Metal-mediated reaction modeled on nature: the activation of isothiocyanates initiated by zinc thiolate complexes

Article abstract of DOI:10.1021/ic101464j

On the basis of detailed theoretical studies of the mode of action of carbonic anhydrase (CA) and models resembling only its reactive core, a complete computational pathway analysis of the reaction between several isothiocyanates and methyl mercaptan acti

Full text of DOI:10.1021/ic101464j

ARTICLE
pubs.acs.org/IC
Metal-Mediated Reaction Modeled on Nature: The Activation of
Isothiocyanates Initiated by Zinc Thiolate Complexes
Wilhelm A. Eger,†,‡ Martin Presselt,§,^ Burkhard O. Jahn,†,z Michael Schmitt,§ J€urgen Popp,§ and
Ernst Anders*,†
†Institut f€ur Organische Chemie und Makromolekulare Chemie, Friedrich-Schiller-Universit€at Jena, Humboldtstrasse 10, 07743 Jena,
Germany
§Institut f€ur Physikalische Chemie, Friedrich-Schiller-Universit€at Jena, Helmholzweg 4, 07743 Jena, Germany
‡School of Molecular Chemistry and Bioscience, University of Queensland, Building 68, St. Lucia, Brisbane, 4072 Queensland, Australia
^Institute of Physics, Ilmenau University of Technology, Weimarer Strasse 32, 98693 Ilmenau, Germany
zInstitute of Organic Chemistry and Biochemistry AV CR, Flemingovo namesti 2, Prague 16610, Czech Republic
S Supporting Information
b
ABSTRACT: On the basis of detailed theoretical studies of the mode of
action of carbonic anhydrase (CA) and models resembling only its reactive
core, a complete computational pathway analysis of the reaction between
several isothiocyanates and methyl mercaptan activated by a thiolate-bearing
model complex [Zn(NH3)3SMe]þ was performed at a high level of density
functional theory (DFT). Furthermore, model reactions have been studied
in the experiment using relatively stable zinc complexes and have been
investigated by gas chromatography/mass spectrometry and Raman spec-
troscopy. The model complexes used in the experiment are based upon
the well-known azamacrocyclic ligand family ([12]aneN4, [14]aneN4,
i-[14]aneN4, and [15]aneN4) and are commonly formulated as ([Zn-
([X]aneN4)(SBn)]ClO4. As predicted by our DFT calculations, all of these
complexes are capable of insertion into the heterocumulene system. Raman
spectroscopic investigations indicate that aryl-substituted isothiocyanates
predominantly add to the CdN bond and that the size of the ring-shaped
ligands of the zinc complex also has a very significant influence on the
selectivity and on the reactivity as well. Unfortunately, the activated
isothiocyanate is not able to add to the thiolate-corresponding mercaptan to invoke a CA analogous catalytic cycle. However,
more reactive compounds such as methyl iodide can be incorporated. This work gives new insight into the mode of action and
reaction path variants derived from the CA principles. Further, aspects of the reliability of DFT calculations concerning the
prediction of the selectivity and reactivity are discussed. In addition, the presented synthetic pathways can offer a completely new
access to a variety of dithiocarbamates.
’ INTRODUCTION
In the past, several attempts were made to explain the mode of
action of CA. Most of them used quantum mechanical, in
particular density functional theory (DFT), calculations to
estimate the activation barriers and to provide insight into the
reaction pathways.7 In parallel, bioanalogue catalytic cycles, in
which a variety of (hetero)cumulenic molecules such as carbonyl
sulfide,8 allenes,9 and isothiocyanates replace CO2, have also
been investigated.10 The aim of these calculations was to
transcribe the reaction principle of CA to synthetic chemistry
using a catalytically active model complex. In a preceding
investigation, we showed that the bioanalogue reaction of CS2
and CA modeling zinc thiolate complexes is considered to be
working.11 Thus, it is of interest to check whether a CA-derived
The class of carbonic anhydrases (CAs) includes one of the
fastest known enzymes. The CAs catalyze the reaction of water
and cumulene carbon dioxide (CO2) to give bicarbonate and a
proton.1 They are remarkably efficient at accelerating this reac-
tion and possess molecular activities up to 106 s-1 2
.
CAs are located both in red blood cells to facilitate CO2
transport3 and in tissues like the gall bladder epithelium or
secretory epithelia in kidneys to accelerate electrolyte secretion.4
CAs belong to the zinc metalloenzymes and possess a Zn2þ
cation,5 which is essential for catalysis. The metal cation is
surrounded by three histidine ligands and one hydroxide ion in
an almost ideal tetrahedral coordination sphere.
For its speed and ability to activate a more or less inert
molecule such as CO2, the paragon CA is a suitable candidate for
biomimetic catalysis.6
Received: July 22, 2010
Published: March 15, 2011
r
2011 American Chemical Society
3223
dx.doi.org/10.1021/ic101464j Inorg. Chem. 2011, 50, 3223–3233
|

Inorganic Chemistry
ARTICLE
Scheme 1. Common Reaction Scheme of the Reaction of Isothiocyanates with Zinc Thiolate Complexes
Figure 1. Benzylzinc thiolate complexes used in this work.
model complex is able to fixate and activate isothiocyanates
analogously to the mode of action given by nature.
proton at the nucleophilic ligand will simplify the reaction
mechanism (cf. the Zinc-Complex-Mediated Reactions section).
Because alkyl thiolate complexes are found to be unstable,16
we chose the benzyl thiolate complexes of the following azama-
crocycles: [12]aneN4 ([Zn([12]aneN4)(SBn)]ClO4, 12Bn),
[14]aneN4 ([Zn([14]aneN4)(SBn)]ClO4, 14Bn), i-[14]aneN4
([Zn(i-[14]aneN4)-(SBn)]ClO4, i14Bn), and [15]aneN4
([Zn([15]aneN4)(SBn)]ClO4, 15Bn) [numbers in brackets
indicate the ring size, whereas the last letter and number mean
the count of possible coordinating atoms; cf. Figure 1]. The
reason for the choice of models with four nitrogen atoms
complexing the zinc cation instead of three is the considerably
better reactivity of these species.17 For economic reasons, we
calculated the pathway using the model [Zn(NH3)3SMe]þ. In
the Computational Methods section, some technical details of
the DFT and ab initio calculations are summarized.
