Improved catalytic activity of ruthenium-arene complexes in the reduction of NAD+

Article abstract of DOI:10.1021/om3006307

A series of neutral Ru^II half-sandwich complexes of the type [(η⁶-arene)Ru(N,N′)Cl] where the arene is para-cymene (p-cym), hexamethylbenzene (hmb), biphenyl (bip), or benzene (bn) and N,N′ is N-(2-aminoethyl)-4-(trifluoromethyl)benzenesulfonamide (TfEn), N-(2-aminoethyl)-4-toluenesulfonamide (TsEn), or N-(2-aminoethyl) methylenesulfonamide (MsEn) were synthesized and characterized. X-ray crystal structures of [(p-cym)Ru(MsEn)Cl] (1), [(hmb)Ru(TsEn)Cl] (5), [(hmb)Ru(TfEn)Cl] (6), [(bip)Ru(MsEn)Cl] (7), and [(bip)Ru(TsEn)Cl] (8) have been determined. The complexes can regioselectively catalyze the transfer hydrogenation of NAD ⁺ to give 1,4-NADH in the presence of formate. The turnover frequencies (TOF) when the arene is varied decrease in the order bn > bip > p-cym > hmb for complexes with the same N,N′ chelating ligand. The TOF decreased with variation in the N,N′ chelating ligand in the order TfEn > TsEn > MsEn for a given arene. [(bn)Ru(TfEn)Cl] (12) was the most active, with a TOF of 10.4 h^-1. The effects of NAD⁺ and formate concentration on the reaction rates were determined for [(p-cym)Ru(TsEn)Cl] (2). Isotope studies implicated the formation of [(arene)Ru(N,N′)(H)] as the rate-limiting step. The coordination of formate and subsequent CO₂ elimination to generate the hydride were modeled computationally by density functional theory (DFT). CO₂ elimination occurs via a two-step process with the coordinated formate first twisting to present its hydrogen toward the metal center. The computed barriers for CO₂ release for arene = benzene follow the order MsEn > TsEn > TfEn, and for the MsEn system the barrier followed bn < hmb, both consistent with the observed rates. The effect of methanol on transfer hydrogenation rates in aqueous solution was investigated. A study of pH dependence of the reaction in D₂O gave the optimum pH* as 7.2 with a TOF of 1.58 h^-1 for 2. The series of compounds reported here show an improvement in the catalytic activity by an order of magnitude compared to the ethylenediamine analogues.

Full text of DOI:10.1021/om3006307

Article
pubs.acs.org/Organometallics
Improved Catalytic Activity of Ruthenium−Arene Complexes in the
Reduction of NAD+
Joan J. Soldevila-Barreda,†,§ Pieter C. A. Bruijnincx,‡,§ Abraha Habtemariam,† Guy J. Clarkson,†
Robert J. Deeth,† and Peter J. Sadler*,†
†Department of Chemistry, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, U.K.
‡Debye Institute of Nanomaterials Science, Inorganic Chemistry & Catalysis, Utrecht University, Universiteitsweg 99, 3584 CG
Utrecht, The Netherlands
S
* Supporting Information
ABSTRACT: A series of neutral RuII half-sandwich complexes of the type
[(η6-arene)Ru(N,N′)Cl] where the arene is para-cymene (p-cym),
hexamethylbenzene (hmb), biphenyl (bip), or benzene (bn) and N,N′ is
N-(2-aminoethyl)-4-(trifluoromethyl)benzenesulfonamide (TfEn), N-(2-
aminoethyl)-4-toluenesulfonamide (TsEn), or N-(2-aminoethyl)-
methylenesulfonamide (MsEn) were synthesized and characterized. X-ray
crystal structures of [(p-cym)Ru(MsEn)Cl] (1), [(hmb)Ru(TsEn)Cl] (5),
[(hmb)Ru(TfEn)Cl] (6), [(bip)Ru(MsEn)Cl] (7), and [(bip)Ru(TsEn)Cl]
(8) have been determined. The complexes can regioselectively catalyze the
transfer hydrogenation of NAD+ to give 1,4-NADH in the presence of
formate. The turnover frequencies (TOF) when the arene is varied decrease
in the order bn > bip > p-cym > hmb for complexes with the same N,N′
chelating ligand. The TOF decreased with variation in the N,N′ chelating
ligand in the order TfEn > TsEn > MsEn for a given arene.
[(bn)Ru(TfEn)Cl] (12) was the most active, with a TOF of 10.4 h−1. The effects of NAD+ and formate concentration on
the reaction rates were determined for [(p-cym)Ru(TsEn)Cl] (2). Isotope studies implicated the formation of
[(arene)Ru(N,N′)(H)] as the rate-limiting step. The coordination of formate and subsequent CO2 elimination to generate
the hydride were modeled computationally by density functional theory (DFT). CO2 elimination occurs via a two-step process
with the coordinated formate first twisting to present its hydrogen toward the metal center. The computed barriers for CO2
release for arene = benzene follow the order MsEn > TsEn > TfEn, and for the MsEn system the barrier followed bn < hmb, both
consistent with the observed rates. The effect of methanol on transfer hydrogenation rates in aqueous solution was investigated.
A study of pH dependence of the reaction in D2O gave the optimum pH* as 7.2 with a TOF of 1.58 h−1 for 2. The series of
compounds reported here show an improvement in the catalytic activity by an order of magnitude compared to the
ethylenediamine analogues.
hepatitis C, diabetic vascular dysfunction, hyperglycemia, and
cancer.3
INTRODUCTION
■
The coenzyme nicotinamide adenine dinucleotide (NAD+) and
its reduced form 1,4-NADH have crucial roles in many cellular
metabolic processes such as regulation of energy metabolism,
antioxidative function, DNA repair and transcription, immuno-
logical functions, and cell death.1 The coenzymes are involved
in many other processes, acting as substrates and cofactors for
enzymes such as NAD+ ligases, oxidoreductases, polymerases,
and deacetylases involved in biosynthesis.
The concentration of NAD+ and the ratio NAD+/NADH
have been shown to be very important for cancer cells. On the
one hand, due to their active metabolism, cancer cells generate
high levels of oxidizing species and, therefore, they are under
constant oxidative stress.4 This makes cancer cells more
dependent on redox regulatory systems and more sensitive to
variations in the NAD+/NADH ratio. On the other hand,
NAD+ is required as a substrate for many enzymatic reactions
such as ADP-ribosylation, which is crucial for genome stability
and DNA repair.3 Due to the up-regulation of some enzymes
required for the biosynthesis of NAD+ in cancer cells, a
The coenzymes NAD+/NADH have been studied intensively
during the past few years. It has been demonstrated that
changes in metabolism result in fluctuations in the ratio NAD+/
NADH or, conversely, changes in the ratio can produce
metabolic changes.2 In some cases alterations in the cellular
redox status have been shown to play an important role in cell
death, and therefore the coenzymes have become possible drug
targets for chronic or autoimmune diseases such as Parkinson’s,
Special Issue: Organometallics in Biology and Medicine
Received: July 6, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
decrease in the NAD+ concentration can cause apoptosis of
cancer cells while having little effect on normal cells.4
From an industrial perspective, NADH regeneration is an
important process due to its high applicability in chiral organic
synthesis and biocatalysis. The coenzymes, for instance, are
required as substrates for many enzymatic reactions used for
stereoselective synthesis, such as formation of D-lactate from
pyruvate or chiral alcohols by alcohol dehydrogenases.5,6
Due to the high cost of NAD+ and NADH, their
stoichiometric use in organic synthesis, biocatalysis, or
enzymatic reactions is not sustainable, hence the development
of nicotinamide adenine dinucleotide models.7 The capability
of regenerating 1,4-NADH has been the subject of intense
studies by enzymatic, chemical, photochemical, or electro-
chemical reactions.8,9
synthesis and purification by avoiding enantiomeric resolution.
The catalytic activity of the complexes toward the reduction of
NAD+ through transfer hydrogenation has been investigated
and compared with reported ethylenediamine RuII arene
complexes. The mechanism of the catalysis has also been
studied both experimentally and computationally in order to
compare with similar known compounds such as those reported
by Steckhan and Fish.32,33 The optimum conditions for the
reaction have also been studied in relation to the potential of
these complexes to act as catalytic anticancer drugs that
modulate NAD+/NADH levels in cells.
