Studies on the coordination chemistry of methylated xanthines and their imidazolium salts. Part 1: Benzyl derivatives

Article abstract of DOI:10.1007/s11243-009-9310-0

New imidazolium salts derived from the natural methylated xanthines theophylline, theobromine and caffeine, namely 1,3-dimethyl-9-benzylxanthinium bromide (tphBzBr, 1a), 3,7-dimethyl-9-benzylxanthinium bromide (tbrBzBr, 2a) and 1,3,7-trimethyl-9-benzylxan

Full text of DOI:10.1007/s11243-009-9310-0

Transition Met Chem (2010) 35:165–175
DOI 10.1007/s11243-009-9310-0
Studies on the coordination chemistry of methylated xanthines
and their imidazolium salts. Part 1: benzyl derivatives
•
Vanessa R. Landaeta Rafael E. Rodrıguez-Lugo
•
´
Eloy N. Rodrıguez-Arias David S. Coll-Gomez
•
•
´
´
´
Teresa Gonzalez
Received: 1 September 2009 / Accepted: 9 November 2009 / Published online: 24 November 2009
Ó Springer Science+Business Media B.V. 2009
Abstract New imidazolium salts derived from the natural
methylated xanthines theophylline, theobromine and caf-
feine, namely 1,3-dimethyl-9-benzylxanthinium bromide
(tphBzBr, 1a), 3,7-dimethyl-9-benzylxanthinium bromide
(tbrBzBr, 2a) and 1,3,7-trimethyl-9-benzylxanthinium
bromide (caffBzBr, 3a), are reported. Also, the disubsti-
tuted analog of 1a, 1,3-dimethyl-7,9-dibenzylxanthinium
bromide (tphBz2Br, 1a0) was identified and characterized
by NMR. The coordination chemistry of ligands 1a–3a
toward palladium, and some theoretical aspects of the
unmodified theophylline, theobromine and caffeine are
studied. Our results prove that the theophylline derivative
has the thermodynamic tendency to form N-bonded spe-
cies, even when an equilibrium between the Pd–NHC and
the ‘‘theophyllinate’’ was observed spectroscopically, due
to the anisotropy of the NHC ligand. To confirm the N-
coordination, the solid state structure of the new ‘‘theo-
phyllinate’’ species PdBr2(tphBz-H)2 (4), derived from 1a,
was determined by X-ray diffraction. The analog with
theobromine, ligand 2a, coordinates to palladium via N1, in
an analogous manner to 1a, and a mixture of the cis/trans
isomers of its palladium complex is obtained. On the other
hand, since there is no possibility of N-coordination in 3a,
this caffeine derivative forms a Pd-NHC compound after
deprotonation with a strong base. Both the theoretical
results and the experimental evidence are in accordance, in
terms of the predicted coordination sites or possibility of
modification of the selected methylated xanthines to obtain
new ligands.
Electronic supplementary material The online version of this
article (doi:10.1007/s11243-009-9310-0) contains supplementary
material, which is available to authorized users.
Introduction
´
V. R. Landaeta (&) ꢀ R. E. Rodrıguez-Lugo ꢀ
The natural methylated xanthines, theophylline, theobro-
mine and caffeine (Fig. 1) are found in tea, cocoa and
coffee. They exhibit biologic activity [1] as stimulants, and
they have also been used for pharmacologic purposes [2].
In this sense, several efforts have been made to study their
coordination chemistry toward ‘‘essential metals’’ [3] and
their interaction with transition metals [3–8]. For instance,
Williams [2] et al. reported an analog to cis-platin with
caffeine, Pt(caff)2Cl2, and there are reports in which the
thermal behavior of these xanthines under physiological
conditions has been studied [3–6]. Also, Taube and
coworkers [7] proposed a C(8) coordination to ruthenium
in the unmodified caffeine and theobromine ligands, and a
mixed N–C, coordination fashion in theophylline.
´
E. N. Rodrıguez-Arias
´
´
´
Departamento de Quımica, Universidad Simon Bolıvar, Valle de
Sartenejas, Baruta. Apartado 89000, Caracas 1080, Venezuela
e-mail: vlandaeta@usb.ve
Present Address:
´
R. E. Rodrıguez-Lugo
Laboratory for Inorganic Chemistry, Swiss Federal Institute
of Technology, ETH Zurich, HCI-H136 Zu¨rich, Switzerland
´
D. S. Coll-Gomez
´
Centro de Quımica, Laboratorio de Quımica Computacional,
Instituto Venezolano de Investigaciones Cientıficas, IVIC,
´
´
Caracas, Venezuela
´
T. Gonzalez
Centro de Quımica, Laboratorio Nacional de Difraccion de
´
Rayos X, Instituto Venezolano de Investigaciones Cientıficas,
IVIC, Caracas, Venezuela
´
´
However, there are only a few reports of inorganic
compounds with functionalized xanthines [9]. Among
123

166
Transition Met Chem (2010) 35:165–175
the modification of these methylated xanthines and their
coordination chemistry toward palladium.
Experimental
Fig. 1 Chemical structure of the selected natural methylated
xanthines
All reactions and manipulations were routinely performed
under dry nitrogen or argon atmosphere using standard
1
Schlenk techniques. H, and 13C{1H} NMR spectra were
recorded on a Jeol Eclipse 400 spectrometer (400.13 and
100.5 MHz). Peak positions are relative to tetramethylsil-
ane and were calibrated against the residual solvent reso-
nance (1H) or the deuterated solvent multiplet (13C). All the
NMR spectra were recorded at room temperature (25 °C)
unless otherwise stated. Infrared spectra were recorded by
Bruker IR 560 FT-IR spectrometer, using samples mulled
in Nujol between KBr plates or in KBr disks. HRMS
analyses were performed using a Bruker Daltonics maxis
ESI-QTOF and a Varian HiResMALDI by ETH Zu¨rich
DCHAB/LOC MS-Service. X-ray diffraction studies were
performed on a Rigaku AFC-7S diffractometer with a
Bruker Smart Apex area detector by the X-ray diffraction
them, 8-alkyl-derivatives from theophylline [9] and
N-methyl caffeine [10–13] have been reported. To the best
of our knowledge, none of those exist with C(8) or N(9)-
substituted modified theobromine.
On the other hand, even when natural methylated xanth-
ines have an azole ring and a bulky backbone attached to it
[14], which makes them potentially interesting to prepare
N-Heterocyclic Carbenes (NHC’s) and their complexes (for
selected references on some of the most recent advances in
the chemistry of NHC’s ligands, see [15–19]), those only
exist with N-methyl caffeine, and examples of its coordi-
nation chemistry are found exclusively with mercury [20],
silver [10], iridium and rhodium [10, 13]. The first 1,3,7,9-
tetramethylxanthine-8-ylidene complex was reported as a
mercury bis-carbene complex [20]. Youngs et al. prepared a
silver (I) bis-carbene complex, which has been evaluated as
antimicrobial agent, and a cationic rhodium (I) complex [10,
11]. Herrmann and coworkers also obtained rhodium (I) and
iridium (I) carbene complexes of the type [M(L)(LCarbene)2]I
and M(L)(LCarbene)(I) (M = Rh, Ir, LCarbene = 1,3,7,9-tetramethyl-
xanthine-8-ylidene, 7,9-dimethylhypoxanthine-8-ylidene,
L = g4-1,5-cyclooctadiene = COD) [13]. To date, to the best
of our knowledge, no NHC’s have ever been reported with
the modified theophylline or theobromine.