In the present paper, we will discuss both the experimental
realization based upon Raman spectroscopy and gas chromatog-
raphy/mass spectrometry (GC/MS) measurements and the
theoretical interpretation using high-level DFT calculations to
investigate the reaction mode of isothiocyanates with zinc
thiolate complexes and of the subsequent transformations
(cf. Scheme 1) with methyl mercaptan. Dithiocarbamates, pos-
sible products of this reaction, are of general increasing interest
because they can be used, for example, as improved linkers to
gold surfaces, thus being an appealing alternative to the common
thiols.12 Furthermore, dithiocarbamate derivatives are widely
used as pesticides and, in particular, as fungicides.13
Calculations. In a recent work, the geometries and energies of
all important states within the postulated catalytic cycle of a CA
model and isothiocyanate derivatives were extensively studied.10
From these investigations, we concluded that it should be
possible to activate the isothiocyanate group and to understand
and finally profit from important steric and electronic properties.
In particular, significant changes in the bond order, bond
strength, and π-electron delocalization as well as changes within
the force constant of a corresponding localized normal mode,
polarizability, and electronic absorption are expected. While we
used the [Zn(NH3)3OH]þ model and water as the nucleophilic
reagent in the preceding work, we now present the DFT
calculations for the reaction of alkyl and aryl isothiocyanates
with methyl mercaptan as a substitute for a HX compound, which
should function as a reactant to allow construction of a catalytic
cycle with [Zn(NH3)3SR]þ as the model complex. To estimate
the influence of different residues, all calculations were carried
out with methyl, phenyl, and p-nitropenyl isothiocyanate.
Standard Reaction. Because isothiocyanates are only soluble
in aprotic solvents, we did not consider water-catalyzed reaction
paths.10 When the free energies of the encounter complexes (U-
1) and the three first transition states (U-2(ts), U-3(ts), and U-
4(ts)) are compared with the enthalpies of the standard reaction
of water with methyl isothiocyanate, a decrease in the energetic
values can be observed (cf. Scheme 2, Table 1, the Supporting
Information, and ref 10).
’ RESULTS AND DISCUSSION
Models. Most of our preceding theoretical studies about CA
analogues and their mechanisms of cumulene activations and
transformations have been performed using a simplified model
structure of the active site, the [Zn(NH3)3OH]þ complex.
Because this species only exists in the gas phase under specific
conditions,14 it is not appropriate to use it for experimental
purposes. Nevertheless, because the most important structural
and electronic properties agree well with those of much larger
structures, it therefore allows reliable DFT-based predictions.
Especially, the structural agreement with complexes built from
well-established ligands (e.g., specific azamacrocycles), which are
commonly used for the construction of such CA-simulating zinc
model complexes and for the investigation of the mode of action
of the enzyme, is very good.15 Because most of these complexes
contain a hydroxyl group to incorporate the natural nucleophilic
ligand, we recently developed catalyst models that include
(nucleophilic) thiolate groups and that were investigated con-
cerning their behavior against cumulenes, e.g., CS2.11,16 For
several reasons (especially with respect to a future application
in syntheses), some advantages appear to be promising for the
use of these thiolate complexes. First, they are soluble in aprotic
solvents, a very important aspect. Second, they do not collapse to
give di- or trimers like the hydroxyl complexes15a and, therefore,
do not shield their catalytic center. Third and finally, the missing
3224
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
Scheme 2. Standard Reaction of Isothiocyanates with Methyl Mercaptan
barrier of all insertion step possibilities, and therefore the C—N
attack is preferred. Varying the residue of the isothiocyanate from
R = methyl to p-nitrophenyl results in a more flattened hypersur-
face, in which the free energies of the transition states and
intermediates, which are higher in enthalpy than the separated
reactants, are lowered while those of the exergonic intermediates
are increased. Solvent corrections using the geometries of the
favored transition state U-2(ts) and the thermodynamically most
stable products U-5 and U-10 show an increase of 30 kJ/mol in
the case of the transition state and a decrease of 10 kJ/mol for the
product. Thus, on the one hand, the reaction is thermodynami-
cally favored, but, on the other hand, it is kinetically more
hindered. Because the activation barriers of U-8(ts) and U-9(ts)
are relatively low, U-10 and U-11 can exist at equilibrium.
The solvent-corrected values can be found in the Supporting
Information.
Zinc-Complex-Mediated Reactions. Analogously to the
standard reaction, we calculated the zinc-complex-mediated
versions of the reaction paths for several isothiocyanates.
In comparison to the hydration reaction catalyzed by a
[Zn(NH3)3OH]þ model,10 the reaction paths are simplified
because the existence of the methyl group excludes complica-
tions resulting from proton shifts (equivalent to the Lipscomb
mechanism; see ref 18). Therefore, only rotational transition
states must be taken into account (equivalent to the Lindskog
mechanism; see Scheme 3 and ref 19). Because two different
transition states resembling the attack on the C—S bond (such as
calculated before in the reaction cycle with water10) can only be
found in the gas phase and with the substrate methyl isothiocya-
nate, this reaction path is not included in the scheme.
Table 1. Free Energies (kJ/mol) Corresponding to Scheme 2
with Different Isothiocyanates Relative to the Separated
Reactants R-Isothiocyanate and Methyl Mercaptan at the
MP2/aug-cc-pVDZ Level of Theory
R = methyl
R = phenyl
R = p-NO2-phenyl
ΔG
ΔΔGaa
ΔG
ΔΔGa
ΔG
ΔΔGa
U-1
13
159
166
167
-15
55
10
151
163
161
-2
37
8
148
155
155
-5
32
U-2(ts)
U-3(ts)
U-4(ts)
U-5
146
153
154
141
153
151
140
147
147
U-6
U-7
30
20
12
U-8(ts)
U-9(ts)
U-10
49
64
30
44
46
29
37
42
30
15
27
25
-10
0
-8
13
-6
11
U-11
a Activation energy relative to the preceding intermediate.
