Syntheses, characterization, density functional theory calculations, and activity of tridentate SNS zinc pincer complexes

Article abstract of DOI:10.1016/j.ica.2011.07.021

A series of tridentate SNS ligand precursors were metallated with ZnCl ₂ to give new tridentate SNS pincer zinc complexes. The zinc complexes serve as models for the zinc active site in liver alcohol dehydrogenase (LADH) and were characterized with single crystal X-ray diffraction, ¹H, ¹³C, and HSQC NMR spectroscopies and electrospray mass spectrometry. The bond lengths and bond angles of the zinc complexes correlate well to those in horse LADH. The zinc complexes feature SNS donor atoms and pseudotetrahedral geometry about the zinc center, as is seen for liver alcohol dehydrogenase. The SNS ligand precursors were characterized with ¹H, ¹³C, and HSQC NMR spectroscopies and cyclic voltammetry, and were found to be redox active. Gaussian calculations were performed and agree quite well with the experimentally observed oxidation potential for the pincer ligand. The zinc complexes were screened for the reduction of electron poor aldehydes in the presence of a hydrogen donor, 1-benzyl-1,4-dihydronicotinamide (BNAH). The zinc complexes enhance the reduction of electron poor aldehydes. Density functional theory calculations were performed to better understand why the geometry about the zinc center is pseudo-tetrahedral rather than pseudo-square planar, which is seen for most pincer complexes. For the SNS tridentate pincer complexes, the data indicate that the pseudo-tetrahedral geometry was 43.8 kcal/mol more stable than the pseudo-square planar geometry. Density functional theory calculations were also performed on zinc complexes with monodentate ligands and the data indicate that the pseudo-tetrahedral geometry was 30.6 kcal/mol more stable than pseudo-square planar geometry. Overall, the relative stabilities of the pseudo-tetrahedral and pseudo-square planar systems are the same for this coordination environment whether the ligand set is a single tridentate SNS system or is broken into three separate units. The preference of a d¹⁰ Zn center to attain a tetrahedral local environment trumps any stabilization gained by removal of constraints within the ligand set.

Full text of DOI:10.1016/j.ica.2011.07.021

Inorganica Chimica Acta 376 (2011) 515–524
Contents lists available at SciVerse ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Syntheses, characterization, density functional theory calculations,
and activity of tridentate SNS zinc pincer complexes
a,
⇑
b
c
a
a
John R. Miecznikowski , Wayne Lo , Matthew A. Lynn , Brianne E. O’Loughlin , Amanda P. DiMarzio ,
a
a
a
a
a
Anthony M. Martinez , Lorraine Lampe , Kathleen M. Foley , Lauren C. Keilich , George P. Lisi ,
Daniel J. Kwiecien , Cristina M. Pires , William J. Kelly , Nathan F. Kloczko , Kaitlyn N. Morio
Department of Chemistry and Biochemistry, Fairfield University, 1073 North Benson Road, Fairfield, CT 06824, USA
Department of Chemistry, Boston College, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, USA
a
a
a
a
a
a
b
c
Department of Science and Mathematics, National Technical Institute for the Deaf, Rochester Institute of Technology, 52 Lomb Memorial Drive, Rochester, NY 14623, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of tridentate SNS ligand precursors were metallated with ZnCl to give new tridentate SNS pincer
2
Received 7 January 2011
Received in revised form 11 July 2011
Accepted 12 July 2011
zinc complexes. The zinc complexes serve as models for the zinc active site in liver alcohol dehydroge-
1
13
nase (LADH) and were characterized with single crystal X-ray diffraction, H, C, and HSQC NMR spec-
troscopies and electrospray mass spectrometry. The bond lengths and bond angles of the zinc complexes
correlate well to those in horse LADH. The zinc complexes feature SNS donor atoms and pseudotetrahe-
dral geometry about the zinc center, as is seen for liver alcohol dehydrogenase. The SNS ligand precursors
Available online 23 July 2011
Keywords:
1
13
were characterized with H, C, and HSQC NMR spectroscopies and cyclic voltammetry, and were found
to be redox active. Gaussian calculations were performed and agree quite well with the experimentally
observed oxidation potential for the pincer ligand. The zinc complexes were screened for the reduction of
electron poor aldehydes in the presence of a hydrogen donor, 1-benzyl-1,4-dihydronicotinamide (BNAH).
The zinc complexes enhance the reduction of electron poor aldehydes. Density functional theory calcu-
lations were performed to better understand why the geometry about the zinc center is pseudo-tetrahe-
dral rather than pseudo-square planar, which is seen for most pincer complexes. For the SNS tridentate
pincer complexes, the data indicate that the pseudo-tetrahedral geometry was 43.8 kcal/mol more stable
than the pseudo-square planar geometry. Density functional theory calculations were also performed on
zinc complexes with monodentate ligands and the data indicate that the pseudo-tetrahedral geometry
was 30.6 kcal/mol more stable than pseudo-square planar geometry. Overall, the relative stabilities of
the pseudo-tetrahedral and pseudo-square planar systems are the same for this coordination environ-
ment whether the ligand set is a single tridentate SNS system or is broken into three separate units.
SNS pincer ligand
Mononuclear Zn complexes
X-ray crystallography
Cyclic voltammetry
Density functional theory calculations
Aldehyde reductions
1
0
The preference of a d Zn center to attain a tetrahedral local environment trumps any stabilization
gained by removal of constraints within the ligand set.
Ó 2011 Elsevier B.V. All rights reserved.
1
. Introduction
Bioinorganic chemists have made contributions toward the
and also catalyzes the reverse reaction: the reduction of a ketone
or an aldehyde to an alcohol [1,3]. The crystal structure of horse
LADH has been solved [4]. The resting enzyme has a zinc(II) metal
center which is pseudo-tetrahedrally ligated with one N-histidine
side chain, two S-cysteine side chains, and one labile water mole-
cule. The nitrogen and sulfur donor atoms are provided by histi-
dine and cysteine residues of a single polypeptide chain [5].
Several structural models for LADH have been reported with the
same donor atoms as the metalloenzyme [6–13]. In some cases,
reactivity data was not reported [14]. Reactivity has also been re-
ported for zinc LADH model complexes with donor atoms that
are different than the enzyme’s active site [15].
understanding of enzymatic activity through the synthesis and sub-
sequent structural and functional characterization of model com-
plexes for metalloenzyme active sites [1]. Model complexes are
low molecular mass systems that seek to mimic the metalloenzyme
in terms of ligand donor atoms, structure, and oxidation states [2].
Bio-inspired model complexes can be prepared to investigate the
structure and function of enzymes through the use of Nature as a
model for the design of highly active and efficient catalysts.
Liver alcohol dehydrogenase (LADH) is a zinc metalloenzyme
that catalyzes the oxidation of alcohols to aldehydes and ketones,
In our efforts to understand the catalytic activity of such a
metalloenzyme, we have chosen to model the structure and
reactivity of the zinc active site through the use of a new family
of robust pincer ligands that provide the same S, N, S donor atoms
⇑
Corresponding author. Tel.: +1 203 254 4000x2125; fax: +1 203 254 4034.
E-mail address: jmiecznikowski@fairfield.edu (J.R. Miecznikowski).
0
020-1693/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2011.07.021

5
16
J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
as the enzyme. Tridentate pincer ligands, first published in 1976,
offer several advantages as compared to monodentate ligands.
First, their metallation is favored due to a less negative delta entro-
py of formation as compared to monodentate ligands, conferring
added stability to their metal complexes [16]. Second, tridentate
pincer ligands are known to inhibit dimerization, a process that
could slow or inhibit catalytic activity. Third, the structural and
electronic properties of the tridentate ligand can be systematically
changed by the use of slightly different materials in their synthetic
preparation. The tridentate pincer ligand usually coordinates to the
metal in a meridional fashion. Sometimes both of the metallocycles
pucker, leading to a pseudo-facial coordination mode.