´
service of Laboratorio Nacional de Difraccion de Rayos X
at IVIC.
Unless otherwise stated, all solvents were distilled just
prior to use from appropriate drying agents. Dichloro-
methane was distilled from CaH2, and tetrahydrofuran
(THF) from sodium/benzophenone. Diethylether and
petroleum ether were dried with sodium. Deuterated sol-
˚
vents were dried over 4-A molecular sieves prior to use. All
other chemicals were commercial products and used as
received without further purification. Literature methods
were employed for the synthesis of (PhCN)2PdCl2 [22].
The lack of results with the latter two, and the few
reports found in the case of caffeine are then interesting.
The difference among them could lie in the fact that caf-
feine, having three of its nitrogen atoms blocked by methyl
groups, is the best candidate to become an NHC precursor.
In contrast, the other two xanthines have an acidic proton
that could compete with C(8) binding either in the modified
or unmodified molecule [7], specially if a base is used to
generate the carbene [21].
Synthesis
1,3-Dimethyl-9-benzylxanthinium bromide,
or theophyllineBzBr (tphBzBr, 1a)
Theophylline (1, 3.00 g, 16.7 mmol) and KI (150 mg,
0.90 mmol) were suspended in acetonitrile (50 mL) and
vigorously stirred. After stirring for 5 min, benzyl bromide
(8.0 mL, 66.8 mmol) was added. The mixture was refluxed
for 4 days. After completion, the crude was cooled and
filtered to eliminate any unreacted theophylline and the
excess of KI. The solution was evaporated to dryness, and
diethyl ether (50 mL) was added. Upon stirring overnight
and filtering, white-off needles were obtained. Yield:
In this sense, three new imidazolium salts from theoph-
ylline (1), theobromine (2) and caffeine (3) bearing the
benzyl group on N(9) are reported, as a part of a wider syn-
thetic study. Also, the coordination chemistry of the xanthine
derivatives toward palladium(II) was investigated and con-
trasted with their unmodified counterparts [2–8], determin-
ing if the N(1,7) or C(8) hapticity is favored. In addition, a
theoretical study of the acidity-basicity and orbital energies
of each xanthine, 1-3, was performed, observing a match
between theoretical results and the experimental evidence in
1
3.65 g, 63%. H NMR (CDCl3, 25 °C, 400.13 MHz): d
3.35 (s, 3H, CH3–N); d 3.53 (s, 3H, CH3–N); d 5.45 (s, 2H,
PhCH2–N); d 7.31–7.29 (m, 5H, aromatics); d 7.56 (s, 1H,
C–H). 13C{1H} NMR (CDCl3, 25 °C, 100.5 MHz): d 28
(CH3–N); d 30 (CH3–N); d 50 (PhCH2–N); d 107 (O=C–
123

Transition Met Chem (2010) 35:165–175
167
C–N); d 128 (aromatic, para); d 128.7 (aromatic, meta); d
129 (aromatic, ortho); d 135 (C ipso); d 141 (N–CH=N); d
149 (N–C–N); d 151 (N–C(=O)–N); d 155 (N–C(=O)–C).
HRMS (ESI): m/z calcd for C14H15N4O2Br: 271.1195
[M?–Br]; found 271.1185. FT-IR t (cm-1): 1701 (C=O),
1655 (C=O).
(aromatic, para); d 128.5 (aromatic, meta); d 129 (aro-
matic, ortho); d 135 (C ipso); d 141 (N–CH=N); d 149 (N–
C–N); d 151 (N–C(=O)–N); d 155 (N–C(=O)–C). HRMS
(ESI): m/z calcd for C15H17N4O2Br: 285.1351 [M?–Br];
found: 285.1346. FT-IR t (cm-1): 1698 (C=O), 1649
(C=O).
Characterization of 1,3-dimethyl-9,7-dibenzylxanthinium
bromide (tphBz2Br, 1a0)
trans-Dibromo bis-7-(1,3-dimethyl-9-benzylxanthine)
palladium(II) (4)
1H NMR (CDCl3, 25 °C, 400.13 MHz): d 3.40 (s, 3H,
CH3–N); d 3.58 (s, 3H, CH3–N); d 5.54 (s, 4H, PhCH2–N);
d 7.30–7.45 (m, 10H, aromatics); d 7.58 (s, 1H, C–H).
HRMS (ESI): m/z calcd for C21H21N4O2Br: 361.1664
[M?–Br]; found 361.1655.
1,3-dimethyl-9-benzylxanthinium bromide (1a, 600 mg,
1.71 mmol) was dissolved in acetonitrile (10 mL). Palla-
dium acetate (190 mg, 0.86 mmol) and sodium bromide
(176 mg, 1.71 mmol) were added, and the mixture was
refluxed for 24 h, cooled and filtered to eliminate the
excess of NaBr and any unreacted 1a. The yellow solution
1
3,7-Dimethyl-9-benzylxanthinium bromide,
or theobromineBzBr (tbrBzBr, 2a)
was evaporated to dryness. Yield: 370 mg, 54%. H NMR
(CDCl3, 25 °C, 400.13 MHz): d 3.46 (s, 1H, CH3–N); d
3.60 (s, 1H, CH3–N); d 4.57 (s, 3H, CH3–N); d 4.72 (s, 3H,
CH3–N); d 5.54 (s, 2H, PhCH2–N); d 5.57 (s, 2H, PhCH2–
N); d 7.41–7.33 (m, 10H, aromatics); d 7.91 (s, 1H, C–H);
8,00 (s, 1H, C–H). 13C{1H} NMR (CDCl3, 25 °C,
100.5 MHz): d 28.5 (CH3–N); d 30 (CH3–N); d 33 (CH3–
N); d 33.4 (CH3–N); d 51.7 (PhCH2–N); d 51.8 (PhCH2–
N); d 108.3 (O=C–C–N); d 108.4 (O=C–C–N); d 128.2; d
128.5; d 129.4; d 129.5; d 129.6 (aromatics); d 133 (C
ipso); d 133.2 (C ipso); d 140 (N–CH=N); d 146.7 (N–C–
N); d 146.9 (N–C–N); d 150.9 (N–C(=O)–N); d 151 (N–
C(=O)–N); d 153.9 (N–C(=O)–C); d 154.1 (N–C(=O)–C).
Theobromine (2, 0.50 g, 2.80 mmol) and KI (23 mg,
0.14 mmol) were suspended in benzyl bromide (5 mL) and
vigorously stirred. The mixture was refluxed for 24 h,
cooled and filtered to eliminate any unreacted xanthine and
KI. Diethyl ether (10 mL) was added and upon stirring
overnight and filtering, a brown powder was obtained.