They result from a lower ring strain; as in the case of the
reaction with water, the ring contains an oxygen and a sulfur atom
instead of two sulfur atoms.10 The slight decrease of the energetic
values going from R = methyl to p-nitrophenyl is in good
agreement with the stabilizing delocalization effects of the phenyl
moieties. This is also due to the higher electron density of the
C—N bond compared to the C—S bond, resulting from the
electron-withdrawing effect increasing from R = methyl to
p-nitrophenyl (cf. the Electronic Properties of Isothiocyanates
section). Thus, the transition state U-2(ts) and the attack of the
C—N bond are preferred throughout. The increase in the free
energy of U-11 in the case of R = phenyl and p-nitrophenyl
compared to methyl is due to hyperconjugation because the
phenyl residues are able to generate a mesomeric system with the
double bond.
Both possibilities of attacking the cumulenic system result in
only one intermediate each (5-A and N5-A; Scheme 3), both of
which are able to transform via rotation of the C—S bond to
either 5-B or N5-B (Scheme 4). Starting from these four
intermediates, a mercaptan (or a HX compound) can add to
each of them via either six- or four-membered cyclic transition
states (e.g., Scheme 4). Thus, eight reaction pathways are
conceivable (cf. the Supporting Information).
Regarding the free energies of these structures (cf. Tables 2
and 3), the influence of the solvent can be estimated. Whereas the
attack on the CdS bond (2(ts)) is thoroughly preferred in the
case of all isothiocyanates in the gas phase, the solvent-corrected
The differences between U-10 and U-11 are the result of steric
hindrance. In summary, this reaction is slightly exergonic relative
to the separated reactants in the cases of U-5 and U-10 (cf.
Scheme 2 and Table 1). Both intermediates are accessible via the
transition state U-2(ts), which possesses the lowest activation
3225
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
Scheme 3. Possible Addition Scenarios of Different Isothiocyanates
Scheme 4. Mechanistic Details of the Coordination Change after the Lindskog Rotation Transition State
Table 2. Free Energies (kJ/mol) Corresponding to Scheme 3 and Several Isothiocyanates Relative to the Separated Reactants
[Zn(NH3)3SMe]þ and R-Isothiocyanate at the B3LYP/6-311þG(d,p) Level of Theory
R = Me
R = Me*a
R = Ph
R = Ph*
R = p-NO2-Ph
ΔΔGa
R = p-NO2-Ph*
ΔG
ΔΔGab
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔG
ΔΔGa
1
6
100
79
40
132
95
6
102
73
43
126
89
25
119
81
41
115
68
2(ts)
3
94
11
92
5
96
13
82
1
94
12
73
8
4(ts)
5-A
90
100
83
86
90
93
76
52
44
76
56
59
N2(ts)
N3
127
78
121
54
128
68
88
84
125
73
119
53
126
72
83
67
142
91
117
38
112
63
70
43
N4(ts)
132
152
126
139
129
106
55
N5-A
24
65
23
66
37
a Columns with stars represent solvent calculations in DMSO. b Activation energy relative to the preceding intermediate.
values do not differ significantly. The same behavior can be
observed in all other cases except the rotation transition states
4(ts) and N4(ts). Contrary to the C—N bond, the rotation
around the C—S bond in 4(ts) is energetically lower because this
bond is longer and the sulfur atom is much softer in terms of the
hard and soft acids and bases principle. As a result, there are
smaller steric and electronic hindrances in the case of 4(ts),
which makes this pathway slightly preferable. Both rotation
bonds possess a significant amount of π molecular orbital and
thus partial double-bond character. Nevertheless, the amount of
this bonding character is nearly equal in both bonds and is
therefore obviously not the reason for the energetic difference.
Interestingly, the energies of nitrophenyl isothiocyanate are
affected significantly more in comparison to the other
3226
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
Table 3. Free Energies (kJ/mol) Corresponding to Scheme 4 and Several Isothiocyanates Relative to the Separated Reactants
[Zn(NH3)3SMe]þ and R-Isothiocyanate at the B3LYP/6-311þG(d,p) Level of Theory
R = Me
R = Me*a
R = Ph
R = Ph*
R = p-NO2-Ph
R = p-NO2-Ph*
ΔG
ΔΔGab
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔΔGa
5-A
52
58
28
63
24
58
63
30
83
86
67
94
65
95
85
74
44
62
28
54
23
62
28
29
76
85
64
95
66
95
87
76
56
74
39
72
37
81
62
45
59
74
38
89
55
91
78
73
5(ts)
6
11
34
2
3
11
30
7
18
10
39
1
9
19
29
12
18
16
44
6
15
30
36
35
5-B
5-A-to-N5-B(ts)
N5-A
N5(ts)
N5-B
5-B-to-N5-A(ts)
a Columns with stars represent solvent calculations in DMSO. b Activation energy relative to the preceding intermediate.
Scheme 5. Two Pathways Leading to the Most Stable Product, U-10
isothiocyanates. An explanation of this effect is the strong
interaction of the nitro residue with the solvent field.
N7-AF(ts) (cf. Scheme 5), while path B starts directly from 5-A
via 5(ts) to 5-B and continues via the six-membered transition
state 7-BS(ts) (cf. Schemes 4 and 5).
Not surprisingly, pathway B is preferred because the less
strained transition state 7-BS(ts) has significantly lower energies,
which is most pronounced in the reaction of the aryl
isothiocyanates.
Although the attack on the CdS double bond is favored in the
gas phase, the calculations are not able to clear the selectivity
problem because the energies of both transition states do not
differ significantly, e.g., in dimethyl sulfoxide (DMSO). A
possible selectivity in the first step would be almost lost after
surmounting the rotation transition states because the following
intermediates are all connected via very low transition states in
the gas phase as well as in DMSO. In a comparison of the only
two ways that lead to the thermodynamically most stable
product, the route via the six-membered cyclic transition state
appears to be the most favored one. Even though the energy of
the last transition state is rather high and might be a problem.
Electronic Properties of Isothiocyanates. To examine the
bonding situation in the cumulenic system, Natural Resonance
Theory (NRT) calculations (as included in the NBO program
package) were performed.20 The NRT calculations resulted in
30% triple-bond character of the C—N bond in the case of
Following the intermediates 5-A and N5-A, either the rota-
tional transition states (5(ts) or N5(ts)) can be generated, which
twist the outstanding methyl group around 180ꢀ, or a coordina-
tion change between the nitrogen and sulfur atoms (A(ts) or
B(ts)) can take place. Because the activation barriers of these
transition states are not very high, it is reasonable to assume that
the pathways easily can change again at this point.