EXPERIMENTAL SECTION
■
Materials. Trihydrated ruthenium(III) trichloride was purchased
from Precious Metals Online (PMO Pty Ltd.) and used as received. 4-
(Trifluoromethyl)benzenesulfonyl chloride, toluenesulfonyl chloride,
di-tert-butyl dicarbonate anhydride (t-BOC), hexamethylbenzene
(hmb), sodium formate, β-nicotinamide adenine dinucleotide hydrate
(NAD+), and β-nicotinamide adenine dinucleotide reduced dipotas-
sium salt (1,4-NADH) were obtained from Sigma-Aldrich. Magnesium
sulfate, potassium hydroxide, sodium chloride, and hydrochloric acid
were obtained from Fisher Scientific. α-Phellandrene was purchased
from SAFC. Methylsulfonyl chloride, toluenesulfonyl chloride, and
triflic acid were obtained from Fluka. The RuII-arene precursor dimers
[(η6-arene)RuCl2]2 where the arene is p-cymene (p-cym), hexame-
thylbenzene (hmb), biphenyl (bip), or benzene (bn) were prepared
following literature methods,34 as were the ligands N-(2-aminoethyl)-
4-(trifluoromethyl)benzenesulfonamide (TfEn)19 and N-(2-amino-
ethyl)-4-toluenesulfonamide (TsEn).28 The solvents used for NMR
spectroscopy were purchased from Sigma-Aldrich and Cambridge
Isotope Laboratories Inc. Nondried solvents used in syntheses were
obtained from Fisher Scientific and Prolabo. Solvents were used as
obtained, except in the case of ethanol, 2-propanol, and methanol,
which were degassed prior to use by bubbling with nitrogen.
Recently, NADH/NAD+ regeneration under mild conditions
in aqueous media has been studied.10,11 Significant attention
has focused on transition metal complexes as catalysts for
regeneration of NADH via hydrogenation10 (H2) or transfer
hydrogenation using 2-propanol,12 glycerol,13 phosphate,14
formic acid together with a base,15 or formate11 as the hydride
19
source in water. Some RuII,16 RhIII,17,18 and IrIII
half-
sandwich complexes can catalyze the hydrogenation or transfer
hydrogenation of ketones,20 aldehydes,21 imines,22 or carbon
dioxide,23 although the optimum conditions for the reactions
are usually not biologically relevant. In some cases the hydride
adducts can be isolated, showing that the mechanism involves
hydride formation.24,25 Of particular interest in this field are
ruthenium complexes that employ Noyori’s chelating ligand Ts-
DPEN (N-[(1R,2R)-2-amino-1,2-diphenylethyl]-4-toluenesul-
fonamide) and its derivatives, which have achieved high
enantioselectivity in transfer hydrogenation of ketones,26
imines,27 and CC double bonds,28 affording up to 97%
conversions. However, the use of similar complexes under
biologically relevant conditions requires high aqueous solubility
and thermodynamic stability. The use of Noyori’s ligand
renders the complexes highly insoluble in water, due to the
presence of the lipophilic phenyl groups, although recently
some water-soluble complexes have been synthesized.21,29,30 In
addition, the presence of two chiral centers, which provide the
complexes with enantioselectivity, makes the acquisition of a
pure enantiomer a challenging process. Given that enantiose-
lectivity is not a requirement in NAD+/NADH catalysis, not
having chiral centers offers an acceptable strategy.
Synthesis of Ruthenium Complexes. [(p-cym)Ru(MsEn)Cl]
(1). [(p-cym)RuCl2]2 (84.2 mg, 0.138 mmol) and MsEn (101.6 mg,
0.35 mmol) were placed in a round-bottom flask to which 2-propanol
(50 mL) and triethylamine (146 μL, 1.047 mmol) were added. The
solution was heated under reflux in a nitrogen atmosphere overnight,
after which the solvent was removed on the rotary evaporator to
obtain a light brown powder. The crude product was redissolved in
dichloromethane and washed with brine, after which the organic layer
was dried over MgSO4 and filtered. A brownish-red powder was
obtained after removal of the solvent in vacuo. Brown crystals were
obtained after recrystallization from methanol/ether (1:10 v/v) by
standing in a freezer for two months at 253 K. Yield: 45.3 mg (40.4%).
1H NMR (400 MHz, methanol-d4): δH 5.59 (s, 2H), 5.42 (d, 2H, J =
We have shown previously that the RuII complexes [(η6-
arene)Ru(en)Cl]+ (arene = hmb, p-cym, indane (ind), and en =
ethylenediamine) can reduce NAD+ to NADH using formate as
the hydride source and that such reactions might be feasible in
cells because they can tolerate millimolar levels of formate.31
However, due to the low catalytic activity of the complex, no
effect was detected on A2780 human ovarian cancer cells in
vitro. In the present work we modified the en chelating ligand
by the addition of a sulfonamide group that resembles Noyori’s
ligand with the main goal of improving the catalytic activity of
the complexes. The synthesized series of RuII complexes of the
type [(η6-arene)Ru(N,N′)Cl], where N,N′ are water-soluble
monosulfonate ethylenediamine ligands, N-(2-aminoethyl)-4-
(trifluoromethyl)benzenesulfonamide (TfEn), N-(2-aminoeth-
yl)-4-toluenesulfonamide (TsEn), or N-(2-aminoethyl)-
methylenesulfonamide (MsEn), derived from Noyori’s ligand
Ts-DPEN (N-[(1R,2R)-2-amino-1,2-diphenylethyl]-4-toluene-
sulfonamide), were envisaged to give rise to good catalytic
properties while maintaining water solubility and facilitating the
5.74 Hz), 2.85 (sep, 1H, J = 7.05 Hz), 2.743 (s, 3H), 2.134 (s, 3H),
1.265 (d, 6H, J = 7.05 Hz). Anal. Calcd for C13H23ClN2O2RuS: C
38.28, H 5.68, N 6.87. Found C 38.27, H 5.65, N 6.82. ESI-MS: calcd
for C13H23ClN2O2RuS (M − Cl)+ m/z 373.0, found m/z 373.0.
[(p-cym)Ru(TsEn)Cl] (2). Complex 2 was obtained following the
method described above for complex 1 using the ligand TsEn (100.6
mg, 0.42 mmol). Recrystallization from methanol resulted in dark red
crystals. Yield: 58.8 mg (36.7%). 1H NMR (400 MHz, acetone-d6): δH
7.73 (d, 2H, J = 8.25 Hz), 7.11 (d, 2H, J = 8.25 Hz), 5.85 (s, 1H), 5.71
(d, 1H, J = 5.59 Hz), 5.64 (d, 1H, J = 5.59 Hz), 5.52 (d, 1H, J = 5.59
Hz), 5.45 (d, 1H, J = 5.59 Hz), 3.24 (s, 1H), 2.94 (septet, 1H, J = 7.07
Hz), 2.84 (m,1H), 2.68 (m, 1H), 2.31 (s, 3H), 2.18 (m, 1H), 2.15 (s,
3H), 1.27 (d, 3H, J = 6.96 Hz), 1.24 (d, 3H, J = 6.96 Hz). Anal. Calcd
for C19H27ClN2O2RuS: C 47.15, H 5.62, N 5.79. Found: C 46.70, H
5.62, N 5.79. ESI-MS: calcd for C19H27ClN2O2RuS (M − Cl)+ m/z
449.0, found m/z 449.0.
[(p-cym)Ru(TfEn)Cl] (3). Complex 3 was obtained following the
method described above for complex 1 using ligand TfEn (125.0 mg,
0.47 mmol). Purification by recrystallization was not successful, and
the product was used as such. Yield: 171.1 mg (85.3%). 1H NMR (400
MHz, acetone-d6): δH 8.00 (d, 2H, J = 8.12 Hz), 7.62 (d, 2H, J = 8.12
B
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Reduction of NAD+ to 1,4-NADH in the Presence of Formate as a Hydride Source
Hz), 5.74 (m, 1H), 5.51 (m, 2H), 5.41 (d, 1H), 4.23 (s, 1H), 3.27 (s,
1H), 3.10 (s, 1H), 2.76 (septet, 1H, J = 6.92), 2.75 (m, 1H), 2.29 (m,
2H), 2.17 (s, 3H), 1.27 (m, 6H). Anal. Calcd for
C19H24ClF3N2O2RuS: C 42.42, H 4.50, N 5.21. Found: C 41.94, H
4.45, N 5.03. ESI-MS: calcd for C19H24ClF3N2O2RuS (M − Cl)+ m/z
503.0, found m/z 503.0.