Yield: 0.94 g, 96%. 1H NMR (DMSO-d6, 80 °C,
400.13 MHz): d 3.35 (s, 3H, CH3–N); d 3.86 (s, 3H, CH3–
N); d 5.46 (s, 2H, PhCH2–N); d 7.33–7.02 (m, 5H, aro-
matics); d 7.90 (s, 1H, C–H); d 10.72 (broad, 1H, N–H).
13C{1H} NMR (DMSO-d6, 80 °C, 100.5 MHz): d 29
(CH3–N); d 34 (CH3–N); d 50 (PhCH2–N); d 108 (O=C–
C–N); d 128.1 (aromatic, para); d 128.4 (aromatic, meta);
d 129 (aromatic, ortho); d 135 (C ipso); d 143 (N–CH=N);
d 150 (N–C–N); d 152 (N–C(=O)–N); d 155 (N–C(=O)–C).
HRMS (MALDI-FT): m/z calcd for C14H15N4O2Br:
350.0378 [M]; found 350.0435. FT-IR t (cm-1): 1685
(C=O).
trans-Dichloro bis(1,3,7-trimethyl-9-benzylxanthine-8-
ylidene) palladium(II) (6)
1,3,7-trimethyl-9-benzylxanthinium bromide (3a, 91 mg,
0.25 mmol) was suspended in THF (1.5 mL), and a solu-
tion of potassium tertbutoxide (28 mg, 0.25 mmol) in THF
(0.5 mL) was slowly added at room temperature. The
mixture was stirred at room temperature for 3 h and filtered
to eliminate KBr, and a brown solution was obtained.
PdCl2(NCPh)2 (34 mg, 0.13 mmol) dissolved in acetoni-
trile (2 mL) was slowly added to the free carbene obtained
from 3a (the brown solution), and the mixture was stirred
for 4 h at room temperature. The bright yellow solution
1,3,7-Trimethyl-9-benzylxanthinium bromide,
or caffeineBzBr (caffBzBr, 3a)
Caffeine (3, 1.00 g, 5.15 mmol) was suspended in benzyl
bromide (10 mL) and vigorously stirred. The mixture was
refluxed for 24 h and then cooled and filtered to eliminate
any unreacted 3. The solvent was reduced under vacuum,
and diethyl ether (50 mL) was added. Upon stirring for 1 h
and filtering, a pale brown powder was obtained. Yield:
1
was evaporated to dryness Yield: 160 mg, 86%. H NMR
(CDCl3,, 25 °C, 400.13 MHz): d 3.36 (s, 3H, CH3–N); d
3.55 (s, 3H, CH3–N); d 3.97 (s, 3H, CH3–N); d 5.52 (s, 2H,
PhCH2–N); d 7.37–7.35 (m, 3H, aromatics); d 7.52 (s, 1H,
aromatic, ortho); d 7.63 (s, 1H, aromatic, ortho). 13C{1H}
NMR (CDCl3, 25 °C, 100.5 MHz): d 27.5 (CH3–N); d 27.6
(CH3–N); d 29 (CH3–N); d 33 (CH3–N); d 50 (PhCH2–N);
d 107 (O=C–C–N); d 128 (aromatic, para); d 128.3 (aro-
matic, meta); d 129 (aromatic, ortho); d 135.9 (C ipso); d
136.0 (C ipso); d 141 (N–CH=N); d 141.4 (N–CH=N); d
1
0.81 g, 43%. H NMR (CDCl3, 25 °C, 400.13 MHz): d
3.39 (s, 3H, CH3–N); d 3.57 (s, 3H, CH3–N); d 3.98 (s, 3H,
CH3–N); d 5.49 (s, 2H, PhCH2–N); d 7.34–7.30 (m, 5H,
aromatics); d 7.58 (s, 1H, C–H). 13C{1H} NMR (CDCl3,
25 °C, 100.5 MHz): d 28 (CH3–N); d 30 (CH3–N); d 33
(CH3–N); d 50 (PhCH2–N); d 107 (O=C–C–N); d 128
123

168
Transition Met Chem (2010) 35:165–175
148.7 (N–C–N); d 148.9 (N–C–N); d 151 (N–C(=O)–N); d
155.2 (N–C(=O)–C); d 155.3 (N–C(=O)–C). HRMS (ESI):
m/z calcd for C30H32N8O4Cl2Pd: 705.1291 [M?–Cl];
found: 705.2322.
Table 1 Summary of crystal data for PdBr2(tphBz-H)2 (4)
4
Formula
C28H28Br2N8O4Pd
806.79
Mol. wt.
Cryst size (mm)
Cryst. syst.
Space group
0.50 9 0.45 9 0.10
Monoclinic
P21/a
In situ reaction of 2a with Pd(OAc)2: Cis and trans-
dibromo bis-1-(3,7-dimethyl-9-benzylxanthine)
palladium(II) (5)
˚
A (A)
8.308 (3)
11.266 (3)
16.358 (5)
93.818 (10)
1527.6 (8)
2
˚
B (A)
3,7-dimethyl-9-benzylxanthinium bromide (2a, 200 mg,
0.570 mmol) was suspended in acetonitrile (20 mL). Pal-
ladium acetate (63.1 mg, 0.281 mmol) and sodium bro-
mide (58 mg, 0.560 mmol) were added. The mixture was
refluxed for 24 h, cooled and filtered to eliminate NaBr and
any unreacted 2a. The orange solution was evaporated to
˚
C (A)
b (°)
3
˚
V (A )
Z
dcalc (mg/m3)
Abs. coeff. (mm-1
F(000)
1.624
1
dryness. Isomers ratio = 2:1. H NMR ((CD3)2CO, 25 °C,
)
3.26
400.13 MHz): d 3.42 (s, 3H, CH3–N, trans); d 3.43 (s, 3H,
CH3–N, trans); d 3.91 (s, 3H, CH3–N, cis); d 3.97 (s, 3H,
CH3–N, cis); d 4.70 (s, 2H, PhCH2–N, cis); d 5.56 (s, 2H,
PhCH2–N, trans); d 7.33–7.31 (m, 5H, aromatics, cis); d
7.35–7.33 (m, 5H, aromatics, trans); d 7.94 (s, 1H, C–H,
cis); 8,12 (s, 1H, C–H, trans). 13C{1H} NMR ((CD3)2CO,
25 °C, 100.5 MHz): d 25 (CH3–N); d 30 (CH3–N); d 50
(PhCH2–N); d 110 (O=C–C–N); d 128.0 (aromatic, para);
d 128.2 (aromatic, meta); d 128.8 (aromatic, ortho); d 137
(C ipso); d 142 (N–CH=N); d 150 (N–C–N); d 151 (N–
C(=O)–N); d 155 (N–C(=O)–C). Both isomers have almost
identical 13C{1H} NMR spectra.