Accordingly, the selectivity at the rate-determining step does
not define the selectivity of the product. Although the insertion
into the C—S bond is energetically lowered, N5-A can be the
overall preferred intermediate of the insertion step because it is in
equilibrium with 5-A.
Analogously to the pathway of CO2 in CA and the reaction of
isothiocyanate with water via a [Zn(NH3)3OH]þ complex, the
attack on comparable intermediates of the reaction with a
[Zn(NH3)3SMe]þ complex continues either via a four-mem-
bered cyclic (F) or a six-membered cyclic (S) transition state (cf.
Scheme 5 and the Supporting Information). Looking at the
products of all possible reaction paths, only one exergonic
structure (U-10) can be found. This reduces the number of
possible pathways to two (cf. Scheme 5). Path A proceeds via N5-
A and reaches U-10 via the four-membered cyclic transition state
3227
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
phenyl isothiocyanate and 0% in the case of methyl isothiocya-
nate. By inspection of the p-nitrophenyl isothiocyanate, we found
100% triple-bond character for the C—N bond. Because of the
higher electron density in the CdN double bond, the addition to
this bond is expected to be facilitated and thus favored over the
C—S addition. Interestingly, this is not supported by calculation
of the zinc-mediated pathway (cf. Schemes 3 and 4); in fact, they
propose a slight preference of the C—S addition. In contrast, the
calculations of the standard reaction path resemble the expected
behavior very well. This is in agreement with the results of an
extensive study concerning the influence of various substituents
on phenyl, stating that a nitro group decreases the π character of
the bond in the para position. Consequently, introduction of the
nitro group leads to a less delocalized π-electron system.21
GC/MS. 1H NMR investigations were performed (cf. Support-
ing Information), but unfortunately this appears not to be a
sufficiently reliable tool for the decision on whether the CdN or
CdS double bond was attacked.
To examine experimentally whether isothiocyanates insert or
not, equimolar amounts of methyl, phenyl, and p-nitrophenyl
isothiocyanate were each reacted with all complexes shown in
Figure 1. After 10 min, they were quenched with methyl iodide to
give zinc iodide complexes and methylated products (A1, B1,
and C1; see Figure 2). In the case of p-nitrophenyl isothiocya-
nate, the solution immediately turns from yellow to red upon the
insertion step and slowly back to yellow when methyl iodide is
added. The existence of the mole peak of the products in the MS
spectrum of the product peak in the GC spectra proves that
methyl iodide reacted with one of the insertion intermediates, as
depicted in Scheme 1, to give compound A1, B1, or C1 (cf.
Figure 2) and not directly with free isothiocyanate.
benzyl mercaptan under the same conditions and observed no
reaction at all.
Furthermore, quenching the insertion intermediate with ben-
zyl mercaptan was not successful at room temperature and higher
temperatures. Although the red color of the solution turned to
yellow again after the addition of a large excess of benzyl
mercaptan at room temperature, no product could be found.
Thus, a ligand exchange and subsequent decomposition of the
inserted product back to the reactants must have occurred.
The problem could be not only the relatively high activation
barrier of the last transition state N7-BS(ts) (cf. Table 4) but also
the missing formation of the thermodynamically very stable
Zn—I bondresulting from reaction withmethyliodide. Moreover,
unfortunately, the applied method is not suitable to distinguish
between the two isomers of the product because they do not
separate well on the column and affords relatively high tempera-
tures, which could initiate rearrangements. Thus, GC/MS is not
the appropriate method for conclusions regarding the selectivity.
Nevertheless, it is possible to give a rough description of the yields
of the reactions because the peaks of the reactants, products, and
decomposition products are all available.
The calculated yields are assembled in Table 5. As expected,
the reaction with methyl isothiocyanate has the poorest yield
because it is not activated as much as aryl isothiocyanates.
However, there is also a correlation between the reactivity and
structure of the zinc thiolate complexes used. While 15Bn and
14Bn have thoroughly good yields, i14Bn is not able to react
with methyl isothiocyanate anymore and 12Bn has very poor
yield except in the case of the reaction with p-nitrophenyl
isothiocyanate.
Raman Spectroscopy. To gain deeper insight into the reac-
tion mechanism, Raman spectroscopy was applied. In particular,
structural changes induced upon insertion into the electron-rich
CdN or CdS bond of the isothiocyanate group are expected to
give rise to significant changes in the Raman spectra. This
assumption is justified by the fact that the vibrational transition
energy, i.e., the wavenumber value of a normal mode involving
the isothiocyanate group, is related to the bond strength as well as
the bond order, while the corresponding Raman intensity is
related to the electron delocalization.22 Consequently, we focus
our discussion on p-nitrophenyl isothiocyanate in the following.
For a profound understanding of the changes in p-nitrophenyl
isothiocyanate during reaction with the zinc complex, a clear
assignment of the Raman bands of p-nitrophenyl isothiocyanate
is necessary. As shown in Figure 3, the DFT-calculated Raman
spectrum of p-nitrophenyl isothiocyanate fits well the corresponding
In addition to the zinc-complex-mediated insertion step, we
examined the standard reaction between each isothiocyanate and
Figure 2. GC/MS results. Conceivable products of the quenching
reaction with methyl iodide: A1, (Z)-benzylmethyl (4-nitrophe-
nyl)carbonimidodithioate; B1, (E)-benzylmethyl (4-nitrophenyl)-
carbonimidodithioate; C1, benzyl methyl(4-nitrophenyl)carbamodi-
thioate.
Table 4. Free Gibbs Enthalpies Referring to Scheme 5 Relative to the Separated Reactants [Zn(NH3)3SMe]þ, R-Isothiocyanate,
and Methyl Mercaptan at the B3LYP/6-311þG(d,p) Level of Theory
R = Me
R = Me*a
R = Ph
R = Ph*
ΔΔGa
R = p-NO2-Ph
R = p-NO2-Ph*
ΔG
ΔΔGab
ΔG
ΔΔGa
ΔG
ΔΔGa
ΔG
ΔG
ΔΔGa
ΔG
ΔΔGa
N6-AF
N7-AF(ts)
N8-AF
6-BS
31
109
5
91
186
59
31
116
19
94
201
71
46
132
36
88
189
71
78
48
95
60
85
63
107
58
86
101
43
96
32
95
51
78
7-BS(ts)
8-BS
91
156
60
95
153
75
103
36
52
128
80
50
5
19
U-10
-10
22
-8
-12
-6
-14
a Columns with stars represent solvent calculations in DMSO. b Activation energy relative to the preceding intermediate.