Bruker Avance 300 MHz NMR spectrometer. Spectra were referred
to the solvent residual peak. Electrospray mass spectrometry was
performed on a Micromass ZQ instrument using nitrogen as the dry-
ing and nebulizing gas. Cyclic voltammetry experiments were per-
formed using a Cypress Electroanalytical System with a silver wire
reference electrode, a glassy carbon working electrode, and a plati-
num counter electrode. The supporting electrolyte for the cyclic vol-
tammetry
experiments
was
tetra-N-butylammonium
tetrafluoroborate. The ferrocenium/ferrocene couple was used as
an internal reference; reduction potential values were corrected by
assigning the ferrocenium/ferrocene couple to 0.40 V versus SCE.
When an inert atmosphere was needed, a M-Braun inert atmosphere
glove box and standard Schlenk techniques were used with thor-
oughly degassed solvents. IR spectra were collected using a Thermo
Nicolet AVATAR 380-FT-IR with a SMART SPECULATR reflectance
adaptor. C, H, N elemental analyses were performed by Atlantic
Microlab Inc. (Norcross, GA). Melting points are uncorrected.
Pincer ligands have been utilized successfully in organometallic
chemistry to prepare highly catalytically active and robust com-
plexes [17]. Interestingly, to the best of our knowledge, such ligand
systems have not been used frequently in bio-inspired modeling
chemistry. The pincer ligand is an excellent one to model biological
2
activity since the N-donor pyridine, like N-histidine, is sp -hybrid-
ized and thioimidazolyl sulfur donors have been reported by
Parkin and Vahrenkamp to model thiol derived ligands in bio-in-
spired zinc chemistry [14a,14b,18,15b,19].
2
.2. Crystallographic analyses
Crystals of 1, 2, and 3 were mounted on a glass fiber or loop and
Toward the goal of synthesizing and understanding bio-inspired
model complexes for the liver alcohol dehydrogenase active site,
we report herein the syntheses, spectroscopic and electrochemical
characterization, density functional theory (DFT) calculations, and
activity screening of tridentate SNS zinc pincer complexes.
placed in a ꢀ80 °C nitrogen stream on a Bruker diffractometer
equipped with a Smart CCD at Boston College (Chestnut Hill,
MA). Crystallographic data were collected using graphite mono-
chromated 0.71073 Å Mo K
a radiation and integrated and cor-
rected for absorption using the Bruker SAINTPLUS software
package. The structures were solved using direct methods and re-
fined using least-square methods on F-squared. All other pertinent
crystallographic details such as h, k, l ranges, 2h ranges, and
R-factors can be found in Table 1.
2
. Experimental
2.1. General procedures
All reagents used are commercially available and were used as re-
2.3. Reactivity
ceived. Isopropyl imidazole, neopentyl imidazole, 2,6-bis(3-buty
limidazol-1-yl)pyridine bromide, 2,6-bis(3-isopropylimidazol-1-
yl)pyridine bromide, and 2,6-bis(3-isopropylimidazol-2-thione-1-
yl)pyridine were prepared as reported previously [20–24] as were
In
0.2 mmol of BNAH, and 0.1 mmol of the zinc complex or 0.2 mmol
ZnCl were dissolved in 3 mL of CDCl . The reaction was heated at
reflux. Aliquots of the reaction were taken at certain times and
a typical reaction, 0.1 mmol of 4-nitrobenzaldehyde,
2
3
+
BNA and BNAH [25]. NMR spectra were recorded at 25 °C on a
Table 1
Crystal and structure refinement data for 1–3.
R = iPr (1)
R = Np (2)
30Cl
R = nBu (3)
4 5 2 2
25Cl N S Zn
Formula
C
17
H22Cl
4
N
5
OS
2
Zn
2
C
21
H
4
5
N OS
2
Zn
2
C
19
H
Formula weight (g/mol)
Temperature (K)
Wavelength (Å)
Crystal system
Space group
a (Å)
649.06
193(2)
0.71073
monoclinic
P2(1)/c
7.475(2)
27.456(8)
12.705(3)
90
98.163(6)
90
2580.9(12)
4
705.16
193(2)
0.71073
triclinic
P1ꢀ
7.6011(19)
13.297(3)
14.952(4)
96.534(5)
97.139(5)
102.721(5)
1447.1(6)
2
660.10
193(2)
0.71073
monoclinic
P2(1)/c
16.5135(6)
12.5308(5)
12.7522(5)
90
94.713(2)
90
2629.86(18)
4
1.667
2.407
1336
b (Å)
c (Å)
a
(°)
b (°)
(°)
c
3
Volume (Å)
Z
q
l
3
(g/cm )
1.670
2.454
1308
1.618
2.196
718
ꢀ
1
(mm
)
F(0 0 0)
Crystal size (mm )
3
0.10 ꢁ 0.02 ꢁ 0.02
0.10 ꢁ 0.05 ꢁ 0.02
0.12 ꢁ 0.02 ꢁ 0.02
h (°)
1.48–28.36
1.39–28.39
1.24–27.57
Refections/Unique
31 809/6362
0.1605
18 174/6979
0.0692
23 665/6059
0.0398
R
int
Absolute correction
none
none
none
Maximum/minimum
Reference method
0.9526/0.7914
Full matrix least squares on F2
6362/0/286
1.122
0.9574/0.8103
Full matrix least squares on F2
6979/0/323
1.092
0.9534/0.7610
Full matrix least squares on F2
6059/0/291
0.790
Data/restraints/parameters
Goodness-of-fit (GOF) on F2
R
1
indices (I > 2
r)
0.0884
0.0636
0.0333
wR
2
0.1577
0.821 and ꢀ0.959
0.1376
0.835 and ꢀ0.742
0.0806
0.577 and ꢀ0.382
Peak/hole (e/Åꢀ3)

J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
517
analyzed using 1H NMR spectroscopy. All data are averages of at
least two runs.
solution was stirred at reflux for 30 min after which not all of the
sodium acetate had dissolved. Then 0.66 g (21 mmol) sulfur was
added to the round-bottom flask with an additional 10 mL of ace-
tonitrile. The solution was stirred at reflux for 3 days. The product
was vacuum filtered to remove the remaining solids. The solvent
was removed under reduced pressure. The yield (0.48 g) was quan-
2.4. Gaussian calculations
GAUSSIAN 03 was used to perform single-point calculations and
DFT geometry optimizations using the B3LYP hybrid functional
with 6-31Gꢂ basis sets as provided with the software [26]. Calcula-
tions were performed on the pincer ligand precursor alone as well
as on the bound Zn(II) system (R = Me in all cases). Symmetry was
imposed on the metal-ligated complex such that one pseudo-tetra-
titative. Electrospray MS (MeOH, 20 V, positive ion mode (m/z): ex-
+
pected: 415.2 (100%) [C21
H
29
N
5
S
2
] , Found: 416.5 (100%)
+
[C21
H
29
N
5
2
S +1] . Anal. Calc. for
C
21
H
29
N
5 2
S ꢃH
2
O (415.62): C,
58.17; H, 7.21; N, 16.15. Found: C, 57.58; H, 6.72; N, 15.80.