736
h range (°)
Index ranges
2.2–28.0
-10 B h B 9,
-14 B k B 14,
-18 B l B 18
16898
Tot. no. of data
No. of unique data,
I C 2r(I)
2151
[R(int)] = 0.059
0.0772
R (all data)
wR = 0.2410
0.99, -1.06
3
Largest diff. peak and hole (e/A )
˚
X-ray crystal structure determination for compound (4)
and R-factors based on ALL data will be even larger. All
structure solutions and refinements were made using
SHELXS97 [24] and SHELXL97 [24].
Data for compound 4 was collected using a Rigaku AFC-
7S diffractometer, provided with graphite monochromated
Crystallographic data for 4 has been deposited with the
Cambridge Crystallographic Data Centre as number CCDC
745624. It contains the supplementary crystallographic
data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
˚
Mo Ka radiation (k = 0.71073 A). Details of the crystal
data and structure refinement are shown in Table 1. A
semi-empirical absorption correction [23] was applied to
the set of data. Crystal structure was solved by direct
methods, and the final models were reached by Fourier
techniques. All e.s.d.’s (except the e.s.d. in the dihedral
angle between two l.s. planes) are estimated using the full
covariance matrix. The cell e.s.d.’s are taken into account
individually in the estimation of e.s.d.’s in distances, angles
and torsion angles; correlations between e.s.d.’s in cell
parameters are only used when they are defined by crystal
symmetry. An approximate (isotropic) treatment of cell
e.s.d.’s is used for estimating e.s.d.’s involving l.s. planes.
Refinement of F2 against ALL reflections. The weighted R-
factor wR and goodness of fit S are based on F2, conven-
tional R-factors R are based on F, with F set to zero for
negative F2. The threshold expression of F2[r(F2) is used
only for calculating R-factors(gt) etc. and is not relevant to
the choice of reflections for refinement. R-factors based on
F2 are statistically about twice as large as those based on F,
Theoretical studies on theophylline (1), theobromine
(2) and caffeine (3)
Nature of active sites can be particularly explored using the
electrostatic potential, V(r), which let us directly to deter-
mine where the electron-rich sites in a molecule or crystal
are localized [25–42]. For the region nearest to the nucleus,
VN dominates and V(r) has similar topology to the electron
density [43, 44], q(r). i.e., positive-valued maxima at the
nuclear site and a positive-valued saddle point between
every pair of bonded atoms. Nevertheless, the existence of
maxima is ruled out via an established result that, barring
the nuclear position, there cannot exist any strict local
maxima in the V(r) map [31–33]. For the region where VE
123

Transition Met Chem (2010) 35:165–175
169
dominates (V(r) is negative) the V(r), topography can be
more complex. The minima of the negative region denote
the zones to which an approaching electrophile may be
attracted. On the contrary, the positive regions do not have
maxima, which might indicate sites for nucleophilic attack.
Nevertheless, Politzer and Sjoberg have shown that by
computing V(r) on the 0.002-electron/bohr3 contour iso-
surface [45] of the molecular electronic density q(r), we
can quantify the susceptibility of molecules to nucleophilic
attack. They demonstrated that relative magnitudes of the
positive electrostatic potential in various regions on this
surface do reveal the sites most susceptible to nucleophilic
attack. Mapping on this isosurface the V(r) values onto
colors allows us to identify the host sites in which nucle-
ophiles (most positive zone) and electrophiles (most neg-
ative zone) should bind. Additionally, the active sites’
susceptibility can be quantified determining the minimum
and maximum V(r) values at the determined host zones
using a Newton–Raphson technique similar to those that
have been previously reported by Aray et al. for the study
of the electronic density [46, 47] and electrostatic potential
topology [40–42, 48].
DMol3 program [49, 50] was used to calculate q(r) and
V(r). DMol3 calculates variational self-consistent solutions
to the DFT equations, expressed in an accurate numerical
atomic orbital basis. The solutions to these equations pro-
vide the molecular electron densities, which can be used to
evaluate the total electrostatic potential of the system. The
correlation and exchange effects were considered using the
gradient generalized approximation (GGA) and the func-
tional PW91 [51]. The numerical double-zeta plus polari-
zation basis set DNP was used in all calculations.
Fig. 2 Methylated xanthines with their representation of V(r). a
Theophilline, b Theobromine, c Caffeine. The white spheres corre-
spond to the H atoms, the red ones to the O atom, the blue ones to the
nitrogen atoms and the gray ones to the C atoms
oxygen atoms, and N(9) (see Fig. 1 for numbering), in each
case. The highest contribution of the electronic distribution
is over the oxygen atoms non-bonded electronic pairs, as
well as the nitrogen lone pairs in N(9). The basic nature of
N(9) in each case becomes then evident. Reported studies
show that, for example, for the purine analog guanine and
its tautomers, the protons on N(7) and N(9) are the most
acidic ones, which reveals the basic nature of the nitrogen
atoms in these cases [59], and is also consistent with our
results.
Results and discussion
Theoretical modeling on theophylline (1), theobromine
(2) and caffeine (3)
The positive sectors are represented with colors between
yellow and red. As can be seen from the figure, the most
positive zones are located over the groups on C(8) and N(7)
in each of the studied xanthines, meaning that these cor-
respond to the most acidic protons in 1, 2 and 3.
Theobromine (2), in particular, also presents an impor-
tant center of acidity on the proton bonded to N(1). In the
case of theophylline (1), the protons from both the C(8) and
N(7) centers are mainly acidic. This contrasts with the
previously reported data for adenine and guanine [59] in
which the C(8) site was much less acidic than N(1), and
explains why N(1) in theobromine is highly acidic. The
acid/base nature of the centers in these molecules must then
become the driving force for any reaction in which they are
involved.
Computational calculations on other oxopurines have
previously been done. For examples on calculations on
oxopurines, see [52–57]. For example, the determination of
relevant lowest energy reaction channels in guanine has
been studied using ab initio methods [53] and the acid/base
behavior of purine analogs was reported, on the basis of
theoretical and spectroscopic NMR data [55, 57]. Molec-
ular orbital aspects regarding caffeine have only been
considered in terms of its adsorption onto polymeric resins
[58].
Figure 2 shows the representation of V(r) over a density
isosurface for theophylline (1), theobromine (2) and caf-
feine (3). The negative zones are represented by the dif-
ferent tones of blue and green, and are located around the
123

170
Transition Met Chem (2010) 35:165–175
In terms of the coordination chemistry of these mole-
On the other hand, the Lewis acid site (LUMO) corre-
sponds to the azole ring, basically around N(7). This might
be due to a partial delocalization of the electron lone pair in
the five-membered ring. The acidic character is actually
remarkable when the substituent on N(7) is a proton (the-
ophylline, 1). This prediction of our model totally agrees
with the experimental observations that have previously
been reported [3, 9], where theophylline can be easily
deprotonated in presence of a basic late transition metal
precursor or a late transition metal precursor in basic
conditions, yielding N(7) coordinated species. The prefer-
ential coordination site at N(7), which has also been the-
oretically studied [3, 61], has previously been proposed not
only for theophylline derivatives but also for analogous
molecules such as guanine and adenine.
cules, it might then be possible to expect their interaction
with transition metals to occur mainly through N(9) in each
case. For 1 and 2, the presence of a base in the reaction
media might then cause a deprotonation to occur, and the
interaction with N(7) for theophylline and N(1) for theo-
bromine should then become possible. Theoretical calcu-
lations have been widely considered to predict the
reactivity of N-ligands toward metals [60].