3228
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
experimentally determined Raman spectra of the DMSO solu-
tion. The normal mode assignment is shown in Table 6.
Because the Raman band at 2111 cm-1 in the Raman spectrum
of the DMSO solution is broad and weak,23 we varied the
solvent and obtained a well-pronounced Raman signal at
2099 cm-1 when using CH3Cl instead of DMSO (cf. the
Supporting Information). In contrast to the well-localized
vibrations of the CtN triple bond, vibrations of the C—S
bond are always associated with the strong stretching vibrations
of the nitro group calculated to result in intense Raman bands at
Furthermore, we analyzed the Raman spectra of the solid and
dissolved zinc complex. Because no significant changes are found
in the Raman spectra of the macrocyclic ligand during the
reaction and the Raman signals of the complex are weak com-
pared to the ones of p-nitrophenyl isothiocyanate, the Raman
spectra of the complexes are not discussed in the present paper.
Focusing on the product formation, predominantly changes of
the above-discussed Raman bands of p-nitrophenyl isothiocya-
nate, which are assigned to the cumulene system, are expected.
Following our NRT calculation described in the previous section,
the C—N bond in the isothiocyanate group loses the triple-bond
character upon addition to the C—N and C—S bonds. Conse-
quently, the CN vibrational mode is not suitable to unravel the
reaction mechanism. Anyhow, in accordance to this prediction,
the 2111 cm-1 mode disappears in the experimental Raman
spectra within 1 or 2 s because of product formation (cf.
Figure 4). Considering the modes involving C—S bond stretch-
ing, i.e., 1343 and 1258 cm-1 in the experimental Raman
spectrum of p-nitrophenyl isothiocyanate, the DFT calculations
predict a vanishing of the latter mode, while the Raman bands
appear between 1200 and 1350 cm-1 in either mode of product
formation. In contrast, the Raman signal referring to the vibration
involving the nitrophenyl group and the C—S bond (1343/
1302 cm-1 in the experimental/DFT-calculated Raman spec-
trum of p-nitrophenyl isothiocyanate) is shifted to 1316 cm-1
and gets strongly intensified because of product formation in the
experimental Raman spectra of 15Bn, i14Bn, and 14Bn, as
shown in Figure 4. Thus, the experimental Raman signal at
1302 and 1288 cm-1
.
Indeed, the wavenumber shift between the latter Raman bands
corresponding to the nitro group is the only discrepancy between
the experiment and theory. This is an artifact of the level of
theory used. A further discussion can be found in the Supporting
Information.
Table 5. Yields of the Zinc-Complex-Mediated Reaction of
Isothiocyanates To Give Dithiocarbamates (%) at Room
Temperature (Derived from the GC Spectra)
substituent
12Bn
i14Bn
14Bn
15Bn
Me
a
>1b
a
21
72
81
18
94
87
Ph
76
88
p-NO2-Ph
72
a These reactions did not take place. b This reaction took place but gave
poor product yield.
Figure 3. Green: Experimental Raman spectrum of p-nitrophenyl isothiocyanate dissolved in DMSO. DMSO Raman bands are marked by asterisks.
Black: DFT-calculated Raman spectrum of p-nitrophenyl isothiocyanate.
3229
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
Table 6. Assignment of the Vibrational Modes in the Raman Spectra of p-Nitrophenyl Isothiocyanate (Calculated and
Experimentally Derived in DMSO and Chloroform)a
ν~calc [cm-1
]
ν~DMSO [cm-1
]
ν~CHCl [cm-1
]
description
3
2093
1602
1302
1288
1181
1097
1019
2111
1589
1343
1258
1174
1108
b
2099
1589
1335
1266
1174
1107
1013
ν(CN)
δip(ph)
ν[C-N(O2)] þ weak ν(CN þ CS)
ν[C-N(O2)] þ strong ν(CN þ CS)
δip(ph)
ν[C-N(O2)] þ δip(ph)
δip(ph)
a δip: in-plane deformation vibrations. ν: stretching vibrations. b Peak is overlapped by a solvent band.
Figure 4. Top: Experimental product Raman spectra of the reaction of p-nitrophenyl isothiocyanate with complexes 15Bn, i14Bn, 14Bn, and 12Bn
dissolved in DMSO (solvent bands areindicated by asterisks). Bottom: DFT-calculated Raman spectra of the products emerging from the different
reaction mechanisms (C—N or C—S insertion). The dotted graphs depict unreacted isothiocyanate.
1316 cm-1 is much more intense than the Raman signal at
1583 cm-1, which corresponds to the calculated Raman spectra
in the case of C—N insertion. Therefore, it seems reasonable
that in the case of these complexes the preferred reaction
mechanism is the insertion of the thiolate ligand into the C—
N bond. This conclusion is further supported by the Raman band
at 1453 cm-1, which is predicted by the calculated Raman spectra
but with a very low intensity. The latter band is assigned to a mode
involving phenyl-N bond stretchings and vibrations of the
azamacrocycle ligands. Apparently, the level of theory has problems
handling the intensities of the macrocyclic ligand bands properly.
12Bn gives a different spectrum because the peak around
1600 cm-1 is broader and slightly shifted to higher wavenumber
values. This change can be explained by a CdN double bond
emerging upon product formation via insertion into the C—S
bond (cf. the Zinc-Complex-Mediated Reactions section).
Furthermore, the DFT calculations predict almost equally in-
tense Raman bands at 1302 and 1590 cm-1 in the case of C—S
insertion, as shown in Figure 4. This is supported by the
experimental Raman spectrum of 12Bn because the Raman
intensity of the mode involving the nitro group (1334 cm-1) is
significantly reduced in comparison to the previously discussed
complexes (15Bn, i14Bn, and 14Bn). Thus, in the case of 12Bn,
the Raman intensity of the mode at 1334 cm-1 approaches the
order of intensity of the Raman band at 1595 cm-1
.