1
H NMR (DMSO-d
6
, 300 MHz) d 8.66 ppm (d, 3J = 8.1 Hz, 2H,
hedral (C
s
) and two pseudo-square planar (C
2
and C2v) structures
pyridine CH), 8.19 ppm (t, 3J = 8.1 Hz, 1H, pyridine CH), 7.83 ppm
(d, 3J = 2.4 Hz, 2H, imidazole CH), 7.33 ppm (d, 3J = 2.7 Hz, 2H,
were examined. In accordance with the molecular point groups,
the SNS ligand set was required to be flat under C2v symmetry,
but could become nonplanar under C
13
imidazole CH), 4.00 ppm (s, 4H, CH
2
), 1.00 ppm (s, 18H, CH
3
),
C
1
2
or C
s
symmetries. Frequency
{ H} NMR (DMSO-d , 75 MHz) d 163.07 ppm (C@S), 148.43 ppm
6
analysis was performed on the optimized structures to determine
whether or not they represented true minima. We found no imag-
inary frequencies for the pseudo-tetrahedral structures, but we did
find several for the pseudo-square planar systems as indicated in
the discussion that follows.
(pyridine C), 139.77 ppm (pyridine CH), 119.99 ppm (imidazole
CH), 117.35 ppm (pyridine CH), 116.19 ppm (imidazole CH),
56.58 ppm (CH
2
), 33.75 (C(CH
3
)
3
3
), 27.86 ppm (CH ). Selected IR
ꢀ1
bands (reflectance):
v
max/cm 1125 (C@S).
Gas-phase geometry optimizations and single-point calcula-
tions using a solvent model were also performed on the unbound
pincer ligand precursor in an effort to model the oxidation poten-
tials that have been determined experimentally. After the struc-
tures of the ligand precursor were optimized in neutral and
singly cationic and anionic states under each of the point groups
described above, single-point SCRF calculations using DMSO via
the CPCM solvent model were performed. The ‘‘radii = uff’’ and
2.5.3. Preparation of 2,6-bis[3-(n-butyl)imidazol-2-thione-1-
yl]pyridine (C19 ) (3b)
A round-bottom flask was charged with 0.48 g (1.0 mmol) of
2,6-bis[3-(n-butyl)imidazol-1-yl]pyridine bromide, 0.24 g
(3.0 mmol) of sodium acetate, and 40 mL of MeCN. The mixture
was heated at reflux. After 30 min, 0.68 g (21 mmol) of sulfur
was added. The reaction was heated at reflux for 7 days. The prod-
uct was vacuum filtered to remove the remaining solids. The sol-
vent was removed under reduced pressure. The yield was 0.38 g
25 5 2
H N S
‘
‘nosymmcav’’ directives were employed. Oxidation and reduction
potentials were then determined by finding the difference in the
total free energies in solution for the neutral and cationic or anio-
(70%). Electrospray MS (MeOH, 20 V, positive ion mode (m/z): ex-
+
pected: 387.2 (100%) [C19
H
25
N
5
S
2
] . Found: 388.6 (100%)
+
nic species. These
D
G values were then referenced to the absolute
[C19
H
25
N
5
S
2
+1] . Anal. Calc. for C19
H
25
N S
5 2
2
ꢃ1.5H O (387.57): C,
SCE potential in DMSO by subtracting 3.83 V (the established cor-
rection [27] to SHE in DMSO) and 0.241 V (the difference between
SHE and SCE).
55.04; H, 6.81; N, 16.89. Found: C, 55.22; H, 6.26; N, 16.73.
1
6
H NMR (DMSO-d , 300 MHz) d 8.68 ppm (d, 3J = 7.8 Hz, 2H,
pyridine CH), 8.20 ppm (t, 3J = 8.1 Hz, 1H, pyridine CH), 7.82 ppm
d, 3J = 2.4 Hz, 2H, imidazole CH), 7.42 ppm (d, 3J = 2.7 Hz, 2H,
imidazole CH), 4.05 ppm (t, 3J = 7.6 Hz, 4H, CH ), 1.73 ppm (m,
), 0.93 ppm (t, 3J = 7.5 Hz, 6H,
, 75 MHz) d 161.26 ppm (C@S),
(
2
2
.5. Syntheses
2
4
2 2
H, CH ), 1.33 ppm (m, 4H, CH
1
3
1
.5.1. Preparation of 2,6-bis(3-neopentylimidazol-1-yl)pyridine
3 6
CH ). C { H} NMR (DMSO-d
bromide (C21
H
31
N
5
Br
2
) (2a)
148.27 ppm (pyridine C), 139.90 ppm (pyridine CH), 118.86 ppm
(imidazole CH), 116.99 ppm (pyridine CH), 116.47 ppm (imidazole
In a round-bottom flask, 3.57 g (25.8 mmol) of neopentyl imid-
azole were added along with 3.20 g (13.5 mmol) of 2,6-dibromo-
pyridine. The solution was brown and oily. The reaction mixture
was heated neat at 160 °C for 18 h. After the reaction mixture
was removed from the heat, the dark brown solid that remained
in the round-bottom flask was dissolved in methanol (15 mL)
and the product was precipitated out of solution with 125 mL of
diethyl ether. The tan crystals were collected by filtration through
a Buchner funnel and were allowed to air dry. The yield was 3.07 g
CH), 46.47 ppm (CH
Selected IR bands (reflectance):
2
), 29.96 (CH
2
), 19.19 (CH
2
), 13.54 ppm (CH
3
).
ꢀ
1
vmax/cm 1126 (C@S).
3
2.5.4. Preparation of chloro(
g
-S,S,N)-[2,6-bis(3-isopropylimidazol-2-
thione-1-yl)]pyridinezinc(II) aquatrichlorozincate
([C17 ZnCl][ZnCl (OH )]) (1)
H
21 5
N S
2
3
2
In a 100 mL round-bottom flask, 0.062 g (0.17 mmol) 2,6-bis(3-
isopropylimidazol-2-thione-1-yl)pyridine was combined with
(
44.3%). Melting point: 317–320 °C (dec). Anal. Calc. for
2 2 2
0.057 g (0.42 mmol) ZnCl and dissolved in 10 mL CH Cl . The solu-
C
21
H
31
N
5
Br
2
ꢃH
2
O (531.33): C, 47.47; H, 6.26; N, 13.18. Found: C,
tion was stirred at reflux for 18 h and the off-white solid was col-
lected by vacuum filtration. The yield for the reaction was 0.070 g
(63%). Single crystals for X-ray diffraction were grown by a slow va-
por diffusion of diethyl ether into a dichloromethane solution con-
4
7.38; H, 6.23; N, 13.08.
1
H NMR (DMSO-d
6
, 300 MHz) d 10.85 ppm (s, 2H, imidazole CH),
8
.93 ppm (s, 2H, imidazole CH), 8.62 ppm (m, 1H, pyridine CH),
8
.34 ppm (d, 3J = 8.1 Hz, 2H, pyridine CH), 8.10 ppm (s, 2H, imidaz-
taining the zinc complex. Electrospray MS (MeOH, 0 V, positive ion
13
1
+
ole CH), 4.25 ppm (s, 2H, CH
2
3
), 1.02 ppm (s, 18H, CH ). C { H} NMR
mode (m/z)): expected: 460.0 (99%) [C17
H
21
N
5
S
2
ZnCl] , Found:
+
(
DMSO-d
6
, 75 MHz) d 145.28 ppm (pyridine CH), 144.64 ppm (pyr-
459.9 (100%) [C17
Zn O (718.17): C, 30.10; H, 4.35; N, 9.75. Found: C,
ꢃMeOHꢃ2H
30.00; H, 3.91; N, 9.48.
21 5 2 4 5 2
H N S ZnCl] . Anal. Calc. for C17H23Cl N OS
idine C), 136.23 ppm (imidazole CH), 125.14 ppm (imidazole CH),
1
5
2
2
19.11 ppm (imidazole CH), 114.29 ppm (pyridine CH),
9.99 ppm (CH ), 32.08 (C(CH ), 26.69 ppm (CH ).
