Hydride Reduction of NAD(P)+ Model Compounds with a Ru(II)-Hydrido Complex

Article abstract of DOI:10.1021/acs.organomet.5b00713

In order to better understand the regioselective hydride transfer of metal hydrido complexes to NAD(P)⁺ model compounds, reactions of [Ru(tpy)(bpy)H]⁺ (Ru-H: tpy = 2,2′:6″,2″-terpyridine, bpy = 2,2′-bipyridine) with various substitue

Full text of DOI:10.1021/acs.organomet.5b00713

Article
pubs.acs.org/Organometallics
Hydride Reduction of NAD(P)+ Model Compounds with a
Ru(II)−Hydrido Complex
Kichitaro Koga,† Yasuo Matsubara,‡ Tatsumi Kosaka,§ Kazuhide Koike,∥ Tatsuki Morimoto,⊥
and Osamu Ishitani*,†
†Department of Chemistry, Graduate School of Science and Engineering, Tokyo Institute of Technology, O-okayama 2-12-1-NE-1,
Meguro-ku, Tokyo 152-8550, Japan
‡Department of Material and Life Chemistry, Kanagawa University, Rokkakubashi 3-27-1, Kanagawa-ku, Yokohama 221-8686, Japan
§Graduate School of Science and Engineering, Saitama University, 255 Shimo-Okubo, Saitama 338-8570, Japan
∥National Institute of Advanced Industrial Science and Technology, Onogawa 16-1, Tsukuba 305-8569, Japan
⊥Department of Applied Chemistry, School of Engineering, Tokyo University of Technology, 1401-1 Katakura, Hachioji,
Tokyo 192-0982, Japan
S
* Supporting Information
ABSTRACT: In order to better understand the regioselective
hydride transfer of metal hydrido complexes to NAD(P)+
model compounds, reactions of [Ru(tpy)(bpy)H]+ (Ru-H: tpy =
2,2′:6″,2″-terpyridine, bpy = 2,2′-bipyridine) with various
substituent NAD(P)+ model compounds were investigated in
detail. All of the NAD(P)+ model compounds accepted hydride
from Ru-H, yielding 1:1 adducts, where the dihydro form(s) of
the model compounds coordinated with the carbamoyl group
to the Ru(II) center of [Ru(tpy)(bpy)]2+, with very different
reaction rates. Some reactions produced the adduct with only
the 1,4-dihydro structure, whereas others produced a mixture
of two adducts, with a 1,4- or 1,2-dihydro structure. In particu-
lar, temperature-dependent adduct formation kinetics studies
provided important information on the transition state(s) of the hydride transfer reactions and factors for determining the
regioselectivity. Most adducts were cleaved to the corresponding free dihydro product(s) with the same distribution of the
regioisomers to the adduct(s).
L = 4-(1H-pyrazol-1-yl-κN2)-benzoate-κC3) can serve as
INTRODUCTION
■
In various biological redox reactions, the coenzyme NAD(P)+
catalysts along with formate or H2 as reductants, selectively
producing the corresponding 1,4-NAD(P)H models. In these
serves as an oxidant and accepts a hydride from various sub-
catalytic reactions, the metal hydrido complexes were presumed
strates, including water and ethanol. This process selectively
to be intermediates, which provided a hydride to the NAD(P)+
yields the 1,4-dihydro product NAD(P)H in the presence
model compound.5c,d Although the production selectivity of
the 1,4-dihydro form during the hydride-transfer process is
assumed to result from the interaction between the carbamoyl
group at the C-3 position of the NAD(P)+ model compound
and the metal center in the transition state,5d,6 direct evidence
and detailed properties of the intermediates have not been
reported to date. We have reported on the photocatalytic and
regioselective hydride reduction of a typical NAD(P)+ model
compound, 1-benzyl-3-carbamoylpyridinium cation (1a), using
[Ru(tpy)(bpy)(NEt3)]2+ (tpy = 2,2′:6″,2″-terpyridine) as a
photocatalyst and NEt3 as a reductant.6,7 In the beginning of
the reaction, [Ru(tpy)(bpy)H]+ is formed via the photoexcita-
tion of [Ru(tpy)(bpy)(NEt3)]2+ (process 1 in Scheme 1).7c,d
of enzymes, including the Fd-NADP+ reductase1 and alcohol
dehydrogenase.2 Alternatively, in artificial reduction systems
without the enzymes, the corresponding NAD(P)+ model com-
pounds are reduced by hydride donors such as NaBH4, to not
only the 1,4-dihydro form but also its structural isomers, i.e.,
1,2- and 1,6-dihydro products (eq 1).3
It is known that NAD(P)+ model compounds can be re-
gioselectively reduced to the corresponding 1,4-dihydro form
via metal hydrido complexes.4,5 For example, [Rh(Cp*)
5c,d
4b,c
(bpy)(H2O)]2+
and [Ir(Cp*)(L)(H2O)]2+
(Cp* =
Received: August 18, 2015
1,2,3,4,5-pentamethylcyclopentadiene, bpy = 2,2′-bipyridine,
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
unsaturated substrate. To use the NAD(P)H model com-
pounds as catalysts, new systems that can recover the NAD(P)
H model compounds with the same chirality by hydride reduc-
tion of the oxidized model compounds are required. Although
the metal hydrido complexes are candidates as the reductant
because they can (photo)catalytically supply a hydride for the
NAD(P)+ model compounds, as described above, reactions
of the metal hydrido complexes with the NAD(P)+ model
compounds that exhibit a substituent at the C-4 position and
various substituents at the N-1 and/or C-3 positions have not
been reported.
Scheme 1. Reaction Mechanism of the Photocatalytic
Reduction of 1a using [Ru(tpy)(bpy)(NEt3)]2+ as a
Photocatalyst and NEt3 as a Reductant6,7
We herein report reactions of Ru-H with various NAD(P)+
model compounds comprising a methyl group at the C-4 posi-
tion and/or substituents at other position(s) to understand
their steric and electronic effects on the regioselectivity and rate
of hydride transfer (Chart 1).
Chart 1. Structures of NAD(P)+ Model Compounds
The product Ru-H rapidly reacts with 1a to give the corre-
sponding 1,4-dihydro product. The reaction between the hydrido
complex Ru-H and 1a was investigated in detail.6,7d In the initial
reaction step, the 1:1 adduct 2a(1,4) producing the 1,4-dihydro
form 3a(1,4) coordinated with its carbamoyl group to the Ru(II)
center of [Ru(tpy)(bpy)]2+ (process 2). 2a(1,4) quantitatively
cleaved at a much slower rate, yielding free 3a(1,4). [Ru(tpy)-
(bpy)(NEt3)]2+ was recovered at the same time (process 3).
The observation that the carbamoyl group coordinated with
the metal center in the intermediate 2a(1,4) is the first clear
evidence that the interaction between the carbamoyl group
and the metal center plays an important role in the hydride
transfer reaction transition state, particularly for inducing the
regioselective formation of the 1,4-dihydro product.7b Because
a “seven-coordinated” Ru(II) complex was produced in the
transition state, where the carbamoyl group of 1a interacts with
Ru(II) and the hydride ligand interacts with the C-4 position of
the pyridinium unit, the hydride could not transfer to the C-6
position, which is farther from the carbamoyl group in com-
parison with the C-4 position. However, it is unclear why the
hydride did not transfer to the C-2 position of the pyridinium
unit of 1a, which is located approximately at the same distance
from the carbamoyl group as the C-4 position. The addition of
steric and/or electronic perturbations to the NAD(P)+ model
compound could help in clarifying this. However, the reactions
of NAD(P)+ model compounds with substituent groups at the
C-2 and/or C-4 positions with Ru-H have not been reported.
Many asymmetric reduction systems of various carbonyl
compounds using chiral NAD(P)H model compounds with
a methyl group at the C-4 position as a reductant have been
reported.8 For example, the NAD(P)+ model compounds
shown in eqs 28a and 38e could transfer the hydride to the sub-
strates in high optical yields.
