Steric and electronic properties of N-heterocyclic carbenes (NHC): A detailed study on their interaction with Ni(CO)4

Article abstract of DOI:10.1021/ja0438821

N-heterocyclic carbene ligands IMes (1), SIMes (2), IPr (3), SIPr (4), and ICy (5) react with Ni(CO)₄ to give the saturated tricarbonyl complexes Ni(CO)₃(IMeS) (8), Ni(CO)₃(SIMeS) (9), Ni(CO)₃(IPr) (10), Ni(CO)₃(SIPr) (11), and Ni(CO) ₃(ICy) (12), respectively. The electronic properties of these complexes have been compared to their phosphine analogues of general formula Ni(CO)₃(PR₃) by recording their v_co stretching frequencies. While all of these NHCs are better donors than tertiary phosphines, the differences in donor properties between ligands 1-5 are surprisingly small. Novel, unsaturated Ni(CO)₂(IAd) (13) and Ni(CO)₂(I ^tBu) (14) compounds are obtained from the reaction of Ni(CO) ₄ with IAd (6) and I^tBu (7). Complexes 13 and 14 are highly active toward substitution of the NHC as well as the carbonyl ligands. This has allowed the determination of Ni-C(NHC) bond dissociation energies and the synthesis of various unsaturated Ni(0) and Ni(II) complexes. Computational studies on compounds 8-14 are in line with the experimental findings and show that IAd (6) and I^tBu (7) are more bulky than IMes (1), SIMes (2), IPr (3), SIPr (4), and ICy (5). Furthermore, a method based on % V_bur values has been developed for the direct comparison of steric requirements of NHCs and tertiary phosphines. Complexes 8-14, as well as NiCl(C ₃H₅)(I^tBu) (16) and NiBr(C₃H ₅)(I^tBu) (17), have been characterized by X-ray crystallography.

Full text of DOI:10.1021/ja0438821

Published on Web 02/02/2005
Steric and Electronic Properties of N-Heterocyclic Carbenes
(NHC): A Detailed Study on Their Interaction with Ni(CO)
4
†
†
†
‡
‡
Reto Dorta, Edwin D. Stevens, Natalie M. Scott, Chiara Costabile, Luigi Cavallo,
§
,†
Carl D. Hoff, and Steven P. Nolan*
Contribution from the Department of Chemistry, UniVersity of New Orleans,
New Orleans, Louisiana 70148, Dipartimento di Chimica, UniVersit a` di Salerno,
Baronissi, SA 84081, Italy, and Department of Chemistry, UniVersity of Miami,
Coral Gables, Florida 33124
Received October 7, 2004; E-mail: snolan@uno.edu
Abstract: N-heterocyclic carbene ligands IMes (1), SIMes (2), IPr (3), SIPr (4), and ICy (5) react with
Ni(CO) to give the saturated tricarbonyl complexes Ni(CO) (IMes) (8), Ni(CO) (SIMes) (9), Ni(CO) (IPr)
10), Ni(CO) (SIPr) (11), and Ni(CO) (ICy) (12), respectively. The electronic properties of these complexes
have been compared to their phosphine analogues of general formula Ni(CO) (PR ) by recording their νCO
stretching frequencies. While all of these NHCs are better donors than tertiary phosphines, the differences
in donor properties between ligands 1-5 are surprisingly small. Novel, unsaturated Ni(CO) (IAd) (13) and
with IAd (6) and I Bu (7). Complexes
3 and 14 are highly active toward substitution of the NHC as well as the carbonyl ligands. This has allowed
the determination of Ni-C(NHC) bond dissociation energies and the synthesis of various unsaturated Ni-
4
3
3
3
(
3
3
3
3
2
t
t
2 4
Ni(CO) (I Bu) (14) compounds are obtained from the reaction of Ni(CO)
1
(
0) and Ni(II) complexes. Computational studies on compounds 8-14 are in line with the experimental
t
findings and show that IAd (6) and I Bu (7) are more bulky than IMes (1), SIMes (2), IPr (3), SIPr (4), and
ICy (5). Furthermore, a method based on %Vbur values has been developed for the direct comparison of
t
steric requirements of NHCs and tertiary phosphines. Complexes 8-14, as well as NiCl(C
3 5
H
)(I Bu) (16)
t
3 5
and NiBr(C H )(I Bu) (17), have been characterized by X-ray crystallography.
5
Introduction
catalyzed hydrosilylation, or C-C coupling reactions using
6
carbene complexes of palladium are now well recognized. In
Since the isolation and crystallographic characterization of
stable N-heterocyclic carbenes (NHCs) by Arduengo et al.,
all of these catalytic processes, monocarbene ligands have
emerged as the most promising systems. The most widely used
NHCs of this class are represented in Figure 1.
1
increasing attention has been focused on using these compounds
as ancillary ligands for transition-metal complexes. Recent
developments in homogeneous catalysis have shown that NHCs
are attractive alternatives to the ubiquitous tertiary phosphine
Although some spectroscopic studies have revealed the close
relationship of NHCs with phosphine ligands, detailed elec-
7
tronic and steric investigations on this ligand class have not
been reported and remain largely unquantified. The importance
of quantifying steric and electronic effects of ligands has been
recognized in pioneering studies by Tolman on phosphines and
has had a major impact on the development of new and
2
ligands. Several beneficial uses of NHC-modified organome-
3
tallic systems in ruthenium-mediated olefin metathesis, iridium-
catalyzed hydrogenation and hydrogen transfer,4 platinum-
†
University of New Orleans.
Universit a` di Salerno.
University of Miami.
8
‡
§
improved phosphine ligands for catalysis. In the case of the
NHC ligands shown in Figure 1, surprising differences in
catalytic activities of the corresponding metal catalysts have
already been observed, but could not be rationalized or explained
(
1) Arduengo, A. J., III; Harlow, R. L.; Kline, M. J. Am. Chem. Soc. 1991,
13, 361-363.
1
(2) (a) Bourissou, D.; Guerret, O.; Gabbai, F. P.; Bertrand, G. Chem. ReV.
2000, 100, 39-91. (b) Herrmann, W. A. Angew. Chem., Int. Ed. 2002, 41,
1290-1309. (c) Yong, B. S.; Nolan, S. P. Chemtracts: Org. Chem. 2003,
16, 205-227.
(5) Mark o` , I. E.; St e´ rin, S.; Buisine, O.; Mignani, G.; Branlard, P.; Tinant, B.;
Declerq, J.-P. Science 2002, 298, 204-208.
(
3) (a) Weskamp, T.; Schattenmann, W. C.; Spiegler, M.; Herrmann, W. A.
Angew. Chem., Int. Ed. 1998, 37, 2490-2493. (b) Huang, J.; Stevens, E.
D.; Nolan, S. P.; Petersen, J. L. J. Am. Chem. Soc. 1999, 121, 2674-2678.
(6) (a) Grasa, G. A.; Viciu, M. S.; Huang, J.; Nolan, S. P. J. Org. Chem. 2001,
66, 7729-7737. (b) Grasa, G. A.; Viciu, M. S.; Huang, J.; Zhang, C.;
Trudell, M. L.; Nolan, S. P. Organometallics 2002, 21, 2866-2873. (c)
Viciu, M. S.; Germaneau, R. F.; Nolan, S. P. Org. Lett. 2002, 4, 4053-
4056. (d) Navarro, O.; Kelly, R. A., III; Nolan, S. P. J. Am. Chem. Soc.
2003, 125, 16194-16195.
(
c) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
9
53-956. For reviews, see: (d) Jafarpour, L.; Nolan, S. P. AdV. Organomet.
Chem. 2001, 46, 181-222. (e) Jafarpour, L.; Nolan, S. P. J. Organomet.
Chem. 2001, 617, 17-27. (f) Trnka, T. M.; Grubbs, R. H. Acc. Chem.
Res. 2001, 34, 18-29. (g) F u¨ rstner, A. Angew. Chem., Int. Ed. 2000, 39,
(7) (a) O¨ fele, K.; Herrmann, W. A.; Mihalios, D.; Elison, M.; Herdtweck, E.;
Scherer, W.; Mink, J. J. Organomet. Chem. 1993, 459, 177-184. (b) Denk,
K.; Sirsch, P.; Herrmann, W. A. J. Organomet. Chem. 2002, 649, 219-
224. (c) Chianese, A. R.; Li, X.; Janzen, M. C.; Faller, J. W.; Crabree, R.
H. Organometallics 2003, 22, 1663-1667.
3
013-3043.
(4) (a) Lee, H. M.; Jiang, T.; Stevens, E. D.; Nolan, S. P. Organometallics
2
001, 20, 1255-1258. (b) Hillier, A. C.; Lee, H. M.; Stevens, E. D.; Nolan,
S. P. Organometallics 2001, 20, 4246-4252. (c) Vasquez-Serrano, L. D.;
Owens, B. T.; Buriak, J. M. Chem. Commun. 2002, 2518-2519.
(8) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.
10.1021/ja0438821 CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 2485-2495
9
2485

A R T I C L E S
Dorta et al.
1
carbonyl ligands. Saturated complexes 9 and 11 result in H
NMR spectra with resonances at 3.15 and 3.59 ppm, respec-
tively, for the ring protons of the imidazole and the expected
signals for the mesityl and isopropylphenyl side chains. For these
1
3
saturated NHC complexes, the C NMR data give rise to
downfield signals for the carbonyl ligands (one resonance) and
one signal for the carbeneic carbon atom at significantly lower
field (ca. 220 ppm) when compared to that for the unsaturated
analogues. Elemental analyses for the complexes confirm their
composition. To unambiguously characterize these complexes
and to get a possible insight into fine structural differences
between unsaturated and saturated NHCs, X-ray quality crystals
were grown for Ni(CO)3(IMes) (8), Ni(CO)3(SIMes) (9), Ni-
Figure 1. Common unsaturated and saturated NHCs.
experimentally. Furthermore, although NHC ligands are believed
to strongly bind to late-transition metals, extensive experimental
determination of metal-NHC bonding energies has not been
carried out so far.
