Steric Effects on the Primary Isotope Dependence of Secondary Kinetic Isotope Effects in Hydride Transfer Reactions in Solution: Caused by the Isotopically Different Tunneling Ready State Conformations?

Article abstract of DOI:10.1021/jacs.5b03085

The observed 1° isotope effect on 2° KIEs in H-transfer reactions has recently been explained on the basis of a H-tunneling mechanism that uses the concept that the tunneling of a heavier isotope requires a shorter donor-acceptor distance (DAD) than that of a lighter isotope. The shorter DAD in D-tunneling, as compared to H-tunneling, could bring about significant spatial crowding effect that stiffens the 2° H/D vibrations, thus decreasing the 2° KIE. This leads to a new physical organic research direction that examines how structure affects the 1° isotope dependence of 2° KIEs and how this dependence provides information about the structure of the tunneling ready states (TRSs). The hypothesis is that H- and D-tunneling have TRS structures which have different DADs, and pronounced 1° isotope effect on 2° KIEs should be observed in tunneling systems that are sterically hindered. This paper investigates the hypothesis by determining the 1° isotope effect on α- and β-2° KIEs for hydride transfer reactions from various hydride donors to different carbocationic hydride acceptors in solution. The systems were designed to include the interactions of the steric groups and the targeted 2° H/D's in the TRSs. The results substantiate our hypothesis, and they are not consistent with the traditional model of H-tunneling and 1° /2° H coupled motions that has been widely used to explain the 1° isotope dependence of 2° KIEs in the enzyme-catalyzed H-transfer reactions. The behaviors of the 1° isotope dependence of 2° KIEs in solution are compared to those with alcohol dehydrogenases, and sources of the observed "puzzling" 2° KIE behaviors in these enzymes are discussed using the concept of the isotopically different TRS conformations. (Figure Presented).

Full text of DOI:10.1021/jacs.5b03085

Article
pubs.acs.org/JACS
Steric Effects on the Primary Isotope Dependence of Secondary
Kinetic Isotope Effects in Hydride Transfer Reactions in Solution:
Caused by the Isotopically Different Tunneling Ready State
Conformations?
Binita Maharjan, Mahdi Raghibi Boroujeni, Jonathan Lefton, Ormacinda R. White, Mortezaali Razzaghi,
Blake A. Hammann, Mortaza Derakhshani-Molayousefi, James E. Eilers, and Yun Lu*
Department of Chemistry, Southern Illinois University Edwardsville, Edwardsville, Illinois 62026, United States
S
* Supporting Information
ABSTRACT: The observed 1° isotope effect on 2° KIEs in H-transfer
reactions has recently been explained on the basis of a H-tunneling mechanism
that uses the concept that the tunneling of a heavier isotope requires a shorter
donor−acceptor distance (DAD) than that of a lighter isotope. The shorter
DAD in D-tunneling, as compared to H-tunneling, could bring about
significant spatial crowding effect that stiffens the 2° H/D vibrations, thus
decreasing the 2° KIE. This leads to a new physical organic research direction
that examines how structure affects the 1° isotope dependence of 2° KIEs and
how this dependence provides information about the structure of the tunneling ready states (TRSs). The hypothesis is that H-
and D-tunneling have TRS structures which have different DADs, and pronounced 1° isotope effect on 2° KIEs should be
observed in tunneling systems that are sterically hindered. This paper investigates the hypothesis by determining the 1° isotope
effect on α- and β-2° KIEs for hydride transfer reactions from various hydride donors to different carbocationic hydride acceptors
in solution. The systems were designed to include the interactions of the steric groups and the targeted 2° H/D’s in the TRSs.
The results substantiate our hypothesis, and they are not consistent with the traditional model of H-tunneling and 1°/2° H
coupled motions that has been widely used to explain the 1° isotope dependence of 2° KIEs in the enzyme-catalyzed H-transfer
reactions. The behaviors of the 1° isotope dependence of 2° KIEs in solution are compared to those with alcohol
dehydrogenases, and sources of the observed “puzzling” 2° KIE behaviors in these enzymes are discussed using the concept of
the isotopically different TRS conformations.
processes is expected.4,10 Thus, the 1° isotope has an effect on
2° KIEs. In the Marcus-like and large-curvature H-tunneling
models, the shorter DAD required for D-tunneling, as
compared to H-tunneling, could bring about spatial crowding
INTRODUCTION
■
The rule of the geometric mean formulated by Bigeleisen in
1955 states that there should be no isotope effects on isotope
effects in chemical reactions.1 For H-transfer reactions the
that in turn stiffens the vibrations of the 2° H and reduces the
2° KIE.12,13,18 This also explains the 1° isotope effect on 2°
KIEs. The model of 1°/2° H coupled motions combined with a
H-tunneling mechanism was proposed in early 1980s and has
been widely used in analyzing enzyme-catalyzed H-transfer
reactions.5,9,23−26 It, however, encountered challenges in
explaining the 2° KIEs that are smaller than the semiclassically
expected value (i.e., deflated 2° KIEs) in D-tunneling processes
of hydride transfer reactions mediated by a series of alcohol
dehydrogenases and mutants.11−13 While both explanations
work for the understanding of the observed unequally inflated
2° KIEs in H- and D-tunneling, the observed deflated 2° KIEs
appear to require using the concept that the DAD is shorter in
D-tunneling resulting in a more pronounced steric isotope
effect.11−13
transferring (“in-flight”) primary (1°) isotopes of H/D/T do
not affect the isotope effects at the “in-place” secondary (2°)
H/D/T positions in the H-donor or acceptor, and vice versa.2
However, violations of the rule were often seen in the enzyme-
catalyzed H-transfer reactions.3−9 Currently, there are two
explanations in vogue, both of which use the concept of 1° H-
tunneling. One explanation uses the model of 1°/2° H coupled
motions in a traditional H-tunneling mechanism,3,9,10 while the
other uses the concept that the donor−acceptor distance
(DAD) is shorter in D-tunneling than in H-tunneling.11−13 The
latter concept has been described in the Marcus-like H-
tunneling model and in the large-curvature H-tunneling
model.14−18 In the coupled motions model, the 2° C−H
bond out-of-plane bending vibrational modes are coupled with
the vibrations of the transferring 1° H thus are a component of
the reaction coordinate. This results in an inflated 2° kinetic
isotope effect (KIE).4,19−22 The inflation of the 2° KIEs is
larger in a tunneling mechanism, and since H-tunneling is more
probable than D-tunneling, more inflation in H-tunneling
Received: March 24, 2015
Published: May 5, 2015
© 2015 American Chemical Society
6653
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
The explanation that uses the concept of a shorter DAD for
D-tunneling than for H-tunneling leads to a new physical
organic research direction. Namely, “how does structure affect
the 1° isotope dependence of 2° KIEs?” and “how, in turn, does
this dependence provide information about the structure of the
tunneling ready states?” The hypothesis is that H-tunneling and
D-tunneling have tunneling ready state structures which have
different DADs, and that pronounced 1° isotope effect on 2°
KIEs should be observed in tunneling systems that are sterically
hindered (Scheme 1). This paper will test the hypothesis by
1,4-dihydronicotinamide (BNAH) to 10-methylacrdinium ion
(MA+) and 9,10-dimethylacrdinium ion (DMA+) (counterions:
BF4 ) in acetonitrile (eq 2); and the hydride transfer reactions
−
from the reduced forms of MA+ and DMA+, i.e., MAH and
a
−
DMAH, to a more reactive tropylium ion (Tr+BF4 ) in the
Scheme 1
same solvent (eq 3). The 1° isotope effects on 2° KIEs at the 9-
α-H/D position of the MA+/or MAH and that at the bulkier 9-
β-CH3/CD3 position of the DMA+/or DMAH will be
determined and compared. Furthermore, our predictions are
inconsistent with the reported pronounced 1° isotope depend-
ence of α-2° KIEs on MA+ in its reaction with another
dihydropyridine hydride donor (5-methyl-5,6-dihydrophenathr-
ine (MPH)) by Ostovic, Roberts and Kreevoy (eq 4), which
a
It is hypothesized that the steric/crowding factors will increase the 1°
isotope effect (i.e., the DAD effect) on the 2° KIEs. The 2° H(D) are
at the α- or the β-position. The oval-shaped area is meant to show the
wave packet of the transferring H/D in the tunneling ready states.