[(hmb)Ru(MsEn)Cl] (4). [(hmb)RuCl2]2 (81.2 mg, 0.12 mmol)
and N-(2-aminoethyl)methylsulfonamide (MsEn) (105 mg, 0.37
mmol) were dissolved in 2-propanol (25 mL). Triethylamine (155
μL, 1.11 mmol) was added, and the solution heated to reflux (363 K)
under a nitrogen atmosphere for 3 h. Solvent was removed on a rotary
evaporator to obtain a dark orange product. The crude product was
redissolved in dichloromethane and washed with brine, after which the
organic layer was dried over MgSO4 and filtered. The solution was
concentrated in vacuo, and the product recrystallized from methanol to
afford an orange powder. Yield: 44.1 mg (41.6%). 1H NMR (400
MHz, acetone-d6): δH 2.11 (s, 18H), 2.32 (m, 2H), 2.72 (s, 3H), 2.67
(m, 2H), 4.58 (s, 2H). Anal. Calcd for C15H27ClN2O2RuS: C 41.32, H
6.24, N 6.43. Found: C 41.92, H 6.31, N 6.32. ESI-MS: calcd for
C15H27ClN2O2RuS (M − Cl)+ m/z 401.0, found m/z 401.0.
[(hmb)Ru(TsEn)Cl] (5). Complex 5 was obtained following the
method described above for complex 4 using ligand TsEn (64.1 mg,
0.267 mmol). The product was recrystallized from methanol, giving
yellow crystals. Crystals suitable for X-ray diffraction were obtained
from a methanol solution stored at 255 K. Yield: 20.5 mg (13.4%). 1H
NMR (400 MHz, acetone-d6): δH 7.83 (d, 2H, J = 8.30 Hz), 7.07 (d,
2H, J = 8.30 Hz), 4.56 (s, 1H), 2.54 (m, 1H), 2.29 (s, 3H), 2.17 (s,
18H). Anal. Calcd for C21H31ClN2O2RuS: C 49.35, H 5.92, N 5.48.
Found: C 49.17, H 6.04, N 5.29. ESI-MS: calcd for C21H31ClN2O2RuS
(M − Cl)+ m/z 477.0, found m/z 477.0.
C 42.10, H 4.48, N 6.55. Found: C 41.88, H 4.49, N 6.46. ESI-MS:
calcd for C15H19ClN2O2RuS (M − Cl)+ m/z 393.0, found m/z 393.0.
[(bip)Ru(TsEn)Cl] (8). Complex 8 was obtained following the
method described above for complex 7 using ligand TsEn (69.9 mg,
0.29 mmol). The product was recrystallized from methanol to give a
dark orange powder. Yield: 39.9 mg (33.9%). Crystals suitable for X-
ray diffraction were obtained from a methanol solution stored at 277
1
K. H NMR (400 MHz, CDCl3): δH 2.35 (s, 3H), 2.51(m,1H), 3.08
(m, 1H), 3.22 (m, 1H), 3.68 (m, 1H), 5.77 (m, 1H), 5.96 (m, 1H),
6.03 (m, 1H), 6.39 (m, 1H), 7.17 (d, 2H, J = 7.71 Hz), 7.53 (m, 3H),
7.71 (m, 2H), 7.77 (d, 2H, J = 7.71 Hz). Anal. Calcd for
C21H23ClN2O2RuS: C 50.04, H 4.60, N 5.56. Found: C 49.57, H
4.6, N 5.44. ESI-MS: calcd for C21H23ClN2O2RuS (M − Cl)+ 469.0 m/
z, found 469.0 m/z.
[(bip)Ru(TfEn)Cl] (9). Complex 9 was obtained following the
method described above for complex 7 using ligand TfEn (40 mg, 0.15
mmol). The product was recrystallized from methanol to give an
1
orange solid. Yield: 45.6 mg (54.3%). H NMR (400 MHz, CDCl3):
δH 2.17 (m, 1H), 2.31 (m, 1H), 2.55 (m, 1H), 3.11 (m, 1H), 3.21 (s,
1H), 3.74 (s, 1H), 5.80 (m, 2H), 5.95 (d, 1H, J = 5.51 Hz), 6.04 (t,
1H, J = 5.30 Hz), 6.34 (t, 1H, J = 5.82 Hz), 7.54 (m, 3H), 7.62 (d, 2H,
J = 8.27 Hz), 7.70 (m, 2H), 8.00 (d, 2H, J = 8.27 Hz). Anal. Calcd for
C21H20ClF3N2O2RuS·0.5 H2O: C 44.49, H 3.52, N 4.94. Found: C
44.22, H 3.59, N 4.81. ESI-MS: calcd for C21H20ClF3N2O2RuS (M −
Cl)+ m/z 522.9, found m/z 522.9.0.
[(bn)Ru(MsEn)Cl] (10). [(bn)RuCl2]2 (64 mg, 0.098 mmol) and
MsEn (70 mg, 0.243 mmol) were placed in a round-bottom flask and
dissolved in methanol (25 mL). Triethylamine (109 μL, 0.783 mmol)
was added, and the solution was stirred at ambient temperature under
nitrogen overnight, after which the solvent was removed on the rotary
evaporator to give a dark brown oil, which was dissolved in
dichloromethane, washed with brine, dried over MgSO4, and filtered.
After removal of solvent on a rotary evaporator, a dark brown oil was
obtained, which was washed with ether to give a brown solid. The
product was recrystallized from methanol to give a brown solid. Yield:
[(hmb)Ru(TfEn)Cl] (6). Complex 6 was obtained following the
method described above for complex 4 using ligand TfEn (87.5 mg,
0.33 mmol). The product was recrystallized from methanol, giving red
crystals. Yield: 82.2 mg (58.9%). Crystals suitable for X-ray diffraction
1
were obtained from a methanol solution stored at 277 K. H NMR
1
(400 MHz, CDCl3): δH 8.05 (d, 2H, J = 8.15 Hz), 7.56 (d, 2H, J =
8.15 Hz), 3.38 (s, 1H), 3.21 (s, 1H), 3.12 (m, 1H), 2.74 (m, 1H), 2.34
(m, 1H), 2.18 (s, 18H). Anal. Calcd for C21H28ClF3N2O2RuS: C
44.56, H 4.99, N 4.95. Found: C 44.38, H 5.04, N 4.87. ESI-MS: calcd
for C21H28ClF3N2O2RuS (M − Cl)+ m/z 531.0, found m/z 531.0.
[(bip)Ru(MsEn)Cl] (7). [(bip)RuCl2]2 (64 mg, 0.098 mmol) and
MsEn (70 mg, 0.243 mmol) were placed in a round-bottom flask and
disolved in methanol (25 mL). Triethylamine (109 μL, 0.783 mmol)
was added, and the solution was stirred at ambient temperature under
nitrogen overnight, after which the solvent was removed on the rotary
evaporator to give a dark brown product. The crude product was
redissolved in dichloromethane and washed with brine, and the
organic layer was dried over MgSO4 and filtered. The solution was
concentrated in vacuo, and the product recrystallized from methanol to
afford red crystals. Yield: 45.2 mg (54.3%). Crystals suitable for X-ray
diffraction were obtained from solution in methanol/ether (10:1 v/v)
stored at 255 K. 1H NMR (400 MHz, CDCl3): δH 7.69 (m, 2H), 7.54
(m, 3H), 6.22 (t, 1H, J = 5.57 Hz), 6.01 (t, 1H, J = 5.57 Hz), 5.82 (m,
3H), 3.78 (s, 1H), 3.23 (s, 1H), 3.09 (m, 1H), 2.81 (s, 3H), 2.64 (m,
1H), 2.45 (m, 1H), 2.15 (m, 1H). Anal. Calcd for C15H19ClN2O2RuS:
30 mg (49.2%). H NMR (400 MHz, CDCl3): δH 5.67 (s, 6H), 2.80
(s, 3H). Anal. Calcd for C9H15ClN2O2RuS: C 30.73, H 4.30, N 7.96.
Found: C 30.29, H 4.14, N 7.76. ESI-MS: calcd for C9H15ClN2O2RuS
(M − Cl)+ m/z 317.0, found m/z 317.0.