(
CO)3(IPr) (10), Ni(CO)3(SIPr) (11), as well as for Ni(CO)3-
To achieve a deeper understanding of the fundamental steric
and electronic factors characterizing the NHC ligand class and
to afford a direct comparison with the large array of data
available for tertiary phosphine ligands, we report here our
studies on the substitution reaction involving Ni(CO)4 and NHC
ligands 1-7. Most of these ligands give rise to the expected,
saturated complexes of general composition Ni(CO)3(NHC). The
electronic properties of these NHCs are compared to those of
tertiary phosphine ligands. The most bulky NHC ligands lead
to the isolation of novel, three-coordinate nickel carbonyl
systems. These unsaturated complexes have allowed, for the
first time, the experimental determination of bond dissociation
energies for this ligand class. All new Ni(0)-NHC complexes
were characterized by X-ray crystallography. In combination
with computational studies, this has enabled the establishment
of a method for the direct comparison of steric requirements of
the NHC and phosphine ligand classes. Furthermore, reactivity
studies showed that while the saturated complexes are essentially
inert toward displacement of the NHC and the carbonyl ligands,
the unsaturated analogues are highly reactive toward ligand
substitution and oxidative addition of substrates. Some of the
results herein have been previously published in a communica-
tion.9
10
(ICy) (12). Ball-and-stick representations are shown in Figure
2
(8 and 9), Figure 3 (10 and 11), and Figure 4 (complex 12),
and a comparison of selected bond distances and angles is
provided in Table 1. All complexes show the expected
tetrahedral geometry around the metal center with bond angles
between 105 and 112°. For all compounds, the Ni-C(NHC)
distances lie in the range of 1.96-1.98 Å, with both complexes
containing the saturated NHC ligands (compounds 9 and 11)
showing slightly shorter bond distances than their corresponding
analogues containing the saturated NHC ligands (1.961 Å for
9
versus 1.971 Å for 8; 1.962 Å for 11 versus 1.979 Å for 10).
These differences are not very significant and lie within
experimental error. The Ni-C(NHC) distance suggests a single-
bond character, in good accordance with their exclusive σ-donor
11
characteristics. As seen for other crystallographically charac-
terized complexes containing NHC ligands with aryl substitu-
ents, the R groups are disposed at an angle of 90° ((10) with
respect to the imidazole backbone plane, resulting in a favorable
steric arrangement with the metal center. In addition, the
saturated imidazole backbones in the SIMes and SIPr com-
3
pounds, due to the presence of two sp carbons in the
heterocyclic ring, show torsion angles of 6.4° (for complex 9)
and 2.6° (for complex 11). Interestingly, this value is signifi-
cantly lower than the one for free SIMes (13.4°).
Results and Discussion
Comparison of Electronic Factors Governing NHC and
Phosphine Ligands. To obtain a direct comparison of the
electronic factors governing NHCs with a large array of
phosphine compounds of general formula Ni(CO)3(PR3), we
recorded the carbonyl stretching frequencies of complexes 8-12.
The values obtained are presented in Table 2 alongside those
of representative Ni(CO)3(PR3) complexes. The data clearly
demonstrate that N-heterocyclic carbene ligands are better
Synthesis and Structure of Saturated Ni(CO)3(NHC).
Reaction of a hexane or THF solution containing IMes (1),
SIMes (2), IPr (3), SIPr (4), or ICy (5) with a slight excess of
Ni(CO)4 led to immediate evolution of CO. Subsequent workup
gave the saturated complexes Ni(CO)3(IMes) (8), Ni(CO)3-
(
(
SIMes) (9), Ni(CO)3(IPr) (10), Ni(CO)3(SIPr) (11), and Ni-
CO)3(ICy) (12) as off-white or beige solids in high yield (eq
1
).
t
σ-donors than even the most basic phosphine, P Bu3. Within
Ni(CO) + NHC f Ni(CO) (NHC) + CO
NHC ) IMes for 8, SIMes for 9, IPr for 10,
SIPr for 11, and ICy for 12
(1)
the NHC ligand class, ICy (5) is expected to be the most basic
ligand due to its alkyl-substituted nitrogen atoms (versus aryl
substitution in the other ligands). Overall, the electronic
differences for NHC ligands are relatively small, especially
compared to the substantial electronic differences seen in
phosphine ligands when moving from aryl- to alkyl-substituted
4
3
The 1H NMR spectrum of complexes 8, 10, and 12,
containing unsaturated NHC ligands, is characterized by a single
resonance at low field for the two imidazole protons (6.50-
6
.80 ppm), as well as signals characteristic for their correspond-
(10) Different crystal batches of 12 were measured but were disordered, and
the X-ray structure determination is therefore somewhat poor (R ) 0.0978
for data with I > 2σ(I) and R ) 0.1169 for all data). While important
information can be obtained for its connectivity, we do not include this
structure in the discussion of bond lengths and angles.
11) (a) Herrmann, W. A.; Elison, M.; Fischer, J.; Kocher, C.; Artus, G. R.
Angew. Chem., Int. Ed. 1995, 34, 2371-2374. (b) McGuinness, D. S.;
Cavell, K. J.; Skelton, B. W.; White, A. H. Organometallics 1999, 18,
1596-1605.
1
3
ing side-chain R groups. The C NMR spectra give rise to
characteristic low-field resonances for the carbeneic carbon at
around 195 ppm and show one low-field signal for the three
(
(
9) Dorta, R.; Stevens, E. D.; Hoff, C. D.; Nolan, S. P. J. Am. Chem. Soc.
2
003, 125, 10490-10491.
2486 J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005

Steric and Electronic Properties of N-Heterocyclic Carbenes
A R T I C L E S
Figure 2. Ball-and-stick representation of complexes 8 (left) and 9 (right). Hydrogen atoms, except for imidazole protons, are omitted for clarity.
Figure 3. Ball-and-stick representation of complexes 10 (left) and 11 (right). Hydrogen atoms, except for imidazole protons, are omitted for clarity.
tion of the bonding picture in LnM-NHC complexes. Our
experimental data suggest that electronic factors related to the
σ-donor abilities of these ligands are unlikely to play a major
role in the differences in catalytic activity for unsaturated versus
saturated NHCs.
Synthesis and Structure of Unsaturated Ni(CO)2(NHC)
Complexes. When reacting a hexane solution containing NHC
t
ligands IAd (6) and I Bu (7) with a slight excess of Ni(CO)4,
we observed the formation of orange-red solutions. Both
complexes were isolated in high yield after workup. Whereas
1
13
H NMR and C NMR spectral data did not permit unequivocal
Figure 4. Ball-and-stick representation of complex 12. Hydrogen atoms,
except for imidazole protons, are omitted for clarity.
structural assignment, significant infrared shifts to lower
frequencies were observed, indicating a possibly different
composition than Ni(CO)3(NHC) for these complexes. Indeed,
elemental analysis data for the orange-red microcrystalline
materials showed the composition to be in line with novel 16
phosphorus ligands. As an example, a difference of νCO of more
-
1
than 12 cm is observed when comparing the stretching
frequencies of Ni(CO)3(L) compounds for L ) PPh3 with those
t
electron complexes, Ni(CO)2(IAd) (13) and Ni(CO)2(I Bu) (14)
t
of the more basic, alkyl-substituted L ) P Bu3 (Table 2).
(eq 2).
Another interesting fact is the very small difference in stretching
frequency between IMes (1) and SIMes (2) in complexes 8 and
Ni(CO) + NHC f Ni(CO) (NHC) + 2 CO
(2)
4
2
9
and IPr (3) and SIPr (4) in complexes 10 and 11. In fact, the
t
saturated NHC ligands are even slightly less electron-donating
than their unsaturated analogues. It is worth noting that we have
recently observed similar trends in the relative bond disruption
enthalpies of ruthenium complexes involving the aforementioned
ligands.12 These experimental results clearly contradict the
common assumption that saturated NHC ligands are more
electron-donating and that metal complexes incorporating them
perform better in catalysis for this reason. The recent work of
NHC ) IAd for 13, and I Bu for 14
Such unsaturated nickel carbonyl systems are exceedingly
rare, and the only example known has been isolated as a mixture
14,15
of Ni(CO) (L) and Ni(CO) (L) compounds.
3
2
To unambigu-
ously confirm the molecular structure of these products, single
crystals of both complexes were grown by slow evaporation of
saturated diethyl ether (for 13) and pentane (for 14) solutions
and were subjected to X-ray crystallographic analyses. Ball-
1
3
Yates on the relative basicity of NHC in various solvents is
on note as some of the theoretical predictions fall in line with
our experimental data in terms of general donor property trends
of NHC but diverge in predictions of overall bonding interac-
tions when saturated and unsaturated NHCs are concerned.
Considering only ligand basicity may prove an oversimplifica-
t
and-stick diagrams of Ni(CO) (IAd) (13) and Ni(CO) (I Bu)
2
2
(14) Petz, W.; Weller, F.; Uddin, J.; Frenking, G. Organometallics 1999, 18,
6
19-626.
(
15) Carbonyl-free, three-coordinate Ni(0) complexes are not uncommon; for
some examples, see: (a) Wilke, G. Angew. Chem., Int. Ed. 1963, 2, 105-
1
1
08. (b) Bogdanovic, B.; Kroener, M.; Wilke, G. Justus Liebigs Ann. Chem.