DAD stands for donor−acceptor distance.
determining the 1° isotope effect on 2° KIEs for hydride
transfer reactions from various hydride donors to different
carbocationic hydride acceptors. Systems are designed to
include close approach of the steric groups and the targeted
2° H/D’s in the tunneling ready states (Scheme 1). Moreover,
these systems are also designed with the idea that they might
facilitate understanding the behaviors of the 1° isotope
dependence of 2° KIEs in alcohol dehydrogenases.
has been cited frequently in the discussion of the 1° isotope
effect on 2° KIEs in literature.19 We will reinvestigate the 2°
KIEs of the reaction both in the 2-PrOH/H2O solvent that
those authors used and in acetonitrile that we use, and compare
the results with observations from our reactions. We will show
that our results, including especially from the unprecedented
study of 1° isotope effect on the β-type 2° KIEs, are consistent
with our hypothesis, and furthermore discuss why our
observations cannot be explained by the traditional model of
1°/2° H coupled motions combined with a tunneling
mechanism. We propose that this study should be extended
to other types of H-transfer reactions to examine how general
the concept of isotopically different tunneling ready state
conformations in explaining the 1° isotope effect on 2° KIEs in
this class of reactions might be. Then the observed behaviors of
the 1° isotope dependence of 2° KIEs are compared to the
previously observed 2° KIEs in the alcohol dehydrogenases,
and the origins of the latter observations are discussed on the
basis of the tunneling DAD concept.
One of our designs uses hydride transfer reactions from 2-
propanol (2-PrOH) to xanthylium ion (Xn+), thioxanthylium
ion (TXn+) and their 9-phenylsubstituted analogues (PhXn+
and PhTXn+) (counterions: BF4− or ClO4 ) in acetonitrile (eq
−
1). These reactions use a sequential hydride-proton transfer
mechanism in which the hydride transfer is rate-determin-
ing.27−29 The 2° KIEs are those on both the 2-propanol (β-H6/
D6) and the (T)Xn+ (α-9-H/D).28,30,31 Since the cations with
the 9-Ph group has a high steric hindrance at the reaction
center, a comparison of the effect of the 1° isotope on the β-2°
KIEs on 2-PrOH in its reactions with Xn+ vs PhXn+ and with
TXn+ vs PhTXn+ can examine the effect of steric hindrance on
the 1° isotope dependence of the β-2° KIEs. Furthermore, since
the relatively bulky β-2° CH3 group has more spatial
interactions than α-2° H with other structures in the reactive
complex, it is expected that the β-2° KIEs will be more sensitive
to the 1° isotope effect. Therefore, a comparison of the 1°
isotope dependence of β-2° KIEs with that of α-2° KIEs can
also provide information about the steric effect on the 1°
isotope effect on the 2° KIEs. To further understand the latter,
we will also study the hydride transfer reactions from 1-benzyl-
RESULTS
■
β-CH3/CD3 2° KIEs Are More Sensitive to the 1°
Isotope Effect than the α-H/D 2° KIEs. Table 1 lists the β-
2° KIEs at the H6/D6 position of 2-PrOH for both the hydride
and deuteride transfers to the (T)Xn+ and Ph(T)Xn+ (1).
Table 2 lists the α-2° KIEs at the 9-H/D position of the Xn(9-
L)+ and TXn(9-L)+ (L = H or D) (1). The α-2° KIEs at the
benzylic H/D position of the benzyl alcohol in its reaction with
NAD+ in an alcohol dehydrogenase are also included in Table 2
6654
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
Table 1. β-H6/D6 2° KIEs on 2-PrOH for Its Reaction with
Table 4. α- and β-2° KIEs on (D)MAH for Their Reactions
a b
,
a b
,
the Carbocations
with the Tropylium Ion (Tr+)
alcohol
CL3CY(OH)CL3
H-donors
MAH(9,9′-Y,L)+
DMAH(9-Y-9′-CL3)+
c
cd
,
PhXn+
cations
Xn+
TXn+
PhTXn+
α-H/D 2° KIE
β-H3/D3 2° KIE
0.97 (0.03)
c
Y = H
Y = D
1.09 (0.04)
Y = H
Y = D
1.05 (0.02)
1.04 (0.02)
1.05 (0.03)
1.00 (0.02)
1.05 (0.03)
1.04 (0.01)
1.04 (0.01)
0.98 (0.01)
c
−
0.91 (0.02)
a
a
In acetonitrile at 35 °C, Y is 1° isotope, L is 2° isotope; numbers in
In acetonitrile at 60 °C; numbers in parentheses are standard
b
b
deviations. Y is 1° isotope; L is 2° isotope (α-C changes from sp3 to
parentheses are standard deviations. On H-donor (9-α-C changes
c
from sp3 to sp2). Estimated indirectly from the rates of the reactions
c
d
sp2). From refs 28, 30. See SI for a detailed analysis of the raw data.
of MAH(9,9′-H,H), MAH(9,9′-H,D) and MAH(9,9′-D,D) by
assuming 2° KIE is 1° isotope independent (see the derivation from
SI) .