One of our objectives was to study the coordination
chemistry of the unmodified xanthines 1, 2 and 3 toward
substitution on N(9) in order to obtain new imidazolium
salts derived from methylated xanthines. Then, the main
objective was to study the coordination chemistry of the
new ligands obtained toward palladium, in an attempt to
synthesize NHC-Pd compounds, and contrast their behav-
ior with the previously known examples of methylated
xanthines-palladium species [2–8].
For caffeine and theobromine, both with identical azole
rings, the frontier orbitals are very similar and there is a
LUMO character on N(9). Thus, an interaction with a rich
metal center is also predicted. In fact, N(9) complexes of
caffeine are well known [2, 60, 62–66], and there are
examples reported with theobromine [3, 67–69]. The latter
one also displays an acidic nature (LUMO) onto N(1),
substituted by a proton. Some N(1) bounded species of
theobromine have previously been reported [68].
In this sense, the acid/base analysis of our theoretical
results indicates that the unmodified xanthines should
interact with transition metals mainly through N(9), which
is exactly the position where the modification will be
introduced to obtain imidazolium salts. It is important then
to study the frontier orbitals of 1, 2 and 3 to determine, on
the orbital level, whether the modification is possible.
The representation of the frontier orbitals of caffeine (3)
is shown in Fig. 3. Analog representations have been
obtained for the frontier orbitals of 1 and 2, and the figures
are available as supplementary material. According to
Fig. 3, and also for molecules 1 and 2, in all cases the
Lewis basic site (HOMO) is located in the six-membered
ring. Theophylline (1) and caffeine (3) have essentially the
same frontier orbitals, as is expected since both of their six-
membered rings are identical. For theobromine (2), the
geometry of the HOMO is slightly different, probably in
response to the presence of a hydrogen atom on N(1),
whose electronic effect is different from that of the methyl
group. For all of the xanthines, a rich electronic density
region is observed on one of the carbonyl groups (C(6)=O)
and N(3).
As it can be seen from the HOMO-like orbitals in 1-3,
the bonding character in the oxo-substituted ring is similar
to that of guanine or other purines [53]. It has been pro-
posed that the reactivity of adenine and guanine might be
due to stretching/twisting of the purine molecules due to
delocalization of their bonding orbitals [53], which is easier
in those than in the cases studied here. Molecules 1-3 are
more constrained due to their planarity and, therefore, it is
more difficult for those to adopt any twisted conformation
in which their reactivity might be enhanced.
Synthesis of benzyl substituted imidazolium salts
from methylated xanthines
The synthesis of imidazolium salts from methylated
xanthines has only been reported for caffeine by different
groups [10–13, 70]. However, only the methyl derivative
has been synthesized. So far, to the best of our knowledge,
there are no other examples of modified xanthines to pro-
duce imidazolium salts.
Our approach is based on the nucleophilic substitution
of benzyl bromide by the methylated xanthines, as shown
in Scheme 1. The reaction is straightforward and produces
1a, 2a and 3a in moderate to good yields. This procedure
has been previously used for the synthesis of a variety of
imidazolium salts and ionic liquids [71, 72].
The new imidazolium salts 1a and 3a are soluble in
common organic solvents such as chloroform or acetoni-
trile. The solubility of 2a in most of the common organic
Fig. 3 Representation of the frontier orbitals for caffeine (3). a
HOMO for 3; b LUMO for 3
123

Transition Met Chem (2010) 35:165–175
171
binding might be enhanced. This will be a matter of future
communications.
Coordination chemistry of 1a, 2a and 3a
toward palladium
Methylated xanthines have previously been studied in
coordination chemistry, and several examples are reported
with the unmodified xanthines [5, 6, 8]. Also, some species
resulting from the interaction between protonated species
of theophylline, theobromine and caffeine and palladium-
halide anions have been reported [3, 4], although these are
not coordination compounds between the corresponding
xanthine and palladium but plain electrostatic attractions.
Our main purpose was to obtain the NHC-Pd complex
from our imidazolium salts 1a, 2a and 3a. However, an
interesting behavior was identified. One of the main
problems in the stabilization of a carbenic species with
theophylline is that one of the nitrogen atoms of the
imidazole ring is substituted by a proton. Although there
are examples in the literature in which Pd-NHC complexes
with hydrogen substituting one of its nitrogen atoms have
been obtained [9], some of those correspond to the product
of an internal acid catalyzed reaction of a diamminopal-
ladium (II) complex [73, 74].
Scheme 1
solvents is poor, but better than its parent compound
theobromine, 2.
The new ligands were characterized using multinuclear
NMR, FTIR and HRMS. As expected, all of the techniques
indicate effective substitution on N(9). This becomes evi-
1
dent in the H NMR spectra from the shift downfield of
about 1 ppm observed in the –CH2– protons of the benzyl
group in all of the cases. This unshielding is indicative of
the new N(9)–CH2– bond formed. Also, H8 shifts in every
case, compared to the parent compounds 1–3. These data,
along with the appearance of the carbonyl groups in the
FTIR and the excellent match between the calculated and
experimental values of HRMS confirms that we have
indeed obtained the new imidazolium salts 1a, 2a and 3a.
Interestingly, the nucleophyllic substitution was carried
out over an aliphatic halide such as benzyl bromide. Even
when there are some examples of this type of reaction for
other imidazoles [71, 72], the previous example of modi-
fied xanthine, i.e. methyl caffeine, was obtained with very
strong alkylating agents, such as dimethylsulphate or MeI
[10–13, 70].
When 1a reacted with Pd(OAc)2 [75–80], in an attempt
to synthesize the Pd-NHC complex, a solvent dependant
behavior was observed. A CDCl3 solution of 1a reacted in
situ (NMR timescale) with Pd(OAc)2, showed a shift in the
signals (downfield) assigned to the ligand. A new signal
attributed to H8 indicates that the interaction has occurred
through N7 instead of C8. In another test, a DMSO-d6
solution of the isolated compound (synthesized according to
the procedure analogous to that of Huynh [81]) shows a
low-field signal in the 13C{1H} NMR spectrum (d 174), and
the disappearance of the H8 proton in the 1H NMR, both of
them indicative of a carbenic interaction between 1a and
Pd(II). Interestingly, a septuplet observed in the 13C{1H}
NMR spectrum is assigned to a DMSO ligand coordinated
to Pd in this species, with a coupling constant of 19.6 Hz,
which is similar to the values reported for this ligand in
similar coordinative environments (JCD = 21.0 Hz) [82].
Over time, the low-field signal in the 13C{1H} NMR
spectrum shifts to higher fields and eventually disappears.
In the 1H NMR spectrum, as the very low field signal shifts,
the signal corresponding to H8 is observed again. This
behavior is indicative of a change in the coordination site
toward Pd(II) and was also observed when the test was
done using CD3CN as solvent.