RESULTS
■
Product Analyses. As an example, in a typical reaction,
Ru-H (8.0 mM) and 1b (8.4 mM) were mixed in MeCN-d3
and monitored using 1H NMR (Figure 1). The signals of Ru-H
Figure 1. 1H NMR spectra of an MeCN-d3 solution containing Ru-H
(8.0 mM) and 1b (8.4 mM), which were kept at room temperature in
○ ●
the dark for (a) 20 min and (b) 20 h. The peaks marked with
and * are attributed to 2b(1,4), 3b(1,4), water, and the solvent,
respectively.
,
, †,
completely disappeared 20 min after mixing. Alternatively, the
1:1 adduct 2b(1,4), where 1b was selectively reduced to the
corresponding 1,4-dihydro product 3b(1,4) and coordinated
However, these reactions require an equimolecular amount
of the NAD(P)H model compounds in comparison with the
B
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
with the Ru(II) center with the carbamoyl group at the C-3
position of the dihydropyridine moiety, free 3b(1,4) and the
solvent complex [Ru(tpy)(bpy)(MeCN-d3)]2+ were also detec-
ted as minor products, which should form via cleavage of the
1:1 adduct 2b(1,4) (eq 4). The reaction solution was held at
and the corresponding free dihydropyridines, with a ratio of
91:9 between 3d(1,4) and 3d(1,2), respectively.9
Similar experiments were conducted using different NAD(P)+
model compounds. The results are summarized in Table 1. All of
the NAD(P)+ model compounds, which were only converted to
the 1:1 adduct with the 1,4-dihydropyridine moiety, had a benzyl
group at the N-1 position of the pyridinium moiety (1a−c). The
only exception was 1d, which had a benzyl group at the N-1
position and methyl groups both at the C-4 position and at the
N atom of the carbamoyl group; this product was converted to a
1:1 adduct with either a 1,4- or a 1,2-dihydropyridine moiety
(2g(1,4), 2g(1,2)) with a 91:9 ratio. In the NAD(P)+ models
with a methyl group (1e) or a 2,2,2-trifluoroethyl group (1f−h),
the mixtures of the two 1:1 adducts, with a 1,4- or 1,2-dihydro-
pyridine moiety, were produced in different ratios. In the cleavage
reaction of the 1:1 adduct(s), isomerization of the dihydro-
pyridine moiety was not observed, except for the reaction of 1g and
Ru-H. The ratio of 2g(1,4) to 2g(1,2) was 29:71 immediately
after mixing 1g and Ru-H. However, the ratio of 3g(1,4) to
3g(1,2) was 43:57 after the cleavage reaction was completed.
Kinetic Analysis of the Reactions of Ru−H with the
NAD(P)+ Model Compounds. The reactions of Ru-H with
the NAD(P)+ model compounds were followed with the stopped-
flow method. As an example, Figure 3 shows the change in the
UV−vis absorption spectra of a DMF solution containing Ru-H
(0.031 mM) mixed with another DMF solution containing an
excess amount of 1b (3.8 mM) at 300 K. During the first 200 s
after mixing (Figure 3a), a set of isosbestic points was observed at
419 and 502 nm, with a decrease in the absorption band at λmax
535 nm attributed to Ru-H and an increase in another absorption
band at λmax 490 nm. As described above, 1H NMR of the reaction
solution clearly indicated that these changes in the absorption
spectra were attributable to the formation of the 1:1 adduct
2b(1,4). Figure 3b shows the spectral changes for longer reaction
times, up to 24 h; the absorption band at λmax 490 nm shifted to
larger wavelengths, and its intensity increased. These changes are
attributed to the cleavage of 2b(1,4) to 3b(1,4) and [Ru(tpy)-
(bpy)(DMF)]2+.
room temperature for 20 h, resulting in the cleavage of 2b(1,4)
to give 3b(1,4) and [Ru(tpy)(bpy)(MeCN-d3)]2+ in a
quantitative yield. Hence, the respective 1,2- and 1,6-dihydro
isomers of 2b(1,4) and 3b(1,4) were not detected at all.
Conversely, when 1d (8.4 mM) was used instead of 1b in a
similar experiment, 1:1 adducts exhibiting both the 1,4-dihydro-
pyridine moiety 2d(1,4) and 1,2-dihydropyridine moiety
2d(1,2) were produced (eq 5 and Figure 2). A reaction time
Figure 4 shows the change in the absorbance at 535 nm,
which enables analysis of the 2b(1,4) reaction rate of forma-
tion. This decay curve can be fitted with pseudo-first-order
kinetics (eq 6; the derivation is shown in Supporting Information,
and the same experiments were repeated six times for each
temperature). The pseudo-first-order reaction rate constant (kobs)
was determined as (1.18 0.01) × 10−2 s−1. Using eq S2 in the
Supporting Information and the values kobs = (1.18 0.01) ×
10−2 s−1 and the initial concentration of 1b [1b]int = 1.9 mM, the
second-order reaction rate constant of formation of 2b(1,4) was
calculated as k = 6.20 0.06 M−1 s−1.11
A = (εRu‐H − ε2b(1,4))[Ru‐H]inte−k + ε2b(1,4)[Ru‐H]int
obst
Figure 2. 1H NMR spectral changes in an MeCN-d3 solution
containing Ru-H (8.0 mM) and 1d (8.4 mM) at room temperature
(6)
where A, ε, t, and [Ru-H]int are the absorbance at 535 nm, the
molar extinction constant, the reaction time after mixing, and the
initial concentration of Ru-H, respectively.
○
△
after 25 min (a−c) and 20 h (d−f) (red , 2d(1,2); black , 2d(1,4);
▲
●
red , 3d(1,2); black , 3d(1,4)). The peaks marked with § and †
were attributed to CH2 at the benzyl group of 1d and the hydrido
ligand of Ru-H, respectively.
Similar measurements at different reaction temperatures in
the range 280−320 K gave linear Eyring plots, as shown in
Figure 5: y = −(4.29 0.06)x + (10.4 0.2). The activation
parameters, ΔH⧧ and ΔS⧧, were found to be 35.7 0.4 kJ mol−1
of 6 h was required to completely remove the signals for 1d and
Ru-H when the product ratios did not change from the initial
state to the end: i.e., [2d(1,4)]:[2d(1,2)] = 91:9. The reaction
mixture was kept at room temperature for 1 day, resulting in the
cleavage of both 1:1 adducts, yielding [Ru(tpy)(bpy)(MeCN-d3)]2+
and −111
1 J mol−1 K−1, respectively. Similar experiments
were conducted using the other NAD(P)+ model compounds
(Figures S8 and S10−S14 in the Supporting Information),
except for 1d, because the formation rate of the adducts was
C
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. Changes in UV−vis spectra after mixing a DMF solution
containing Ru-H (0.031 mM) and a DMF solution containing 1d
(3.8 mM) at 300 K (a) for 0−200 s and (b) for 3−24 h.
Figure 4. Decay of the absorbance measured at 535 nm after a DMF
solution containing Ru-H (0.031 mM) and a DMF solution containing
○
1b (3.8 mM) were mixed at 300 K ( ). The red solid curve is the
result of fitting with eq 6 (y = 0.0793 × e−0.0120t + 0.0777).
Figure 5. Eyring plots for the formation of 2b in DMF measured at
280−320 K. The same measurements were repeated 6 times.
too slow to be measured with the stopped-flow method.
The derived k values and the activation parameters are sum-
marized in Table 1.