1
2
3
)
3
3
6
H NMR (DMSO-d , 300 MHz) d 8.64 ppm (d, 3J = 8.1 Hz, 2H, pyr-
idine CH), 8.19 ppm (t, 3J = 8.1 Hz, 1H, pyridine CH), 7.84 ppm (d,
3J = 2.7 Hz, 2H, imidazole CH), 7.52 ppm (d, 3J = 2.7 Hz, 2H, imidaz-
ole CH), 5.02 ppm (m, 2H, isopropyl CH), 1.35 ppm (d, 3J = 6.6 Hz,
2
.5.2. Preparation of 2,6-bis(3-neopentylimidazol-2-thione-1-
yl)pyridine (C21 ) (2b)
100 mL round-bottom flask was charged with 0.52 g
1.0 mmol) 2,6-bis(3-neopentylimidazol-1-yl)pyridine bromide,
.20 g (2.4 mmol) sodium acetate, and 40 mL of acetonitrile. The
29 5 2
H N S
1
3
1
A
12H, isopropyl CH
3
).
6
C { H} NMR (DMSO-d , 75 MHz) d
(
0
160.53 ppm (C@S), 148.16 ppm (pyridine C), 140.00 ppm (pyridine
CH), 117.48 ppm (imidazole CH), 117.17 ppm (pyridine CH),

5
18
J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
1
(
15.25 ppm (imidazole CH), 48.12 ppm (isopropyl CH), 21.00 ppm
isopropyl CH ).
H NMR (MeOD-d
imidazole CH), 4.16 ppm (m, 4H, CH
2
), 1.83 ppm (m, 4H, CH
2
),
3
1.42 ppm (m, 4H, CH ), 1.03 ppm (m, 6H, CH
2
3
).
1
3
, 300 MHz) d 8.72 ppm (d, 3J = 8.1 Hz, 2H,
8
=
3
.15 ppm (m, 1H), 7.70 ppm (d, 3J = 2.7 Hz, 2H), 7.33 ppm (d, 3J
2.7 Hz, 2H), 5.21 ppm (m, 2H, isopropyl CH), 1.19 ppm (d,
J = 7.2 Hz, 12H, isopropyl CH
3
).
3. Results and discussion
.1. Syntheses and spectroscopy
The syntheses of the tridentate SNS ligand precursors and zinc
pincer complexes 1–3 were accomplished following Scheme 1.
The alkyl imidazoles were either prepared following known routes
or were commercially available [20,22]. Different R groups were
employed because Crabtree has reported that modification of such
substituents affects the solubility and catalytic activity of the metal
complexes [23]. The alkyl imidazoles react neatly with 2,6-dib-
romopyridine to form the ligand precursor salts 1a, 2a, or 3a
3
3
2.5.5. Preparation of chloro(
g
-S,S,N)-[2,6-bis(3-neopentylimidazol-2-
thione-1-yl)]pyridinezinc(II) aquatrichlorozincate
ZnCl][ZnCl (OH )]) (2)
In a round-bottom flask, 0.14 g (0.35 mmol) of 2,6-bis(3-neo-
pentylimidazol-2-thione-1-yl)pyridine was added to 0.10 g
0.73 mmol) of ZnCl . The reactants were dissolved in 8 mL of
dichloromethane. The yellow solution was heated at reflux for
8 h. An off-white precipitate was present at the end of the reac-
(
[C21
H
27
N
S
5 2
3
2
(
2
1
tion. The solvent was removed under reduced pressure. Single
crystals for X-ray diffraction were grown by a slow vapor diffusion
of pentane into a dichloromethane solution containing the zinc
complex. The yield for the synthesis was 0.24 g (quantitative).
[
28], which are soluble in DMSO, methanol, acetonitrile, and water.
Compounds 1a, 2a, and 3a, react with a mild base, sodium ace-
tate, and elemental sulfur in refluxing acetonitrile to form the bis-
thione ligand precursors 1b, 2b, and 3b, respectively [24]. Instead
of sodium acetate, a stronger base, potassium t-butoxide, could
also be utilized [29]. As determined by NMR spectroscopy, 1b,
Electrospray MS (MeOH, 0 V, positive ion mode (m/z): expected:
+
5
14.0
(100%)
[C21
H
27
N
S
5 2
ZnCl] .
Found:
OS
706.21): C, 39.48; H, 5.40; N, 10.23. Found: C, 39.21; H, 5.21; N,
0.75.
1H NMR (DMSO-d
pyridine CH), 8.19 ppm (t, 3J = 8.1 Hz, 1H, pyridine CH), 7.83 ppm
514.3
(100%)
+
[
C
21
H
27
N
S
5 2
ZnCl] . Anal. Calc. for C21
H31Cl
4
N
5
2
Zn
2
ꢃCH
3
2
CNꢃEt O
(
1
2
b, or 3b can be purified by filtering a dichloromethane solution
containing this compound through alumina. These compounds
are soluble in DMSO, dichloromethane, chloroform, acetone, aceto-
nitrile, and methanol.
6
, 300 MHz) d 8.66 ppm (d, 3J = 8.1 Hz, 2H,
(
d, 3J = 2.4 Hz, 2H, imidazole CH), 7.33 ppm (d, 3J= 2.1 Hz, 2H, pyr-
The bis-thione ligand precursors subsequently react with ZnCl
in refluxing CH Cl to afford compounds 1–3. The driving force for
the metallation is the formation of zinc complexes 1–3, which are
sparingly soluble in CH Cl . Off-white crystals that were suitable
2
13
1
2 3
idine CH), 4.00 ppm (s, 4H, CH ), 1.00 ppm (s, 18H, CH ). C { H}
NMR (DMSO-d
dine C), 139.32 ppm (pyridine CH), 120.04 ppm (imidazole CH),
2
2
6
, 75 MHz) d 163.08 ppm (C@S), 148.46 ppm (pyri-
2
2
1
17.42 ppm (pyridine CH), 116.24 ppm (imidazole CH),
6.61 ppm (CH ), 33.78 (C(CH ), 27.89 ppm (CH ).
H NMR (MeOD-d , 300 MHz) d 8.64 ppm (d ( J = 8.1 Hz), 2H,
pyridine CH), 8.14 ppm (m, 1H, pyridine CH), 7.69 ppm (d
for X-ray diffraction were grown by allowing diethyl ether (1) or
pentane (2, 3) to slowly diffuse into a dichloromethane solution
containing 1, 2, or 3. All of the reactions could be carried out in
air, and the reactions proceeded in 63% to quantitative yield. The
zinc complexes 1–3 are soluble in dimethyl sulfoxide, acetonitrile,
methanol, and water and are sparingly soluble in dichloromethane
and chloroform.
5
2
3
)
3
3
1
3
3
3
(
J = 2.4 Hz), 2H, imidazole CH), 7.23 ppm (d, 3J = 2.1 Hz, 2H, imid-
2 3
azole CH), 4.08 ppm (s, 4H, CH ), 1.05 ppm (s, 18H, CH ).
3
2.5.6. Preparation of chloro(
g
-S,S,N)-{2,6-bis[3-(n-butyl)imidazol-2-
Ligand precursors 1a, 1b, 2a, 2b, 3a, and 3b were characterized
1
13
thione-1-yl]}pyridinezinc(II) aquatrichlorozincate
ZnCl][ZnCl ]) (3)
In a round-bottom flask, 0.15 g (0.38 mmol) of 2,6-bis[3-(n-
butyl)imidazol-2-thione-1-yl]pyridine were combined with ZnCl
.099 g (0.73 mmol), and dissolved in 10 mL of CH Cl . The solu-
using H, C, and HSQC NMR spectroscopy. For these ligand pre-
cursors and zinc complexes 1–3, only one set of resonances was
detected indicating that the two halves of each molecule are sym-
(
[C19
H
25
N
S
5 2
3
1
3
2
,
metry-related. The C NMR spectra of 1b, 2b, and 3b, show reso-
nances at ꢄ161 ppm that are consistent with C@S formation [17].