In fact, it is observed that the alkylation reaction occurs
more easily on theophylline than on theobromine or caf-
feine. This observation could easily be explained by our
theoretical studies, since, as it was explained before, it
becomes obvious that the orbitals which interact with the
alkylating agent are less available for caffeine than for
theophylline or theobromine. As we have confirmed
experimentally, the nucleophyllic substitution was only
possible under harsh conditions: refluxing the xanthines in
the neat alkylating agent (approximately 200 °C) for long
periods of time (1–4 days).
On the other hand, when the synthesis of 1a was
attempted using strong reaction conditions, we also
detected the dialkylated imidazolium salt (i.e. both N(9)
and N(7) have been substituted by a benzyl group) as a
minor reaction byproduct that could be characterized. The
1,3-dimethyl-9,7-dibenzylxanthinium bromide (tphBz2Br,
1a0) or other analog N(7) substituted species could be very
interesting as potential carbene precursors, since a C(8)
We propose an equilibrium between the carbene and the
theophyllinate species, the kinetic product being the car-
bene and the thermodynamic product the theophyllinate, as
depicted in Scheme 2. The dynamic behavior observed can
123

172
Transition Met Chem (2010) 35:165–175
Scheme 2
be slowed with the solvent used. In fact, the ‘‘theophyllinate’’
species seems to be favored by poor coordinating solvents
(CDCl3, for example), since in those the carbenic species is
almost undetectable. When the product is analyzed in
strongly coordinating solvents such as DMSO-d6, the car-
bene can be stabilized, and detected spectroscopically.
The behavior observed is possible, since the coordina-
tion of DMSO restrains the free rotation around the
M-NHC complex due to its anisotropy [75–78]. In fact, in
chloroform, the possibility of stabilizing the species
through coordination of the solvent does not exist, and
rotation becomes possible, explaining the double amount of
signals observed on the NMR for the ligand in this complex
(see Experimental section, synthesis of 4). In this sense, in
solution, the interconversion between the carbene and the
N-coordinated species probably depends on the anisotropy
of 1a as NHC [75–78].
Fig. 4 Solid state structure of PdBr2(tphBz-H)2 (4)
Complex 4 crystallizes in a monoclinic system with a
space group P21/a. As shown in Fig. 4, ligand 1a is
bounded to Pd through N(7) in an almost perfect square
planar geometry, with a small deviation of 0.5° from the
ideal 90° angle. In the solid state, the new ‘‘theophyllinate’’
complex 4 is absolutely co-planar, since both the Br1–Pd1–
Br1i and N1i–Pd1–N1 angles are of 180 degrees, showing a
In an attempt to stabilize the carbene, the synthesis was
repeated in acetonitrile using NaBr. Either with or without
NaBr, the species obtained is the same, the N-coordination
product, or ‘‘theophyllinate’’. The formation of the ‘‘theo-
phyllinate’’ species is confirmed by X-ray diffraction and
was either identified on NMR timescale and individually
synthesized from Pd(OAc)2 and 1a. Figure 4 displays the
solid state structure for PdBr2(tphBz-H)2, 4. This complex is
analogous to the previously reported PdCl2(tph)2 [6, 8]
containing the unmodified theophylline ligand. Tables 1 and
2 show a summary of the crystal data and a selection of bond
distances and angles respectively, for PdBr2(tphBz-H)2 (4).
˚
high level of symmetry. The Pd–Br distance is 2.4261A,
typical length for this kind of bonds as reported previously
˚
for analogous species [81]. The Pd-N bond, 2.023A, is also
in the range of reported distances for Pd-Nsp2 interactions
[9]. The new ‘‘theophyllinate’’ complex is the trans isomer,
probably in order to minimize the sterical hindrance
between the xanthine ligands. In fact, this is also reflected
in the transoid configuration adopted by the benzyl groups,
with each one back and forward respectively in terms of the
plane orthogonal to the one containing both of the nitrogen
atoms in the PdBr2N2 unit. Regarding its spatial configu-
ration and since the compound only presents one symmetry
element, an improper S2 axis, it can be classified as a Ci.
It has been observed that theophylline has a particular
reactivity due to the presence of an acidic proton on the
nitrogen [3, 9]. In most cases, theophylline coordinates as
theophyllinate anionic species through this nitrogen [67].
For instance, Romerosa et al. have obtained a complex
where both theophylline and 8-(methylthio)-theophylline
are coordinated as monoanionic ligands to palladium (II) [9].
The interaction between theophylline and palladium
complexes has previously been reported [63]. By reaction
˚
Table 2 Selected bond distances (A) and angles (°) for PdBr2(tphBz-
H)2 (4)
˚
Distances (A)
Pd1–N1i
2.023(6)
2.023(6)
Pd1–Br1i
Pd1–Br1
2.4261(11)
2.4261(11)
Pd1–N1
Angles (°)
N1i–Pd1–N1
N1i–Pd1–Br1
N1–Pd1–Br1
180.000(2)
90.5(2)
N1i–Pd1–Br1i
N1–Pd1–Br1i
Br1–Pd1–Br1i
89.5(2)
90.5(2)
180.0
89.5(2)
The ‘‘i’’ atoms are generated by an improper S2 axis, i.e., a reflection
plus a rotation
123

Transition Met Chem (2010) 35:165–175
173
of theophylline with K2PdCl4, two kinds of metallic com-
plexes are formed, and either one depends on the acidity
and the solvent used. Pneumatikakis et al. [63] identified
two coordination modes for theophilline: the first one as a
neutral ligand through N(9), obtaining K[Pd(theophyl-
line)Cl3], and the second one involving a monoanionic
bidentate chelate through the N(7) and O(6), yielding a
dimer such as [Pd(N^O-theophylline)Cl]2 [63].
However, in those examples theobromine coordinated as a
monoanionic ligand to palladium and platinum. To the best
of our knowledge, this is the first example of an imidazo-
lium salt with theobromine, and as a consequence, the first
coordination compound reported of it.
Since both the theophylline and the theobromine benzyl
derivatives tend to interact with palladium as N-coordi-
nated species, but in the case of 1a it was possible to detect
a carbenic species, it becomes interesting to contrast these
ligands with the analogous 1,3,7-trimethyl-9-benzyl-
xanthinium bromide (3a) in which the possibility of
N-coordination is blocked, since N1, N3 and N7 are
substituted by methyl groups.
Then, the imidazolium salt 1a, obtained from theophyl-
line, acts mainly as an N-bounded ligand, confirming
the nature of the binding site previously observed for
the unmodified theophylline and other derivatives of it
[9, 62, 63].