In the case of 1d, the adduct formation reaction was much
slower, while the rate of solvolysis was not much changed: this
leads to competing kinetics that could be monitored on the
NMR time scale. This gave us a chance to follow the formation
processes of each adduct, 2d(1,2) or 2d(1,4), separately using
1H NMR. Figure 6 shows time-dependent changes in the sub-
strates, adducts, and final products when reacting Ru-H (8.0 mM)
with an equal amount of 1d in DMF-d7 at 300 K. They can be
reasonably modeled with a global fitting method on the basis of
D
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
with the methyl group at the C-4 position, the reaction rate was
much slower than that in the case of 1a by 1/470 at 300 K. In
the case of 1c, the reaction rate was reduced by 1/5 in
comparison to that in the case of 1a. It has been proposed that
the reaction of Ru-H with 1a proceeds via the transition state
with a seven-coordinated six-membered-ring structure, where
the carbamoyl group interacts with the Ru center and the
hydride ligand interacts with the carbon atom at the C-4
position (Scheme 3).10,7b,11 For 2b(1,4) formation, ΔH⧧ was
Scheme 3. Schematic Structures in the Transition State in
a
the Reaction of Ru−H with 1a
●
●
□
Figure 6. Time dependencies of Ru-H (black ), 2d(1,2) (green ),
□
●
2d(1,4) (red ), 3d(1,2) (green ), and 3d(1,4) (red
)
concentrations in the reaction of Ru-H (8.0 mM) and 1d (8.0 mM)
in DMF-d7 at 300 K. The fitting curves using eqs 7−11 are also shown.
Scheme 2. Reaction of Ru-H with 1d in DMF-d7
a
The difference in the structures is the coordination atom of 1a to the
Ru center: oxygen in I(a-1) and nitrogen in I(a-2).10
much larger than that for 2a(1,4), by 11.8
0.5 kJ mol−1
(Table 1), mainly because of the electron-donating property
of the methyl group, which should be the major reason the
reaction between 1b and Ru-H was much slower. In con-
trast, the difference in the entropic term was relatively lower
(Δ(−TΔS⧧)2b(1,4)‑2a(1,4) = 3.6 0.5 kJ mol−1 at 300 K). On the
other hand, for 2c(1,4) formation, the increase in ΔH⧧
eqs 7−11 (Scheme 2), which gave the adduct formation rates for
2d(1,2) and 2d(1,4) as k2d(1,2) = 1.9 × 10−2 M−1 s−1 and
k2d(1,4) = 1.8 × 10−1 M−1 s−1, respectively.
(ΔΔH⧧
= 1.6 0.4 kJ mol−1) was significantly less
d[Ru‐H]
2c(1,4)‑2a(1,4)
= −(k2d(1,4) + k2d(1,2))[Ru‐H][1d]
than that for 2b(1,4) formation (ΔΔH⧧
= 11.8
2b(1,4)‑2a(1,4)
(7)
dt
0.5 kJ mol−1), whereas the increase in − TΔS⧧ at 300 K from the
formation of 2a(1,4) was similar to that in the case of 1b
(−TΔS⧧ = 29.6 0.3 kJ mol−1 for the formation of 2a(1,4);
d[2d(1,4)]
= k2d(1,4)[Ru‐H][1d] − k3d(1,4)[2d(1,4)]
dt
33.2
0.4 kJ mol−1 for 2b(1,4); 32.0
0.2 kJ mol−1 for
(8)
2c(1,4)). The introduction of the methyl groups into the car-
bamoyl group at the C-3 position should induce both an increase
in the nucleophilicity of the amide group and a decrease in the
positive charge of the carbon atom at the C-4 position of
the pyridinium ring. The former should be advantageous for
transition state formation in the reaction of Ru-H with 1c,
whereas the latter should be disadvantageous. The introduction
of the methyl group(s) into the carbamoyl group or into the C-4
position caused an increase in −TΔS⧧ (Table 1). This is likely
due to steric restrictions of the methyl group(s) on the seven-
coordinated transition-state structure (Scheme 3).
We already clarified that the N atom of the carbamoyl
group coordinates the central ruthenium ion in the oxidized
adduct which was constructed with the deprotonated 1a and
[Ru(tpy)(bpy)]2+ using spectroscopic and X-ray crystallographic
analyses; on the other hand, the O atom of the deprotonated
carbamoyl group coordinates the central ruthenium ion in the
reduced adduct which was constructed with the deprotonated
3a(1,4) and [Ru(tpy)(bpy)]2+, i.e., deprotonated 2a(1,4).6
Unfortunately, the X-ray crystallographic data of 2a(1,4) itself
have not been obtained because of its instability and spectro-
scopic data of 2a(1,4) could not give clear evidence for deter-
mining the coordinating atom to the Ru(II) center.7b Using the
DFT method, therefore, we calculated the stabilities of “2a(1,4)”
d[2d(1,2)]
= k2d(1,2)[Ru‐H][1d] − k3d(1,2)[2d(1,2)]
dt
(9)
d[3d(1,4)]
= k3d(1,4)[2d(1,4)]
(10)
dt
d[3d(1,2)]
= k3d(1,2)[2d(1,2)]
(11)
dt
Eyring plots of 2d(1,2) and 2d(1,4) formation showed
good linearity at 300−320 K (Figure S15 in the Supporting
Information), which was independently used to derive the
activation parameters for both adducts (Table 1).
DISCUSSION
■
The reaction of Ru-H with 1b,c, which have methyl group(s)
either at the C-4 position or on the N atom of the carbamoyl
group at the C-3 position, gave only the adduct with the 1,4-
dihydropyridine structure, without the formation of its posi-
tional isomers: i.e., the corresponding adducts with the 1,2- or
1,6-dihydro structure (Table 1 and eq 4). These reaction rates
were slower than the reaction rate of Ru-H with 1a, although
the product distribution was the same. In the case of 1b, particularly
E
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
and “2c(1,4)” with a Ru−O or Ru−N bond as typical examples
with a carbamoyl group or a dimethyl carbamoyl group (Figure S16
in the Supporting Information); this indicated that the O atom
is clearly more favorable for coordination to the Ru center than
that with the N atom in the case of 2c(1,4) with the dimethyl
carbamoyl group because of steric hindrance of the methyl groups
on the N atom (ΔΔG(Ru−N)−(Ru−O) = 74.1 kJ mol−1). In the case
of 2a(1,4) with the carbamoyl group, on the other hand, the N
coordination was thermodynamically more stable even though
the difference between the two coordination modes was much
smaller (ΔΔG(Ru−N)−(Ru−O) = −9.66 kJ mol−1). These results
suggest that the coordination mode between the Ru center and
the NAD(P)H model in 2, and possibly in the transition state,
might change depending on the structure of the carbamoyl group.
In the case of 1d with the methyl groups in both the C-4
position and the carbamoyl group, the main product was still
the 1,4-dihydro adduct 2d(1,4). However, 2d(1,2) with the
1,2-dihydro pyridine structure was also produced in a 9% yield
(eq 12).
explains the decrease in the reaction rate in comparison with
that for 1c because of the stronger electron donating ability of
the methyl group in comparison to the benzyl group. Note that
the 2e(1,2) adduct formed even though the corresponding
adduct with the 1,2-dihydropyridine moiety was not produced
from 1c and that the introduction of the methyl group to the
N-1 position should reduce the hydride-accepting ability of the
carbon at the C-2 position, which is closer to the N-1 position
than to the C-4 position. This result strongly indicates that a
decrease in steric hindrance of a substituent at the N-1 position
enables hydride transfer to the C-2 position. In other words, the
steric hindrance of the benzyl group at the N-1 position plays
an important role in the 1,4-selectivity during the hydride transfer
reaction of Ru-H to the NAD(P)+ model compounds. This was
also supported by the fact that the decrease of ΔS⧧ in the
reaction of Ru-H with 1e (−(9.57 0.07) × 10 J mol−1 K−1)
was relatively smaller than that with 1c (−(1.07
0.01) ×
102 J mol−1 K−1). Furthermore, we never detected the adduct
with the 1,2-dihydropyridine moiety of the reaction between
Ru-H and 1b, although the methyl group was at the C-4 posi-
tion of 1b, which significantly lessens the hydride-accepting
ability of the C-4 position. One of the main reasons for this is
likely the steric hindrance of the benzyl group at the N-1
position.