These results indicate a preference of insertion into the CdN
double bond in most cases. This correlates well with the
prediction of the higher electron density in this bond. However,
3230
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
the influence of the azamacrocyclic ligand becomes significant in
the example of 12Bn because in this case the alternative insertion
into the C—S bond seems to be preferred. Further vibrational
assignments of the Raman bands calculated for the different
reaction mechanisms are given in the Supporting Information.
This combination serves as a powerful tool to develop novel and
efficient syntheses modeled on nature.
’ EXPERIMENTAL SECTION
Materials and Methods. All reagents used were of analytical
purity. GC/MS data were recorded on a Shimadzu GC-2010 mass
spectrometer (GCMS-QP2010S), Ultra Alloyþ -5 capillary column,
30 m ꢀ 0.25 mm i.d., 0.25 μm film (Frontier Laboratories), and a
standard program at 2 min at 40 ꢀC, followed by a temperature increase
of 16 ꢀC/min, and held at 250 ꢀC for 5 min.
Raman spectra of solid samples were recorded with a Jobin-Yvon
Labram HR spectrometer, equipped with an Olympus BX41 inverse
microscope. The 830 nm output from a diode laser was used as the
Raman excitation source, and the scattered light was detected with a
CCD detector. A grating with 300 grooves/mm was applied, leading to a
spectral resolution of roughly 1 cm-1. The measurements were carried
out in dry DMSO at room temperature. The Raman spectra of unreacted
isothiocyanates were also measured in dry CHCl3.
Preparation of Azamacrocycles. The azamacrocycles
[12]aneN4, [14]aneN4, i-[14]aneN4, and [15]aneN4 were synthesized
using the usual Richman-Atkins method.24 Some changes in the
procedure developed in our work group were applied to this synthesis.16
Preparation of the Zinc Thiolate Complexes. In general, the
thiolate complexes were synthesized as described previously.16 How-
ever, some changes were made to achieve a better yield in the case of
12Bn and 15Bn.
’ CONCLUSION
In this work, we examined the translation of an enzymatic
mode of action principle to a synthetic scenario. Therefore,
exhaustive DFT calculations were performed to predict possible
reaction paths. Subsequently, a variety of potential catalysts was
synthesized to identify the conditions to follow the desired
reaction paths and to obtain the aspired products.
Because of the unsymmetrical cumulated bond system in
isothiocyanates, the number of conceivable reaction paths ex-
pands into a huge manifold of similar alternatives, which is quite
difficult to handle compared to the natural zinc-complex-
mediated standard reaction of the initial CA/CO2/H2O system.
Nevertheless, from these investigations, we conclude that the
natural paragon, the activation and hydration of CO2 and COS
by specific zinc enzymes (CAs), can be applied to a much wider
number of cumulenic substrates and zinc complex variants than
previously assumed because the zinc thiolate complex definitely
inserts into the cumulene system of isothiocyanate at room
temperature. We demonstrated that these [LnZn-SR]þ com-
plexes are able to induce reaction mechanisms similar to those of
their oxygen analogues, the nucleophilic [LnZn-OH]þ cation
complexes, i.e., the activation of a variety of isothiocyanates to
open further addition reaction paths. Unfortunately, the activa-
tion of isothiocyanates by [LnZn-SR]þ complexes is not suffi-
ciently reactive to allow the addition of the corresponding
mercaptans, which would initiate a catalytic cycle analogously
to the natural “CA/CO2/H2O” mode of action. Further, at this
point the predictions of the calculations are not reliable through-
out because they result in a surmountable transition state. The
selectivity problem (CdN vs CdS) can also not be solved
completely by DFT calculations alone, but this is due to relatively
small energy differences. In contrast, our Raman spectroscopic
investigations verify that the CdN bond of the aryl isothiocya-
nate structures is the preferred moiety to be attacked by the
cationic but nucleophilic [LnZn-SR]þ complexes. This is in good
agreement with the more stable intermediate N5-A, which is in
equilibrium with the CdS insertion product 5-A. The activation
barrier of the most probable pathway, i.e., 2(ts) (cf. Scheme 3), is
reduced by 60-30 kJ/mol compared to the nonactivated (or
standard) reaction path (cf. Figure 2) and depends on the
isothiocyanate used. Another unfavored alternative via N2(ts)
is conceivable but can lead to the same product U-10 (cf.
Schemes 4 and 5). Additionally, we could reproduce the reactiv-
ity of the zinc thiolate complexes depending on the ring size of
their ligands found in the reaction with CS2.
The yield of [Zn([15]aneN4)(SBn)]ClO4 (15Bn) could be in-
creased from 35% to 76% by using 8 mmol of the ligand, 40 mL of
a 0.2 M solution of Zn(ClO4)2 6H2O in methanol, and 8 mL of a
3
0.5 M solution of KOH in methanol. The solution was kept in the
freezer at -40 ꢀC to allow it precipitate properly. In the case of
[Zn([12]aneN4)-(SBn)]ClO4 (12Bn), it was possible to increase the
yield from 35% to 73% by using 8 mmol of the ligand, 80 mL of a 0.1 M
solution of Zn(ClO4)2 6H2O in methanol, and 8 mL of a 0.5 M solution
3
of KOH in methanol.
Reaction of the Zinc Thiolate Complexes with Isothiocya-
nates To Give Dithiocarbamates. All reactions were performed in
dry DMSO under standard conditions and at room temperature. A total
of 0.5 mL of a 0.1 M solution of isothiocyanate was slowly added into
0.5 mL of a 0.1 M solution of the thiolate complex. After 10 min of
stirring, an equimolar amount of methyl iodide was added, and the
mixture was filtered and immediately examined via GC/MS. All reaction
solutions have a pale-yellow color; in the case of p-nitrophenyl iso-
thiocyanate, the color is strong yellow and turns into a deep red
immediately after the addition of isothiocyanate to the zinc thiolate
complex. After the addition of methyl iodide, the color slowly returns to
yellow.
Regarding the Raman measurement, the same procedure, except the
addition of methyl iodide, was used.
