The full Newton-
Raphson (FULLNR) orbital update algorithm is used. For
single-point CASSCF energy calculations, the geometries were
optimized using Gaussian 03 with the B3LYP functional. A
mixed basis set of LANL2DZ for Cu and 6-31G(d) for other
atoms is used in the geometry optimization.
KIE Calculations. Kinetic isotope effect calculations were done
using Gaussian 03. The B3LYP functional and a mixed basis set of
LANL2DZ for Cu and 6-31G(d) for other atoms was used. A
scaling factor of 0.9613 was used to correct the calculated frequen-
cies.35 The kinetic isotope effects on the C-H insertion transition
states singlet- and triplet-TS14 were calculated, and the results are
shown in Table S1. The kH/kD are calculated from the differences
of the Gibbs free energies of the H/D-substituted transition states
at 298 K. KIEs with tunneling corrections (kH/kD(W)) were
calculated using Wigner’s tunneling correction.36 These calcula-
tions suggest that the tunneling effects in both singlet and triplet
transition states are small (+20%). The calculated KIEs without
tunneling corrections were used in this article.
Acknowledgment. We thank Prof. Karl Trindle
(U. Virginia), who aided in the initial phase of the
computational studies begun by K.N. during a sabbatical
leave at UCLA. K.N. and D.B. are grateful for financial
support provided by the National Science Foundation
(CHE-0848591). P.L. and K.N.H. also thank the National
Science Foundation for financial support (CHE-0548209).
Calculations were performed on the National Science
Foundation Terascale Computing System at the NCSA and
the UCLA ATS and IDRE clusters.
Aminonosylation of cis-1-tert-Butyl-4-phenylcyclohexane. The
amination reaction of cis-1-tert-butyl-4-phenylcyclohexane (5-C)
was carried out in the same way as with the trans isomer. NMR
analysis of crude product showed a 1:1 ratio of isomers 9-C/9-T
from the integration of the tert-butyl proton peaks.
Amination of 1-Benzyl-2-phenylcyclopropane. Commercial Cu
(CH3CN)4PF6 (9.0 mg, 0.024 mmol), diimine 7 (7.0 mg, 0.024
mmol), and 5 mL of dry CH3CN were added to a round-bottom
flask containing dry molecular sieves (4 A, ca. 200 mg) under
argon. To the well-stirred suspension was added 1-benzyl-2-
phenylcyclopropane (10, 50 mg, 0.24 mmol). Anhydrous chlor-
amine-T (94 mg, 0.36 mmol) was added to the reaction vessel
after 30 min, and the mixture was stirred at room temperature
overnight. The mixture was then filtered through Celite, and the
solvent was evaporated under reduced pressure. The crude resi-
due was purified by preparative TLC (3:1 hexane/ethyl acetate)
Supporting Information Available: Spectroscopic data for all
new compounds, optimized geometries and energies, diagrams
of orbitals in the active space for singlet- and triplet-13 and
singlet- and triplet-TS14, details of IRC calculations on the
singlet- and triplet-TS14, singlet/triplet splitting calculations
with a full ligand, and complete ref 33. This material is available
free of charge via the Internet at http://pubs.acs.org.
(34) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
Su, S. J.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput.
Chem. 1993, 14, 1347.
(35) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391.
(36) Wigner, E. P. Z. Phys. Chem. 1932, B19, 203.
(33) Frisch, M. J.; et al. Gaussian 03, Rev. D.01; Gaussian, Inc.:
Wallingford, CT, 2004.