In the reaction between Ru-H and 1f with a 2,2,2-trifluo-
roethyl group, which is an electron-withdrawing group, at the
N-1 position and two methyl groups at the carbamoyl group,
the 1,2-dihydro adduct was produced along with the 1,4-dihydro
adduct in the ratio [2f(1,4)]:[2f(1,2)] = 73:27 (eq 14).
Because the reaction of Ru-H with 1d was very slow, the
formation of these adducts could be separately monitored by
1H NMR. The main reason for the much slower rates than that
of 1a−c is the drastic increase in ΔH⧧ (Table 1). Interestingly,
however, the decrease of the activation entropies of 2d(1,4)
(ΔS⧧ = −(6.76 0.34) × 10 J mol−1 K−1) and 2d(1,2) (ΔS⧧ =
−(7.46 0.52) × 10 J mol−1 K−1) were much smaller than the
activation entropies of 2a(1,4), 2b(1,4), and 2c(1,4) (ΔS⧧ ≈
−100 J mol−1 K−1). This indicates that approach of Ru-H to 1d
to form the transition state was much more difficult than that to
1a−c because of synergistic effects of the steric hindrances of
the methyl groups introduced into both the C-4 position and
the carbamoyl group. The benzyl group at the C-1 position also
had an important steric influence on transition state formation,
which is discussed in more detail later. This suggests that the
interactions of the hydride ligand with the C-4 position and
the carbamoyl group with the Ru center, in the transition state
between the reaction of Ru-H and 1d, proceed at a distance
longer than that in the cases of 1a−c. This process likely
governs the significantly greater adduct formation activation
enthalpies from 1d in comparison to 1b. The 1d transition
states probably have looser conformations, which likely explain
the loss in the hydride transfer C-4 position selectivity into the
pyridinium moiety: i.e., the production of the byproduct 2d(1,2).
Some NAD(P)+/NAD(P)H model compounds exhibiting a
benzyl group at the N-1 position, e.g., 1a, so-called BNA+, have
been used in many studies. To elucidate the substituent effect at
the N-1 position, the reaction between Ru-H and 1e, with
methyl groups at both the N-1 and carbamoyl group, was
compared with that between Ru-H and 1c, which has a benzyl
group at the N-1 position and a methyl group at the car-
bamoyl group. The reaction with 1e produced 2e(1,4) with a
1,4-dihydropyridine moiety as a main product. However, the
isomer 2e(1,2) with a 1,2-dihydropyridine moiety also formed
as a byproduct. The ratio between the two during the reaction
was [2e(1,4)]:[2e(1,2)] = 96:4 (eq 13).
The reaction rate with 1f was 21 times faster than that with
1c at 300 K. The addition of the trifluoroethyl group to the
N-1 position, instead of the benzyl group, should increase the
hydride accepting ability, mainly of the C-2 position. This
caused ΔH⧧ to be lower by 1.6 0.3 k Jmol−1 in the case of 1f.
The ower − TΔS⧧ value in the reaction of 1f (25.9 0.2 J mol−1)
in comparison with that of 1c (32.0
0.2 J mol−1) more
effectively promoted the increase in the reaction rate at 300 K. An
increase in a degree of freedom in the transition state, which can
give both 2f(1,4) and 2f(1,2), should increase the activation
entropy of the reaction. A similar analysis can be more clearly
conducted in a comparison of reactivity between 1f and 1a.
Although ΔH⧧ of the reaction between Ru-H and 1f was as large
as that for 1a, the reaction rate with 1f was 4 times faster than that
with 1a at 300 K. This can be explained if we consider that the
adduct with the 1,4-dihydropyridine moiety was only produced in
the case of 1a. For 1f, the adduct having the 1,2-dihydropyridine
moiety was produced in approximately 30% yield, which reduced
−TΔS⧧.
The main product was the 1,2-adduct 2g(1,2) in the case of
1g, which exhibited a trifluoroethyl group at the N-1 position
and a methyl group at the C-4 position (eq 15). This high
2g(1,2) yield may be caused by a number of reasons, including
less steric hindrance of the trifluoroethyl group at the N-1
position, the electron-withdrawing ability of the trifluoroethyl
The reaction rate for 1e was approximately 1/8 slower at
300 K in comparison to that for 1c. An increase in ΔH⧧ by
8.5 0.3 kJ mol−1 in the reaction between Ru-H and 1e mainly
F
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
several hours, up to around 20 h. Initially, only 2(1,4) was
produced from 1a−c, following which only the corresponding
1,4-dihydronicotinamide 3(1,4) was obtained (eq 17). Other
cases (1d−f,h) produced both the 1,2- and 1,4-adducts. The
corresponding 1,2- and 1,4-dihydronicotinamides 3(1,2) and
3(1,4) were produced in the same relative amounts as the adducts
(Table 1). Therefore, during the cleavage reactions, isomerization
of the dihydronicotinamide moiety should not proceed. Only one
exceptional case was the reaction of 1g with Ru-H, whereby the
ratio of the adducts was [2g(1,4)]:[2g(1,2)] = 29:71, whereas
[3g(1,4)]:[3g(1,2)] = 43:57 (Figure 7). Because these ratios did
group, and steric and electronic effects of the methyl group
at the C-4 position, which should cause a slower hydride
transfer rate to the C-4 position. The transition state leading to
the 1,2-dihydro adduct likely had a more compact structure
than the transition states leading to 2d(1,4) and 2d(1,2)
formation, owing to the lesser steric hindrance of the
trifluoroethyl group in comparison to that of the benzyl group,
which probably enhances the −TΔS⧧ value in the 1g reaction in
comparison with that in the 1d reaction. Excluded volumes of
two model compounds having the benzyl (1c) and CH2CF3 (1f)
groups at the 1-positions in DMF were calculated as 2573 and
1541 Å3, respectively, with the assumption that they are freely
rotating. This is also supported by the fact that the benzyl group
gave greater steric repulsion than the CH2CF3 group.
The introduction of another methyl group to the C-2
position drastically affected the product distribution. For the 1h
molecule with methyl groups at the C-2, C-4, and carbamoyl
group positions, where the difference in comparison with 1g
was just the methyl substituent at the C-2 position, the main
product was the 1,4-dihydro adduct with 96% selectivity (eq 16).
The introduction of an electron-donating methyl group made the
hydride insertion less feasible, particularly to the C-2 position.
Interestingly, the −TΔS⧧ value of the reaction with 1h was
similar to that with 1g but much larger than that with 1d,
although its product distribution was similar to that with 1h.
This also suggests something about the properties of the benzyl
group at the N-1 position. For 1d, the benzyl group at the N-1
position and the methyl groups at the C-4 position and on the
carbamoyl group efficiently prevented 1d from approaching
Ru-H in the transition state, which induced a −TΔS⧧ value
significantly smaller than that for the other NAD(P)+ model
compounds. On the other hand, such an effect was not
observed in the case of 1h, which has methyl groups not only at
the C-4 position and carbamoyl group but also at the C-2
position. This strongly indicates that the benzyl group at the
N-1 position hindered the approach by Ru-H in the transition
state.
Figure 7. 1H NMR spectra of the methyl group at the C-4 position of
the dihydronicotinamide moiety at (a) 10 min and (b) 20 h after
mixing 1g (8.4 mM) with Ru-H (8.0 mM) in an MeCN-d3 solution at
△
○
▲
room temperature: (red ) 2g(1,2); (black ) 2g(1,4); (red )
●
3g(1,2); (black ) 3g(1,4).
not change during the reaction, the following two kinetic expla-
nations for the cleavage reactions of 2g(1,4) and 2g(1,2) might
be considered.
(1) Rapid isomerization between 2g(1,4) and 2g(1,2) occurred
1
within the H NMR measurements, and the cleavage reac-
tions of each adduct proceeded with different reaction rates.