Table 2. α-2° H/D KIEs on Carbocations for Their
Reactions with Alcohols and That on Benzyl Alcohol for Its
a b
,
Oxidation by NAD+
Table 5. Comparison of the α-H/D 2° KIEs on MA+ for Its
Reduction by MPH-6,6′-Y,Y with Those Reported in
alcohols
CH3CY(OH)CH3
PhCYLOH
a b
,
Literature
c
d
cations
Xn(9-L)+
TXn(9-L)+
NAD+
e
e
f
Y = H
Y = D
0.99(0.01)
0.98(0.03)
0.97(0.02)
1.20
solvents
Y = H
acetonitrile
2-PrOH/H2O (4/1, v/v)
e
e
f
c
c
0.96(0.03)
1.12
0.98 (0.02)
0.99 (0.03)
1.00 (0.03)
1.01 (0.03)
1.05
0.96
a
In acetonitrile at 60 °C; numbers in parentheses are standard
Y = D
b
c
d
a
deviations. Y is 1° isotope; L is 2° isotope. From ref 30. Converted
At 22 °C (unless otherwise indicated), numbers in parentheses are
b
standard deviations; Y is 1° isotope, L is 2° isotope. On MA(9-L)+
from the observed H/T KIEs from ref 13 using KIE(H/T) = [KIE(H/
D)]1.42,2 catalyzed by the yeast alcohol dehydrogenase. At the 9-H/D
e
c
which reaction center changes from sp2 to sp3. From ref 19, at 25 °C.
f
positions of Xn+ or TXn+ (9-α-C changes from sp2 to sp3). At the
benzylic H/D position of PhCYLOH (α-C changes from sp3 to sp2).
plexes. Activation parameters for H-transfer processes in eqs 1
to 3 are listed in Table 6. The observed higher activation
enthalpies in the crowded systems of PhXn+, DMA+, DMAH
(vs Xn+, MA+, MAH) suggest that the steric effects most likely
play a role in decreasing the rate of these reactions. The
observed large negative activation entropies especially in the
crowded systems suggest that the transition states are tightly
associated complexes, which in turn suggests that the steric
factor likely hinders/stiffens the 2° H’s vibrations in the reactive
complexes.
The tightly associated reactive complexes may result from the
n−π and π−π interactions between the O in 2-PrOH (in eq 1)
or between the π-bonds on the dihydropyridine rings (in eq 2
to 4) with the electron-deficient heteroaromatic carbocations.
In order to acquire the geometric structure of the tunneling
reactive complex (also called tunneling ready state or TS for H-
tunneling) and to gain information regarding how the group
interaction affects the vibrations of the 2° H/D during the
formation of the tunneling ready state, we carried out gas-phase
calculations to obtain the classical TS structures for the 2-
PrOH/Xn+ and 2-PrOH/PhXn+ systems (Figure 1 and SI for
details). Note that the differences caused by the phenyl for H
substitution in the classical TS’s is expected to be similar for the
tunneling ready states (even though the two differ in DAD, the
hybridization of the reaction centers and the trajectory of the
H-transfer).32 Calculation confirms that these are tight
complexes in the sense that the electron clouds from the
reacting entities are mutually penetrating even in areas that are
not forming covalent bonds (i.e., the OH group with the central
ring of the (Ph)Xn+). Importantly, there is apparent steric
interaction between the Ph-group in PhXn+ and the β-CH3
groups in 2-PrOH, but little steric interaction can be seen in the
reaction of Xn+.
for comparison (see later discussions in this paper).13 All the α-
2° KIEs in solution reactions are 1° isotope independent. The
β-2° KIE on 2-PrOH is 1° isotope independent in the reactions
of both Xn+ and TXn+, but is 1° isotope dependent in the
reactions of both of their bulky 9-Ph substituted derivatives.
This shows that the steric factor enhances the effect of the 1°
isotope on β-2° KIEs in these systems. The trends are consistent
with the observations in the hydride transfer eqs 2−4. Table 3
Table 3. α- and β-2° KIEs on (D)MA+ for Their Reactions
ab
,
with BNAH
H-donor
cations
BNAH-4,4′-Y,Y
c
d
MA(9-L)+
DMA(9-CL3)+
α-H/D 2° KIE
0.99 (0.02)
β-H3/D3 2° KIE
1.07 (0.03)
Y = H
Y = D
0.99 (0.01)
0.99 (0.01)
a
In acetonitrile, Y is 1° isotope, L is 2° isotope; numbers in
b
parentheses are standard deviations. On cation (9-α-C changes from
c
d
sp2 to sp3). At 24 °C. At 32 °C.
shows that 1° isotope has no effect on α-2° KIEs at the 9-H/D
position of the MA+ and significant effect on the β-2° KIE at
the 9-CH3/CD3 positon of the DMA+ in their reactions with
BNAH (2). The significant effect of the 1° isotope on the β-2°
KIEs at the 9-CH3/CD3 positon of DMAH in its hydride
transfer reactions to Tr+ (3) are included in Table 4. Table 5
shows no 1° isotope effect on the α-2° KIEs on MA+ in its
reaction with MPH both in acetonitrile and in 2-PrOH/H2O
(4/1, v/v). The literature results from Ostovic, Roberts and
Kreevoy that deviate significantly from ours are also included in
Table 5 for comparison. The 1° KIEs of eqs 1−3 are listed in
the following Table 6 to indicate that the hydride transfers are
involved in the rate-limiting steps.
The formation of π−π charge-transfer complex between
nitrogen heterocycles and electron deficient cyclic π systems
similar to the substrates in eqs 2−4 have been reported.33−36
These strongly suggest tightly associated tunneling ready state
structures in our dihydropyridine oxidation reactions.