[(bn)Ru(TsEn)Cl] (11). Complex 11 was obtained following the
method described above for complex 10 using ligand TsEn (100 mg,
0.417 mmol). The crude product was recrystallized from methanol to
1
give an orange powder. Yield: 47 mg (52.4%). H NMR (400 MHz,
CDCl3): δH 7.75 (d, 2H, J = 7.79 Hz), 7.28 (d, 2H, J = 7.79 Hz), 5.71
(s, 6H), 4.89 (s, 1H), 3.14 (s, 1H), 3.05 (s, 1H), 2.75 (s, 1H), 2.35 (m,
1H), 2.35 (s, 3H). Anal. Calcd for C15H19ClN2O2RuS: C 42.10, H
4.48, N 6.55. Found: C 41.83, H 4.48, N 6.74. ESI-MS: calcd for
C15H19ClN2O2RuS (M − Cl)+ m/z 393.0, found m/z 393.0.
[(bn)Ru(TfEn)Cl] (12). Complex 12 was obtained following the
method described above for complex 10 using ligand TfEn (100 mg,
0.38 mmol). The crude product was recrystallized from methanol to
1
give a reddish powder. Yield: 53.6 mg (59.0%). H NMR (400 MHz,
CDCl3): δH 7.99 (d, 2H, J = 8.2 Hz), 7.62 (d, 2H, J = 8.2 Hz), 5.73 (s,
6H), 4.71 (s, 1H), 3.12 (m, 2H), 2.79 (m,1H), 2.25 (m, 2H). Anal.
Calcd for C15H16ClF3N2O2RuS: C 37.39, H 3.35, N 5.81. Found: C
C
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
37.00, H 3.39, N 5.81. ESI-MS: calcd for C15H16ClF3N2O2RuS (M −
Cl)+ m/z 446.9, found m/z 446.9.
nitrogens, which were located in a difference map. Anisotropic
displacement parameters were used for all non-H atoms; H atoms
were given an isotropic displacement parameter equal to 1.2 (or 1.5 for
methyl and NH H atoms) times the equivalent isotropic displacement
parameter of the atom to which they are attached. The data were
processed by the modeling program Mercury 1.4.1.
NAD+ Reduction. Complexes 1−12 were dissolved in D2O or
MeOD/D2O (5:1 v/v) (1.4 mM, 4 mL) in a glass vial. Solutions of
sodium formate (35 mM, 4 mL) and NAD+ in D2O (2.8 mM, 2 mL)
were also prepared and then incubated at 310 K. In a typical
experiment an aliquot of 200 μL from each solution was added to a 5
mm NMR tube, and the pH* adjusted to 7.2 0.1, bringing the total
volume to 0.635 mL (final concentrations were Ru complex 0.44 mM;
NAD+ 0.88 mM; NaHCO2 11.02 mM; molar ratio 1:2:25). A 1H
NMR spectrum was recorded at 310 K every 162 s until the
completion of the reaction.
X-ray crystallographic data for complexes 1, 5, 6, 7, and 8 are
available as Supporting Information and have been deposited in the
Cambridge Crystallographic Data Centre (CCDC reference numbers
885977, 885979, 885978, 885976, and 885980, respectively).
Computational Details. All calculations used the ORCA program
version 2.8.37 Minimum-energy structures and transition states were
located using the OPBE functional, spin restricted, in conjunction with
the def2-TZVP basis sets and the resolution of identity approx-
imation.38,39 Solvation effects were included via the conductor-like
screening model implemented in ORCA with water as the solvent.
Stationary points were confirmed as local minima or transition states
via numerical frequency calculations. Final free energies were
computed using the statistical mechanics corrections computed from
the OPBE optimizations with B3LYP energies using the def2-TZVP
basis sets plus the COSMO solvation corrections40 and Grimme’s
empirical correction for dispersion.41
Molar ratios of NAD+ and NADH were determined by integrating
the peaks corresponding to HB (Scheme 1) of NAD+ (9.33 ppm) and
HB of 1,4-NADH (6.96 ppm). The turnover number (TON) for the
reaction was calculated as follows:
[NAD+]0
I6.96
TON =
I6.96 + I9.93 [Catalyst]
where In is the integral of the signal at n ppm and [NAD+]0 is the
concentration of NAD+ at the start of the reaction.
A series of experiments were performed where complex 2 (2.71 mg,
0.0056 mmol) was dissolved in D2O (1.4 mM, 4 mL) in a vial.
Following the same procedure used above, the kinetics of the reaction
was studied using different concentrations of NAD+ (1, 2, 3, 4, and 6
molar equiv). A second series of experiments using different
concentrations of sodium formate (2, 5, 10, 25, 50, 100, and 200
molar equiv) were also performed. 1H NMR spectra were recorded at
310 K every 162 s until completion of the reaction in both cases.
The optimum pH range for the catalytic reaction was studied by a
series of kinetic experiments following the same procedure described
above. In each experiment the pH* was adjusted to 6, 6.5, 7, 7.2, 7.5, 8,
and 10. The effect of added methanol on the reaction rate was studied
by six experiments containing different quantities of methanol (0, 5,
10, 23.3, 50, 66.6% v/v of d4-MeOD in D2O).
RESULTS
■
Synthesis and Characterization. RuII complexes 1−12
were synthesized (Chart 1) using a similar procedure. Typically,
triethylamine (4−6 molar equiv) and the ligand (2−2.5 molar
equiv) were added to an alcoholic solution of the ruthenium
dimer [(η6-arene)RuCl2]2, and the reaction mixture was stirred
under a N2 atmosphere under reflux or at ambient temperature.
The details for individual reactions are described in the
Experimental Section. All the synthesized complexes were
characterized by elemental analysis (CHN), mass spectrometry
1
(ESI-MS), and H NMR spectroscopy. X-ray crystal structures
for complexes 1 and 5−8 were obtained. 1H NMR spectra were
fully assigned although peaks corresponding to the CH2 of the
chelating ethylenediamine backbone, expected between 2 to 4
ppm, were broad.
1
NMR Spectroscopy. H NMR spectra were acquired in 5 mm
NMR tubes at 298 or 310 K on either a Bruker DPX-300, AV-400,
DRX-500, or AV III 600 spectrometer. Data processing was carried out
using XWIN-NMR version 3.6 (Bruker U.K. Ltd.). 1H NMR chemical
shifts were internally referenced to TMS via 1,4-dioxane in D2O (δ =
3.75) or residual DMSO (δ = 2.52 ppm), acetone (δ = 2.05 ppm),
MeOH (δ = 3.31 ppm), or CHCl3 (δ = 7.26 ppm). 1D spectra were
recorded using standard pulse sequences. Typically, data were acquired
with 16 transients into 32 k data points over a spectral width of 14
ppm and, for the kinetic experiment, 32 transients into 32 k data
points over a spectral width of 30 ppm using a relaxation delay of 2 s.
pH* Measurements. pH values were measured at ambient
temperature using a Minilab IQ125 pH meter equipped with a
ISFET silicon chip pH sensor and referenced in KCl gel. The electrode
was calibrated with Aldrich buffer solutions of pH 4, 7, and 10. pH*
values (pH meter reading without correction for the effect of
deuterium on a glass electrode) of NMR samples in D2O were
measured at 310 K. pH* values were adjusted with KOH or HClO4
solutions in D2O.
Chart 1. Ruthenium Arene Complexes Studied in This Work
Elemental Analysis. Elemental analyses were performed by
Warwick Analytical Service using an Exeter Analytical elemental
analyzer (CE440).
Electrospray Ionization Mass Spectrometry (ESI-MS). Positive
ion electrospray mass spectra were obtained on a Bruker Daltonics
Esquire 2000 ion trap mass spectrometer. All samples were prepared in
water/acetonitrile (20:80 v/v) or methanol (100%). Data were
processed using Data-Analysis version 3.3 (Bruker Daltonics).