(2) Following adduct cleavage, chemical equilibrium between
3g(1,4) and 3g(1,2) was achieved within the time scale
Cleavage Reaction of the Adducts 2. The 1:1 adducts
(2) produced via the reactions of Ru-H with each NAD(P)+
model compound 1 were cleaved and yielded the correspond-
ing dihydro products, and the Ru complex coordinated to the
solvent quantitatively (eqs 17 and 18).
1
of the H NMR measurements because of their rapid
isomerization.
CONCLUSION
■
Formation processes of the 1:1 adducts (2) by the reactions
of various NAD(P)+ model compounds (1) with Ru-H were
discussed in detail using production distributions, kinetics, and
activation parameters. The main reason the hydride did not
transfer to the C-6 position from Ru-H is that, in the transition
state for the production of 2, the interaction between the Ru
center and the carbamoyl group of 1 played an important role
and this interaction promotes the hydride ligand to a location
unfavorable for interaction with the carbon at the C-6 position.
One of the primary factors leading to the formation of the 1,4-
dihydro isomer of the NAD(P)H model compound (3) as the
All types of 2 produced by the reactions between 1 and Ru-H,
except 2g, were quantitatively separated into the correspond-
ing dihydronicotinamide derivative and the solvent complex for
G
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
warmed to room temperature and stirred for 2 h. The solution was
poured into 40 mL of water, followed by extraction (3 × 30 mL) with
CH2Cl2. The organic layer was dried with anhydrous sodium sulfate and
evaporated. After the solvent was evaporated, the target compound was
distilled under low pressure (111 °C at 0.3 mmHg) as a colorless oil.
The yield was 23% (276 mg, 1.68 mmol). 1H NMR (400 MHz, MeCN-
d3): δ (ppm) 8.44 (d, J = 5.0 Hz, 1H, 6-H), 8.33 (s, 1H, 2-H), 7.23 (d,
J = 5.0 Hz, 1H, 5-H), 3.05 (s, 3H, NMe2), 2.79 (s, 3H, NMe2), 2.25
(s, 3H, 4-Me). 13C NMR (101 MHz, MeCN-d3): δ (ppm) 169.2 (CO),
150.4 (6-C), 147.4 (2-C), 144.7 (4-C), 134.5 (3-C), 126.2 (5-C), 38.7
(NMe2), 34.7 (NMe2), 18.6 (4-Me).
final product is the steric hindrance of the benzyl group at the
N-1 position of 1, which prevents hydride transfer to the carbon
at the C-2 position. When a methyl group was introduced into
the pyridinium ring at the C-4 and/or C-2 positions, the reaction
rate drastically decreased, mainly because decreasing the elec-
trophilic character of the carbon at the position increased ΔH⧧
for the formation of 2. In addition, the steric hindrance of the
substituents significantly decreased the reaction rate because of
the separation of Ru-H from 1 in the transition state, and this
distance affected the product distribution in 2.
1-Benzyl-3-(N,N-dimethylcarbamoyl)-4-methylpyridinium Hexa-
fluorophosphate (1d). This compound was synthesized using a
method similar to that used for 1c, but instead of using N,N-dimethyl-
nicotinamide, 3-(N,N-dimethylcarbamoyl)-4-methylpyridine was used.
The yield was 81% (596 mg, 1.49 mmol). Mp: 190 °C dec. Anal. Calcd
for C16H19F6N2OP: C, 48.01; H, 4.78; N, 7.00. Found: C, 47.75;
EXPERIMENTAL SECTION
■
General Procedures. 1H and 13C NMR spectra were recorded on
a JEOL AL300, AL400, or ECAII400 or Bruker AC500 spectrometer.
The residual proton of the deuterated solvent was used as an internal
standard. Melting points were measured using a Stuart Scientific
Co. Ltd. SMP3 melting point apparatus. Measurements using stopped-
flow techniques were performed on a Union Giken RA-401 stopped-flow
spectrophotometer combined with an Otsuka Electronic Co. MCPD-5000
or MCPD-9000 multichannel photodiode array system. Changes in both
absorbance at 535 nm (recorded using the stopped-flow method) and
1
H,4.68; N, 7.10. H NMR (298 MHz, MeCN-d3): δ (ppm) 8.56 (d,
J = 6.6 Hz, 1H, 6-H), 8.55 (s, 1H, 2-H), 7.88 (d, J = 6.6 Hz, 1H, 5-H),
7.51−7.40 (m, 5H, Ph), 5.64 (s, 2H, 1-CH2−), 3.06 (s, 3H, NMe2),
2.79 (s, 3H, NMe2), 2.52 (s, 3H, 4-Me). 13C NMR (101 MHz, MeCN-
d3): δ (ppm) 164.3 (CO), 158.0 (4-C), 144.2 (6-C), 141.9 (2-C),
138.6 (3-C), 133.7 (Ph), 131.2 (5-C), 130.9 (Ph), 130.5 (Ph), 130.3
(Ph), 65.0 (1-CH2−), 38.7 (NMe2), 35.1 (NMe2), 20.1 (4-Me).
1-Methyl-3-(N,N-dimethylcarbamoyl)pyridinium Hexafluoro-
phosphate (1e). A 3 mL MeCN solution containing 3-(N,N-dimethyl-
carbamoyl)-4-methylpyridine (2.0 g, 13.3 mmol) and iodomethane
(3.06 g, 21.6 mmol) were refluxed for 13 h. After the solvent was
evaporated, the residue was dissolved into a small amount of methanol,
followed by dropping a saturated methanol solution of NH4PF6. The
white precipitates were recrystallized with methanol/ethanol. The
yield was 57% (2.37 g, 7.64 mmol). Mp: 124−125 °C. Anal. Calcd for
C9H13F6N2OP: C, 34.85; H, 4.22; N, 9.03. Found: C, 34.84; H, 4.00;
N, 8.94. 1H NMR (400 MHz, MeCN-d3): δ (ppm) 8.69 (s, 1H, 2-H),
8.63 (d, J = 6.4 Hz, 1H, 6-H), 8.47 (d, J = 8.1 Hz, 1H, 4-H), 8.04 (dd,
J = 6.4, 8.1 Hz, 1H, 5-H), 4.30 (d, J = 8.1 Hz, 3H, 1-Me), 3.07 (s, 3H,
NMe2), 2.94 (s, 3H, NMe2). 13C NMR (101 MHz, MeCN-d3): δ
(ppm) 164.8 (CO), 146.6 (6-C), 145.0 (2-C), 144.4 (4-C), 138.0
(3-C), 129.2 (5-C), 49.5 (1-Me), 39.6 (NMe2), 35.7 (NMe2).
1-(2,2,2-Trifluoroethyl)-3-(N,N-dimethylcarbamoyl)pyridinium
Hexafluorophosphate (1f). A 2 mL MeCN solution containing N,N-
dimethylnicotinamide (1.00 g, 6.66 mmol) and 2,2,2-trifluoroethyl
trifluoromethanesulfonate (1.90 g, 8.19 mmol) was refluxed for 12 h.
After the solvent was evaporated, the residue was dissolved into a small
amount of water, followed by dropping a saturated aqueous solution of
NH4PF6. The white precipitates were recrystallized with water. The
yield was 87% (2.19 g, 5.79 mmol). Mp: 182−183 °C. Anal. Calcd for
C10H15F6N2OP: C, 31.76; H, 3.20; N, 7.41. Found: C, 31.66; H, 3.36;
N, 7.41. 1H NMR (298 MHz, MeCN-d3): δ (ppm) 8.87 (s, 1H, 2-H),
8.80 (d, J = 6.2 Hz, 1H, 6-H), 8.70 (d, J = 8.0 Hz, 1H, 4-H), 8.22 (dd,
J = 8.0, 6.2 Hz, 1H, 5-H), 5.36 (q, J = 7.9 Hz, 2H, 1-CH2−), 3.07 (s,
3H, NMe2), 2.97 (s, 3H, NMe2). 13C NMR (101 MHz, MeCN-d3): δ
(ppm) 164.1 (CO), 147.6 (4-C), 147.5 (6-C), 146.1 (2-C), 139.0
(3-C), 130.4 (5-C), 123.0 (q, JC−F = 280 Hz, CF3), 60.6 (q, JC−F = 36
Hz, 1-CH2−), 39.7 (NMe2), 35.8 (NMe2).