The theobromine derivative, 2a, was also studied. One
of the main problems with this one is that its poor solubility
in most of the solvents makes it difficult to choose a
reaction medium. Nevertheless, an in situ reaction was
performed on the NMR timescale. Two sets of identical
The coordination study of CaffBzBr (3a) was initially
performed in situ on the NMR timescale. A reaction
between Pd(OAc)2 and 3a in CDCl3 was studied, observing
that the ligand does not react completely, but a clear set of
shifted signals lacking the one assigned to H8 appears in
1
1
signals are observed in the H NMR spectrum, in a 2:1
the H NMR. The species is stable even after 24 h. How-
ratio, both of them shifted from those of 2a. Interestingly,
the acid proton assigned to N1-H is not observed. From the
13C{1H} NMR spectrum, it is possible to observe the signal
corresponding to C8 basically at the same chemical shift
from the parent compound 2a, and no carbenic interactions
are observed.
ever, even when this behavior was observed, the amount of
ligand that reacts under the same conditions when the
reaction is scaled up is negligible, and very low yields of
the carbene are obtained and the unreacted ligand is
recovered almost completely. This might be a consequence
of the strength of the acetate, which acts as an internal base
and should deprotonate C8 to generate the free carbene.
In the case of 3a, the use of a stronger base to deprotonate
C8 is possible, since the only acidic proton in the ligand is
H8, and no further modifications of the chemical structure
would occur, as in 1a or 2a. Using KOtBu as deprotonating
agent, and PdCl2(NCPh)2 as palladium precursor, it was
possible to obtain the Pd-NHC complex from 3a, in good
yields (86%). The pale yellow product was characterized
using multinuclear NMR. The 1H NMR spectrum lacks the
signal assigned to H8, and all of the other signals shift
downfield. Based on these results, it is possible to propose a
From these spectroscopic data, it is possible to propose a
neutral coordination of 2a to palladium through N1, with
the general formula of PdBr2(tbrBz-H)2. With the infor-
mation available, it is not possible to assign a specific
spatial configuration of the isomers or which isomer is the
major one. The structures proposed for these species are
shown in Fig. 5. This reaction is currently being scaled up,
and further studies on the characteristics of these species
will be performed.
There are some examples of N-coordinated complexes
from the unmodified theobromine in the literature [63].
Fig. 5 Proposed structures for the N-coordinated product PdBr2(tbrBz-H)2 (5)
123

174
Transition Met Chem (2010) 35:165–175
promoting this scientific activity, and Laboratorios FARMA (Caracas,
Venezuela) for a generous loan on theophylline. We also thank ETH
Zu¨rich DCHAB/LOC MS-Service for performing the HRMS analyses
and Lic. Giuseppe Lubes for some previous synthetic tests.
References
1. Lusty JR, Lee PF (1984) Inorg Chim Acta 91:L47–L49
2. Goodgame DML, Hayman PB, Riley RT, Williams DJ (1984)
Inorg Chim Acta 91:89–93
´
3. Colacio-Rodrıguez E, Salas-Peregrin JM (1984) Thermochim
Acta 74:45–54
´
4. Moreno-Vida MI, Colacio-Rodrıguez E, Moreno-Carretero MN,
Salas-Peregrin JM, Simard M, Beauchamp AL (1989) Inorg
Chim Acta 157:201–207
Fig. 6 Proposed structure for PdCl2(CaffBz)2 (6)
´
5. Moreno-Vida MI, Colacio-Rodrıguez E, Moreno-Carretero MN,
Ruiz-Sanchez J, Salas-Peregrin JM (1987) Thermochim Acta
´
biscarbene complex, such as PdCl2(caffBz)2 (6) (Fig. 6).
Steric hindrance might become the driving force for the
proposed configuration. The available information does not
provide further evidence to assign a final stereochemistry.
However, the proposed configuration is based on the pre-
vious examples [13] of NHC’s from methylated caffeine, an
imidazolium salt analogous to 3a.
115:45–55
6. Salas-Peregrin JM, Colacio-Rodrıguez E, Romero-Molina MA,
´
´
´
Sanchez-Sanchez MP (1983) Thermochim Acta 69:313–321
7. Clarke MJ, Taube H (1975) J Am Chem Soc 97:1397–1403
8. Umapathy P, Harnesswala RA, Dorai CS (1985) Polyhedron
4:1595–1602
´
9. Romerosa A, Varela JS, Hidalgo MA, Avila-Roson JC, Colacio E
(1997) Inorg Chem 36:3784–3786
10. Kascatan-Nebioglu A, Panzner MJ, Garrison JC, Tessier CA,
Youngs WJ (2004) Organometallics 23:1928–1931
11. Kascatan-Nebioglu A, Melaiye A, Hindi K, Durmus S, Panzner
MJ, Hogue LA, Mallett RJ, Hovis CE, Coughenour MH, Crosby
SD, Milsted A, Ely DL, Tessier CA, Cannon CL, Youngs WJ
(2006) J Med Chem 49:6811–6818
12. Melaiye A, Simons RS, Milsted A, Pingitore F, Wesdemiotis C,
Tessier CA, Youngs WJ (2004) J Med Chem 47:973–977
13. Herrmann WA, Schu¨tz J (2004) J Organomet Chem 689:2995–
2999
14. Chung CK, Grubbs RH (2008) Org Lett 10:2693–2696
15. Poyatos M, McNamara W, Incarvito C, Clot E, Peris E, Crabtree
RH (2008) Organometallics 27:2128–2136
16. Ray L, Shaikh MM, Ghosh P (2008) Inorg Chem 47:230–240
17. Radius U, Bickelhaupt FM (2009) Coord Chem Rev 253:678–686
18. Jacobsen H, Correa A, Poater A, Costabile C, Cavallo L (2009)
Coord Chem Rev 253:687–703
Conclusions
Three new imidazolium salts based on natural methylated
xanthines bearing the benzyl group onto N(9) were syn-
thesized and characterized. The coordination chemistry
toward palladium of these potential NHC’s precursors was
studied. 1a acts both as a carbenic and an N-bounded
ligand, in a behavior strongly dependent on the coordina-
tive properties of the reaction media. An equilibrium is
proposed between both species, the carbene being the
kinetic product, while the N-coordination is favored by the
thermodynamics. 2a acts essentially as an N-bounded
ligand, and the C(8) coordination was not detected spec-
troscopically, and a mixture of isomers is proposed.
Finally, 3a is bounded to palladium as expected, as an
NHC, i.e., via C(8), since in this case there are no other
potential coordination sites.
19. Buscemi G, Biffis A, Tubaro C, Basato M (2009) Catal Today
140:84–89
20. Beck W, Kottmair N (1976) Chem Ber 109:970–993
21. Herrmann WA (2002) Angew Chem Int Ed 41:1290–1309
22. Doyle JR, Slade PE, Jonassen HB (1960) Inorg Synth 6:216–219
23. North ACT, Phillips DC, Mathews FS (1968) Acta Cryst A.
24:351–359
The two theoretical models described for 1, 2 and 3,
agree completely with our synthetic experiences and with
the experimental facts previously reported. Both approa-
ches could be valuable tools in future xanthines’ chemistry
studies, to predict reactivities and trends based on acid–
base considerations. In fact, the synthesis of other imi-
dazolium salts from methylated xanthines and their com-
plexes is currently under study and will be matter of a
future communication.