The bis-thione ligand precursors, 1b, 2b, and 3b, lacked the acidic
0
2
2
tion was refluxed for 18 h. An off-white precipitate was present
at the end of the reaction. The solvent was removed under reduced
pressure. Single crystals for X-ray diffraction were grown by a slow
vapor diffusion of pentane into a dichloromethane solution con-
taining the zinc complex. The yield was 0.25 g (quantitative). Elec-
trospray MS (MeOH, 0 V, positive ion mode (m/z): expected: 486.0
1
C–H proton at d ꢄ11 ppm in the H NMR spectra, whereas this pro-
ton resonance was present for the precursors 1a, 2a, and 3a.
1
13
Complexes 1–3 were analyzed with H, C, and HSQC NMR
1
13
spectroscopy in DMSO-d6. The H and C NMR spectra of the zinc
complexes 1–3 are largely the same with very little shift in the res-
onances when compared to the respective ligand precursors 1b, 2b,
and 3b. This was expected since no hydrogen or carbon atoms are
displaced or added upon metallation with zinc(II) chloride. In addi-
+
+
(
100%) [C19
H
25
N
5
S
2
ZnCl] , Found: 486.4 (100%) [C19
25 5 2
H N S ZnCl] .
Anal. Calc. for C19
H
27Cl OS Zn CN (678.15): C, 34.38; H,
4
N
5
2
2
ꢃ0.5CH
3
4
.11; N, 11.03. Found: C, 34.33; H, 4.17; N, 10.54.
1
1
H NMR (DMSO-d
pyridine CH), 8.20 ppm (t, 3J = 8.1 Hz, 1H, pyridine CH), 7.82 ppm
d, 3J = 2.1 Hz, 2H, imidazole CH), 7.42 ppm (d, 3J = 2.1 Hz, 2H,
imidazole CH), 4.06 ppm (t, 3J = 7.2 Hz, 4H, CH ), 1.73 ppm (m,
), 0.93 ppm (t, 3J = 7.2 Hz, 6H,
, 75 MHz) d 161.26 ppm (C@S),
6
, 300 MHz) d 8.68 ppm (d, 3J = 8.1 Hz, 2H,
tion, the H NMR spectra of complexes 1–3 were acquired in a less
polar and weakly coordinating solvent, MeOD-d3, to verify that the
SNS pincer ligands cannot be displaced from the coordination
sphere of zinc. Electrospray mass spectrometry data, described be-
(
2
4
H, CH
2
), 1.34 ppm (m, 4H, CH
2
low, also verifies this result. Zinc complexes 1–3 are sparingly sol-
1
3
1
1
CH
3
). C { H} NMR (DMSO-d
6
uble in MeOD-d
3
. When compared to the H NMR spectra of
1
(
48.27 ppm (pyridine C), 139.93 ppm (pyridine CH), 118.88 ppm
imidazole CH), 117.04 ppm (pyridine CH), 116.50 ppm (imidazole
CH), 46.48 ppm (CH ), 29.97 (CH ), 19.19 (CH ), 13.55 ppm (CH ).
H NMR (MeOD-d , 300 MHz) d 8.64 ppm (d ( J = 8.1 Hz), 2H,
pyridine CH), 8.17 ppm (m, 1H, pyridine CH), 7.70 ppm (d
complexes 1–3 that were acquired in DMSO-d
6
, there was a slight
shift (d ꢄ0.05–0.10) ppm of all of the resonances in the proton
1
2
2
2
3
NMR spectra that were acquired in MeOD-d
tra that were acquired in MeOD-d , when compared to the H NMR
spectra acquired in DMSO-d , some of the resonances shifted
upfield while others shifted downfield.
3
. In the H NMR spec-
1
3
1
3
3
6
3
(
J = 2.7 Hz), 2H, imidazole CH), 7.26 ppm (d, 3J = 2.7 Hz, 2H,

J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
519
Scheme 1. Preparation of 1–3.
ESI-MS spectra for compounds 1–3 were collected with cone
Under the various point groups considered, the computed
ues are 1297 mV (C ), 1304 mV (C
three of these potentials are close to that found experimentally
(1289 mV), the G value found when the ligand is modeled under
symmetry is the closest. As can be seen from the contour plot of
the HOMO shown in Fig. 3, the electron is removed from a -type
orbital that is largely based on the thioimidazolyl portions of the
molecule. Upon oxidation, the main structural change observed
in the optimized structures is an increase in the dihedral angle be-
tween the thioimidazolyl and pyridinyl sections of the molecule.
D
G val-
voltages of 0 and 70 V. The predominant feature in the spectra of
these systems at the higher cone voltage is that of the fully ligated
zinc complex, indicating that the compound is stable and suggest-
ing that it is unlikely that the ligand is displaced when dissolved in
s
2
), and 1420 mV (C2v). While all
D
C
s
a polar solvent. In negative ion mode, the expected m/z values were
p
ꢀ
seen for [ZnCl
3
] . The isotopic patterns in the mass spectrometry
data fit the assigned structures.
3
.2. X-ray crystallography
s
Under C symmetry, the dihedral angle is determined to be 21.3°
for the neutral system and 38.5° for the cation, which indicates
that the sulfur atoms are closer to each other in the oxidized com-
pound than in the neutral system. Given that the sulfurs of the
The solid-state molecular structures of 1, 2, and 3 are shown in
Fig. 1. Complexes 1–3 feature SNS donor atoms and pseudotetrahe-
dral geometry about the zinc center, as is seen for liver alcohol
dehydrogenase. The bond lengths and bond angles for 1–3 are
listed and are compared to the active site in horse LADH–CNAD
in Table 2. CNAD is an isoteric C-glycosidic analog of NADH con-
taining a neutral pyridine ring, which is a potential inhibitor of
LADH. During catalysis, it is proposed that cofactor and substrate
are near the zinc(II) active site [32]. In the LADH–CNAD structure,
the CNAD mimics the cofactor binding to LADH. Therefore, we se-
lected this crystal structure because this structure contained a co-
factor mimic and an ethanol molecule in the active site. The bond
lengths and bond angles for 1–3 compare reasonably well to
LADH–CNAD. The carbon–sulfur bond lengths of ꢄ1.7 Å are be-
tween what is normally associated with a C–S single bond,
thioimidazole rings are closest to each other under C
s
symmetry
than under the C or C2v point groups and that similar systems
2
are known to form disulfides upon oxidation [33], the excellent
agreement between the experimental oxidation potential
(
1289 mV) and the value calculated by Gaussian (1297 mV) indi-
cates an acceptable level of modeling for this system.
As for the one-electron reduction of the pincer ligand, the com-
puted
D
G values (ꢀ5734 mV (C
s
), ꢀ5720 mV (C ), and ꢀ5748 mV
2
(
(
C2v)) do not match the potential observed experimentally
ꢀ2367 mV), suggesting that the reduction wave does not corre-
spond with a simple one-electron reduction of the ligand precur-
sor. The contour plot of the LUMO, shown in Fig. 3, indicates that
the electron is added into a p-type orbital that is distributed across
all portions of the molecule. Upon reduction, the ligand is flat with
a calculated thioimidazolyl–pyridinyl dihedral angle of 0.0°, which
1
.83 Å, and a C@S double bond, 1.61 Å [30]. For 1 and 2, the coun-
ꢀ
ꢀ
ter-anion is [ZnCl
(H
3 2
O)] and for 3, the counter ion is [ZnCl
3
] .