1-(2,2,2-Trifluoroethyl)-3-(N,N-dimethylcarbamoyl)-4-methylpyri-
dinium Hexafluorophosphate (1g). 3-(N,N-Dimethylcarbamoyl)-4-
methylpyridine (276 mg, 2.02 mmol) was dissolved in 2,2,2-trifluo-
roethyl trifluoromethanesulfonate (1.88 mg, 8.11 mmol), and the solu-
tion was heated to 100 °C for 12 h. The following processes were the
same as those for 1f. The yield was 52% (411 mg, 1.05 mmol). Mp:
199−200 °C dec. Anal. Calcd for C17H18F9N2OP: C, 33.69; H, 3.60;
N, 7.1. Found: C, 33.68; H, 3.60; N, 7.04. 1H NMR (400 MHz,
MeCN-d3): δ (ppm) 8.61 (d, J = 6.8 Hz, 1H, 6-H), 8.60 (s, 1H, 2-H),
8.04 (d, J = 6.8 Hz, 1H, 5-H), 5.26 (q, J = 8.1 Hz, 2H, 1-CH2−), 3.09
(s, 3H, NMe2), 2.85 (s, 3H, NMe2), 2.60 (s, 3H, 4-Me). 13C NMR
(101 MHz, MeCN-d3): δ (ppm) 163.6 (CO), 161.2 (4-C), 145.8
(6-C), 143.5 (2-C), 139.0 (3-C), 132.0 (5-C), 123.1 (q, JC−F = 280 Hz,
CF3), 60.0 (q, JC−F = 35 Hz, 1-CH2−), 38.7 (NMe2), 35.2 (NMe2),
20.6 (4-Me).
1
peaks attributed to the product (observed in the H NMR spectra) with
the passage of time were analyzed using Igor Pro 6 (WaveMetrics, Inc.)
and Mathematica 9.0 (Wolfram) programs.
Materials. N,N-Dimethylformamide (DMF) and DMF-d7 were
dried over molecular sieves 4A and then distilled under reduced pressure
(10−20 mmHg). All other reagents were reagent-grade quality and were
used without further purification. [Ru(tpy)(bpy)H](PF6) (Ru-H),12
3-carbamoyl-4-methylpyridine,13 1-benzyl-3-carbamoylpyridinium hexa-
fluorophosphate (1a),14 and 2,4-dimethyl-3-carboxypyridine hydro-
chloride were prepared according to the reported methods.
Syntheses. 1-Benzyl-3-carbamoyl-4-methylpyridinium Hexa-
fluorophosphate (1b). A 15 mL MeCN solution containing 4-methyl-
nicotinamide (267 mg, 1.96 mmol) and 2 mL of benzyl chloride was
refluxed for 18 h. After a small amount of diethyl ether was added to
the solution, the precipitated brown solids were filtered and then
dissolved in ethanol. The solution was treated with charcoal. After the
solvent was evaporated, the residue was dissolved in a small amount of
water and a NH4PF6-saturated aqueous solution was added dropwise.
The precipitates were recrystallized with water. The yield was 61%
(445 mg, 1.19 mmol). Mp: 133 °C. Anal. Calcd for C14H15F6N2OP: C.
1
45.17; H. 4.06; N. 7.53. Found: C. 45.26; H. 3.95; N. 7.51. H NMR
(298 MHz, MeCN-d3): δ (ppm) 8.74 (s, 1H, 2-H), 8.56 (d, J = 5.1 Hz,
1H, 6-H), 7.88 (m, 5H, Ph), 7.28 (d, J = 5.1 Hz, 1H, 5-H), 6.78 (s, 1H,
NH2), 6.58 (s, 1H, NH2), 5.63 (s, 2H, 1-CH2−), 2.65 (s, 3H, 4-Me).
13C NMR (101 MHz, MeCN-d3): δ (ppm) 165.3 (CO), 159.5
(4-C), 144.7 (6-C), 143.2 (2-C), 137.5 (3-C), 133.8 (Ph), 131.4
(5-C), 130.9 (Ph), 130.5 (Ph), 130.1 (Ph), 64.9 (1-CH2−), 20.8 (4-Me).
1-Benzyl-3-(N,N-dimethylcarbamoyl)pyridinium Hexafluoro-
phosphate (1c). An MeCN solution (2 mL) containing N,N-dimethyl-
nicotinamide (1.00 g, 6.66 mmol) and benzyl chloride (1.00 mg,
7.10 mmol) was refluxed for 12 h. After the solvent was evaporated,
the residue was dissolved into a small amount of water. An NH4PF6-
saturated aqueous solution was dropped to the solution. The white
precipitates were recrystallized with water. The yield was 80% (2.05 g,
5.31 mmol). Mp: 164 °C. Anal. Calcd for C15H17F6N2OP: C, 46.64; H,
4.44; N, 7.25. Found: C, 46.78; H, 4.15; N, 7.29. 1H NMR (298 MHz,
MeCN-d3): δ (ppm) 8.80 (s, 1H, 2-H), 8.74 (d, J = 6.4 Hz, 1H, 6-H),
8.50 (d, J = 7.8 Hz, 1H, 4-H), 8.06 (dd, J = 7.8, 6.4 Hz, 1H, 5-H), 7.49
(m, 5H, Ph), 5.73 (s, 2H, 1-CH2−), 3.05 (s, 3H, NMe2), 2.91 (s, 3H,
NMe2). 13C NMR (101 MHz, MeCN-d3): δ (ppm) 164.7 (CO),
145.7 (6-C), 145.2 (4-C), 144.1 (2-C), 138.4 (3-C), 133.4 (Ph), 131.1
(Ph), 130.6 (Ph), 130.5 (Ph), 129.7 (5-C), 65.8 (1-CH2−), 39.6
(NMe2), 35.7 (NMe2).
3-(N,N-Dimethylcarbamoyl)-4-methylpyridine. A DMF solution
(12 mL) containing 3-carbamoyl-4-methylpyridine (1.00 g, 7.35 mmol)
and sodium hydride (528 mg, 22.0 mmol) was stirred at 0 °C for 1 h,
and iodomethane (2.30 g, 16.1 mmol) was added. The solution was
H
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
2,4-Dimethyl-3-(N,N-dimethylcarbamoyl)pyridine. 2,4-Dimethyl-
3-carboxypyridine hydrochloride (2.43 g, 13.0 mmol) was dissolved
in thionyl chloride (13 mL), and the solution was heated to 80−90 °C
for 1 h. After thionyl chloride was evaporated under reduced pressure
(12 mmHg at 80 °C), the residue was dissolved in a mixed dichloro-
methane (5 mL) and triethylamine (4.4 mL) solution. A THF solution
(7 mL) containing dimethylamine (13.1 mmol) was added dropwise
to the mixed solution for 20 min and then stirred for 1.5 h at room
temperature. After evaporation of the solvent, the residue was dis-
solved in a 35% HCl aqueous solution (11.3 mL). Active charcoal was
added to the solution, and the solution was heated for about 2 min.
After the active charcoal was filtered out, the filtrate was neutralized
with NaHCO3 up to pH 7, and the compound was extracted with
dichloromethane four times. The combined organic layer was washed
with an aqueous solution containing NaHCO3. The solution was dried
over Na2SO4 and then evaporated under reduced pressure. The colorless
oil was obtained by distillation under reduced pressure (0.5 mmHg, at
100 °C). The yield was 77% (1.79 g, 10.0 mmol). 1H NMR (400 MHz,
MeCN-d3): δ (ppm) 8.31 (d, J = 5.1 Hz, 1H, 6-H), 7.04 (d, J = 5.1 Hz,
3H, 5-H), 3.06 (s, 3H, NMe2), 2.75 (s, 3H, NMe2), 2.34 (s, 3H, 2-Me),
2.18 (s, 3H, 4-Me). 13C NMR (101 MHz, MeCN-d3): δ (ppm) 169.7
(CO), 154.3 (2-C), 149.4 (6-C), 144.1 (4-C), 133.5 (3-C), 123.5
(5-C), 37.6 (NMe2), 34.2 (NMe2), 22.2 (2-Me), 18.6 (4-Me).