966, 699, 1-23. (c) Jonas, K.; Heimbach, P.; Wilke, G. Angew. Chem.,
(
12) Hillier, A. C.; Sommer, W. J.; Yong, B. S.; Peterson, J. L.; Cavallo, L.;
Nolan, S. P. Organometallics 2003, 22, 4322-4326.
Int. Ed. 1968, 7, 949-950. (d) Gosser, L. W.; Tolman, C. A. Inorg. Chem.
1970, 9, 2350-2353. (e) Gabor, B.; Krueger, K.; Marczinke, B.; Mynott,
R.; Wilke, G. Angew. Chem., Int. Ed. 1991, 30, 1666-1668. (f) Mindiola,
D. J.; Hillhouse, G. L. J. Am. Chem. Soc. 2002, 124, 9976-9977.
(13) Magill, A. M.; Cavell, K. G.; Yates, B. F. J. Am. Chem. Soc. 2004, 126,
8
717-8724.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005 2487

A R T I C L E S
Dorta et al.
Table 1. Selected Bond Lengths and Angles for Ni(CO)3(IMes) (8), Ni(CO)3(SIMes) (9), Ni(CO)3(IPr) (10), and Ni(CO)3(SIPr) (11)
Complex 8
Complex 9
Complex 10
Complex 11
Ni1-C23
Ni1-C22
Ni1-C24
Ni1-C10
1.785(4)
1.797(4)
1.801(3)
1.971(3)
Ni26-C49
Ni26-C48
Ni26-C47
Ni26-C35
1.7706
1.7807
1.7916
1.9602
Ni1-C29
Ni1-C28
Ni1-C30
Ni1-C15
1.783(3)
1.796(2)
1.801(3)
1.979(2)
Ni1-C29
Ni1-C28
Ni1-C30
Ni1-C15
1.771(5)
1.787(4)
1.810(4)
1.962(4)
C23-Ni1-C22
C23-Ni1-C24
C22-Ni1-C24
C23-Ni1-C10
C22-Ni1-C10
C24-Ni1-C10
109.76(15)
109.89(16)
111.46(16)
108.46(14)
107.99(14)
109.22(14)
C49-Ni26-C48
C49-Ni26-C47
C48-Ni26-C47
C49-Ni26-C35
C48-Ni26-C35
C47-Ni26-C35
108.3
C29-Ni1-C28
C29-Ni1-C30
C28-Ni1-C30
C29-Ni1-C15
C28-Ni1-C15
C30-Ni1-C15
109.01(11)
109.59(11)
111.46(10)
105.34(9)
111.03(9)
110.22(9)
C29-Ni1-C28
C29-Ni1-C30
C28-Ni1-C30
C29-Ni1-C15
C28-Ni1-C15
C30-Ni1-C15
109.30(19)
106.25(17)
111.63(18)
108.01(17)
111.79(17)
109.66(16)
111.1
107.4
108.3
110.8
111.0
Table 2. IR Values for the Carbonyl Stretching Frequencies in Ni(CO)3(L) and Ni(CO)2(L) Complexes
ν
CO (A
1
);
νCO (E);
νCO (A
1
);
νCO (E);
ligand
complex
CH Cl
2 2
CH Cl
2
2
hexane
hexane
IMes (1)
SIMes (2)
IPr (3)
Ni(CO)3(IMes) (8)
Ni(CO)3(SIMes) (9)
Ni(CO)3(IPr) (10)
Ni(CO)3(SIPr) (11)
Ni(CO)3(ICy) (12)
Ni(CO)2(IAd) (13)
2050.7
2051.5
2051.5
2052.2
2049.6
1969.8
1970.6
1970.0
1971.3
1964.6
2054.0
2054.7
2055.1
2056.0
2052.4
2007.2
2009.7
1978.2, 1971.0
1980.0, 1972.3
1980.4, 1972.3
1982.6, 1973.7
1973.4, 1968.0
1926.3
SIPr (4)
ICy (5)
IAd (6)
t
t
I Bu (7)
Ni(CO)2(I Bu) (14)
1928.6
t
a
t
P Bu3
Ni(CO)3(P Bu3)
2056.1
2056.4
2068.9
1971
1973
1990
a
PCy3
PPh3
Ni(CO)3(PCy3)
Ni(CO)3(PPh3)
a
a
From ref 8.
(14), alongside selected bond lengths and angles, are presented
in Figures 5 (for 13) and 6 (for 14). Both complexes show only
slightly distorted trigonal structures with angles between 117
and 123°. The Ni-C(NHC) bond distances of 1.953(2) (13)
and 1.957(1) Å (14) are marginally shorter than those in the
saturated complexes 8-12, but significantly longer than for
those of the crystallographically characterized, homoleptic Ni-
(
1
IMes)2 compound, which shows bond lengths of 1.827 and
.830 Å. One of the carbon atoms from both side-chain R
1
6
Figure 5. Ball-and-stick representation of complex 13. Hydrogen atoms
are omitted for clarity. Selected bond lengths (Å) and angles (deg): Ni1-
C1 1.7583(19), Ni1-C2 1.761(2), Ni1-C3 1.9528(16); C1-Ni1-C2
123.01(10), C1-Ni1-C3 116.91(9), C2-Ni1-C3 119.93(10).
groups in 13 and 14 is approaching the metal center. Although
these Ni‚‚‚C(R) distances lie outside the range seen for agostic
1
7
interactions with first-row metal centers (Ni‚‚‚C(R) g 3 Å),
the structures give an explanation for the higher steric require-
ments of I Bu (6) and IAd (7). In fact, these two ligands are the
according to eq 4. Variable temperature spectroscopic data for
eq 4 are shown in Figure 7.
t
only ones in the series to contain R-carbon atoms in the side
chains that are bound to three additional C atoms. Whereas the
aromatic IMes (1), SIMes (2), IPr (3), and SIPr (4), as well as
ICy (5), can orient their R groups perpendicularly to the
imidazole plane and therefore minimize steric interactions with
the metal center, such an arrangement cannot be reached for
Ni(CO) (NHC) + 2 CO h Ni(CO) + NHC
(4)
2
4
t
NHC ) IAd or I Bu
Although eq 4 is most readily approached through the
intermediate Ni(CO) (NHC), no spectroscopic sign of this
3
t
IAd (6) and I Bu (7).
intermediate could be found for complexes of either IAd or
t
Experimental Determination of Ni-C(NHC) Bond Dis-
sociation Enthalpies (BDE). This work began as an attempt
to spectroscopically detect the saturated tricarbonyl adducts of
I Bu. This implies that Ni(CO)3(NHC) under CO has only a
t
fleeting existence for IAd and I Bu, and that the NHC ligands
are readily displaced by CO to form Ni(CO)4.
t
the I Bu and IAd compounds under CO pressure, illustrated for
In contrast, the stable complexes, Ni(CO)3(IMes) and Ni-
(CO)3(SIMes), are not carbonylated to Ni(CO)4 even under
pressure of CO (eq 5). Spectroscopic data under argon and under
pressure of CO [550 psi CO for (9) and 350 psi CO for (8)] are
shown in Figure 8.
Ni(CO)2(IAd) in eq 3 below:
Ni(CO) (IAd) + CO h Ni(CO) (IAd)
(3)
2
3
Instead of observing the equilibrium associated with eq 3,
t
net displacement of either I Bu or IAd by 2 mol of CO occurred
(
16) Arduengo, A. J., III; Gamper, S. F.; Calabrese, J. C.; Davidson, F. J. Am.
Chem. Soc. 1994, 116, 4391-4394.
(17) (a) Brookhart, M.; Green, M. L. H.; Wong, L. Prog. Inorg. Chem. 1988,
The complexes are essentially unchanged under these condi-
tions. Solutions left under pressure overnight show no sign of
36, 1-124. (b) Crabtree, R. H. Angew. Chem., Int. Ed. 1993, 32, 789-
05.
8
2488 J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005

Steric and Electronic Properties of N-Heterocyclic Carbenes
A R T I C L E S
C-Ni(CO)3. It should be pointed out that previous calorimetric
studies have shown that the enthalpy of binding for IMes
2
0
21
compared to that for IAd in both metals, and sulfur, is more
exothermic in the order of 10 kcal/mol. That trend is in keeping
with the qualitative observation for the nickel tricarbonyl
complexes, as well. Furthermore, these qualitative observations
are supported by theoretical calculations discussed later.
The sum of the first two bond dissociation energies for Ni-
(CO)4 (52 kcal/mol) can be combined with variable temperature
equilibrium data for the formation of Ni(CO)2(NHC) (approach-
ing the equilibrium in eq 4 from the right) to generate absolute
estimates for the Ni-C(NHC) bond strength in the dicarbonyl
complexes. Plots of ln Keq versus 1/T for formation of Ni(CO)2-
Figure 6. Ball-and-stick representation of complex 14. Hydrogen atoms,
except for imidazole protons, are omitted for clarity. Selected bond lengths
(
1
Å) and angles (deg): Ni1-C1 1.7489(13), Ni1-C2 1.7535(13), Ni1-C3
(
NHC) are shown in Figure 9 and lead to derived thermody-
namic data for the formation of Ni(CO)2(IAd) [∆H ) 10.2 (
.3 kcal/mol and ∆S ) 41 ( 3 cal/mol deg] and for the
.9569(11); C1-Ni1-C2 120.10(6), C1-Ni1-C3 117.25(5), C2-Ni1-
C3 122.36(5).
0
t
formation of Ni(CO)2(I Bu) [∆H ) 14.6 ( 0.4 kcal/mol and
∆
S ) 53 ( 5 cal/mol deg].