X-ray Crystallography. Diffraction data were collected on an
Oxford Diffraction Gemini four-circle system with a Ruby CCD area
detector. All structures were refined by full-matrix least-squares against
F2 using SHELXL 9735 and were solved by direct methods using
SHELXS36 (TREF) with additional light atoms found by Fourier
methods. Hydrogen atoms were added at calculated positions and
refined using a riding model, except the hydrogens on the NH
X-ray Crystal Structures. Crystallographic data for
complexes 1 and 5−8 are listed in Table S1, and selected
bond lengths and angles in Tables 1 and 2. Their molecular
structures are depicted in Figure 1. These complexes adopt the
familiar pseudo-octahedral “piano-stool” geometry with the η6-
bonded arene occupying one face of the complex. Complexes 1,
5, 7, and 8 were obtained in an enantiomerically pure form.
The chelating ligand appeared to be deprotonated and
D
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
X-ray diffraction studies for complex 6 showed three
crystallographically independent structures in the asymmetric
unit (Figure S1): 6a and 6a′ correspond to one enantiomer;
6a′ forms an H-bond to MeOH and 6b to the other
enantiomer.
Hydrolysis and pKa of Aqua Complexes. Hydrolysis of
complex 2 was studied since the formation of an aqua complex
is a possible first step in the catalytic cycle. When the complex
was dissolved in D2O (1.4 mM), no changes were observed by
1H NMR after 24 h. The peaks in the 1H NMR spectra can be
assigned to the aqua complex since the chemical shifts of the
resonances correlate to those obtained for the aqua complex
(prepared by the addition of silver nitrate (1 molar equiv) in
water). The rate of aquation was too fast to be determined by
1H NMR.
Table 1. Selected Bond Lengths (Å) and Angles (deg) for
Complexes 1, 5, 7, and 8
a
a
b
b
5
8
1
7
Ru1−Na
Ru1−N−
Ru1−Cl1
2.141(3)
2.129(3)
2.4146(8)
1.6742
2.122(3)
2.096(4)
2.4444(11)
1.6742
2.1331(15)
2.1339(13)
2.4425(4)
1.672
2.1276(19)
2.1073(16)
2.4331(6)
1.672
Ru1−arene
(centroid)
N−−Ru1−Na
Na−Ru1−Cl1
N−−Ru1−Cl1
78.00(11)
89.44(8)
85.63(8)
78.54(14)
85.44(11)
82.97(11)
78.74(6)
87.91(4)
77.80(7)
82.57(6)
88.61(5)
84.53(5)
a
b
N− corresponds to N4; Na corresponds to N1. N− corresponds to
N3; Na corresponds to N6.
Table 2. Selected Bond Lengths (Å) and Angles (deg) for the
Three Crystallographically Independent Molecules of
Complex 6
The changes in the 1H NMR chemical shifts of the p-cymene
protons in the aqua adduct of complex 2 (1.4 mM) were
followed by changing the pH* ranging from 2 to 12. The data
were fitted to the Henderson−Hasselbalch equation, which
resulted in a pKa* of 9.06 (Figure S2).
a
b
c
6a
6a′
6b
Ru1−N−
Ru1−Na
Ru1−Cl1
Ru1−arenecentroid
N−−Ru1−Na
Na−Ru1−Cl1
N−−Ru1−Cl1
2.116(2)
2.151(2)
2.4344(6)
1.679
2.1234(18)
2.134(2)
2.4229(6)
1.675
2.133(2)
2.142(2)
2.4240(6)
1.674
Kinetics of Transfer Hydrogenation Reactions. Cata-
lytic transfer hydrogenation of nicotinamide adenine dinucleo-
tide (NAD+) in aqueous media using complexes 1−12 as
catalyst and formate as a hydride source (25 molar equiv) was
studied by 1H NMR at 310 K and pH* 7.2 0.1. Spectra were
recorded every 162 s until completion of the reaction. In order
to compare the catalytic activity of the complexes and due to
the poor solubility of complexes 6−9, the experiments were
also performed in the mixed solvent MeOD/D2O (2:9 v/v).
The reaction was regioselective, giving 1,4-NADH in 93%
yield when complex [(hmb)Ru(MsEn)Cl] (4) was used, and a
similar regioselectivity was achieved with complexes containing
hexamethylbenzene (5 and 6). When using p-cymene
complexes 1, 2, and 3, the 1,6-NADH adduct was also
produced (up to 16%), with complex [(p-cym)Ru(TfEn)Cl]
(3) giving the lowest regioselectivity from the three p-cymene
77.58(8)
90.39(6)
84.74(6)
77.69(8)
87.49(5)
84.81(6)
77.54(8)
89.43(5)
84.64(6)
a
b
N− corresponds to N104; Na corresponds to N101. N− corresponds
c
to N204; Na corresponds to N201. N− corresponds to N304; Na
corresponds to N301.
coordinated as a monoanionic bidentate N,N′-ligand (amine
and amide coordination). A chloride anion completes the
coordination sphere around ruthenium. Especially interesting
for further discussion are the distances Ru−Cl 2.4444(11) Å for
[(bip)Ru(TsEn)Cl] (8), 2.4331(6) Å for [(bip)Ru(MsEn)Cl]
(7), 2.4425(4) Å for [(p-cym)Ru(MsEn)Cl] (1), 2.4146(8) Å
for [(hmb)Ru(TsEn)Cl] (5), and 2.4344(6), 2.4229(6), and
2.4240(6) Å for [(hmb)Ru(TfEn)Cl] (6).
Figure 1. ORTEP diagrams for complexes (A) 7, (B) 8, (C) 1, (D) 6, and (E) 5. Ellipsoids are shown at the 50% probability level. All hydrogen
atoms and solvent molecules have been omitted for clarity.
E
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
complexes. Finally complexes 10, 11, and 12, containing
benzene as arene, gave 23−25% of 1,6-NADH.
Molar ratios of NAD+ and NADH (Scheme 1) were
determined by integrating the signals corresponding to HB of
NAD+ (9.33 ppm) and HB of 1,4-NADH (6.96 ppm; Scheme
1). The turnover numbers for the reactions were determined as
described in the Experimental Section. An important effect of
adding methanol to the reaction mixture on the TON was
discovered (vide infra). The turnover frequencies (TOF, the
increment of turnover number over time), Table 3, showed a
Table 3. Turnover Frequencies for Transfer Hydrogenation
Reactions Using Catalysts 1−12
a
complex R1 (arene)
R2
TOF (h−1) D2O/MeOD
TOF (h−1) D2O
1
p-cym
p-cym
p-cym
hmb
hmb
hmb
bip
Ms
Ts
Tf
1.25
2.88
5.77
0.15
0.38
1.47
3.04
4.28
7.45
4.09
6.76
10.39
1.11
1.58
3.06
0.14
0.34
Figure 2. Dependence of the turnover number on time for the
reduction of NAD+ by complex 2 using formate as a hydride source
(mol ratio 2:1:X, X = 2, 5, 10, 25, 50, 100, or 200. pH* 7.4, 310 K,
D2O).
2
3
4
Ms
Ts
Tf
5
b
__
b
__
b
__
b
__
6
a plot of turnover frequency versus formate concentration.
Thus, from the double-reciprocal plot (Figure 3) the Michaelis
constant (KM = 27.8 mM) and the maximum turnover
frequency (TOFmax = 6.4 h−1) were calculated.
7
Ms
Ts
Tf
8
bip
9
bip
10
11
12
bn
Ms
Ts
2.94
4.50
6.62
bn
bn
Tf
a
b
23% v/v MeOD/D2O. Not soluble in D2O.
marked dependence of the catalytic activity on the arene.
Complexes of the type [(arene)Ru(XEn)Cl], where X is Tf, Ts,
or Ms, are more active when the arene is benzene, and activity
decreases in the order bn > bip > p-cym > hmb. It was also
evident from the turnover frequencies that catalytic activity
depends on the N,N′-ligand. The complex with the electron-
withdrawing chelated ligand sulfonamide substituent (TfEn)
exhibited the highest activity, while the complex with the least
electron-withdrawing sulfonamide substituent (MsEn) was the
least active (Table 3). [(bn)Ru(TfEn)Cl] (12), with a TOF of
10.4 h−1, was the most active.
Mechanistic Studies. The rate-limiting step of the catalytic
cycle for NAD+ reduction was investigated by studying the
reaction of complex 3 with deutero-formate (25 molar equiv) as
a deuteride source. The turnover frequency (1.96 h−1) was 2.4
times smaller than that obtained using formate as the hydride
source.