Computational Methods. All optimizations and frequency cal-
culations reported in this Article were performed using the Gaussian03
program package.25 Atomic charges and hyperconjugative interaction
energies were obtained using version 5.1 of the natural bond orbital
(NBO) analysis of Weinhold et al. in Gaussian03, which is not included
in Gaussian by default.20,26 Default convergence criteria were used for all
calculations. All energies reported are Gibbs free energies (ΔG values
relative to the separated reactants) and thus contain zero-point, thermal,
and entropy effects at 298 K and 1 atm pressure.
These possibilities are, of course, structurally and conceptually
far beyond the natural mode of action of the CAs as catalysts,
which stand out because of their efficiency in transforming the
substrates CO2 and COS. However, the aim of this work was not
to reinvestigate the reaction principles of CAs but rather to
understand and finally to translate the basic reaction sequences in
order to estimate whether they might be applicable in synthesis
or not. Therefore, we combined theoretical studies (DFT and ab
initio calculations, if possible under inclusion of solvent effects)
with experimental investigations (GC/MSand Raman spectroscopy).
The hybrid density functional B3LYP27 with the 6-311þG(d,p) basis
set28 was used throughout for the zinc-complex-mediated reaction.
Previous investigations have demonstrated reliable results on this level
of theory for systems like [Zn(NH3)3OH]þ.8-10 The standard reaction
was calculated at the MP2 level29 using the aug-cc-pVDZ30 basis due to
3231
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
the smaller system. Full geometry optimizations, i.e., without con-
straints, as well as frequency calculations on the stationary points thus
obtained were performed for all species on the hypersurface in the gas
phase as well as in the presence of a dielectric field as described by the
C-PCM model for selected structures.31 In this model, the species of
interest are embedded in a cavity of molecular shape surrounded by a
polarizable continuum whose field modifies the energy and physical
properties of the solute. The solvent reaction field is described by
polarization charges distributed on the cavity surface. This procedure is
known to reproduce experimental solvation energies quite well. We
chose the solvent DMSO used in the experimental investigations.
Because of the better PCM model implemented in Gaussian09,32 the
solvent corrections have been calculated using this newer version of that
program. The gas-phase reactions have been calculated with Gaussian03
only for historical reasons and are fully comparable to the Gaussian09
PCM calculations.
Weston, J.; Anders, E. ChemBioChem 2001, 2, 190. (e) Merz, K. M.;
Banci, L. J. Phys. Chem. 1996, 100, 17414. (f) Miscione, G.; Stenta, M.;
Spinelli, D.; Anders, E.; Bottoni, A. Theor. Chem. Acc. 2007, 118, 193.
(8) (a) Schenk, S.; Kesselmeier, J.; Anders, E. Chem.—Eur. J. 2004,
10, 3091. (b) Notni, J.; Schenk, S.; Protoschill-Krebs, G.; Kesselmeier, J.;
Anders, E. ChemBioChem 2007, 8, 530.
(9) (a) Jahn, B. O.; Eger, W. A.; Anders, E. J. Org. Chem. 2008,
73, 8265. (b) Jahn, B. O.; Eger, W. A.; Anders, E. Z. Naturforsch. 2010,
65b, 425.
(10) Eger, W. A.; Jahn, B. O.; Anders, E. J. Mol. Model. 2009, 15, 443.
(11) Notni, J.; Schenk, S.; Roth, A.; Plass, W.; G€orls, H.; Uhlemann,
U.; Walter, A.; Schmitt, M.; Popp, J.; Chatzipapadopoulos, S.; Emmler,
T.; Breitzke, H.; Leppert, J.; Buntkowsky, G.; Kempe, K.; Anders, E. Eur.
J. Inorg. Chem. 2006, 14, 2783.
(12) (a) von Wrochem, F.; Gao, D.; Scholz, F.; Nothofer, H.-G.;
Nelles, G.; Wessels, J. M. Nat. Nanotechnol. 2010, 5, 618. (b) Vickers,
M. S.; Cookson, J.; Beer, P. D.; Bishop, P. T.; Thiebaut, B. J. Mater. Chem.
2006, 16, 209. (c) Zhao, Y.; Pꢀerez-Segarra, W.; Shi, Q.; Wei, A. J. Am.
Chem. Soc. 2005, 127, 7328. (d) Schulz, P.; Zangmeister, C. D.; Zhao, Y.-
L.; Frail, P. R.; Saudari, S. R.; Gonzalez, C. A.; Kagan, C. R.; Wuttig, M.;
Zee, R. D. v. J. Phys. Chem. C 2010, 114, 20843.
For calculation of the Raman spectra, we used BP86/TZVP because
this combination turned out to give accurate molecular structures,
frequencies, and Raman intensities for medium-sized molecules.33
Furthermore, there is no need to scale the obtained wavenumbers of
the vibrational normal modes to adjust them to experimental values.22
(13) (a) Halls, D. J. Microchim. Acta 1969, 57, 62. (b) Malik, A. K.;
Faubel, W. Pestic. Sci. 1999, 55, 965.
(14) Schr€oder, D.; Schwarz, H.; Schenk, S.; Anders, E. Angew. Chem.
2003, 115, 5241.
’ ASSOCIATED CONTENT
(15) (a) Kimura, E.; Shiota, T.; Koike, T.; Shiro, M.; Kodama, M.
J. Am. Chem. Soc. 1990, 112, 5805. (b) Zhang, X.; van Eldik, R. Inorg.
Chem. 1995, 34, 5606. (c) Zhang, X.; van Eldik, R.; Koike, T.; Kimura, E.
Inorg. Chem. 1993, 32, 5749.
(16) Notni, J.; G€orls, H.; Anders, E. Eur. J. Inorg. Chem. 2006,
7, 1444.
(17) Notni, J.; G€unther, W.; Anders, E. Eur. J. Inorg. Chem. 2007,
7, 985.
(18) (a) Liang, J. Y.; Lipscomb, W. N. Biochemistry 1987, 26, 5293.
(b) Sola, M.; Lledos, A.; Duran, M.; Bertran, J. J. Am. Chem. Soc. 1992,
114, 869. (c) Tautermann, C. S.; Loferer, M. J.; Voegele, A. F.; Liedl,
K. R. J. Phys. Chem. B 2003, 107, 12013.
S
Supporting Information. Cartesian coordinates of all
b
stationary points as well as relative free energies not listed in
the paper, all total energies, and a detailed discussion of the
Raman spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: ernst.anders@uni-jena.de.