These data lead directly to estimates of the Ni-C(NHC) bond
strength in these dicarbonyls of 42 [for Ni-C(IAd)] and 37 kcal/
t
mol [for Ni-C(I Bu)]. The positive entropies of reaction are in
keeping with the release of 2 mol of CO upon uptake of the
carbene. The less positive value for binding of IAd compared
t
to that of I Bu may be due to a relatively greater loss of rotational
freedom in binding of IAd. Thus, the greater enthalpic preference
for binding of IAd is partially negated by a corresponding less
favorable entropy of binding.
t
In view of the fact that both I Bu and IAd were in equilibrium
with Ni(CO)4, reaction 7 was investigated.
Figure 7. Variable temperature FT-IR spectral data for Ni(CO)4 + IAd f
-
1
Ni(CO)2(IAd) + 2 CO. The band near 2046 cm is due to Ni(CO)4 and
t
t
increases with decreasing temperature.
Ni(CO) (I Bu) + IAd h Ni(CO) (IAd) + I Bu
(7)
2
2
-
1
Ni(CO)4 with its characteristic band at 2046 cm . In contrast,
In principle, since both carbene complexes establish an
equilibrium through intermediary Ni(CO)4, direct measurement
of eq 7 was deemed possible. In practice, even under argon
atmosphere with no sign of intermediate Ni(CO)4, eq 7 is
established within minutes at room temperature. The reaction
depicted in eq 7 can be approached from either side and yields
an estimate of Keq = 5 at 20 °C, in good agreement with
comparative data calculated from data in Figure 9.
t
Ni(CO)2(IAd) and Ni(CO)2(I Bu) are rapidly and quantitatively
converted to Ni(CO)4 under pressure of CO.
These observations allow some bracketing of the Ni-C(NHC)
bond strength in the tricarbonyl Ni(CO)3(IMes) complex and
the unstable proposed intermediate, Ni(CO)3(IAd). The most
recent gas-phase data for sequential Ni-CO bond strengths
are summarized in eq 6 (numbers are in kcal/mol):
1
8
The relatively rapid establishment of the equilibrium in eq 7
for the carbene exchange as well as the absence of detectable
Ni(CO)4 implies that an associative exchange as shown in eq 8
may be operative.
2
4
28
47
Ni(CO)4
98 CO + Ni(CO)3 98 CO + Ni(CO) 98
2
41
CO + Ni(CO)
98 CO + Ni (6)
The sum of the gas-phase bond strengths in Ni(CO)4 of 141.8
kcal/mol is based on combustion calorimetric measurements.
t
t
1
9
Ni(CO) (I Bu) + IAd h Ni(CO) (I Bu)(IAd) h
2 2
t
This value leads to an average Ni-CO bond dissociation energy
of ≈35 kcal/mol. There is more individual error in the
partitioning of these bond strengths than there is in the average,
but it is the initial first two dissociations that are most important
in the current system.
Using the value in eq 6 for the first bond dissociation energy
of Ni(CO)4 can provide qualitative insight into the magnitude
of the Ni-C(NHC) bond strength in Ni(CO)3(NHC) complexes.
The observation that Ni(CO)3(IMes) resists carbonylation under
CO pressure and that no evidence can be found for Ni(CO)3-
Ni(CO) (IAd) + I Bu (8)
2
Thus, despite severe steric crowding in a complex, such as
Ni(CO)2(I Bu)(IAd), the unsaturated Ni(CO)2(NHC) compounds
t
seem to be capable of undergoing associative exchange reac-
tions. This reactivity pattern might also be operative in the
interaction of Ni(CO)2(NHC) compounds with olefins (see
discussion below).
Steric Requirements of NHCs: Molecular Modeling
Studies for Ni(CO)2(NHC) versus Ni(CO)3(NHC). The strik-
t
(IAd) clearly implies that the (CO)3Ni-C(IMes) g 24 g (IAd)-
ing difference in reactivity of NHC ligands IAd and I Bu toward
(
18) Data taken from: Sunderlin, L. S.; Wang, D.; Squires, R. R. J. Am. Chem.
Soc. 1992, 114, 2788-2796.
(20) Huang, J.; Schanz, H.-J.; Stevens, E. D.; Nolan, S. P. Organometallics
1999, 18, 2370-2375.
(21) Huang, J.; Schanz, H.-J.; Nolan, S. P.; Capps, K. B.; Bauer, A.; Hoff, C.
D. Inorg. Chem. 2000, 38, 1042-1045.
(
19) Fischer, A. K.; Cotton, F. A.; Wilkinson, G. J. Am. Chem. Soc. 1957, 79,
2
044-2046.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005 2489

A R T I C L E S
Dorta et al.
Figure 8. High-pressure FT-IR experiments of Ni(CO)3(IMes) (8, left) and Ni(CO)3(SIMes) (9, right) with CO.
Table 3. CO and L BDE Values (in kcal/mol) in the Ni(CO)2(L)
and Ni(CO)3(L) Complexes (L ) NHC ligand or phosphine) and
Calculated %Vbur of the Free NHC and PR3 Ligands
BDE of CO
in Ni(CO) (L)
BDE of L
BDE of L
ligand
3
in Ni(CO)
3
(L)
in Ni(CO)
2
(L)
%Vbur
t
13.3 (15.6)a
10.3 (13.1)a
I Bu
24.0
21.8
20.4
41.1
40.2
38.5
38.0
39.6
22.7
26.7
44.3
45.1
46.5
46.5
47.2
45.4
46.1
46.3
25.7
30.0
37
38
37
26
27
29
30
23
17
22
26
30
t
SI Bu
IAd
IMes
SIMes
IPr
SIPr
ICy
PH3
PPh3
PCy3
P Bu3
a
7.6 (14.1)
28.3 (27.2)a
a
26.8 (26.4)
26.7 (27.2)a
a
25.6 (25.8)
27.0 (26.1)a
30.7
Figure 9. Van’t Hoff plots for Ni(CO)2(IAd) (13, above) and Ni(CO)2(It-
Bu) (14, below).
30.4
t
27.4
28.0
34.3
Ni(CO)4, producing unsaturated dicarbonyl complexes, seems
22
to be based on higher steric requirements of these two ligands.
a
Values in parentheses were obtained with the QM/MM technique.
To rationalize the experimental data, and to possibly afford a
direct steric comparison between NHCs and phosphine ligands,
di- and tricarbonyl nickel complexes were investigated by DFT
calculations. Geometries and energies of the Ni(CO)2(NHC) and
Ni(CO)3(NHC) structures were optimized for R ) I Bu, SI Bu,
IAd, IMes, SIMes, IPr, SIPr, and ICy (both with full QM and
with QM/MM techniques). For the phosphine-based systems,
full QM geometry optimizations were performed for the Ni-
t
t
(
CO)2(PR3) and Ni(CO)3(PR3) structures, with R ) H, Ph, Cy,
t
and Bu. The calculated BDE values of CO and L in Ni(CO)3-
L) and Ni(CO)2(L) systems (L ) NHC or PR3) are reported in
Table 3.
The calculated BDE values indicate that CO is weakly bound
(
Figure 10. Bond distances and angles for unsaturated (a) and saturated
t
(
b) complexes with I Bu.
t
t
to the I Bu-, SI Bu-, and IAd-based complexes (BDE ≈ 10 kcal/
mol), whereas CO is much more strongly bound to the IPr, SIPr,
ICy, IMes, and SIMes complexes (BDE > 25 kcal/mol). These
results are in good qualitative agreement with the experimental
fact that Ni(CO)2(NHC) complexes are observed for NHC )
t
ligand in the Ni(CO)2(IAd) and Ni(CO)2(I Bu) complexes (46.5
and 44.3 kcal/mol, respectively) are in excellent agreement with
the experimental values (43 ( 3 and 39 ( 3 kcal/mol,
respectively).
t
IAd and I Bu, while Ni(CO)3(NHC) complexes are observed
To rationalize the different behavior shown by the two groups
of NHC-based complexes, we calculated the BDE of CO in the
Ni(CO)3(NHC) systems with a combined QM/MM approach.
for the other NHC ligands. The BDEs calculated for the NHC
ligands show a similar trend since the BDEs of the NHC ligand
in the I Bu-, SI Bu-, and IAd-based Ni(CO)3(NHC) complexes
are roughly 20 kcal/mol lower than that in the IMes, SIMes,
IPr, SIPr, and ICy complexes. The BDEs calculated for the NHC
t
t
t
Within this approach, all the unsaturated (IAd, I Bu, ICy, IPr,
t
and IMes) and saturated (SI Bu, SIPr, and SIMes) NHC ligands
share the same QM part (see the Computational Details), and
different behavior can be related to their different steric
properties. The numbers in Table 3 indicate that the QM/MM
approach substantially replicates the full QM calculations. Since
the QM/MM BDEs are very close to QM values, the difference
between the two groups of NHC ligands can be attributed to
steric factors exclusively.
(
22) The concept of “steric pressure” noted by Power and others in the use of
m-terphenyl ligands (isosteric with some NHC presented here) should be
brought to the reader’s attention. Similar steric pressures, as the one
t
described here for IAd and I Bu, lead to the stabilization and formation of
monomeric versus dimeric or dimeric versus trimeric architectures in main-
group-m-terphenyl compounds. For example, see: Hino, S.; Olmstead, M.;
Phillips, A. D.; Wright, R. J.; Power, P. P. Inorg. Chem. 2004, 43, 7346-
7
352 and references cited therein.
2
490 J. AM. CHEM. SOC. VOL. 127, NO. 8, 2005
9

Steric and Electronic Properties of N-Heterocyclic Carbenes
A R T I C L E S
Figure 11. Bond distances and angles for unsaturated (a) and saturated (b) complexes with IPr.