Figure 3. Plot of the reciprocal of the TOF against formate
concentration for the reduction of NAD+ in the presence of various
molar equiv of formate, catalyzed by complex 2. For a reaction
following Michaelis-type kinetics, TOF = TOFmax[S]/(Km + [S]),
where TOFmax is the turnover frequency at infinite substrate (formate)
concentration, [S] is the substrate concentration, and Km is the
Michaelis constant. Hence, TOF−1 = (Km/TOFmax)(1/[S]) + (1/
TOFmax), and Km and TOFmax can be obtained from the gradient and y
intercept, respectively, of the double-reciprocal plot.
The dependence of the rate of catalytic activity on
concentration of NAD+ and formate was also determined
using complex 2 as a model system. Accordingly five
experiments were performed with complex 2, sodium formate,
and NAD+ in the ratio of 1:25:X, respectively, where X = 1, 2, 3,
1
4, or 6 molar equiv, and the H NMR spectra were recorded
The pH* dependence of the catalytic reaction ranging from 6
to 10 was also investigated via a series of experiments using
complex 2, sodium formate, and NAD+ (ratio 1:25:2,
respectively) in D2O. A strong dependence on pH* was
found. The highest TOF (1.58 h−1) was achieved at pH* 7.2,
followed by a decrease in the activity when the pH* increased
(Figure S3).
The catalytic activity of all the complexes was initially studied
in D2O, but complexes 6−9 could not be dissolved to the same
concentration (1.4 mM). For this reason experiments were also
performed in a D2O/methanol (9:2 v/v) mixture in order to
increase the solubility. The addition of methanol in the reaction
every 2 min until 100% conversion was achieved. The analysis
of the experiments gave an unchanged turnover frequency
(2.88 h−1).
A second series of experiments were performed in order to
investigate the dependence of the reaction rate on the formate
concentration. Experiments with complex 2, sodium formate,
and NAD+ in the ratio 1:X:2, respectively, where X = 2, 5, 10,
25, 50, 100, or 200 molar equiv of formate, were performed.
The plot of the data obtained shows a strong dependence of the
turnover frequency on formate concentration (Figure 2).
Typical Michaelis−Menten kinetic behavior is apparent from
F
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Table 4. Computed DFT(B3LYP) Free Energy Barriers ΔG⧧
mixture resulted in higher turnover frequencies (Table 3) for
catalysts 1−5 and 10−12 compared to in D2O alone. In order
to study the effect of methanol in the reaction, five experiments
with catalyst 2, sodium formate, and NAD+ (ratio 1:25:2, pH*
7.2, 310 K) were performed in D2O/MeOD (5, 10, 23.3, 50, or
66.6% v/v). The higher percentage of methanol resulted in an
increase in the rate of the reaction and therefore higher
turnover frequency (Figure S4).
(kcal mol−1) Leading to CO2 Elimination
complex
ΔG⧧ twist
ΔG⧧ CO2 elim
10
11
12
4
5.9
4.2
4.2
13.7
12.9
12.6
18.1
DFT Calculations. A DFT mechanistic study was
performed initially with the aqua complex [(bn)Ru(MsEn)-
(H2O)]+ (10) in the presence of formate. The energy of the
formate complex plus water was found to be lower than the
energy of the aqua complex plus formate, so the formate
complex was taken as the reference point for determining
barriers to CO2 elimination.
Ru(TsEn)Cl] (8) were obtained and analyzed by X-ray
diffraction. The crystal structure of complex [(hmb)Ru(TfEn)-
Cl] (6) shows an asymmetric unit containing three crystallo-
graphically independent molecules. One of the ruthenium
molecules (6a) establishes an intermolecular hydrogen bond
with a methanol molecule (Table S4). The other two molecules
(6a′ and 6b) are enantiomers with similar bond lengths and
angles (Table 3), which are slightly different from those of
complex 6a.
Formate coordinates via a single oxygen with the carbonyl
moiety making an H-bond with an NH proton (Figure 4). It
Comparison with the analogous ethylenediamine com-
plex42,43 shows no significant difference between the Ru−
NH2 bond length and that of Ru−N− despite the higher
electron density on the imino nitrogen: 2.130(3) Å for [(p-
cym)Ru(en)Cl]+, 2.1331(15) Å for [(p-cym)Ru(MsEn)Cl]
(1), 2.109(3) Å for [(bip)Ru(en)Cl]+, and 2.096(4) and
2.1073(16) Å for complexes 8 and 7, respectively. However,
comparing the complexes containing hmb, the Ru−N− bond
length is slightly shorter (ca. 0.02−0.03 Å). The respective Ru−
Cl bond lengths in complexes 1, 5, 6, 7, and 8 (2.4146(8)−
2.4444(11) Å) are within the range displayed by other
compounds of the type [(η6-arene)Ru(en)Cl]+ (2.41−2.44
Å).42
When the Ru−N− and the Ru−Cl bond lengths of the
complexes with different η6-arene but the same N,N′ chelating
ligand (1, 7, 5, and 8) are compared, each differs by ca. 0.03
and 0.01−0.03 Å, respectively. Similarly, comparing the
influence of the chelating ligand when the same arene is
used, an increase in the electron-withdrawing power of the R2
group (Tf > Ts > Ms) appears to increase the Ru−Cl bond
length, whereas the Ru−N− bond length decreases.
Figure 4. Calculated DFT structures for the formate adduct of
{(bn)Ru(MsEn)}+ (from 10). (A) Intermediate complex with formate
bound via a single oxygen with an H-bond between the second
formate oxygen and NH from the chelating ligand. (B) Transition
state for the rotation of formate that begins to orient the hydrogen
toward the metal center. (C) Intermediate showing the hydrogen now
oriented toward the Ru center, and finally (D) the transition state for
the formation of the ruthenium−hydride bond and CO2 elimination.
Imaginary frequencies for B and D were 148i and 329i cm−1,
respectively.
Hydrolysis and Acidity of the Aqua Complex. Aquation
of the complexes in D2O was confirmed by comparison of the
1H NMR spectra of the solution of complex 2 in D2O and the
solution obtained after removal of the chloride by reaction with
silver ions (AgNO3). Hydrolysis of 2 appeared to be rapid
under biologically relevant conditions (310 K, pH* 7.2), since
equilibrium was reached by the time the first 1H NMR
spectrum was recorded (<6 min). Consequently the turnover
frequency for the transfer hydrogenation to form NAD+ does
not depend on whether the chlorido or aqua adduct is used.
then rotates via a low-energy transition state to an intermediate
with the formate H now oriented toward the Ru center. From
here, the Ru−H bond is formed via a higher energy transition
state (TS). The energetics of this process and for CO2
elimination for the formate adducts of the bn complexes
{(bn)Ru(MsEn)}+ (10), {(bn)Ru(TsEn)}+ (11), and {(bn)-
Ru(TfEn)}+ (12) and for the formate adduct of the hmb
complex {(hmb)Ru(MsEn)}+ (4) are collected in Table 4.
1
A H NMR pH* titration of 2 gave a pKa* value of 9.06 for
the aqua ligand (Figure S3). These results suggest that complex
2 would exist mainly as the aqua adduct over the pH* range
used during the experiment, which was close to physiological
pH (7.4).
Mechanism of the Catalytic Reduction of NAD+. Use of
deuterated formate and complex 3 gave a turnover frequency
2.4 times smaller than that with formate. Since the rate of
formation of the hydride (β-elimination) with the heavier atom
is lower, the difference in the turnover frequency between the
two reactions suggests that the rate-limiting step of the reaction
is the formation of Ru−H. To determine the factors that
influence the reaction rate, a series of experiments with complex
2 were performed by varying the concentration of the different
DISCUSSION
■
Structures of the Complexes. The ligands TsEn, TfEn,
and MsEn were selected in order to compare the catalytic
properties of complexes of the type [(arene)Ru(XEn)Cl],
where X is Ms, Ts, or Tf, with an increasing electron-
withdrawing power of the sulfonamide. Enantiomerically pure
single crystals of complexes [(p-cym)Ru(MsEn)Cl] (1),
[(hmb)Ru(TsEn)Cl] (5), [(bip)Ru(MsEn)Cl] (7), and [(bip)-
G
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 2. Proposed Mechanism for Reduction of NAD+ to 1,4-NADH via Transfer Hydrogenation Using Formate as the
a
Hydride Source
a
R1 = Ms (SO2CH3), Ts (SO2(C6H4)CH3), or Tf (SO2(C6H4)CF3); Arene-R2 is p-cym, hmb, bip, or bn.
reactants. When the NAD+ concentration was changed, no
significant alterations in the rate were observed, implying that
the reaction rate does not depend on the NAD+ concentration.