The counter-ion is seen even when either one or two molar equiv-
alents of ZnCl are used in the preparation of 1–3.
can be understood through consideration of the N–C
p bonding
2
nature between the thioimidazolyl and pyridinyl fragments of
the molecule.
3.3. Cyclic voltammetry of ligand precursors and comparison
of cyclic voltammetry results to Gaussian calculations
3.4. Density functional theory calculations to understand
The bis-thione ligand precursor, 2b, was studied by cyclic vol-
tammetry, as part of its characterization. This compound shows a
reversible oxidation wave at 1289 mV and a reduction wave at
geometry about the zinc center
Because tridentate pincer ligand precursors are known to coor-
dinate to metals in either a meridional or facial fashion, we were
interested in understanding the energy difference between a pseu-
do-tetrahedral complex (facial coordination of pincer ligand) and a
pseudo-square planar complex (meridional coordination of pincer
ligand) of this zinc pincer system. Pincer ligands have been re-
ꢀ
2367 mV by cyclic voltammetry in DMSO (Fig. 2). The oxidation
and reduction occurred at the same potential when two consecu-
tive cycles were collected. Thus, the ligand precursor itself is stable
with respect to repeated oxidations and reductions. We performed
Gaussian calculations to further understand our cyclic voltamme-
try results.
8
ported to coordinate in a meridional fashion to d Pd(II) and Pt(II)
Our Gaussian calculations agree quite well with the experimen-
tally observed oxidation potential for the pincer ligand precursor.
centers to yield systems that possess a pseudo-square planar envi-
ronment at the metal center [17]. Given the ability of the ligand

5
20
J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
Our calculations show that a pseudo-tetrahedral geometry is
1
0
preferred for a d zinc system with a ligand precursor set that is
analogous to that found in complexes 1–3 (Fig. 1 and Table 1).
The system with the Zn center in a pseudo-square planar environ-
ment and C
2
molecular geometry overall is calculated to be
1
8.3 kcal/mol less stable than the pseudo-tetrahedral structure as
depicted in Fig. 4. The pseudo-square planar system with C2v sym-
metry is 43.8 kcal/mol less stable than the pseudo-tetrahedral
structure. Analysis of the C
2
system finds that it has one vibration
with an imaginary frequency (28.03i), the C2v system possesses
ꢀ1
s
four (44.16i, 62.88i, 123.07i, and 138.96i cm ), and the C system
has none in agreement with the observation of pseudo-tetrahedral
systems for compounds 1–3. Visualization with GaussView of sev-
eral of the imaginary frequencies for the C2v system shows that
they represent torsion of the ligand set away from being planar
and toward the geometry adopted in the pseudo-tetrahedral
structure.
(
1)
Various bond lengths and angles found in the three optimized
structures are worth note. The Zn–S, Zn–Cl, and Zn–N bond lengths
of the pseudo-tetrahedral structure are in good agreement with
those given for 1–3 in Table 1. Constraining the pincer precursor
ligand to be planar to enforce a pseudo-square planar environment
around the Zn center under C2v symmetry requires the Zn–N bond
length to elongate to 3.32 Å, which is nearly 1.2 Å longer than was
calculated for the pseudo-tetrahedral structure. Furthermore, the
bond angles at the Zn center (N–Zn–S 76.9°) show that the donor
atoms are reasonably well aligned with the filled Zn dx
2ꢀy2
orbital,
thereby providing instability to this structure. The pseudo-square
planar system with C symmetry has a Zn–N bond of a similar
length (3.27 Å), but given that the ligand can flex more freely than
2
(
2)
it can under C2v symmetry, the optimized N–Zn–S bond angle of
6
5.1° allows relaxation of the
r
-donor atoms away from the Zn
dx
2ꢀy2
orbital and therefore a more stable structure than found
for the C2v system.
We next chose to calculate the energy differences between the
pseudo-tetrahedral and pseudo-square planar systems with the li-
gand set split into three monodentate fragments. Doing so allows
1
0
us to separate the influence of the d electronic population of
the metal center from the coordination environment imposed by
a unified tridentate SNS ligand precursor set. Shown in Fig. 5 are
the optimized pseudo-tetrahedral and pseudo-square planar struc-
tures in which the pyridine and imidazole fragments have been al-
lowed to attain the most stable orientations possible within the
constraints of the imposed molecular symmetries. In these struc-
tures, the pyridine rings are not connected to the imidazoles. No
imaginary frequencies were found for the pseudo-tetrahedral C
s
ꢀ1
structure while five (18.32i, 21.91i, 24.28i, 28.34i, 45.73i cm
)
were found for the C2v pseudo-square planar system and four
ꢀ
1
(
7.77i, 9.81i, 19.52i, 52.57i cm ) were located for the C
2
pseudo-
square planar system.
Comparison of the systems that contain monodentate ligands
with those that have a tridentate ligand reveals interesting struc-
tural information. The calculated Zn–Cl, Zn–S, and Zn–N bond
lengths for the pseudo-tetrahedral systems are quite similar under
both ligand environments. However, the C2v pseudo-square planar
system has comparatively short Zn–N and long Zn–S bonds in the
tridentate ligand environment while the structural constraints of
the single tridentate ligand impose a long Zn–N bond and relatively
short Zn–S bonds on the molecule. The geometric requirements of
a single, flat SNS ligand precursor therefore impose an additional
13.2 kcal/mol penalty (43.8–30.6 kcal/mol) upon the C2v pseudo-
square planar structure relative to the Cs pseudo-tetrahedral
system.
(
3)
Fig. 1. Solid-state structures of complexes 1–3.
precursor systems 1b, 2b, and 3b to establish p-conjugation across
multiple rings when flat, we were interested in understanding
whether such electronic influences would enforce a square planar
environment upon the metal center or if the preference of d
1
0
On the whole, the relative stabilities of the pseudo-tetrahedral
and pseudo-square planar systems are the same for this coordina-
tion environment about zinc whether the ligand set is a single
Zn(II) to attain
predominate.
a tetrahedral coordination sphere would

J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
521
Table 2
Selected bond lengths and angles (esd) for 1–3 and comparison to LADH–CNAD [31].
R = iPr (1)
R = neopentyl (2)
R = nBu (3)
LADH–CNADa
Zn(1)–N(1) (Å)
Zn(1)–Cl(1) (Å)
Zn(1)–S(1) (Å)
Zn(1)–S(2) (Å)
S(1)–C(6) (Å)
2.091(6)
2.210(2)
2.303(2)
2.307(2)
1.710(8)
2.101(5)
2.2301(15)
2.2937(16)
2.3146(16)
1.706(5)
2.145(2)
2.1930(9)
2.2822(9)
2.3431(9)
1.710(3)
2.0
2.2
2.3
2.3
a
N(1)–Zn(1)–Cl(1)
N(1)–Zn(1)–S(2)
Cl(1)–Zn(1)–S(2)
N(1)–Zn(1)–S(1)
Cl(1)–Zn(1)–S(1)
S(2)–Zn(1)–S(1)
111.96(17)
96.66(17)
117.01(8)
95.53(17)
120.44(9)
110.36(8)
111.80(13)
97.83(13)
118.76(13)
94.98(13)
117.41(6)
111.46(6)
113.92(7)
94.07(7)
108.92(4)
96.01(7)
123.85(4)
115.26(3)
136
106
a
97
97
a
98
125
a
Instead of Cl atom in LADH–CNAD, this entry refers to N5
N
atom of CNAD pyridine ring.