To quantify the steric factors characterizing these ligands,
we measured the amount of volume of a sphere centered on
the metal, buried by overlap with atoms of the various NHC
and phosphine ligands, %Vbur. The volume of this sphere would
represent the space around the metal atom that must be shared
by the different ligands upon coordination. Of course, the bulkier
a specific ligand, the larger the amount of that sphere will be
occupied by the ligand, that is, greater %Vbur. To have an aseptic
estimate of the bulkiness of the various ligands, we examined
the DFT optimized geometries of the free ligands and positioned
the putative metal atom 2 Å from the coordinating C atom for
the NHC ligand class and 2.28 Å from the coordinating P atom
for the PR3 ligands.23 The radius of the sphere was chosen to
be 3 Å. A compilation of %Vbur values is presented in Table 3.
agreement with the common assumption that NHC ligands bind
more tightly to LTMs than phosphines.
Most importantly, calculation of the %Vbur of the PR3 systems
allows a direct comparison of the steric requirements of the NHC
and phosphine ligand families. The calculated values for %Vbur
t
show that sterically, the very bulky phosphine ligand, P Bu3, is
best compared to IPr and SIPr; PCy3 gives a %Vbur value close
to the ones found for IMes and SIMes, whereas PPh3 has a
%Vbur similar to that for ICy. Incidentally, the steric requirements
for tertiary phosphines and NHCs we find here coincide with
catalytic data in the literature. For example, it has been shown
that for palladium-catalyzed coupling reactions of aryl chlorides,
where 12-electron L-Pd(0) compounds are involved, bulky
t
ligands, such as P Bu3 and IPr/SIPr, outperform PCy3 and IMes/
t
t
The larger values of %Vbur of the I Bu, SI Bu, and IAd ligands,
relative to the IMes, SIMes, IPr, SIPr, and ICy ligands, are in
qualitative agreement with the different BDEs we calculated
for the two groups of NHC ligands.
6,24
SIMes. On the other hand, in ruthenium-catalyzed metathesis
reactions, where the active catalyst is thought to involve a (L)-
RuCl2(dCR2) species, the ligands performing best are PCy3 and
SIMes/IMes, with more bulky phosphine and NHC ligands
Geometrical inspection of the calculated structures for the
3d-g
being considerably less active.
Sterically most demand-
t
t
exemplary cases, Ni(CO)2(I Bu) and Ni(CO)3(I Bu), supports the
t
t
ing ligands, I Bu, SI Bu, and IAd, are all significantly more
bulky than P Bu3, a fact that in turn might lead to different
t
previous findings. As reported in Figure 10, Ni(CO)3(I Bu)
t
shows several steric interactions, indicated by short distances,
behavior in reactivity and catalysis of metals incorporating these
between the tert-butyl group and the CO molecules, relative to
25
ligands.
t
t
Ni(CO)2(I Bu). These strong steric interactions in Ni(CO)3(I -
Bu) result in the severe distortion of one of the Ni-C-O angles
from the ideal value of 180° to assume a value of 169.4°. In
contrast, Ni(CO)3(IPr) is characterized by very few short
distances between the NHC ligand and the CO molecules, which
explains its stability (Figure 11). Incidentally, the calculated
geometries are in an overall good agreement with the corre-
sponding X-ray structures.
Reactivity of Ni(CO)n(NHC) Complexes. Reactivity toward
substitution of the carbonyl ligands in the Ni(CO)n(NHC) (n )
2
or 3) complexes was also investigated by exploring their
interaction with different olefins. The results obtained for the
saturated Ni(CO)3(NHC) systems were not surprising, showing
that these complexes are essentially inert toward CO substitution
at room temperature with the substrates used (see below). The
unsaturated Ni(CO)2(NHC) compounds 13 and 14 did also not
react with olefins, such as styrene and tetracyanoethylene. In
With regard to a comparison between NHC- and PR3-based
systems, we note that the CO BDE values in the PR3 systems
of Table 3 are quite similar, although the group R of PR3 goes
from extremely small, for R ) H, to extremely bulky, for R )
t
contrast, Ni(CO)2(I Bu) (14) reacted readily with the small and
electron-poor trifluoropropene. When an orange-red pentane
t
solution of Ni(CO)2(I Bu) (14) was saturated with C3F3H3
t
Bu. This clearly indicates that steric effects play almost no role
(excess), an immediate color change to bright yellow was
in determining the stability of the Ni(CO)3(PR3) systems.
Moreover, the BDEs of the CO in the PR3 systems and in the
less bulky NHC systems are similar, which suggests a high
stability of the PR3-based tricarbonyl systems. The BDEs we
calculated for the NHC ligands are remarkably higher than the
BDEs we calculated for the PR3 ligands, a finding that is in
observed with concomitant release of CO. Workup yielded the
unsaturated, monocarbonyl complex Ni(CO)(C3H3F3)(I Bu) (15)
t
selectively and in high yield. The complex arising from double
t
substitution of CO, Ni(C3H3F3)2(I Bu) (eq 9), was not formed.
-1
The carbonyl stretching frequency of 15 is found at 1982 cm .
1
13
19
H, C, and F NMR spectra are consistent with its formulation,
(
23) This value reflects the bond distances found in crystallographically
characterized complexes of this type: (a) Pickardt, J.; Rosch, L.; Schumann,
H. Z. Anorg. Allg. Chem. 1976, 426, 66-76. (b) Baumgartner, T.; Moors,
P.; Nieger, M.; Niecka, E. Organometallics 2002, 21, 4919-4926. (c)
Gudat, D.; Bajorat, V.; Hap, S.; Niegler, M.; Schroder, G. Eur. J. Inorg.
Chem. 1999, 1169-1174.
t
(24) For a recent review on the use of P Bu
3
in these transformations, see: Littke,
A. F.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 4176-4211.
(25) We have very recently taken advantage of these findings in a Rh-NHC
system: Dorta, R.; Stevens, E. D.; Nolan, S. P. J. Am. Chem. Soc. 2004,
126, 5054-5055.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005 2491

A R T I C L E S
Dorta et al.
relevant bond lengths and angles, are found in Figures 12 (for
1
6) and 13 (for 17). In both cases, distorted square-planar
coordinations around the nickel center are observed. The halide
anions are located cis to the carbene ligand, and the allyl moiety
3
is η -coordinated to the nickel with one terminal carbon trans
to the carbene and the second terminal carbon trans to the halide
ligand. In both complexes, the Ni-Callyl bonds opposite the NHC
moiety are elongated as a result of a strong trans effect of the
NHC ligand. Finally, the Ni-C(NHC) bonds are somewhat
shorter than those for the Ni(0) complexes described above.
Figure 12. Ball-and-stick representation of complex 16. Hydrogen atoms
omitted for clarity. Selected bond lengths (Å) and angles (deg): Ni1-C1
1.926(2), Ni1-C15 1.981(3), Ni1-C16 1.993(2), Ni1-C14 2.058(3), Ni1-
Cl1 2.2322(7); C1-Ni1-C15 132.46(11), C1-Ni1-C16 95.92(11), C15-
Ni1-C16 41.06(12), C1-Ni1-C14 167.35(12), C15-Ni1-C14 39.48(12),
C16-Ni1-C14 72.60(12), C1-Ni1-Cl1 97.88(7), C15-Ni1-Cl1 126.75-
(9), C16-Ni1-Cl1 166.04(9), C14-Ni1-Cl1 93.46(9).
Conclusions
To gain insight into electronic and steric factors governing
NHC ligands and to compare them with the widely used
phosphine ligand class, we have performed a detailed experi-
mental and computational study on the interaction of the most
commonly used monodentate NHC ligands with Ni(CO)4. In
most of the cases, saturated Ni(CO)3(NHC) were obtained. IR
carbonyl stretching frequencies clearly showed that NHCs are
better electron donors than the most basic tertiary phosphines.
Within the NHC ligand class, saturated NHC ligands are not
better donors than their unsaturated analogues and alkyl-
substituted NHCs are only marginally more electron-donating
than their aryl-substituted counterparts. Reaction of IAd and
Figure 13. Ball-and-stick representation of complex 17. Hydrogen atoms
omitted for clarity. Selected bond lengths (Å) and angles (deg): Br1-Ni1
t
I Bu with Ni(CO)4 gave rise to unsaturated Ni(CO)2(NHC)
2
.3970(8), Ni1-C1 1.957(4), Ni1-C15 1.987(5), Ni1-C16 2.021(5), Ni1-
compounds, indicating that both these ligands are substantially
C14 2.096(5); C1-Ni1-C15 132.8(2), C1-Ni1-C16 95.8(2), C15-Ni1-
C16 41.3(3), C1-Ni1-C14 167.4(2), C15-Ni1-C14 38.3(3), C16-Ni1-
C14 72.4(3), C1-Ni1-Br1 97.61(11), C15-Ni1-Br1 126.76(19), C16-
Ni1-Br1 166.49(17), C14-Ni1-Br1 94.2(2).
t
more bulky than P Bu3. Due to the unsaturated nature of these
Ni(CO)2(NHC) complexes, both the NHC and the carbonyl
ligands could be displaced. Reacting Ni(CO)2(NHC) with CO
afforded the starting Ni(CO)4 complex quantitatively and
allowed for the determination of Ni-C(NHC) bond dissociation
energies. Substitution of one carbonyl ligand was observed when
reacting Ni(CO)2(NHC) with trifluoropropene and yielded the
unsaturated Ni(CO)(C3H3F3)(NHC) compound selectively. Both
CO ligands were displaced when Ni(CO)2(NHC) was treated
with either allyl chloride or allyl bromide. The resulting allyl
halide complexes of general formula NiX(C3H5)(NHC) were
obtained through oxidative addition of these substrates.
and elemental analysis confirmed the composition of Ni(CO)-
t
(
C3H3F3)(I Bu) (15).