However the reaction rate decreased with lower concentrations
of the hydride source (sodium formate), suggesting that
formate is involved in the rate-determining step (Figure 2).
The plot of turnover frequency against formate concen-
tration showed Michaelis−Menten behavior (Figure 3). The
maximum turnover frequency for complex 2 (TOFmax = 6.4
h−1) is about 4 times higher than that previously determined for
the en complex [(hmb)Ru(en)Cl]+ (TOFmax = 1.46 h−1),31 so
achieving the desired increased efficiency in the reaction.
On the basis of the data obtained, we can propose a plausible
catalytic cycle for the reduction of NAD+ via transfer
hydrogenation with formate as a hydride source (Scheme 2).
This cycle is comparable with that established by Fish et al.33
for the reduction of NAD+ via transfer hydrogenation using
[(Cp*)Rh(bipy)H2O]2+ as the catalyst and formate as hydride
source. In the first step, the precatalyst chlorido complex is
converted to the aqua adduct. The reaction of the aqua
complex with HCOO− generates the formate adduct [(arene)-
Ru(N,N′)(OHCO)]. The formate coordinates via a single
oxygen with the carbonyl moiety making an H-bond with an
NH proton. It then rotates to form an intermediate with the
formate H now oriented toward the Ru center. From here, the
formate adduct undergoes decomposition to the hydride
complex [(arene)Ru(N,N′)(H)], through a β-elimination
reaction and generation of CO2.33 Transfer of hydride to
produce 1,4-NADH regenerates the aqua adduct.
The low Michaelis constant (Km = 27.8 mM) indicates a
stronger affinity of the complex for formate compared to the
complexes [(hmb)Ru(en)Cl]+ (Km = 58 mM)31 and [(Cp*)-
Rh(bipy)H2O]2+ (Km = 140 mM),33 while the maximum
turnover frequency (TOFmax = 6.4 h−1) of the reaction
indicated an increase in the catalytic activity compared to
[(hmb)Ru(en)Cl]+ (TOFmax = 1.46 h−1),31 but lower catalytic
activity compared with the complex [(Cp*)Rh(bipy)H2O]2+
(TOF = 77.5 h−1).33 Several other Ru, Rh, and Ir complexes
have previously been reported to catalyze the regioselective
reduction of NAD+ to 1,4-NADH using formate as a hydride
source.44,45 The complexes studied in the present work show
comparable TOFs (ca. 0.2−7 h−1) to reported [(η6-arene)Ru-
(NN′)Cl] complexes (ca. 0.0056−10 h−1), although the TOFs
displayed are much lower than those for some Rh or Ir
complexes.44,45 The maximum activity for complex 2 was
observed at pH* 7.2 (Figure S3). At higher pH* the
concentration of OH− inhibits the reaction, and at pH* 10
or higher the complex decomposes. The experiments with
complexes 1−12 as catalysts were performed in D2O and in a
mixture of D2O/methanol (9:2 v/v). The addition of methanol
resulted in an increase in the turnover frequency compared to
in D2O alone.
The DFT energetics correlate well with other experimental
data and predict increasing barriers (and hence slower
reactions) for bn complexes in the order TfEn (12) < TsEn
(11) < MsEn (10), although it should be noted that the
difference between barriers is small (ca. 1.1 kcal mol−1). For
MsEn, the barrier for {(hmb)Ru(MsEn)}+ (4) is substantially
larger than for {(bn)Ru(MsEn)}+ (10), which also agrees with
H
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
experiment. Overall, the computational protocol appears to
provide a very good description of these systems. DFT thus
could provide, in the future, a useful tool to help the a priori
design of new complexes with desired properties.
< hmb, both consistent with the observed rates. Finally
optimum pH* was found to be 7.2, which is close to
physiological pH. The series of compounds studied here
showed improved catalytic activity that is an order of
magnitude higher than that of the ethylenediamine analogues,
which could be significant if such catalytic reactions are to be
used to modulate NADH levels in cells.
Kinetics of Transfer Hydrogenation Reactions. The
catalytic reduction of NAD+ under biologically relevant
conditions (310 K, pH* 7.2) using complexes 1−12 as
catalysts and formate as a hydride source showed a general
tendency in which the more electron-withdrawing sulfonamides
on the chelating ligand achieved higher catalytic activity.
Accordingly, complexes using TfEn as chelating ligand are more
active than those with the TsEn ligand, and, in turn, both are
more active than the complexes containing the MsEn ligand.
Such a trend has been previously reported in the catalytic
transfer hydrogenation of aldehydes46 and quinoxalines.19
The nature of the arene had a dramatic effect on the reaction
rate. Electron-poor arene ligands such as benzene are reported
to accept electron density from the central RuII. Consequently
the lower electron density on the RuII center for complexes
[(bn)Ru(N,N′)Cl] (10, 11, and 12) compared to complexes
[(hmb)Ru(N,N′)Cl] (4, 5, and 6) stabilizes the coordination
of the negatively charged formate to Ru, making the hydride
transfer from formate more difficult.47,48 However, the
experimental data show higher catalytic activity when the
arene is more electron deficient (Table 3), [(bn)Ru(N,N′))Cl]
being the more active complexes while [(hmb)Ru(N,N′)Cl]
are less active. This unexpected trend may be due to steric
hindrance since the more sterically hindered arene (hmb)
shows lower activity while the less sterically hindered arene
(bn) shows higher activity.
ASSOCIATED CONTENT
■
S
* Supporting Information
Ligand synthesis, titration of complex 2 (Figure S2), pH-
dependent study of the reaction (Figure S3), methanol effect
(Figure S4), and crystallographic data (Figure S1, Tables S1−
S6). This material is available free of charge via the Internet at
http://pubs.acs.org. X-ray crystallographic data in CIF format
are available from the Cambridge Crystallographic Data Centre
(http://www.ccdc.cam.ac.uk).
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: p.j.sadler@warwick.ac.uk.
Author Contributions
§These authors contributed equally.
Funding
We thank ERDF/AWM (Science City), NWO (Rubicon
grant), EPSRC (grant no. EP/F042159/1), and ERC (grant
no. 247450) for support of this work, and J.A.S.C. for a
scholarship for J.J.S.B.
Notes
The authors declare no competing financial interest.
CONCLUSIONS
■
In this work we have studied a series of water-soluble RuII half-
sandwich complexes of the type [(η6-arene)Ru(N,N′)(Cl)]
where N,N′ are monosulfonamide chelating ligands derived
from ethylenediamine. The X-ray crystal structures of
complexes 1 and 5−8 showed the familiar piano stool geometry
common to all [(η6-arene)Ru(X,Y)(L)] compounds. Partic-
ularly interesting is the effect of the more electron-withdrawing
sulfonamide groups from the chelating ligand on weakening the
Ru−Cl bond. Longer Ru−Cl distances appear to correlate with
an increase in catalytic activity. The ability of the complexes to
act as catalysts for the regioselective transfer hydrogenation of
NAD+ to form 1,4-NADH was studied. The catalytic activity
decreases in the order TfEn > TsEn > MsEn, showing that
more electron-withdrawing sulfonamides increase the activity.
The nature of the arene significantly influences the catalytic
activity. Benzene complexes are the most active, while the
activity decreases in the order biphenyl > p-cymene >
hexamethylbenzene. Complex [(bn)Ru(TfEn)Cl] (12) was
the most active, with a TOF of 10.4 h−1. Mechanistic studies on
complex 2 [(p-cym)Ru(TsEn)(Cl)] showed a dependence of
the reaction rate on formate, but not on NAD+ concentration.
The formation of the hydride complex was shown to be the
rate-limiting step through deuterium isotope experiments. DFT
calculations were performed for the coordination of formate
and subsequent CO2 elimination to generate the hydride
complex. The elimination of CO2 occurs via a two-step process
with the coordinated formate first twisting to present its
hydrogen toward the metal center followed by β-elimination
and Ru−H formation. The computed barriers for complexes
using η6-benzene follow the order MsEn > TsEn > TfEn, and
for the MsEn chelating system the barrier follows the order bn
ACKNOWLEDGMENTS
■
We thank Dr. Ivan Prokes, Dr. Lijiang Song, and Mr. Philip
Aston (University of Warwick) for their excellent assistance
with the NMR and MS instruments.