Activation Parameters and Calculation Provide
Evidence about the Steric Effect on the Targeted 2°
H’s Vibrations in Tightly Associated Reactive Com-
6655
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
Table 6. 1° KIEs and Activation Parameters for the Hydride Transfer Reactions in Acetonitrile
2-PrOH/Xn+
(eq 1)
2-PrOH/PhXn+
(eq 1)
BANH/MA+
(eq 2)
BNAH/DMA+
(eq 2)
MAH/Tr+
(eq 3)
DMAH/Tr+
(eq 3)
reaction system
temp (°C)
1° KIE
22.5−68.5
22.0−67.0
4.5−29.5
17.0−48.0
15.0−55.0
15.0−55.0
2.57 0.31 (68.5 °C) 3.03 0.12 (67.0 °C) 3.96 0.03 (29.5 °C) 3.49 0.18 (48.0 °C) 4.02 0.16 (55.0 °C) 3.42 0.20 (55.0 °C)
ΔH‡ (kJ/mol) 44.8 0.6
56.9 0.8
21.2 1.1
25.5 1.0
41.7 1.0
47.3 1.1
ΔS‡ (J/mol K) −123.1 1.9
−141.9 2.4
−137.9 3.7
−148.2 3.2
−92.4 3.4
−110.5 3.4
The β-2° KIEs on 2-PrOH are generally normal (>1) and those
on the cations are inverse (<1), consistent with the semi-
classical predictions for the donor hybridization change from
sp3 to sp2 and the acceptor hybridization change from sp2 to
sp3. However, the magnitudes of the 2° KIEs are “unusual”. In a
classical mechanism, 2° KIEs fall within the range from unity to
the corresponding equilibrium isotope effect (EIE).42−44 When
the 2° KIE is close to unity an early TS is suggested, and when
it is close to the EIE a late TS is suggested. We have estimated
the β-H6/D6 2° EIE for the conversion from CL3CH(OH)CL3
to CL3C+(OH)CL3 to be 1.51, and we take the EIE of 0.89
previously reported for the C-4 H/D position of the NAD+ (for
its conversion to NADH) to be the EIE for the α-9-H/D in the
conversion from (T)Xn+ to (T)XnH.30,31 Thus, both the β-H6/
D6 2° KIEs on 2-PrOH and the α-H/D KIEs on Xn+ and TXn+
are close to unity and far away from the respective EIEs,
suggesting that the TS’s are early in terms of the
rehybridization of the reaction centers. This, however, is
inconsistent with the endothermic nature of the hydride
transfer processes that produce a highly reactive CL3C+(OH)-
CL3 carbocation product, for which a late TS is predicted by
the Hammond’s Postulate. It is also inconsistent with the large
negative Hammett constant (ρ = −2.67) reported by us31 for
the reactions of substituted benzyl alcohols with PhXn+, which
also suggests a late TS. Therefore, the observed β-H6/D6 2°
KIEs on 2-propanol are deflated with respect to the
semiclassical prediction of close to 1.51, and the observed α-
H/D KIE on (T)Xn+ are inflated with respect to the
semiclassical prediction of close to 0.89. These 2° KIE results
are inconsistent with a classical H-transfer mechanism.
Within the Marcus-like H-tunneling model, the observed 2°
KIEs (close to unity for both 2-PrOH and the (T)Xn+ cations)
can be explained in terms of the small degree of
rehybridization/reorganization of the reaction centers during
the formation of the tunneling ready state. In the reactions of
Xn+ and TXn+, the steric hindrance effect at both reaction
centers is insignificant; so a decrease in DAD resulted from the
change from H- to D-tunneling does not change the degree of
rehybridization/reorganization, and hence the size of 2° KIEs,
to an extent that can be observed experimentally (cf., Figure 1,
the left structure) (Scheme 2 (a)). Therefore, the 2° KIEs at
both the α-H/D position of the (T)Xn+ and that at the β-H6/
D6 position of the 2-PrOH show little dependence on the 1°
isotope. In the reactions of Ph(T)Xn+, however, the steric
hindrance effect between the Ph-group of the cations and the β-
CH3 groups of 2-PrOH becomes significant, thus hindering the
reorganization of the 2° CH3 groups that are expected to swing
toward the Ph-group during the sp3 to sp2 conversion (cf., the
right structure in Figure 1). Hence, a decrease in DAD for the
reactions involving the Ph-containing cations will hinder the β-
CH3(CD3) reorganization (or stiffen the β-H/D vibrations to
increase the 2° H/D zero point energy difference) more than in
the case of (T)Xn+. As a result, the decrease of the 2° KIEs
from H- to D-tunneling is pronounced (Scheme 2 (b)). The
Figure 1. Calculated classical TS structures for the reactions of 2-
PrOH with Xn+ (left) and PhXn+ (right) in gas phase. The tunneling
ready state structures are expected to have similar geometric
differences. Steric interaction between the β-CH3 groups in 2-PrOH
and the 9-Ph group in PhXn+ is apparent.
DISCUSSION
■
Theoretical H-Tunneling Models That Require a
Shorter DAD for Transfer of a Heavier Isotope. Prior to
the discussion of the structural effects on the 1° isotope
dependence of 2° KIEs, it is necessary to describe the newly
established Marcus-like H-tunneling model that invokes the
DAD concept. This full tunneling model was established on the
basis of the Marcus theory of electron tunneling to include the
short DAD sampling necessary for H-tunneling to
occur.12,14−17,37−40 Achieving a tunneling-ready state requires
that (i) the system must reach a point of energy degeneracy
between the reactant and product, and (ii) the system must
reach a DAD short enough that the barrier to tunneling is
sufficiently narrow. Both processes are assisted by heavy atom
motions. Process (i), sometimes referred to as reorganization,
has skeletal changes in both the donor and the acceptor. This
involves hybridization changes and thus determines the 2°
KIEs. Process (ii), known as gating, samples a range of
conformations at various DADs. It determines the extent of 1°
H wave function overlap between the reactant and product
states, and thus the extent of H-tunneling and the 1° KIE. Since
the D wave function is more localized than the H wave
function, a shorter DAD is required for D-tunneling. This DAD
difference results in different tunneling ready state structures
for the hydrogen isotopes.13 A schematic description of the new
Marcus-like model that clearly shows the different DADs in H-
and D-tunneling processes can be found in ref 13. It should be
emphasized that this DAD difference is also required in the
large-curvature H-tunneling model.18 Therefore, the effect of
the 1° isotope on 2° KIEs is brought about by the effect of the
DAD on 2° KIEs within both models. Note that calculations
following the Marcus-like model that reproduce the 2° KIEs on
benzyl alcohol oxidation by the yeast alcohol dehydrogenase
have shown that the DAD difference between H-tunneling and
D-tunneling is about 0.2 Å.41
“Unusual” 2° KIEs and the Steric Effect on the 1°
Isotope Dependence of β-Type 2° KIEs. Hydride Transfer
from 2-PrOH to (T)Xn+ and Ph(T)Xn+ (Eq 1, Tables 1 and 2).
6656
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
explained in terms of a release of the spatial crowding
interaction between vibrations of its 9-CH3 group and 4,5-H
groups during a sp2 to sp3 conversion (structure (A) in Scheme
3). The release of the crowding effect is more for a longer C−H
Scheme 2. Schematic Description of the Tunneling Ready
States for the Alcohol Oxidation
a
a
Scheme 3.
a
Release of intramolecular spatial crowding effect between vibrations
of the 9-CH(D)3 group and the 4,5-H’s in DMA+ for an sp2 (structure
(A)) to sp3 conversion results in a normal β-2° KIE. Conversely, a gain
of the steric effect for an sp3 (in DMAH) to sp2 (DMA+) conversion
will result in an inverse 2° KIE (see Discussion for the reaction of
DMAH). No such steric release effect is involved between vibrations
of 9-H and 4,5-H’s in MA+ (structure (B)).
a
(a) The vibrations of the β-CH3 bonds of 2-PrOH and the 9-α-H of
(T)Xn+ do not interfere with each other in either H- or D-tunneling.