24. Sheldrick GM, SHELX-97 (Release 97-2) (1997) Program for
structure refinement. University of Gottingen, Gottingen
25. Leboeuf M, Koster M, Jug K, Salahub DR (1999) J Chem Phys
111:4893–4905
26. Gadre SR, Shirsat RN (2000) Electrostatics of atoms and mole-
cules. Universities Press, Hyderabad
27. Pingale SS, Gadre SR, Batolotti LJ (1998) J Phys Chem A
102:9987–9992
28. Deshmukh MM, Sastry NV, Gadre SR (2004) J Chem Phys
121:12402–12410
29. Joshi KA, Gejji SP (2005) J Mol Struc Theochem 724:87–93
30. Tachikawa H, Iyama T, Kawabata H (2005) J Mol Struc Theo-
chem 718:117–122
Acknowledgments The authors thank the Decanato de Investiga-
´
31. Pathak RK, Gadre SR (1990) J Chem Phys 93:1770–1773
´ ´
cion y Desarrollo de la Universidad Simon Bolıvar (DID-USB) for
123

Transition Met Chem (2010) 35:165–175
175
32. Gadre SR, Kulkarni SA, Shrivastava IH (1992) J Chem Phys
96:5253–5260
33. Gadre SR, Pathak RK (1989) Proc Ind Acad Sci (Chem Sci)
102:189–192
58. Dutta Saikia M, Dutta NN (2006) Col Surfa A 280:163–168
59. Huang Y, Kenttamaa H (2004) J Phys Chem A 108:4485–4490
60. Blanco F, Alkorta I, Zborowski K, Elguero J (2007) Struct Chem
18:965–975
34. Alhambra C, Luque FJ, Orozco M (1995) J Phys Chem 99:3084–
3092
61. Baik M, Friesner RA, Lippard SJ (2003) J Am Chem Soc
125:14082–14092
35. Kornelak P, Michalak A, Najbar M (2005) Catal Today 101:175–
183
62. Pneumatikakis G (1984) Inorg Chim Acta 93:5–11
63. Pneumatikakis G, Yannopoulos A, Markopoulos J, Angelopoulos
C (1988) Inorg Chim Acta 152:101–106
64. Rajendran S, Amalraj AJ, Joice MJ, Anthony N, Trivedi DC,
Sundaravadivelu M (2004) Corr Rev 22:233–248
36. Michalak A (2004) Chem Phys Lett 386:346–350
37. Politzer P, Truhlar DG (1982) Chemical applications of atomic
and molecular electrostatic potentials. New York, Plenum
38. Murray JS, Sen KD (1996) Molecular electrostatic potential:
concepts and applications. Elsevier, Amsterdam
39. Orozco M, Luque FJ (1996) Theor comput Chem Ser 3:181
˘
65. Kolaylı S, Ocak M, Ku¨c¸u¨k M, Abbasoglu R (2004) Food Chem
84:383–388
66. Nafisi S, Sadjadi AS, Zadeh SS, Damerchelli M (2003) Biomol
Struct & Dynam 21:289–296
67. Mikulski CM, Kurlan MK, Bayne M, Gaul M, Karayannis NM
(1988) J Coord Chem 17:267–275
68. Mikulski CM, Grossman S, Lee CJ, Karayannis NM (1987) Trans
Met Chem 12:21–25
69. Mikulski CM, Grossman S, Bayne ML, Gaul M, Kanach D, Udell
K, Karayannis NM (1990) Inorg Chim Acta 173:31–35
70. Ivanov EI, Kalayanov GD, Yaroshchenko IM, Stepanov DE
(1989) Chem Heterocyc Comp 25:1318
´
40. Aray Y, Marquez M, Rodrıguez J, Coll S, Simon-Manso Y,
Gonzalez C, Weitz DA (2003) J Phys Chem B 107:8946–8952
´
´
41. Aray Y, Marquez M, Rodrıguez J, Coll S, Simon-Manso Y,
Gonzalez C, Weitz DA (2004) J Phys Chem B 108:2418–2424
´
´
42. Aray Y, Rodrıguez J, Coll S, Gonzalez C, Marquez M (2004) J
Phys Chem B 108:18942–18948
43. Tal Y, Bader RFW, Erkku J (1980) Phys Rev A 21:1–11
44. Keith TA, Bader RFW, Aray Y (1996) Int J Quantum Chem
57:183–198
45. Sjoberg P, Politzer P (1990) J Phys Chem 94:3959–3961
´
71. Dupont J, Consorti CS, Suarez PAZ, de Souza RF (2004) Org
Synth 10:184–187
46. Aray Y, Rodrıguez J (2007) J Mol Catal A 265:32–41
´
47. Aray Y, RodrIguez J, Vega D (2006) In: Matta CF, Boyd RJ (eds)
72. Creary X, Willis E (2005) Org Synth 82:166–167
73. Ito Y, Hirao T, Tsubata K, Saegusa T (1978) Tetrahedron Lett
18:1535–1538
The quantum theory of atoms in molecules: from DNA to solid
and drug design. Wiley-VCH, Weinheim
´
´
48. Aray Y, Rodrıguez J, Coll S, Rodrıguez-Arias E, Vega D (2005)
J. Phys. Chem. B 109:23564–23570
49. Delley B (1990) J Chem Phys 92:508
74. Fehlhammer WP, Blib T, Fuchs J, Holzmann G (1992) Zeitschrift
Fur Naturforschung. 47B:79–89
75. Glorius F (2007) Top Organomet Chem 21:1–20
´
´
76. Dıez-Gonzalez S, Nolan SP (2007) Top Organomet Chem 21:47–
50. Delley B (2000) J Chem Phys 113:7756
51. Perdew JP, Wang Y (1992) Phys Rev B 45:13244–13249
52. Chen J, Chen L, Xu L, Zheng K, Ji L (2008) J Phys Chem B
112:9966–9974
82
77. Peris E (2007) Top Organomet Chem 21:83–116
78. Tekavec TN, Louie J (2007) Top Organomet Chem 21:159–192
79. Herrmann WA, Schwarz J, Gardiner MG, Spiegler M (1999)
J Organomet Chem 575:80–86
80. Loch JA, Albrecht M, Peris E, Mata J, Faller JW, Crabtree RH
(2002) Organometallics 21:700–706
´
´
53. Serrano-Andres L, Merchan M, Borin AC (2008) J Am Chem Soc
130:2473–2484
54. Ramaekers R, Dkhissi A, Adamowicz L, Maes G (2002) J Phys
Chem A 106:4502–4512
55. Major DT, Laxer A, Fischer B (2002) J Org Chem 67:790–802
56. Virta P, Koch A, Roslund MU, Mattjus P, Kleinpeter E, Kronberg
81. Huynh HV, Han Y, Ho JHH, Tan GK (2006) Organometallics
25:3267–3274
¨
L, Sjoholm R, Klika KD (2005) Org Biomol Chem 3:2924–2929
57. Gonella NC, Roberts JD (1982) J Am Chem Soc 104:3162–3164
82. Gottleib HE, Kotlyar V, Nudelman A (1997) J Org Chem
62:7512–7515
123