To shed light on the experimental results, we performed DFT
calculations on a series of Ni(CO)2(L) and Ni(CO)3(L) com-
pounds (L ) NHC or PR3). In line with experimental results,
the calculations showed that both Ni(CO)3(IAd) and Ni(CO)3-
Substitution of both CO ligands is achieved when the
unsaturated compounds 13 and 14 are reacted with a slight
excess of allyl chloride or allyl bromide. In fact, these substrates
add readily to these complexes, giving the allyl halide Ni(II)
complexes 16-19 in quantitative yield via oxidative addition
t
(I Bu) are inherently unstable, as seen in the BDE values for
both the CO and NHC ligands. Furthermore, the Ni-PR3 BDE
values obtained are generally lower than those for Ni-NHC,
in agreement with the common assumption that NHC ligands
bind more tightly to LTMs. Most importantly, we have
established a method to quantify the steric factors characterizing
NHCs and phosphines, enabling a direct comparison between
these ligand classes. The values for the different ligands are
measured as %Vbur. All values obtained are in excellent
t
t
of the allyl halides. NiCl(C3H5)(I Bu) (16), NiBr(C3H5)(I Bu)
(
17), NiCl(C3H5)(IAd) (18), and NiBr(C3H5)(IAd) (19) were
synthesized according to eq 10 and were fully characterized by
NMR spectroscopy and elemental analysis. In addition, crystals
suitable for an X-ray diffraction analysis for complexes 16 and
7 were grown by slow evaporation of concentrated acetone
solutions. Ball-and-stick views of these complexes, including
1
2492 J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005

Steric and Electronic Properties of N-Heterocyclic Carbenes
A R T I C L E S
t
agreement with experimental data and show that IAd and I Bu
are substantially more bulky than P Bu3 and other NHC ligands.
Synthesis of Ni(CO)
3
(IPr) (10). A THF solution (5 mL) of Ni-
t
(CO) (143 mg, 0.84 mmol) was added to a stirred THF solution (5
4
mL) containing IPr (271 mg, 0.70 mmol). The resulting pale yellow
solution was left stirring at room temperature for 2 h, and then the
volatiles were removed in vacuo, yielding a beige solid. The compound
was washed with cold hexane (-50 °C, 2 × 3 mL), affording complex
10 as an off-white solid. X-ray quality crystals were obtained by cooling
a saturated hexane solution to -50 °C. Yield: 320 mg (86%). Anal.
Experimental Section
General Considerations. All reactions were carried out using
standard Schlenk techniques under an atmosphere of dry argon or in
MBraun gloveboxes containing dry argon and less than 1 ppm oxygen.
Solvents were distilled from appropriate drying agents or were passed
through an alumina column in an MBraun solvent purification system.
Other anhydrous solvents were purchased from Aldrich and degassed
prior to use by purging with dry argon and were kept over molecular
sieves. Solvents for NMR spectroscopy were degassed with argon and
dried over molecular sieves. NMR spectra were collected on a 400
MHz Varian Gemini spectrometer. Infrared spectra were recorded on
a PE 2000 FT-IR spectrometer. Elemental analyses were performed
by Desert Analytics, Tucson, AZ, and by Robertson Microlit Labs.
Carbene ligands 1-7 were synthesized following literature procedures.26
Calcd for C30
Found: C, 67.55; H, 6.76; N, 5.09. H NMR (C
36 2 3
H N NiO (MW 531.31): C, 67.82; H, 6.83; N, 5.27.
1
6 6
D , 400 MHz, δ):
7.02-7.16 (m, 6H, IPr-CH-Ar), 6.50 (s, 2H, NCHdCHN), 2.70
(septet, 4H, IPr-CH), 1.25 (d, 12H, IPr-CH3), 0.94 (d, 12H, IPr-
CH3). 13C NMR (C
6 6
D , 100.6 MHz, δ): 198.21 (s, N-C-N), 197.64
(s, CO), 146.36 (s, IPr-C), 138.09 (s, IPr-C), 130.58 (s, IPr-C),
124.57 (s, IPr-C), 123.73 (s, NCHdCHN), 29.05 (s, IPr-CH), 25.79
-
1
(s, IPr-CH
3
), 23.09 (s, IPr-CH ), 17.93. IR νCO (hexane, cm ):
3
2 2
2055.1 (s), 1980.4 (vs), 1972.3 (vs). IR νCO (CH Cl
, cm-1): 2051.5
(s), 1970.0 (vs).
Ni(CO)
been taken while manipulating the EXTREMELY TOXIC Ni(CO)
manipulations have been carried out in the glovebox, using additional
protective gloves. Ni(CO) was constantly maintained at -50 °C. The
solutions containing the NHC ligands were cooled to -50 °C, and Ni-
CO) was added via syringe at the same temperature. The slight excess
of Ni(CO) at the end of each reaction (carried out in Schlenk glassware
4
was graciously supplied by Dupont CR&D. Special care has
3 4
Synthesis of Ni(CO) SIPr (11). A THF solution (5 mL) of Ni(CO)
4
. All
(143 mg, 0.84 mmol) was added to a stirred THF solution (5 mL)
containing SIPr (272 mg, 0.70 mmol). The resulting pale yellow solution
was left stirring at room temperature for 2 h, and then the volatiles
were removed in vacuo, yielding a beige solid. The compound was
washed with cold hexane (-45 °C, 2 × 3 mL), affording 312 mg (84%
yield) of the desired complex. X-ray quality crystals were obtained by
cooling a saturated hexane solution to -50 °C. Anal. Calcd for
C H N NiO (MW 533.33): C, 67.56; H, 7.18; N, 5.25. Found: C,
4
(
4
4
inside the glovebox) was evaporated and trapped into a toluene solution
containing phosphines.
3
0
38
2
3
1
Synthesis of Ni(CO)
1000 mg, 3.290 mmol) was added dropwise to a hexane solution (4
mL) of Ni(CO) (674 mg, 3.948 mmol). Evolution of CO was observed
3
(IMes) (8). A hexane solution (20 mL) of IMes
67.31; H, 7.20; N, 5.08. H NMR (C
(m, overlapping with C , 6H, SIPr-CH-Ar), 3.59 (s, 4H,
NCH CH N), 3.32 (septet, 4H, SIPr-CH), 1.52 (d, 12H, J ) 7.2 Hz,
3 6 6
SIPr-CH ), 1.25 (d, 12H, J ) 6.8 Hz, SIPr-CH ). C NMR (C D ,
6 6
D , 400 MHz, δ): 7.22-7.37
(
6 6
D
4
2
2
13
with concomitant formation of a partial off-white precipitate. The pale
suspension was stirred at room temperature for 1 h, and the volatiles
were removed in vacuo. The solid was collected, washed with cold
pentane (-50 °C, 2 × 5 mL), and dried in vacuo to give complex 8 as
a beige solid. Crystals were obtained by slow evaporation of a saturated
3
100.6 MHz, δ): 223.16 (s, small, N-C-N), 198.16 (s, CO), 147.47
(s, SIPr-C), 138.65 (s, SIPr-C), 129.82 (s, SIPr-C), 125.09 (s, SIPr-
C), 53.62 (s, NCH
24.10 (s, SIPr-CH
1973.7 (vs). IR νCO (CH
Synthesis of Ni(CO)
mL) containing ICy (240 mg, 1.033 mmol) to a hexane solution (5
mL) of Ni(CO) (229 mg, 1.343 mmol) under stirring led to the
2
CH
). IR νCO (hexane, cm ): 2056.0 (s), 1982.6 (vs),
Cl
, cm-1): 2052.2 (s), 1971.3 (vs).
(ICy) (12). Addition of a hexane solution (5
2
3
N), 29.09 (s, SIPr-CH), 26.25 (s, SIPr-CH ),
-
1
3
pentane solution. Yield: 1397 mg (95%). Anal. Calcd for C24
NiO (MW 447.15): C, 64.46; H, 5.41; N, 6.26. Found: C, 64.06; H,
.58; N, 5.86. H NMR (C
CH-Ar), 6.25 (s, 2H, NCHdCHN), 2.12 (s, 6H, IMes-CH
24 2
H N -
2
2
3
3
1
5
6
D
6
, 400 MHz, δ): 6.81 (s, 4H, IMes-
), 2.02
, 100.6 MHz, δ): 198.73 (s,
N-C-N), 194.30 (s, CO), 139.22 (s, IMes-C), 138.15 (s, IMes-C),
3
4
1
3
(
s, 12H, IMes-CH
3
). C NMR (C
6
D
6
evolution of CO, and the resulting brownish solution was stirred at
room temperature for 2 h. Slow evaporation of the solvent resulted in
a color change to blue. The solid was washed with cold pentane (-50
°C, 2 × 3 mL), affording complex 12 as a blue powder. X-ray quality
crystals were obtained by cooling a saturated pentane solution to -50
1
2
2
35.76 (s, IMes-C), 129.75 (s, IMes-C), 122.13 (s, NCHdCHN),
-
1
3
1.39 (s, IMes-CH ), 17.93 (s, IMes-CH
3
). IR νCO (hexane, cm ):
054.0 (s), 1978.2 (vs), 1971.0 (vs). IR νCO (CH
2
Cl
2
, cm-1): 2050.7
(s), 1969.8 (vs).
°C. Yield: 340 mg (88%). Anal. Calcd for C18
H
24
N
2
3
NiO (MW
3
7
74.11): C, 57.64; H, 6.45; N, 7.47. Found: C, 57.86; H, 6.58; N,
3
Synthesis of Ni(CO) (SIMes) (9). Addition of a hexane solution (3
1
.43. H NMR (C
D
6 6
, 400 MHz, δ): 6.53 (s, 2H, NCHdCHN), 4.68
-Cy), 1.54 (d, 4H,
-Cy), 1.44 (d, 2H, J ) 12.4 Hz, CH -Cy), 1.27
-Cy), 1.12 (m, 4H, CH -Cy), 0.88 (m, 2H, CH -Cy).