REFERENCES
■
(1) Matthew, C. K.; Ahern, K. G.; Van Holde, K. E. Biochemistry, 3rd
ed.; Pearson Educacion: Madrid, 2002.
(2) Ying, W. Antioxid. Redox Signaling 2008, 10, 179−206.
(3) Khan, J. A.; Forouhar, F.; Tao, X.; Tong, L. Expert Opin. Ther.
Targets 2007, 11, 695−705.
(4) Hileman, E. O.; Liu, J.; Albitar, M.; Keating, M. J.; Huang, P.
Cancer Chemother. Pharm. 2004, 53, 209−219.
(5) Weckbecker, A.; Gro, H.; Hummel, W. Adv. Biochem. Eng./
Biotechnol. 2010, 120, 195−242.
(6) Held, M.; Schmid, A.; B., V. B. J.; Witholt, B. Pure Appl. Chem.
2000, 72, 1337−1343.
(7) Chen, Q.; Gao, K.; Duan, Y.; Ye, Z.; Shi, L.; Yang, Y.; Zhou, Y. J.
Am. Chem. Soc. 2012, 134, 2442−2448.
(8) Dibenedetto, A.; Stufano, P.; Macyk, W.; Baran, T.; Fragale, C.;
Costa, M.; Aresta, M. ChemSusChem 2012, 5, 373−378.
(9) Kim, Y. H.; Yoo, Y. J. Enzyme Microb. Technol. 2009, 44, 129−
134.
(10) Maenaka, Y.; Suenobu, T.; Fukuzumi, S. J. Am. Chem. Soc. 2012,
134, 367−374.
(11) Romain, C.; Gaillard, S.; Elmkaddem, M. K.; Toupet, L. c.;
Fischmeister, C. d.; Thomas, C. M.; Renaud, J.-L. Organometallics
2010, 29, 1992−1995.
́
́ ́
(12) Carrion, M. C.; Sepulveda, F.; Jalon, F. l. a.; Manzano, B. R.;
Rodríguez, A. M. Organometallics 2009, 28, 3822−3833.
(13) Wolfson, A.; Dlugy, C.; Shotland, Y.; Tavor, D. Tetrahedron Lett.
2009, 50, 5951−5953.
I
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(14) Grau, M. M.; Poizat, M.; Arends, I. W. C. E.; Hollmann, F. Appl.
Organomet. Chem. 2010, 24, 380−385.
(47) Wang, F.; Habtemariam, A.; van der Geer, E. P. L.; Fernandez,
́
R.; Melchart, M.; Deeth, R. J.; Aird, R.; Guichard, S.; Fabbiani, F. P.;
Lozano-Casal, P.; Oswald, I. D. H.; Jodrell, D. I.; Parsons, S.; Sadler, P.
J. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 18269−18274.
(15) Matharu, D. S.; Morris, D. J.; Clarkson, G. J.; Wills, M. Chem.
Commun. 2006, 3232−3234.
(48) Clayden, J.; Greeves, N.; Warren, S.; Wothers, P. Organic
Chemistry, 1 ed.; Oxford University Press: Oxford, 2001.
(16) Noyori, R. Angew. Chem., Int. Ed. 2002, 41, 2008−2022.
(17) Fish, R. H. Aust. J. Chem. 2010, 63, 1505−1513.
(18) Leiva, C.; Lo, H. C.; Fish, R. H. J. Organomet. Chem. 2010, 695,
145−150.
(19) Tan, J.; Tang, W.; Sun, Y.; Jiang, Z.; Chen, F.; Xu, L.; Fan, Q.;
Xiao, J. Tetrahedron 2011, 67, 6206−6213.
(20) Watanabe, M.; Kashiwame, Y.; Kuwata, S.; Ikariya, T. Eur. J.
Inorg. Chem. 2012, 504−511.
(21) Zhou, Z.; Ma, Q.; Zhang, A.; Wu, L. Appl. Organomet. Chem.
2011, 556−561.
(22) Wang, C.; Villa-Marcos, B.; Xiao, J. Chem. Commun. 2011, 47,
9773−9785.
(23) Thai, T.; Therrien, B.; Suss-fink, G. J. Organomet. Chem. 2009,
̈
694, 3973−3981.
(24) Sandoval, C. A.; Bie, F.; Matsuoka, A.; Yamaguchi, Y.; Naka, H.;
Li, Y.; Kato, K.; Utsumi, N.; Tsutsumi, K.; Ohkuma, T.; Murata, K.;
Noyori, R. Chem. Asian J. 2010, 5, 806−816.
(25) Creutz, C.; Chou, M. H. J. Am. Chem. Soc. 2009, 131, 2794−
2795.
(26) Robertson, A.; Matsumoto, T.; Ogo, S. Dalton Trans. 2011, 40,
10304−10310.
(27) Martins, J. E. D.; Clarkson, G. J.; Wills, M. Org. Lett. 2009, 11,
847−850.
(28) Li, X.; Li, L.; Tang, Y.; Zhong, L.; Cun, L.; Zhu, J.; Liao, J.;
Deng, J. J. Org. Chem. 2010, 75, 2981−2988.
(29) Wu, J.; Wang, F.; Ma, Y.; Cui, X.; Cun, L.; Zhu, J. Chem.
Commun. 2006, 1766−1768.
(30) Zhou, Z.; Ma, Q.; Sun, Y.; Zhang, A. Heteroat. Chem. 2010, 21,
505−514.
(31) Yan, Y. K.; Melchart, M.; Habtemariam, A.; Peacock, A. F.;
Sadler, P. J. J. Biol. Inorg. Chem. 2006, 11, 483−488.
(32) Ruppert, R.; Herrmann, S.; Steckhan, E. Chem. Commun. 1988,
1, 1150−1151.
(33) Lo, H. C.; Leiva, C.; Buriez, O.; Kerr, J. B.; Olmstead, M. M.;
Fish, R. H. Inorg. Chem. 2001, 40, 6705−6716.
(34) Habtemariam, A.; Melchart, M.; Fernandez, R.; Parsons, S.;
Oswald, I. D. H.; Parkin, A.; Fabbiani, F. P.; Davidson, J. E.; Dawson,
A.; Aird, R. E.; Jodrell, D. I.; Sadler, P. J. J. Med. Chem. 2006, 49,
6858−6668.
(35) Sheldrick, G. M. SHELXL97; University of Gottingen:
Gottingen, Germany, 1997.
(36) Sheldrick, G. M. Acta Crystallogr. A 2007, 64, 112−122.
(37) Neese, F. WIREs Comput. Mol. Sci. 2012, 2, 73−78.
(38) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97,
̈
2571−2577.
(39) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
(40) Sinnecker, S.; Rajendran, A.; Klamt, A.; Diedenhofen, M.;
Neese, F. J. Phys. Chem. A 2006, 110, 2235−2245.
(41) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H.; Grimme, S.;
Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104−
154117.
(42) Morris, R. E.; Aird, R. E.; Murdoch, S.; Chen, H.; Cummings, J.;
Hughes, N. D.; Parsons, S.; Parkin, A.; Boyd, G.; Jodrell, D. I.; Sadler,
P. J. J. Med. Chem. 2001, 6, 3616−3621.
́
(43) Peacock, A. F. A.; Habtemariam, A.; Fernandez, R.; Walland, V.;
Fabbiani, F. P. A.; Parsons, S.; Aird, R. E.; Duncan, I.; Sadler, P. J. J.
Am. Chem. Soc. 2006, 128, 1739−1748.
̌
̌ ̌
(44) Canivet, J.; Suss-Fink, G.; Stepnicka, P. Eur. J. Inorg. Chem.
̈
2007, 4736−4742.
(45) Haquette, P.; Talbi, B.; Barilleau, L.; Madern, N.; Fosse, C.;
Salmain, M. Org. Biomol. Chem. 2011, 9, 5720−5727.
(46) Wu, X.; Liu, J.; Li, X.; Zanotti-Gerosa, A.; Hancock, F.; Vinci,
D.; Ruan, J.; Xiao, J. Angew. Chem., Int. Ed 2006, 45, 6718−6722.
J
dx.doi.org/10.1021/om3006307 | Organometallics XXXX, XXX, XXX−XXX