1
0
5
0
5
-3000
-2500
-2000
-1500
-1000
-500
0
500
1000
1500
2000
-
-
-
-
10
15
20
Potential (mV)
Fig. 2. Cyclic voltammogram of 2b in DMSO (2 mM) with 0.2 M TBAF. The scan rate was 100 mV/s with ferrocene (E1/2 = 400 mV) used as an internal standard.
tridentate SNS system or is broken into three separate units. The
1
0
preference of a d Zn center to attain a tetrahedral local environ-
ment trumps any stabilization gained by removal of constraints
within the ligand set.
3.5. Reactivity
Having established a synthetic protocol for these metalloen-
zyme models and gained an understanding of their structural char-
acteristics, we turned our attention to probing their stoichometric
activity. We chose 4-nitrobenzaldehyde, an electron-poor alde-
hyde, to screen the ability of 1–3 to reduce such a system in the
presence of the hydrogen donor BNAH (Eq. (1)). BNAH was pre-
pared following a known literature procedure and is the reagent
1
of choice to model NADH [25]. H NMR spectroscopy was used to
follow the disappearance of the aldehyde proton (d 10.2 ppm)
and the shifting of the aromatic C–H protons in the alcohol product
(
d 8.24 ppm). The aromatic proton resonance of the alcohol prod-
1
uct did not overlap with the H NMR resonances of either the start-
ing materials or the products or the zinc complex. In all of the
reactivity experiments, 0.1 mmol of aldehyde, 0.1 mmol zinc pre-
cursor or 0.2 mmol of ZnCl
ZnCl was added, the ZnCl
reaction mixture. In the control reactions, two equivalents of ZnCl
2
, and 0.2 mmol BNAH were used. When
2
2
was the only zinc compound in the
2
were used because in complexes 1–3, there are two zinc ions (one
in the cation and one in the anion) ion per neutral compound. The
1
products were detected by H NMR by comparison with authentic
material. In no case was there any indication of reduction of nitro
substituents. Table 3 illustrates the reactivity data for 1–3 as well
Fig. 3. Orbital contour plots for the LUMO (top) and HOMO (bottom) of the methyl-
substituted analog of 2b.

5
22
J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
ð1Þ
Fig. 4. Optimized pseudo-tetrahedral and pseudo-square planar structures from DFT calculations. Select optimized bond distances are provided.
Fig. 5. Optimized pseudo-tetrahedral and pseudo-square planar structures from DFT calculations with the pyridine ring and imidazole moieties in optimal configurations.
Select optimized bond distances are provided.

J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
523
as for ZnCl
2
and the ligand precursor.
tyl wingtip group, when compared to neopentyl or isopropyl, may
shield the zinc metal center from interacting with BNAH and sub-
strate. This is another example that illustrates that the identity of
the wingtip group is a very important variable to consider when
screening metal complexes for activity [23].
As shown in Table 3, zinc complexes 1–3 enhance the rate of
the reaction for the reduction of 4-nitrobenzaldehyde when com-
pared to ZnCl or ligand precursor. Based on a mechanism for LADH
2
proposed by Berreau and co-workers, there is a hydrogen transfer
between the co-factor and the substrate, which is coordinated to
We also tried to reduce another electron poor aldehyde, pyri-
dine-2-carboxaldehyde, in the presence of a stoichiometric amount
the zinc active site [32]. We also learned that ZnCl
2
reacts stoichio-
2
of 1 or ZnCl (Eq. (2))
ð2Þ
metrically with electron poor aldehydes such as 4-nitrobenzald-
hyde to yield alcohol product to small extent (ca. 18%
conversion). We suppose that Zn acts as a Lewis acid catalyst in
1H NMR was used to follow the disappearance of the aldehyde
proton (d 10.2 ppm) and the shifting of the aromatic C–H protons
in the alcohol product (d 8.4 ppm). The aromatic proton resonance
of the alcohol product did not overlap with the H NMR resonances
of either the starting materials or the products or the zinc complex.
Table 4 illustrates the reactivity data for 1 as well as for ZnCl .
2
As shown in Table 4, zinc complex 1 enhances the rate of the
reaction for the reduction of pyridine-2-carboxaldehyde when
a
2+
is utilized. Perhaps, the Zn2 in the count-
+
1
the reaction where ZnCl
eranion contributes to the reactivity that is shown for 1–3 in Table
. Experiments are underway to prepare tridentate zinc SNS pincer
complexes with a counter-anion that does not contain a zinc ion.
We also wondered if an excess of ZnCl would enhance the rate
of conversion of 4-nitrobenzaldehyde. We saw 42% conversion of
-nitrobenzaldehyde after 20 h when excess ZnCl (10 equiv.)
and BNAH could be utilized to
2
3
2
2
compared to ZnCl .
4
2
As seen in Tables 3 and 4, enhancement for the reduction of 4-
nitrobenzaldehyde or pyridine-2-carboxaldehyde was observed for
complexes 1–3. Stoichiometric conversion to the alcohol product
was not observed for either 4-nitrobenzaldehyde or pyridine-2-
carboxaldehyde. Based on the mechanism proposed by Berreau
and co-workers [32], the low activity of complexes 1–3 could be
due to the slow hydrogen transfer between the co-factor and the
substrate, which is coordinated to the zinc active site. More impor-
tantly, the alcohol product may inhibit the reaction. As the alcohol
is formed, it may coordinate to the zinc metal center as the reac-
tion progresses, and thereby hinder the reaction.
was used. Thus, an excess of ZnCl
reduce 4-nitrobenzaldehyde.
2
Based on the data presented in Table 3, it appears that the
choice of the R group is important as isopropyl and neopentyl
groups gave a higher percent conversion than n-butyl. Larger R
groups are needed to prepare zinc complexes that have increased
solubility in organic solvents. The more sterically demanding n-bu-
Table 3
Reactivity data for 1–3.
Entry
Zn Complex
Time (h)
Conversion (%)
1
2
3
4
5
6
7
8
9
None
20
20
5
20
20
20
20
20
20
<5
<5
15
13
18
42
48
33
42
4. Conclusions
2b (bis-thione ligand precursor)
ZnCl
ZnCl
ZnCl
1
2
2
2
(2 equiv.)
(2 equiv.)
(2 equiv.)
a
A series of Zn(II) compounds containing the SNS facial coordina-
tion of a tridentate pincer ligand was prepared and characterized.
The tridentate SNS pincer ligand precursors and zinc complexes
used in this work have provided new insights in the field of bioin-
organic modeling chemistry. The zinc complexes serve as models
for the zinc active site in liver alcohol dehydrogenase. The SNS zinc
pincer complexes adopt pseudo-tetrahedral geometry and have a
SNS coordination environment about the zinc center like LADH
and react with BNAH to reduce electron poor aldehydes. The reac-
tivity of 1–3 is not optimal when compared to LADH. It remains a
challenge to synthesize a neutral zinc complex with a tridentate li-
gand with SNS donor atoms that yield reactivity that is comparable
to LADH.
2
3
ZnCl
2
(10 equiv.)
a
2
Reaction was run under an N environment. The reaction was performed in an
inert atmosphere glove box.
Table 4
Reactivity data for 1 for the reduction of pyridine-2-carboxaldehyde.
Entry
Zn Complex
Time (h)
Conversion (%)
Gaussian calculations were performed to prove various struc-
tural and electronic properties of these compounds. The computed
oxidation potentials matches well with what is observed experi-
mentally while the calculated reduction potential indicates that
1
2
3
None
1
20
20
20
<5
35
19
ZnCl
2

5
24
J.R. Miecznikowski et al. / Inorganica Chimica Acta 376 (2011) 515–524
the experimental reduction wave does not correspond to a simple
one-electron reduction of the ligand precursor without other reac-
tivity occurring. DFT calculations were also performed to better
understand why the geometry about the zinc center is pseudo-tet-
rahedral rather than pseudo-square planar, which is seen for most
pincer complexes. Overall, the relative stabilities of the pseudo-tet-
rahedral and pseudo-square planar systems are the same for this
coordination environment whether the ligand set is a single triden-
tate SNS system or is broken into three separate units. The prefer-
[13] D.T. Corwin, R. Fikar, S.A. Koch, Inorg. Chem. 26 (1987) 3079.