(b) The Ph group in Ph(T)Xn+ significantly hinders the vibrations of
the sterically readily accessible β-CH3 group in 2-PrOH more in the D-
tunneling ready state that requires a shorter DAD.
bond (than for a C−D bond) thus resulting in an inflated β-
CH3/CD3 2° KIE. So the observed inflated 2° KIE of 1.07
suggests that the KIE caused by the intramolecular steric
hindrance release54 (should be >1) wins over the effect of the
hyperconjugative loss (<1) in determining the 2° KIE. In the
reaction of MA+, however, the 9-H group on MA+ does not
have an intramolecular crowding effect with 4,5-H’s (structure
(B) in Scheme 3), so an expected inverse 2° KIE for an sp3 to
sp2 conversion is observed.
decreased 2° KIEs due to the steric effect (here brought about
by a decrease in DAD) can also be understood using the
concept that the 2° C−D bond is slightly shorter than the C−H
bond (by about 0.001 Å) due to the difference in the vibrational
extent of the C−H(D) bonds.45−47 That is, the shorter 2° C−D
bond encounters less steric hindrance during the formation of
the tunneling ready state, making the rate decrease due to the
steric effect less than that for the 2° C−H substrate reaction
and resulting in a diminished 2° H/D KIE.47−51 Note that this
isotopic bond-length difference is too small to differentiate the
DADs for the 1° H- and D-tunneling and thus to affect the 1°
KIEs significantly (the latter DAD difference is about 0.2 Å (see
above)). The above inference has been recently substantiated
by our computations (using the Marcus-like H-tunneling
model) that reproduce the 2° KIE behaviors in the reactions
of both Xn+ and PhXn+.32,52
In the reaction of DMA+ the significantly decreased β-2° KIE
on D-tunneling (0.99) as compared to H-tunneling (1.07) can
be explained as a consequence of a decreased DAD in D-
tunneling. The intermolecular steric hindrance with BNAH
becomes important, somehow hindering the reorganization of
the β-2° CH(D)3 group in the DMA moiety of the tunneling
ready state, decreasing the 2° KIE (Scheme 4(a), Sub =
BNAH). There is no such steric effect between the
inaccessible/“hidden” α-9-H on MA+ and the BNAH structure.
Therefore, a change in DAD does not affect the corresponding
α-2° KIE on MA+, resulting in no 1° isotope effect on 2° KIEs
(Scheme 4(b)).
Hydride Transfer from BNAH to MA+ and DMA+ (Eq 2,
Table 3). The hydride affinities of BNA+ (the oxidized form of
BNAH) and MA+ from literature are 243 and 318 kJ/mol (in
DMSO), respectively.53 This indicates that the reaction is
highly exothermic which, according to the Hammond’s
Postulate, suggests an early TS. The hybridization of the
reaction center at the MA+/DMA+ changes from sp2 to sp3.
Therefore, the 2° KIEs are expected to be inverse and close to
unity. The observed α-2° KIE on MA+ for the hydride transfer
process is indeed inverse (<1) and close to unity, but the β-2°
KIE on DMA+ is inflated to be >1 (1.07), and it becomes
inverse for deuteride transfer. While the loss of the conjugative
effect between the β-CH(D)3 and the 9-C+ as a result of
conversion from sp2 to sp3 should lead to an inverse 2° KIE, the
2° KIE behaviors in these systems appear to be inconsistent
with the classical H-transfer process.
Hydride Transfer from MAH and DMAH to Tr+ (Eq 3, Table
4). These reactions use the reduced forms of MA+ and DMA+,
i.e., MAH and DMAH, as hydride donors. The hybridization
changes on the donor centers are from sp3 to sp2, which is
opposite to those in the reductions of (D)MA+ in eq 2.
Therefore, including a comparison of the 2° KIE behaviors in
the two systems can avoid the overexplanation that might result
from considering only one reaction direction. From literature,
the hydride affinity of the hydride acceptor of Tr+ is 690 kJ/mol
(in DMSO).53 A comparison of this value with that of MA+
(632 kJ/mol) suggests that the reactions are highly exothermic,
which also suggests an early TS. Therefore, the 2° KIE is
expected to be normal and close to unity for the sp3 to sp2
conversion from (D)MAH to (D)MA+. The observed α-2° KIE
on MAH (1.09) is indeed normal but far from unity, whereas
the β-2° KIE on DMAH (0.97 for H-transfer) is even inverse
showing significant deflation. As expected, the “signs” (normal
or inverse) of the 2° KIEs are opposite to those observed in eq
2. While both reactions are highly exothermic, the α-2° KIE on
MAH in eq 3 is close to the reaction EIE of 1.12, whereas the
The α-2° KIE on MA+ (inverse and close to unity) can be
rationalized in terms of a small degree of rehybridization/
reorganization at the reaction center during the formation of
the tunneling ready state, similar to the explanation for the
observed small α-2° KIEs on (T)Xn+ in their oxidation of
alcohols. Knowing the size difference between the 9-H of MA+
and the 9-CH3 of DMA+, the inflation in the β-2° KIE (from an
expected inverse value to 1.07) on the latter cation may be
6657
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
Hydride Transfer from MPH to MA+ (Eq 4, Table 5).
Ostovic, Roberts and Kreevoy reported that the α-2° KIE on
MA+ is 10% higher in H-transfer than in D-transfer in this
reaction in 2-PrOH/H2O (4/1, v/v).19 The significant effect of
the 1° isotope on the 2° KIEs reported was very influential at
the time. It was interpreted as results from the solution
reactions strongly supporting the model of 1°/2° H coupled
motions and H-tunneling, which had been proposed on the
basis of enzymatic reaction studies and supported by a
subsequent theoretical investigation.3,10 Unfortunately, the
results were not reproducible in our lab. Our results show
that the 2° KIEs are independent of the 1° isotope not only in
the aqueous 2-PrOH solvents that those authors used but also
in acetonitrile. Our results are consistent with the trends
observed in eqs 1−3, which is that the α-type 2° KIEs are not
sensitive to the 1° isotope effect in solution reactions.
Scheme 4. Schematic Description of the Tunneling Ready
States for the Hydride Transfers from Substrate BNAH to
a
DMA+ and MA+
1°/2° H Coupled Motion Model Cannot Explain the
Observed Effect of 1° Isotope on 2° KIEs. As described in
the Introduction, currently there are two models that have been
used to explain the 1° isotope effect on 2° KIEs. The need to
use the tunneling DAD concept to explain the effect of 1°
isotope on 2° KIEs exposes the difficulties inherent to a
possible analysis based on the traditional model of H-tunneling
and 1°/2° H coupled motions.10,20,22,55,56 Within the coupled
motions model, the 1° isotope independence of α-2° KIEs
would be explained in terms of little or no 1° H-tunneling,
whereas the observed pronounced 1° isotope dependence of β-
2° KIEs would be interpreted in terms of a greater degree of
tunneling. According to this analysis, it can be concluded that
the reactions of BNAH/MA+, MAH/Tr+ and 2-PrOH/Xn+ take
place with little or no tunneling effect, which would imply that
the temperature dependence of 1° KIEs will not fall outside of
the semiclassical limits predicted by the traditional Bell H-
tunneling theory that has been used in combination with the
coupled motions model to explain the 1° isotope effect on 2°
KIEs, whereas those of BNAH/DMA+, DMAH/Tr+ and 2-
PrOH/PhXn+ take place with significant tunneling effects,
which would imply that the temperature dependence of 1°
KIEs will be most likely outside of the semiclassical limits.