, 100.6 MHz, δ): 199.48 (s, N-C-N), 188.04 (s, CO),
17.39 (s, NCHdCHN), 60.10 (s, CH-Cy), 34.35 (s, CH -Cy), 26.08
mL) containing SIMes (300 mg, 0.980 mmol) to a hexane solution (3
mL) of Ni(CO) (198 mg, 1.159 mmol) under stirring led to the
(m, 2H, CH-Cy), 1.85 (d, 4H, J ) 10.0 Hz, CH
2
4
J ) 12.8 Hz, CH
2
2
evolution of CO. The transparent solution was stirred at room
temperature for 2 h, resulting in partial precipitation of a white solid.
Subsequently, the volatiles were removed in vacuo, yielding a white
solid, and the compound was washed with cold pentane (-50 °C, 2 ×
(
m, 4H, CH
2
2
2
1
3
6 6
C NMR (C D
1
2
-
1
(s, CH
2
-Cy), 25.87 (s, CH
2
-Cy). IR νCO (hexane, cm ): 2052.4 (s),
3
mL), affording complex 9 as a white powder. X-ray quality crystals
1
973.4 (vs), 1968.0 (vs). IR νCO (CH
2
Cl
2
, cm-1): 2049.6 (s), 1964.6
were obtained from a saturated hexane solution. Yield: 400 mg (91%).
Anal. Calcd for C24 NiO (MW 449.17): C, 64.18; H, 5.83; N,
.24. Found: C, 63.97; H, 5.70; N, 6.06. H NMR (C
δ): 6.87 (s, 4H, SIMes-CH-Ar), 3.15 (s, 4H, NCH
(
vs).
Synthesis of Ni(CO)
(15 mL) containing IAd (1768 mg, 5.253 mmol) to a hexane (15 mL)
solution of Ni(CO) (1166 mg, 6.829 mmol) under stirring led to
H N
26 2
3
1
6
6
D
6
, 400 MHz,
-CH N), 2.20
). C NMR (C
00.6 MHz, δ): 220.40 (s, N-C-N), 198.73 (s, CO), 138.54 (s,
SIMes-C), 138.35 (s, SIMes-C), 136.67 (s, SIMes-C), 130.17 (s,
SIMes-C), 50.67 (s, NCH CH N), 21.38 (s, SIMes-CH ), 18.02 (s,
SIMes-CH ). IR νCO (hexane, cm ): 2054.7 (s), 1980.0 (vs), 1972.3
vs). IR νCO (CH Cl
, cm-1): 2051.5 (s), 1970.6 (vs).
2
IAd (13). Slow addition of a hexane solution
2
2
1
3
(
1
s, 12H, SIMes-CH
3
), 2.12 (s, 6H, SIMes-CH
3
6 6
D ,
4
evolution of CO and a color change to orange. The resulting solution
was stirred at room temperature for 2 h during which time an orange
precipitate formed. The solvent was evaporated, and the solid was
washed with cold pentane (-50 °C, 2 × 7 mL), affording complex 13
as an orange powder. Yield: 2257 mg (95%). X-ray quality crystals
were obtained by slow evaporation of a saturated diethyl ether solution
2
2
3
-
1
3
(
2
2
containing 13. Anal. Calcd for C25
H, 7.15; N, 6.21. Found: C, 66.21; H, 7.48; N, 6.17. H NMR (C D ,
32 2 2
H N NiO (MW 451.23): C, 66.54;
(
26) Arduengo, A. J., III; Krafczyk, R.; Schmutzler, R.; Craig, H. A.; Goerlich,
1
J. R.; Marshall, W. J.; Unverzagt, M. Tetrahedron 1999, 55, 14523-14534.
6 6
J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005 2493

A R T I C L E S
Dorta et al.
4
1
1
00 MHz, δ): 6.75 (s, 2H, NCHdCHN), 2.28 (s, 12H, IAd-CH
2
),
CHN), 5.02 (sept, J ) 6.5 Hz, 1H), 3.74 (d, J ) 7.0 Hz, 1H), 2.87 (d,
1
3
t
.93 (s, 6H, IAd-CH), 1.50 (m, 12H, IAd-CH
2
). C NMR(C
6
D
6
,
J ) 14.0 Hz, 1H), 2.32 (d, J ) 6.5 Hz, 1H), 1.90 (s, 9H, I Bu-CH
3
),
6
,
t
13
00.6 MHz, δ): 198.17 (s), 188.05 (s), 116.12 (s), 57.56 (s), 43.39
1.63 (s, 9H, I Bu-CH
3
), 1.29 (d, J ) 12.0 Hz, 1H). C NMR (C
6
D
-
1
(s), 36.41 (s), 30.42 (s). IR νCO (hexane, cm ): 2007.2 (s), 1926.3
100.6 MHz, δ): 179.36 (s, N-C-N), 119.65 (s, NCHdCHN), 119.59
t
t
(vs).
(s, NCHdCHN), 105.61(s), 64.88 (s), 58.91 (s, I Bu-C), 58.51 (s, I -
t
t
t
Synthesis of Ni(CO)
2
I Bu (14). Dropwise addition of a hexane
Bu-C), 46.50 (s), 32.36 (s, I Bu-CH
Synthesis of NiCl(C )(IAd) (18). A pentane solution (10 mL) of
Ni(CO) (IAd) (13) (60 mg, 0.133 mmol) was added to allyl chloride
3
), 31.97 (s, I Bu-CH
3
).
t
solution (17 mL) containing I Bu (1000 mg, 5.54 mmol) to a hexane
solution (3 mL) of Ni(CO) (1230 mg, 7.20 mmol) under stirring led
3 5
H
4
2
to the evolution of CO and a color change from colorless to orange.
The solution was stirred at room temperature for 2 h, resulting in partial
precipitation of a red solid. Subsequently, the volatiles were removed
in vacuo, yielding a red solid, and the compound was washed with
cold pentane (-50 °C, 2 × 10 mL), affording the desired complex 14
as a red solid. X-ray quality crystals were obtained by slow evaporation
of a saturated pentane solution containing 14. Yield: 1.57 g (96%).
(31 mg, 0.399 mmol) leading to slow formation of a flaky, yellowish
precipitate. The suspension was stirred under argon at room temperature
for 6 h, during which time a pale precipitate formed. The pentane
solution was decanted, the solid washed with additional pentane (3 ×
5 mL), and dried in vacuo, yielding the desired complex 18 as a pale
1
orange solid. Yield: 61 mg (98%). H NMR (C
6 6
D , 400 MHz, δ): 6.82
(s, 1H, NCHdCHN), 6.74 (s, 1H, NCHdCHN), 5.29 (m, 1H), 3.70
Anal. Calcd for C13
H
20
N
2
NiO
2
(MW 295.00): C, 52.93; H, 6.83; N,
(d, J ) 7.6 Hz, 1H), 3.07-1.50 (different d and m, 32H), 1.37 (d, J )
1
13
9
.50. Found: C, 52.08; H, 6.94; N, 9.38. H NMR (C
6
D
6
, 400 MHz,
). C NMR
12.0 Hz, 1H). C NMR (C
D
6 6
, 100.6 MHz, δ): 179.01 (s, N-C-N),
t
13
δ): 6.60 (s, 2H, NCHdCHN), 1.40 (m, 18H, I Bu-CH
3
118.03 (d, J ) 4.8 Hz, NCHdCHN), 105.68 (s), 65.52 (s), 59.78 (s,
IAd-C), 59.30 (s, IAd-C), 44.93 (s, IAd-CH), 44.56 (s, IAd-CH),
(
C D , 100.6 MHz, δ): 198.17 (s), 189.48 (s), 117.38 (s), 57.16 (s),
6 6
-
1
3
0.24 (s). IR νCO (hexane, cm ): 2009.7 (s), 1928.6 (vs).
44.47 (s), 36.60 (s, IAd-CH
CH ), 30.66 (s, IAd-CH ).
Synthesis of NiBr(C H )(IAd) (19). Addition of a pentane solution
2
), 36.53 (s, IAd-CH ), 30.77 (s, IAd-
2
t
2
2
Synthesis of Ni(CO)(C
3
H
3
F
3
)(I Bu) (15). A pentane solution (12
t
mL) of Ni(CO)
2
(I Bu) (14) (200 mg, 0.678 mmol) was treated with an
3
5
(5 mL) containing allyl bromide (48 mg, 0.399 mmol) to a pentane
excess of gaseous trifluoropropene. The solution turned yellow im-
mediately with concomitant evolution of CO. The yellow solution was
stirred for 30 min. The volatiles were evaporated in vacuo, yielding a
yellow powder which was washed with small portions of cold pentane
to give the desired complex 15 as a yellow solid. Yield: 214 mg (87%).