1-(2,2,2-Trifluoroethyl)-2,4-methyl-3-(N,N-dimethylcarbamoyl)-
pyridinium Hexafluorophosphate (1h). 2,4-Dimethyl-3-(N,N-
dimethylcarbamoyl)pyridine (600 mg, 3.37 mmol) was dissolved in
2,2,2-trifluoroethyl trifluoromethanesulfonate (2.35 g, 10.1 mmol), and
the solution was heated to 100 °C for 14 h. The following processes
were the same as those for 1f. The yield was 31% (430 mg, 1.20 mmol).
Mp: 203−204 °C dec. Anal. Calcd for C12H16F9N2OP: C, 35.48; H,
3.97; N, 6.90. Found: C, 35.68; H, 3.90; N, 6.94. 1H NMR (400 MHz,
MeCN-d3): δ (ppm) 8.51 (d, J = 6.7 Hz, 1H, 6-H), 7.86 (d, J = 6.7 Hz,
1H, 5-H), 5.28 (qd, J = 8.0, 2.0 Hz, 2H, 1-CH2−), 3.11 (s, 3H, NMe2),
2.80 (s, 3H, NMe2), 2.68 (s, 3H, 2-Me), 2.52 (s, 3H, 4-Me). 13C NMR
(MeCN-d3, 101 MHz): δ (ppm) 164.7 (CO), 159.9 (4-C), 153.6
(2-C), 146.5 (6-C), 139.9 (3-C), 129.2 (5-C), 123.4 (q, JCF = 280 Hz,
CF3), 56.9 (q, JCF = 35 Hz, 1-CH2−), 37.7 (NMe2), 34.8 (NMe2), 20.8
(4-Me), 18.6 (2-Me).
and 3d were about 90% on the basis of Ru-H consumed during the
reaction.
Computational Calculations. All quantum chemical calculations
were carried out using the ORCA program package (version 3.0.3).15
Geometry optimizations under implicit solvation were carried out
using density functional theory (DFT) with the spin-restricted PW6B95
hybrid meta-GGA functional,16 which was supplemented with the atom-
pairwise dispersion correction with Becke-Johnson damping (D3BJ).17
In SCF and gradient calculations, a combination (RIJCOSX)18 of the
“resolution of the identity” approximation19 for the Coulomb parts and
the “chain-of-spheres exchange” algorithm for the exchange parts were
employed in order for the efficient computation of the Fock matrix. The
single polarized double-ζ basis set (def2-SVP)20 was used along with an
appropriate density-fitting basis set (def2-SVP/J)21 for C, H, N, and O
atoms. For the Ru atom, the single polarized double-ζ valence basis set
(def2-SVP)20 along with the corresponding density-fitting basis set
(def2-SVP/J)21 were accompanied by the Stuttgart−Dresden effective
core potentials (ECP, 28MWB)22 for taking the relativistic effect into
account. Implicit solvation by DMF was implemented in a dielectric
continuum (ε = 38.30, n = 1.430) using a conductor-like screening
model (COSMO).23 All local stationary structures were confirmed by
numerical vibrational frequency calculations, using the same level of the
theory as for the geometry optimizations. For their Gibbs energies,
electronic SCF energies were calculated first at the geometries optimized
by the above method, but using the single polarized triple-ζ basis set
(def2-TZVP)20 along with an appropriate density-fitting basis set (def2-
TZVP/J)21 for C, H, N, and O atoms and the single polarized triple-ζ
valence basis set (def2-TZVP)20 with the corresponding density-fitting
basis set (def2-TZVP/J)21 under the same ECP (28MWB) for the Ru
atom. Each of the Gibbs free energies was calculated by addition of this
electronic SCF energy to a zero-point energy, thermal corrections, and
entropy terms calculated using the vibrational calculations under the
ideal gas, rigid rotor, and harmonic oscillator approximations at a
temperature and pressure of 298.15 K and 1 atm, respectively. In the
case of adducts, they were finally corrected by the counterpoise method
of Boys and Bernardi24 to compensate for the basis set superposition
error (BSSE) between the [Ru(tpy)(bpy)]2+ and dihydropyridine
moieties. For evaluating the excluded volume of a solvent (DMF) by a
substituent at the N-1 position of a NAD(P)+ model, a solvent-accessible
surface of a molecule, where each atom has its own vdW radius, by DMF
(effective radius 2.64 Å) was calculated by employing the Winmostar
program (version 5.003) at a geometry optimized by the above method.
Then, a series of such surfaces was calculated at the geometries optimized
except a dihedral angle defined by the pyridine ring and the substituent,
where these angles were varied from 0° to 360° with an interval of 20°.
Finally, an envelope of this series of the surfaces was taken as the
excluded volume.
Pursuit of the Reaction between the NAD(P)+ Model
Compound and Ru-H. 1H NMR Spectroscopy Analysis. An
MeCN-d3 solution (0.5 mL) was bubbled with Ar in an NMR tube
for 30 min, and Ru-H (4.0 μmol) and an NAD(P)+ model compound
(4.2 μmol) were then dissolved in it, while the mixture was con-
tinuously bubbled with Ar. The reaction solution was bubbled with Ar
for an additional 5 min. After the solution was kept in the dark for a
1
suitable time period, the H NMR spectrum was measured. This pro-
cedure was repeated until the formation of the corresponding NAD(P)H
model compound or compounds was completed.
Determination of Rate Constants. For stopped-flow measurements,
each DMF solution containing Ru-H (0.031 mM) or the NAD(P)+
model compound (0.38 mM) was bubbled with Ar for 20 min. These
solutions were instantaneously mixed using the stopped-flow apparatus,
where the initial concentrations of Ru-H ([Ru-H]int) and the NAD(P)+
model compound ([1]int) were 0.0155 and 1.9 mM, respectively. After
mixing, UV−vis absorption spectral changes in the reaction solution
were obtained using a photodiode detector in the cases of 1b,g,,h or the
absorbance at 535 nm was obtained using a photomultiplier detector for
relatively fast reaction (1a,c,e,f). The same experiments were repeated
6−30 times, and the data were well fitted using a single exponential
function. For details, see the Supporting Information.
For determining the reaction rate of 1d with Ru-H, each 0.5 mL of
DMF-d7 solution containing 1d (8 mM) or Ru-H (8 mM) was mixed
under an Ar atmosphere, and the mixed solutions were kept at each
temperature. Second-order kinetics was directly applied to obtain k
using the 1H NMR data. The global fitting method was used with the
average integration values of the peaks at 9.81 and 8.68 ppm for
decrease of Ru-H and those of the peaks for formation and decrease
of 2d(1,4) (5.78, 5.64, 3.96, and 0.64 ppm), 2d(1,2) (6.24 and
0.78 ppm), 3d(1,4) (6.46, 6.05, 4.58, 4.43, 2.97, and 0.94 ppm), and
3d(1,2) (4.18, 3.72, and 1.59 ppm) (Figure 6). The total yields of 2d
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.organomet.5b00713.
1
Derivation of eq 6 and rate constants, H NMR spectra,
UV−vis spectra, decays of the absorbance, Eyring plots,
and most stable structures of 2a,c calculated using the
DFT method (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail for O.I.: ishitani@chem.titech.ac.jp.
Notes
The authors declare no competing financial interest.
I
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(13) Bobbitt, J. M.; Scola, D. A. J. Org. Chem. 1960, 25, 560−564.
(14) Fukuzumi, S.; Kondo, Y.; Tanaka, T. J. Chem. Soc., Perkin Trans.
2 1984, 673−679.