However, the temperature dependence of 1° KIEs that we
observed in these systems (Table S2 in SI) is not consistent
with such an analysis. The reactions of BNAH/MA+ and
BNAH/DMA+ (2) give rise to similar Arrhenius isotopic
preexponential ratios (AH/AD = 0.2 vs 0.1), and these are well
outside of the semiclassical limits (The semiclassical limits of
AH/AD = 0.7 to 1.4).16,56,57 So do the reactions of MAH/Tr+
and DMAH/Tr+ (3) (AH/AD = 0.4 vs 0.4). In contrast, the
reactions of 2-PrOH/Xn+ and 2-PrOH/PhXn+ (1) give rise to
AH/AD values (0.9 and 1.4), neither of which fall outside of the
semiclassical limits. These results show that the extent of H-
tunneling is similar in the three pairs of reactions irrespective of
α-2° H or β-2° CH3 groups and the steric factors in the
substrates. (Note that the magnitude of the AH/AD values in
the Marcus-like full-tunneling model is ascribed to the effect of
thermal motions on the ease of the tunneling DAD samplings.
The analysis of the temperature dependence of the 1° KIEs
using this model is not a focus of discussion in this paper.)
Moreover, within the model of H-tunneling and 1°/2° H
coupled motions, the 2° KIEs are expected to be inflated with
respect to the values predicted by the semiclassical KIE theory.
This is not always true in our observations. For example, in the
reactions of 2-PrOH with (T)Xn+ and Ph(T)Xn+, the β-H6/D6
KIEs on 2-PrOH are in the range of 0.98 to 1.05 for the
a
(a) The vibrations of the sterically readily accessible 9-β-CH3 in
DMA interfere with the other substrate BNAH more in D-tunneling
than in H-tunneling. (b) The interference does not happen for the
reaction of MA+ where the isotope effect of a “hidden” 9-α-H is
concerned.
observed 0.99 on MA+ in the reverse hybridization change in eq
2 is far away from the corresponding EIE of 0.89. Note again,
the EIEs of 1.12/0.89 are taken from previous reports at the C-
4 H/D position of the NADH/NAD+ in their interconversion
process. The unexpectedly large α-2° KIE on MAH is not
consistent with the classical H-transfer mechanism. However, it
can be explained in terms of a large degree of rehybridization/
reorganization of the reaction center during the formation of
the tunneling ready state in a tunneling mechanism.
The observed deflated, even inverse, β-2° KIE (0.97) on
DMAH can result from the spatial crowding between its 9-
CH3/CD3 group and 4,5-H groups, which increases in the sp3
to sp2 activation process (cf. structure (A) in Scheme 3); i.e.,
gain of the steric effect stiffens the CH3 vibrations thus reducing
the 2° KIE. Therefore, the intramolecular isotope effect that
results from increased hindrance (2° KIE < 1) wins over the
hyperconjugative effect (>1 due to sp3 to sp2 conversion) in
determining the 2° KIE. This is consistent with the explanation
for the inflated normal β-2° KIE on DMA+ in its oxidation of
BNAH (2) where the hybridization change is in the opposite
direction.
As compared to the H-tunneling, the more inverse β-2° KIE
(0.91) in the D-tunneling can be explained in terms of a larger
intermolecular steric effect between the DMAH and Tr
moieties due to a decreased DAD, which adds to the
intramolecular steric effect between vibrations of 9-CH3 and
4,5-H’s that arises in the tunneling ready state. That is, the
intermolecular steric hindrance somehow hinders the reorgan-
ization of the β-2° CH(D)3 group in the DMAH moiety in the
tunneling ready state, making the 2° KIE more inverse (cf.
Scheme 4(a) where Sub = Tr). In the reaction of MAH,
however, the 9-H group on MAH does not have an
intramolecular crowding effect with 4,5-H’s (cf. structure (B)
in Scheme 3) and does not have an intermolecular crowding
effect with the Tr+; so the α-2° KIE is normal and does not
depend upon the 1° isotope (cf. Scheme 4(b) where Sub = Tr).
6658
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
hydride/deuteride transfer processes (Table 1). As discussed
earlier, however, the expected values should be close to an EIE
of 1.51 because of the endothermic nature of the reactions. So
all of these observed 2° KIEs, in either H- or D-tunneling
process, are actually deflated. Moreover, while the β-2° KIEs on
2-PrOH in its reactions with (T)Xn+ are deflated, the α-2° KIEs
on (T)Xn+ (0.97 and 0.99, Table 2) are inflated (also see the
earlier discussion). It is not possible that the 1°/2° H coupled
motions occur with a 2° H in one reactant but not with a 2° H
in the other. Also, in the exothermic reactions between 1,4-
dihydropyridines and carbocations, the observed α-2° KIE on
MA+ in its reaction with BNAH (Table 3) and the β-2° KIE on
DMAH in its reaction with Tr+ (Table 4) appear not to be
inflated either, even though the temperature dependence of 1°
KIEs study strongly suggests a traditional H-tunneling
mechanism (AH/AD = 0.2 and 0.4, respectively).
α-2° KIEs: 1° Isotope Independent in Solution vs
Dependent in Alcohol Dehydrogenases. Interestingly, the
1° isotope effect on the α-type 2° KIEs is not significant in the
solution reactions, but it is significant in the hydride transfer
reactions catalyzed by alcohol dehydrogenases.11−13 Klinman
and co-workers determined the α-2° KIEs on the oxidation of
benzyl alcohol (PhCL2OH, L = isotopes of H) by NAD+ for H-
and D-transfers catalyzed by both wild-type alcohol dehydro-
genases and mutants. It was observed that the α-2° H/T KIEs
on H-transfers were inflated and invariant no matter what type
of alcohol dehydrogenases were used, but the 2° D/T KIEs on
D-transfers were deflated to different degrees when using
different enzymes.5−8,41 Roston and Kohen recently deter-
mined the 2° H/T KIEs on both H- and D-transfers catalyzed
by the yeast alcohol dehydrogenase, directly showing a
significantly deflated 2° KIE with D-transfer.13 Table 2 lists
the latter 1° isotope dependence of α-2° KIE results. Our
results show that steric hindrance from the reaction system
deflates the 2° KIEs on the sterically readily accessible β-CH3/
CD3 groups but not on the “hidden” α-H/D groups. Therefore,
the observed significant 1° isotope effect on the α-2° KIEs in
the alcohol dehydrogenase reactions implies that the
reorganization of the “hidden” α-2° H/D’s encounters a
significant resistance during the D-transfer rehybridization
process which can be attributed to the restrictive protein
environment of the active site rather than to steric hindrance
between the two moieties of the tunneling ready state itself. It
has been proposed from the alcohol dehydrogenases data that
the H- and D-transfers have isotopically different tunneling
ready state conformations and a shorter DAD necessary for D-
tunneling would induce a less optimal fit in the active site thus
increasing the environmental crowding effect and decreasing
the α-2° KIEs.11−13 Our results confirm the proposition.