solution (5 mL) of Ni(CO) (IAd) (13) (60 mg, 0.133 mmol) led to the
2
precipitation of a flaky, yellow solid within 20 min. The suspension
was stirred at room temperature for 5 h, decanted, and washed with
pentane (3 × 5 mL). Subsequently, the volatiles were removed in vacuo,
yielding the desired complex 19 as a pale orange solid. Yield: 64 mg
23 2 3
Anal. Calcd for C15H N NiOF (MW 363.05): C, 49.63; H, 6.39; N,
1
1
9
(93%). H NMR (C D , 400 MHz, δ): 6.83 (s, 1H, NCHdCHN), 6.75
7
.72. Found: C, 49.69; H, 6.55; N, 7.44. F NMR (C
6
D
6
, 376.3 MHz,
, 500 MHz, δ): 6.59 (s, 1H, NCHd
CHN), 3.67 (dd, J ) 12.0 and 8.7 Hz, 1H, CH dCHCF ), 2.06 (d, J
12.5 Hz, 1H, CH dCHCF ), 1.71 (d, J ) 9.0 Hz, 1H, CH
.34 (s, 9H, I Bu-CH ), 1.27 (s, 9H, I Bu-CH
6
6
1
(s, 1H, NCHdCHN), 5.20 (septet, J ) 6.4 Hz, 1H), 3.76 (d, J ) 6.4
Hz, 1H), 2.94 (d, J ) 14.0 Hz, 1H), 3.07-1.50 (different d and m,
6 6
δ): 40.55 (s). H NMR (C D
2
3
3
1H), 1.44 (d, J ) 12.4 Hz, 1H). 13C NMR (C
D
6 6
, 100.6 MHz, δ):
)
1
2
3
2
dCHCF
3
),
t
t
13
178.59 (s, N-C-N), 118.32 (s, NCHdCHN), 105.11 (s), 64.49 (s),
3
3
). C NMR (C
6
D
6
,
5
4
9.78 (s, IAd-C), 59.34 (s, IAd-C), 47.23 (s), 44.82 (s, IAd-CH),
4.51 (s, IAd-CH), 36.56 (s, IAd-CH ), 36.50 (s, IAd-CH ), 30.73
125.6 MHz, δ): 196.13 (s), 190.18 (s), 118.29 (d, J ) 10.8 Hz, NCHd
CHN), 57.63 (d, J ) 12.8 Hz), 49.93 (q, J ) 34.5 Hz), 38.84 (d, J )
2
2
t
t
(s, IAd-CH ), 30.64 (s, IAd-CH ).
3
.1 Hz), 31.28 (s, I Bu-CH
3
), 31.14 (s, I Bu-CH
3
). IR νCO (CH
2
Cl
2
,
2
2
-
1
Computational Details. The Amsterdam Density Functional (ADF)
cm ): 1982.0 (vs).
2
7,28
t
program was used to obtain all of the results discussed herein.
The
Synthesis of NiCl(C
3
H
5
)(I Bu) (16). Addition of a pentane solution
electronic configuration of the molecular systems was described by a
triple-ú STO basis set on nickel for 3s, 3p, 3d, 4s, and 4p (ADF basis
set IV).25 Double-ú STO basis sets were used for phosphorus (3s, 3p),
oxygen, nitrogen, and carbon (2s,2p) and hydrogen (1s) augmented
with single 4d, 3d, 3d, 3d, and 2p functions, respectively (ADF basis
set III).25 The inner shells on nickel and phosphorus (including 2s, 2p),
oxygen, nitrogen, and carbon (1s), were treated within the frozen core
approximation. Energies and geometries were evaluated using the local
(
4 mL) containing allyl chloride (154 mg, 2.034 mmol) to a pentane
t
2
solution (12 mL) of Ni(CO) (I Bu) (14) (200 mg, 0.678 mmol) led to
instant precipitation of a flaky, yellow solid. Leaving the flask under
a stream of argon led to evolution of CO and to a heavy orange
precipitate within 5 min. The suspension was stirred at room temperature
for 2 h, decanted, and washed with pentane (2 × 3 mL). Subsequently,
the volatiles were removed in vacuo, yielding the desired complex 16
as an orange solid. X-ray quality crystals were obtained by slow
evaporation of an acetone solution. Yield: 207 mg (97%). Anal. Calcd
29
exchange-correlation potential by Vosko et al., augmented in a self-
3
0
consistent manner with Becke’s exchange gradient correction and
for C14
2
H25ClN Ni (MW 330.54): C, 53.29; H, 7.99; N, 8.87. Found:
Perdew’s3 correlation gradient correction.
1,32
1
C, 52.85; H, 8.15; N, 8.49. H NMR (C
1
)
1
6
D
6
, 400 MHz, δ): 6.62 (s,
The ADF program was modified by one of us3 to include standard
molecular mechanics force fields in such a way that the QM and MM
parts are coupled self-consistently.3 The simple model QM systems
and the full QM/MM and QM systems are displayed in Figure 14. The
partitioning of the systems into QM and MM parts only involves the
NHC ligand. Specifically, the N-bonded R groups are treated as MM
atoms.
3,34
H, NCHdCHN), 6.55 (s, 1H, NCHdCHN), 5.11 (m, 1H), 3.67 (d, J
10.4 Hz, 1H), 2.98 (d, J ) 18.8 Hz, 1H), 2.14 (d, J ) 8.8 Hz, 1H),
2,35
t
t
.94 (s, 9H, I Bu-CH
3
), 1.66 (s, 9H, I Bu-CH
3
), 1.22 (d, J ) 16.8
13
Hz, 1H). C NMR (C D , 100.6 MHz, δ): 179.93 (s, N-C-N), 119.32
6 6
t
t
(s, I Bu-NCHdCHN), 106.20 (s), 65.97 (s), 58.98 (s, I Bu-C), 58.55
t
t
t
(
s, I Bu-C), 43.77 (s), 32.49 (s, I Bu-CH
3
3
), 32.08 (s, I Bu-CH ).
t
Synthesis of NiBr(C
H
3 5
)(I Bu) (17). Addition of a pentane solution
(
4 mL) containing allyl bromide (246 mg, 2.034 mmol) to a pentane
(
27) ADF 2003, User’s Manual; Vrije Universiteit Amsterdam: Amsterdam,
t
2
solution (12 mL) of Ni(CO) (I Bu) (14) (200 mg, 0.678 mmol) led to
The Netherlands, 2003.
instant precipitation of a flaky, yellow solid. After a few seconds,
evolution of CO and precipitation of a yellow-orange solid were
observed. The suspension was stirred at room temperature for 2 h,
decanted, and washed with pentane (2 × 3 mL). Subsequently, the
volatiles were removed in vacuo, yielding the desired complex 17 as
an orange solid. X-ray quality crystals were obtained by slow
(28) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; Van
Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001,
22, 931-967.
(29) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200-1211.
(
(
(
30) Becke, A. D. Phys. ReV. A 1988, 38, 3098-3100.
31) Perdew, J. P. Phys. ReV. B 1986, 33, 8822-8824.
32) Perdew, J. P. Phys. ReV. B 1986, 34, 7406-7408.
(33) Cavallo, L.; Woo, T. K.; Ziegler, T. Can. J. Chem. 1998, 76, 1457-1466.
(
34) Woo, T. K.; Cavallo, L.; Ziegler, T. Theor. Chim. Acta 1998, 100, 307-
13.
(35) Maseras, F.; Morokuma, K. J. Comput. Chem. 1995, 16, 1170-1179.
1
evaporation of an acetone solution. Yield: 114 mg (93%). H NMR
(
3
6 6
C D , 500 MHz, δ): 6.60 (s, 1H, NCHdCHN), 6.53 (s, 1H, NCHd
2
494 J. AM. CHEM. SOC. VOL. 127, NO. 8, 2005
9

Steric and Electronic Properties of N-Heterocyclic Carbenes
A R T I C L E S
The minima were localized by full optimization of the starting
structures. Convergence criterion in the geometry optimizations was
-
3
set to 2 × 10 a.u. on the maximum Cartesian gradient.
Acknowledgment. The National Science Foundation is grate-
fully acknowledged for financial support of this work. Dr.
Kenneth Moloy and DuPont CR&D are especially thanked for
the gift of Ni(CO)4. We thank Eli Lilly and Co. and the Lonza
Group for generous donation of amines. L.C. thanks the
University of Salerno for financial support (Grant Piccole di
Ateneo 2002).
Figure 14. Representation of the partitioning of the systems into QM and
MM regions.
As for the connection between the QM and MM parts, this occurs
by means of the so-called “capping” dummy atoms, which are replaced
in the real system by the corresponding “linking” atom.3 H atoms
were used as capping atoms for all of the N-C linking bonds. In the
QM/MM optimizations, the ratio between the N-C bonds crossing the
QM/MM border and the corresponding optimized N-H distance was
Supporting Information Available: Crystallographic infor-
mation files (CIF) of complexes 8-11, 16, and 17. Crystal-
lographic data have been deposited at the CCDC, 12 Union
Road, Cambridge CB2 1EZ, U.K., and can be obtained on
request, free of charge, by quoting the publication citation and
the deposition number 252129 (8), 252130 (9), 252131(10),
252132(11), 252133(16), 252134(17). This material is available
free of charge via the Internet at http://pubs.acs.org.
2,33
3
6
fixed to 1.36. The AMBER95 force field was used for the MM
potentials, except for Ni, which was treated with the UFF force field.37
To eliminate spurious stabilizations from the long-range attractive part
of the Lennard-Jones potential,38,39 we used an exponential expression
4
0,41
JA0438821
fitted to the repulsive part of the Lennard-Jones potential.
(
36) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M. J.;
Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kolmann,
P. A. J. Am. Chem. Soc. 1995, 117, 5179-5197.
(38) Sauers, R. R. J. Chem. Educ. 1996, 73, 114-116.
(39) Lee, K. J.; Brown, T. L. Inorg. Chem. 1992, 31, 289-294.
(40) Woo, T. K.; Ziegler, T. Inorg. Chem. 1994, 33, 1857-1863.
(41) Guerra, G.; Cavallo, L.; Corradini, P.; Longo, P.; Resconi, L. J. Am. Chem.
Soc. 1997, 119, 4394-4403.
(37) Rapp e´ , A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A., III; Shiff,
W. M. J. Am. Chem. Soc. 1992, 114, 10024-10035.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 8, 2005 2495