(19) Lindskog, S.; Engberg, P.; Forsman, C.; Ibrahim, A. S.; Jonsson,
H. B.; Simonsson, I.; Tibell, L. Ann. N.Y. Acad. Sci. 1987, 429, 61.
(20) (a) Glendening, E. D.; Badenhoop, J. K.; Reed, A. E.; Carpenter,
J. E.; Bohmann, J. A.; Morales, C. M.; Weinhold, F. NBO; University of
Wisconsin: Madison, WI, 2001. (b) Reed, A. E.; Curtiss, L. A.; Weinhold,
F. Chem. Rev. (Washington, DC, U.S.) 1988, 88, 899.
(21) (a) Presselt, M.; Schnedermann, C.; Schmitt, M.; Popp, J.
J. Phys. Chem. A 2009, 113, 3210. (b) Presselt, M.; Dietzek, B.; Schmitt,
M.; Rau, S.; Winter, A.; J€ager, M.; Schubert, U. S.; Popp, J. J. Phys. Chem.
A 2010, 114, 13163.
’ ACKNOWLEDGMENT
Financial support by the Deutsche Forschungsgemeinschaft,
Collaborative Research Center 436 Metal Mediated Reactions
modeled on Nature and the Fonds der Chemischen Industrie,
Germany, is gratefully acknowledged.
’ REFERENCES
(22) (a) Presselt, M.; Dietzek, B.; Schmitt, M.; Popp, J.; Winter, A.;
Chiper, M.; Friebe, C.; Schubert, U. S. J. Phys. Chem. C 2008, 112, 18651.
(b) Winter, A.; Friebe, C.; Chiper, M.; Schubert, U. S.; Presselt, M.;
Dietzek, B.; Schmitt, M.; Popp, J. ChemPhysChem 2009, 10, 787.
(23) (a) Joo, S.-W. Surf. Interface Anal. 2006, 38, 173. (b) Stephenson,
C. V.; Coburn, W. C.; Wilcox, W. S. Spectrochim. Acta 1961, 17, 933.
(24) Richman, J. E.; Atkins, T. J. J. Am. Chem. Soc. 1974, 96, 2268.
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. J. A.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa,
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.;
Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
(1) (a) Bertini, I.; Luchinat, C. Acc. Chem. Res. 1983, 16, 272. (b)
Christianson, D. W.; Fierke, C. A. Acc. Chem. Res. 1996, 29, 331. (c)
Coleman, J. E. J. Biol. Chem. 1967, 242, 5212. (d) Lindskog, S.
Pharmacol. Ther. 1997, 74, 1. (e) Lindskog, S.; Liljas, A. Curr. Opin.
Struct. Biol. 1993, 3, 915.(f) Silverman, D. N. In Enzyme Kinetics and
Mechanism Part D: Developments in Enzyme Dynamics; Purich, D. L., Ed.;
Academic Press: New York, 1995; Vol. 249, p 479. (g) Silverman, D. N.;
Lindskog, S. Acc. Chem. Res. 1988, 21, 30.
(2) Khalifah, R. G. J. Biol. Chem. 1971, 246, 2561.
(3) Pocker, Y.; Stone, J. T. J. Am. Chem. Soc. 1965, 87, 5497.
(4) (a) Sly, W. S.; Hu, P. Y. Annu. Rev. Biochem. 1995, 64, 375. (b)
Supuran, C. T.; Scozzafava, A.; Casini, A. Med. Res. Rev. 2003, 23, 146.
(5) Hewett-Emmett, D.; Tashian, R. E. Mol. Phylogenet. Evol. 1996,
5, 50.
(6) Jahn, B. O.; Eger, W. A.; Anders, E. In Biomimetics Learning from
Nature; Mukherjee, A., Ed.; INTECH: Vienna, Austria, 2010; p 167.
(7) (a) Bottoni, A.; Lanza, C. Z.; Miscione, G. P.; Spinelli, D. J. Am.
Chem. Soc. 2004, 126, 1542. (b) Br€auer, M.; Anders, E.; Sinnecker, S.;
Koch, W.; Rombach, M.; Brombacher, H.; Vahrenkamp, H. Chem.
Commun. 2000, 1, 647. (c) Br€auer, M.; Perez-Lustres, J. L.; Weston,
J.; Anders, E. Inorg. Chem. 2002, 41, 1454. (d) Mauksch, M.; Br€auer, M.;
3232
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233

Inorganic Chemistry
ARTICLE
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.
Gaussian03, revision C.01; Gaussian Inc.: Wallingford, CT, 2004.
(26) (a) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1983, 78, 4066. (b)
Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735.
(27) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(28) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem.
Phys. 1980, 72, 650. (b) McLean, A. D.; Chandler, G. S. J. Chem. Phys.
1980, 72, 5639.
(29) (a) Frisch, M. J.; Head-Gordon, M.; Pople, J. A. Chem. Phys.
Lett. 1990, 166, 275. (b) Head-Gordon, M.; Pople, J. A.; Frisch, M. J.
Chem. Phys. Lett. 1988, 153, 503.
(30) (a) Kendall, R. A.; Dunning, J. T. H.; Harrison, R. J. J. Chem.
Phys. 1992, 96, 6796. (b) Woon, D. E.; Dunning, J. T. H. J. Chem. Phys.
1993, 98, 1358.
(31) (a) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995. (b)
Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003,
24, 669. (c) Klamt, A.; Sch€u€urmann, G. J. Chem. Soc., Perkin Trans. 2
1993, 799.
(32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.;
Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
€
S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. Gaussian09, revision A.01; Gaussian Inc.: Wallingford, CT,
2009.
(33) (a) Neugebauer, J.; Hess, B. A. J. Chem. Phys. 2003, 118, 7215.
(b) Reiher, M.; Brehm, G.; Schneider, S. J. Phys. Chem. A 2004, 108, 734.
(c) Reiher, M.; Neugebauer, J.; Hess, B. A. Z. Physik. Chem. 2000,
118, 7215.
3233
dx.doi.org/10.1021/ic101464j |Inorg. Chem. 2011, 50, 3223–3233