ACKNOWLEDGMENTS
■
This work was partially supported by a Grant-in-Aid for Scientific
Research on Innovative Areas “Artificial photosynthesis
(AnApple)” (No. 24107005).
(15) Neese, F. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73−
78.
(16) (a) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2005, 109, 5656−
5667. (b) Goerigk, L.; Grimme, S. Phys. Chem. Chem. Phys. 2011, 13,
6670−6688.
(17) (a) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys.
2010, 132, 154104. (b) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput.
Chem. 2011, 32, 1456−1465.
(18) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Chem. Phys.
2009, 356, 98−109.
(19) Neese, F. J. Comput. Chem. 2003, 24, 1740−1747.
(20) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
3297−3305.
(21) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057−1065.
(22) (a) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123−141. (b) Flores-Moreno, R.; Alvarez-
Mendez, R. J.; Vela, A. K. A. M. J. Comput. Chem. 2006, 27, 1009−
1019.
REFERENCES
■
(1) (a) Batie, C. J.; Kamin, H. J. Biol. Chem. 1984, 259, 11976−
11985. (b) Karplus, P. A.; Bruns, C. M. J. Bioenerg. Biomembr. 1994,
26, 89−99. (c) Carrillo, N.; Ceccarelli, E. A. Eur. J. Biochem. 2003, 270,
1900−1915. (d) Karplus, P. A.; Faber, H. R. Photosynth. Res. 2004, 81,
303−315. (e) Medina, M. FEBS J. 2009, 276, 3942−3958.
(2) (a) Dołęga, A. Coord. Chem. Rev. 2010, 254, 916−937. (b) Plapp,
B. V. Arch. Biochem. Biophys. 2010, 493, 3−12.
(3) (a) Paglietti, G. S. P.; Nuvole, A.; Soccolini, F.; Acheson, R. M. J.
Chem. Res. Miniprint 1983, 2326−2342. (b) Eisner, U.; Kuthan, J.
Chem. Rev. 1972, 72, 1−42. (c) Stout, D. M.; Meyers, A. I. Chem. Rev.
1982, 82, 223−243. (d) Lavilla, R. J. Chem. Soc., Perkin Trans. 1 2002,
1141−1156. (e) Wagenknecht, P. S.; Sambriski, E. J. Recent Res. Dev.
Inorg. Chem. 2003, 3, 35−50. (f) McSkimming, A.; Colbran, S. B.
Chem. Soc. Rev. 2013, 42, 5439−5488. (g) Quinto, T.; Kohler, V.;
Ward, T. R. Top. Catal. 2014, 57, 321−331.
̈
(23) (a) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2
1993, 799−805. (b) Klamt, A. J. Phys. Chem. 1995, 99, 2224−4435.
(c) Klamt, A.; Jonas, V. J. Chem. Phys. 1996, 105, 9972−9981.
(d) Sinnecker, S.; Rajendran, A.; Klamt, A.; Diedenhofen, M.; Neese,
F. J. Phys. Chem. A 2006, 110, 2235−2245.
(4) (a) Soldevila-Barreda, J. J.; Bruijnincx, P. C. A.; Habtemariam, A.;
Clarkson, G. J.; Deeth, R. J.; Sadler, P. J. Organometallics 2012, 31,
5958−5967. (b) Maenaka, Y.; Suenobu, T.; Fukuzumi, S. J. Am. Chem.
Soc. 2012, 134, 9417−9427. (c) Maenaka, Y.; Suenobu, T.; Fukuzumi,
S. J. Am. Chem. Soc. 2012, 134, 367−374. (d) Haquette, P.; Talbi, B.;
Barilleau, L.; Madern, N.; Fosse, C.; Salmain, M. Org. Biomol. Chem.
2011, 9, 5720−5727. (e) Canivet, J.; Suss-Fink, G.; Stepnicka, P. Eur. J.
Inorg. Chem. 2007, 2007, 4736−4742. (f) Yan, Y. K.; Melchart, M.;
Habtemariam, A.; Peacock, A. F.; Sadler, P. J. JBIC, J. Biol. Inorg. Chem.
2006, 11, 483−488.
(24) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553−566.
(5) (a) Ruppert, R.; Herrmann, S.; Steckhan, E. J. Chem. Soc., Chem.
Commun. 1988, 1150−1151. (b) Steckhan, E.; Herrmann, S.; Ruppert,
R.; Dietz, E.; Frede, M.; Spika, E. Organometallics 1991, 10, 1568−
1577. (c) Lo, H. C.; Buriez, O.; Kerr, J. B.; Fish, R. H. Angew. Chem.,
Int. Ed. 1999, 38, 1429−1432. (d) Lo, H. C.; Leiva, C.; Buriez, O.;
Kerr, J. B.; Olmstead, M. M.; Fish, R. H. Inorg. Chem. 2001, 40, 6705−
6716.
(6) Kobayashi, A.; Konno, H.; Sakamoto, K.; Sekine, A.; Ohashi, Y.;
Iida, M.; Ishitani, O. Chem. - Eur. J. 2005, 11, 4219−4226.
(7) (a) Ishitani, O.; Inoue, N.; Koike, K.; Ibusuki, T. J. Chem. Soc.,
Chem. Commun. 1994, 367. (b) Kobayashi, A.; Takatori, R.; Kikuchi,
I.; Konno, H.; Sakamoto, K.; Ishitani, O. Organometallics 2001, 20,
3361−3363. (c) Matsubara, Y.; Konno, H.; Kobayashi, A.; Ishitani, O.
Inorg. Chem. 2009, 48, 10138−10145. (d) Matsubara, Y.; Koga, K.;
Kobayashi, A.; Konno, H.; Sakamoto, K.; Morimoto, T.; Ishitani, O. J.
Am. Chem. Soc. 2010, 132, 10547−10552.
(8) (a) Ohno, A.; Ikeguchi, M.; Kimura, T.; Oka, S. J. Am. Chem. Soc.
1979, 101, 7036−7040. (b) Ohno, A.; Goto, T.; Nakai, J.; Oka, S. Bull.
Chem. Soc. Jpn. 1981, 54, 3478−3481. (c) Meyers, A. I.; Oppenlaender,
T. J. Am. Chem. Soc. 1986, 108, 1989−1996. (d) Meyers, A. I.; Brown,
J. D. J. Am. Chem. Soc. 1987, 109, 3155−3156. (e) De Kok, P. M. T.;
Bastiaansen, L. A. M.; Van Lier, P. M.; Vekemans, J. A. J. M.; Buck, H.
M. J. Org. Chem. 1989, 54, 1313−1320. (f) Okamura, M.; Mikata, Y.;
Yamazaki, N.; Tsutsumi, A.; Ohno, A. Bull. Chem. Soc. Jpn. 1993, 66,
1191−1196. (g) Mikata, Y.; Hayashi, K.; Mizukami, K.; Matsumoto, S.;
Yano, S.; Yamazaki, N.; Ohno, A. Tetrahedron Lett. 2000, 41, 1035−
1038. (h) Mikata, Y.; Aida, S.; Yano, S. Org. Lett. 2004, 6, 2921−2924.
(9) In a DMF-d7 solution instead of MeCN-d3, a similar experiment
was examined. The ratio was [3d(1,4)]:[3d(1,2)] = 91:9, which was
very similar to that obtained in MeCN-d3 solution.
(10) The coordinating atom to the Ru(II) center, i.e., N or O, has
not yet been determined in the transition state.
(11) Matsubara, Y.; Kosaka, T.; Koga, K.; Nagasawa, A.; Kobayashi,
A.; Konno, H.; Creutz, C.; Sakamoto, K.; Ishitani, O. Organometallics
2013, 32, 6162−6165.
(12) Konno, H.; Sakamoto, K.; Ishitani, O. Angew. Chem., Int. Ed.
2000, 39, 4061−4063.
J
DOI: 10.1021/acs.organomet.5b00713
Organometallics XXXX, XXX, XXX−XXX