Our findings include the following: (1) most 2° KIE values
do not conform to the KIE values predicted from the
semiclassical transition state theory; (2) the steric effect can
augment the 1° isotope dependence of 2° KIEs, and the β-
CH3/CD3 2° KIE is more sensitive to the 1° isotope effect than
is the α-H/D 2° KIE; (3) the 1° isotope dependence of 2° KIEs
can imply information about the structure of the tunneling
ready states; (4) the significant 1° isotope dependence of α-2°
KIEs reported by Ostovic, Roberts and Kreevoy in 1983 for a
hydride transfer reaction similar to ours (and which has been
frequently cited in the 1° isotope dependence of 2° KIE studies
in biological H-transfer reactions) is not reproduced in our
work; (5) both the 2° KIE values and their 1° isotope
dependences cannot be explained by the traditional model of
H-tunneling and 1°/2° H coupled motions. Our results support
our hypothesis. Further studies should be extended to other
types of H-transfer reactions including proton and hydrogen
atom transfers.
Sources of the observed “puzzling” 1° isotope dependence of
α-type 2° KIEs in alcohol dehydrogenases were also analyzed. It
has been proposed in literature that D-tunneling that requires a
shorter DAD induces a densely packed protein active site
environment, which significantly hinders the α-2° H/D’s
reorganization and results in deflated 2° KIEs and hence the
effect of the 1° isotope on the α-2° KIEs. Our results are
consistent with this idea.
EXPERIMENTAL SECTION
■
General Procedures. Alcohol Oxidations. 9-Phenylthioxanthy-
−
lium tetrafluoroborate (PhTXn+BF4 ) and thioxanthylium perchlorate
(TXn(9-H)+ClO4−) and 9-deuterated thioxanthylium perchlorate
((TXn(9-D)+ClO4−)) were synthesized from the dehydration
reactions of the corresponding alcohol precursors with HBF4 or
HClO4 in propionic anhydride or acetic anhydride according to a
published procedure.58 The procedure is the same as the one that we
used to synthesize the 9-phenylxanthylium tetrafluoroborate
−
−
(PhXn+BF4 ) and the xanthylium perchlorate (Xn(9-H)+ClO4 ,
30
−
Xn(9-D)+ClO4 ). The 9-phenyl-thioxanthen-9-ol (PhTXnOH) was
synthesized by reacting the thioxanthone with the Grignard reagent
PhMgBr in dry THF. The 9H(D)-thioxanthen-9-ol (TXn(9-H)OH or
TXn(9-D)OH) was synthesized from the reduction of thioxanthen-9-
one by NaBH4 or NaBD4 according to a literature procedure.59 2-
Propanol-β-d6 ((CD3)2CHOH), 2-propanol-α-d1 ((CH3)2CDOH)
and 2-propanol-α-d1-β-d6 ((CD3)2CDOH) were prepared by the
reduction of acetone or acetone-d6 with NaBH4 or NaBD4 according
to a literature procedure.28,60 In order to make a direct kinetic
comparison, the normal 2-propanol was also synthesized by using the
same procedure, i.e., reduction of acetone with NaBH4. The densities
of the alcohols were determined (see Table S1 in SI). Acetonitrile was
distilled over KMnO4/K2CO3 and then P2O5 in nitrogen atmosphere
before use.
1,4- and 1,2-Dihydropyridine Oxidations. 1-Benzylnicotinamide
bromide (BNA+), 1-benzyl-1,4-dihydronicotinamide (BNAH) and its
deuterated analogue (BNAH-4,4′-d,d, 96% D content (from NMR
data)) were prepared according to the commonly used procedure.53,61
CONCLUSIONS
■
The 1° isotope effects on 2° KIEs were determined in hydride
transfer reactions from various hydride donors to different
carbocationic hydride acceptors in acetonitrile. The purposes of
the study are (i) to test the hypothesis that H-tunneling and D-
tunneling have tunneling ready state structures which have
different donor−acceptor distances and thus different con-
formations, and that therefore a pronounced 1° isotope effect
on 2° KIEs should be observed in the tunneling systems that
are sterically hindered, and (ii) to compare the 2° KIE
behaviors in solution with those in alcohol dehydrogenases in
order to provide insight into the mechanism of the enzyme
catalyzed H-transfer reactions.
−
The normal 10-methylacridinium ion (MA(9-H)+BF4 ) and its 9-
−
deuterated analogue (MA(9-D)+BF4 ) were synthesized from the
oxidation of 10-methylacridan (MAH) and 9,9′-dideuterio-10-
methylacridan (MAH-9,9′-d,d) by slight excess of the tropylium
−
tetrafluoroborate (Tr+BF4 ) (molar ratio = 1:1.1) in acetonitrile and
−
recrystallized in an acetonitrile−ether mixture. The Tr+BF4 was
commercially available and was recrystallized from acetonitrile-ether
for both synthesis and kinetic study uses. The 10-methylacidan
(MAH) was synthesized from the reaction of NaBH4 with MA+I− that
was made from the reaction of acridine with CH3I in acetone.53 The
MAH-9,9′-d,d was prepared from the reduction of the commercially
available 10-methylacridone by LiAlD4 according to a literature
6659
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
method.53 Likewise, the 9,10-dimethylacridinium ion (DMA(9-
half-lives to determine the rate constant. The obtained −ln(A − A∞)
∼ t data were plotted and the slope of the linear correlation (R2 >
0.999) was taken as the kpfo. Determination of the effect of [BNAH]
on kpfo showed that the reaction follows the second-order rate law.
Second-order rate constants (k) and the β-d3 2° KIE on DMA+ were
calculated.