14] (a) C. Kimblin, T. Hascall, G. Parkin, Inorg. Chem. 36 (1997) 5680;
[
(
(
b) C. Kimblin, B.M. Bridgewater, D.G. Churchill, G. Parkin, Chem. Commun.
1999) 2301.
[15] (a) R. Walz, H. Vahrenkamp, Inorg. Chim. Acta 314 (2001) 58;
b) Y.H. Zhang, H. Vahrenkamp, Inorg. Chim. Acta 351 (2003) 201.
16] (a) M. Albrecht, G. van Koten, Angew. Chem., Int. Ed. 40 (2001) 3750;
b) A.T. Normand, K.J. Cavell, Eur. J. Inorg. Chem. (2008) 2781.
(
[
(
[17] D. Morales-Morales, C.M. Jensen (Eds.), The Chemistry of Pincer Compounds,
Elsevier, New York, 2007.
[
18] (a) C. Bergquist, G. Parkin, Inorg. Chem. 38 (1999) 422;
b) G. Parkin, Chem. Rev. 104 (2004) 699;
(
1
0
ence of a d Zn center to attain a tetrahedral local environment
trumps any stabilization gained by removal of constraints within
the ligand set.
(c) B.M. Bridgewater, T. Fillebeen, R.A. Friesner, G. Parkin, J. Chem. Soc., Dalton
Trans. (2000) 4494;
(
(
d) J.G. Melnick, A. Docrat, G. Parkin, Chem. Commun. (2004) 2870;
e) C. Kimblin, B.M. Bridgewater, D.G. Churchill, T. Hascall, G. Parkin, Inorg.
Chem. 39 (2000) 4240.
[
19] (a) M. Tessmer, M. Shu, H. Vahrenkamp, Inorg. Chem. 40 (2001) 4022;
Acknowledgements
(
(
b) J. Seebacher, M. Shu, H. Vahrenkamp, Chem. Commun. (2001) 1026;
c) M.M. Ibrahim, M. Shu, H. Vahrenkamp, Eur. J. Inorg. Chem. (2005) 1388;
This work was supported by generous funding from The Fair-
field University Science Institute (J.R.M.), Fairfield University
Start-up Funding (J.R.M.), a Fairfield University Research Grant
(d) R. Walz, H. Vahrenkamp, Inorg. Chim. Acta 314 (2001) 58;
(e) M. Rombach, J. Seebacher, M. Ji, G. Zhang, G. He, M. Ibrahim, B. Benkmil, H.
Vahrenkamp, Inorg. Chem. 45 (2006) 4571;
(
(
f) M. Ibrahim, J. Seebacher, G. Steinfeld, H. Vahrenkamp, Inorg. Chem. 44
2005) 8531.
(
(
J.R.M.), and an NTID Faculty Evaluation and Development Grant
M.A.L.). J.R.M. thanks Prof. Matthew A. Kubasik, Prof. L. Kraig Stef-
[20] A.A. Gridnev, I.M. Mihaltseva, Synth. Commun. 24 (1994) 1547.
21] J.A. Loch, M.A. Albrecht, E. Peris, J. Mata, J.W. Faller, R.H. Crabtree,
Organometallics 21 (4) (2002) 700.
[
fen, and Prof. Swapan Jain for helpful suggestions.
[
[
22] E. Kuwano, M. Kikuchi, M. Eto, Agric. Biol. Chem. 52 (6) (1988) 1619.
23] M. Albrecht, J.R. Miecznikowski, A. Samuel, J.W. Faller, R.H. Crabtree,
Organometallics 21 (17) (2002) 3596.
Appendix A. Supplementary data
[
24] W.-G. Jia, Y.-B. Huang, G.-X. Jin, J. Organomet. Chem. 694 (25) (2009) 4008.
The H, 13C, and HSQC NMR spectra of 2a, 2b, 3b, and 1–3 and
mass spectra of 1–3 are provided. IR spectra for 2b and 3b and
crystallographic details of 1–3 are also given.
1
[25] J. Lutz, F. Hollmann, T.V. Ho, A. Schnyder, R. Fish, A. Schmid, J. Organomet.
Chem. 689 (2004) 4783.
[
26] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox,
H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E.
Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y.
Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S.
Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J.
Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M.
Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A.
Pople, Gaussian 03, Revision E.01, Gaussian, Inc., Wallingford, CT, 2004.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ica.2011.07.021.
References
[
[
[
1] R.H. Holm, P. Kennepohl, E.I. Solomon, Chem. Rev. 96 (1996) 2239.
2] J.A. Ibers, R.H. Holm, Science 209 (1980) 223.
3] (a) W.N. Lipscomb, N. Sträter, Chem. Rev. 96 (1996) 2375;
b) E. Kimura, T. Koike, M. Shionoya, Struct. Bond. 89 (1997) 1.
4] K.K. Kannan, B. Nostrand, K. Fridborg, S. Lovgren, A. Ohlsson, M. Petef, Proc.
Natl. Acad. Sci. USA 72 (1975) 51.
(
[
[
[
[
5] A. Meibner, W. Haehnel, H. Vahrenkamp, Chem. Eur. J 3 (1997) 261.
6] A. Dolega, Coord. Chem. Rev. 254 (2010) 916.
7] A. Dolega, A. Pladzyk, K. Baranowska, J. Jezierska, Inorg. Chim. Acta 362 (2009)
[27] (a) W.R. Fawcett, Langmuir 24 (2008) 9868;
(b) L.E. Roy, E. Jakubikova, G. Guthrie, E.R. Batista, J. Phys. Chem. A 113 (2009)
6745.
5
085.
[28] E. Peris, J.A. Loch, J. Mata, R.H. Crabtree, Chem. Commun. (2001) 201.
[29] A. Caballero, E. Díez-Barra, F.A. Jalón, S. Merino, J. Tejeda, J. Organomet. Chem.
617–618 (2001) 395.
[
[
8] L. Cronin, P.H. Walton, Chem. Commun. (2003) 1572.
9] L.M. Berreau, M.M. Makowska-Grzyska, A.M. Arif, Inorg. Chem. 40 (2001) 2212.
[
[
10] R.M. Kellogg, R.P. Hof, J. Chem. Soc., Perkin Trans. 1 (1996) 1651.
11] S.C. Shoner, K.J. Humphreys, D. Barnhart, J.A. Kovacs, Inorg. Chem. 34 (1995)
[30] B.V. Trzhtsinskaya, N.D. Abramova, J. Sulfur Chem. 10 (4) (1991) 389.
[31] H. Li, W.H. Hallows, J.S. Punzi, K.W. Pankiewicz, K.A. Watanabe, B.M. Goldstein,
Biochemistry 33 (1994) 11734.
[32] M.M. Makowska-Grzyska, P.C. Jeppson, R.A. Allred, A.M. Arif, L.M. Berreau,
Inorg. Chem. 41 (2002) 4872.
5
933.
12] (a) B. Kaptein, G. Barf, R.M. Kellogg, F. Van Bolhuis, J. Org. Chem. 55 (1990)
890;
b) B. Kaptein, L. Wang-Griffen, G. Bart, R.M. Kellogg, J. Chem. Soc., Chem
Commun. (1987) 1457.
[
1
(
[33] A. Suszka, Pol. J. Chem. 54 (1980) 2289.