−
−
CH3)+BF4 ) and its 9-CD3 analogue (DMA(9-CD3)+BF4 ) were
synthesized from the hydride transfer reactions from 9,10-
dimethylacridan (DMAH) and 9-deuterated methyl-10-methylacrdan
−
to Tr+BF4 . The latter two 9-methyl substituted acridans were
prepared from the reactions of MA+I− with Grignard reagents CH3MgI
and CD3MgI in dry ether, following a literature procedure.62 The 5-
methyl-5,6-dihydrophenanthridine (MPH-6,6′-h,h) was synthesized by
the hydride reduction of 5-methylphenanthridinium iodide (MP+I−)
with NaBH4. The latter salt was directly prepared by the reaction of
phenanthridine with CH3I in ether. The MPH-6,6′-d,d was
synthesized by reduction of the 5-methyl-phenanthridin-6(5H)-one
using LiAlD4 in dry ether, and the ketone was prepared by the direct
oxidation of MP+I− by KO2 according to a literature procedure.63
Kinetic and KIE Measurements. Alcohol Oxidations. The
procedure for the kinetic and KIE determinations of the reactions of
PhXn+ and Xn+ in pure acetonitrile was reported previously.28,30 The
pseudo-first-order rate constant (kpfo) was determined by following the
decay of the cations spectroscopically. The observed effect of [2-
propanol] on the kpfo of the reactions, in which the [2-propanol] range
was chosen for the convenience of kinetic measurements, are
consistent with the second-order rate law within the experimental
error.28 The second-order rate constant (k) was calculated by kpfo
dividing by [2-propanol] and the KIE was calculated from the k values
(= kH/kD). Use of the normal 2-propanol that we synthesized
following the same procedure to synthesize its isotopologues
minimizes the errors in KIEs determined.
For the fast MA+ reaction kinetic determinations, equal volumes of
6.00 × 10−3 M BNAH solution and 6.00 × 10−4 M MA+ solution
(both in acetonitrile) were mixed in a specially designed thermostated
cuvette connected to a rapid stopped-flow mixer equipped with a
pneumatic drive unit (a Hi-Tech Scientific SFA-20 fast kinetic
determination kit). The decay in absorbance at 440 nm due to MA+
with time (t) was recorded. The obtained −ln(A − A∞) ∼ t data were
plotted (R2 is within the range of 0.9980−0.9999) and the kpfo was
obtained. Second-order rate constants (k) and the α-D 2° KIE on MA+
were calculated.
The measurements were repeated three times, and the whole
procedure was further repeated three times using the new solutions.
(D)MAH Oxidations by Tr+. The kinetics of the (D)MAH
oxidations by Tr+ were determined spectroscopically by following
the formation of the (D)MA+ at 373 nm. Detailed kinetic
determination procedures are similar to those for the oxidations of
BNAH and MPH except that the formation of the cations was
monitored. In the reaction of MAH/Tr+, [MAH] = 7.32 × 10−5 M,
[Tr+] = 2.00 × 10−2 M; in the reaction of DMAH/Tr+, [DMAH] =
7.32 × 10−5 M, [Tr+] = 8.00 × 10−2 M, at 35 °C. The A∞ was
determined and −ln(A∞ − A) ∼ t was plotted to derive the kpfo. The
reactions follow the second-order rate law. The second-order rate
constants and the KIEs on (D)MA+ were calculated. The measure-
ments were repeated three times, and the whole procedure was further
repeated three times using the new solutions.
Calculation of the TS Geometry. The gas-phase TS (activated
complex) structure was determined by Hartree−Fock calculations with
a 6-31G** basis set. It is only meant to show the qualitative geometry
of the TS and infer the information about the spatial interactions of
groups in the actual tunneling ready state. The detailed calculation
procedure is included in the SI.
The kinetic and KIE determinations of the reaction of PhTXn+ used
the same procedure for the kinetic determination of the PhXn+
reaction.28,30 Typically, 40 μL of 0.207 M stock solution of PhTXn+
in acetonitrile was added to 4 mL of acetonitrile solution of large
excess of 2-propanol (0.314 M) in a sealed 10 mL reaction vial placed
in a water bath set at 60 °C. Aliquots of about 0.25 mL were
periodically withdrawn from the reaction into the sample vials which
were precooled in ice. The samples were immediately placed in a
freezer (∼ −20 °C). For each kinetic experiment, 7 to 8 reaction
aliquots were collected within 1−3 half-lives of the reaction. For UV−
vis spectral analysis, 60 μL aliquots of the reactions were diluted into
1.94 mL acetonitrile. The corresponding UV−vis spectra at different
reaction times, i.e., the kinetic scans, were obtained (SI). The
absorbance (A) decrease at 372 nm due to the PhTXn+ consumption
was recorded as a function of time (t). The obtained A vs t data were
fitted to the integrated first-order rate equation, −ln(A) = kobs·t +
constant, to give the kpfo (R2 > 0.999). The kinetic solutions of a pair
of isotopologues were placed side by side in the water bath. The rate
measurements were repeated 4 times and the standard deviations are
small.
ASSOCIATED CONTENT
■
S
* Supporting Information
Kinetic scans, densities of the 2-PrOH and its isotopologues,
temperature dependence of 1° KIEs, effect of [2-PrOH] and
[H+] on the β-2° KIEs on 2-PrOH in its reaction with PhXn+,
reaction rates and 2° KIEs from typical determinations, raw
kinetic data examples to show the method to derive the
standard deviations in 2° KIEs, and computational method and
procedure are included. The Supporting Information is
available free of charge on the ACS Publications website at
DOI: 10.1021/jacs.5b03085.
The kinetic and KIE determinations of the reaction of TXn+ used
the same procedure as used for the reaction of Xn+. 2 mL 0.0120 M
alcohol solution in acetonitrile was placed in a cuvette sitting in a
thermostated cell holder with temperature maintained at 60 °C. 32 μL
of 0.00250 M TXn+ stock solution (in acetonitrile) was rapidly
transferred to the alcohol solution and the decay in absorbance at 373
nm due to TXn+ with time (t) was recorded for about 2−3 half-lives.
The obtained −ln(A) − t data were plotted, and the slope of the linear
correlation (R2, mostly >0.999) was taken as the kpfo. The
measurement was repeated three to four times, and the whole
procedure was further repeated at least four times using the new
solutions.
AUTHOR INFORMATION
■
Corresponding Author
*yulu@siue.edu
Notes
The authors declare no competing financial interest.
Oxidations of BNAH and MPH by (D)MA+. The kinetics of the
BNAH oxidation by MA+ and DMA+ were determined spectroscopi-
cally by following the decay of the cations at 440 nm. For the
oxidations of MPH (1.20 × 10−2 M) by MA+ (6.00 × 10−4 M), the
determination was made by following the decay of the cation at 373
nm. A typical determination of the kinetics for the reaction of DMA+
with BNAH is as follows. To the 1.90 mL 3.99 × 10−3 M BNAH
solution in acetonitrile with certain temperature in a cuvette, 100 μL of
4.70 × 10−3 M DMA+ stock solution (in acetonitrile) was rapidly
added. The decay in absorbance at 440 nm due to DMA+ with time (t)
was recorded for at least 10 half-lives to determine A∞ and for 2−3
ACKNOWLEDGMENTS
■
This work was supported by the SIUE’s Seed Grant for
Transitional and Exploratory Project and the Petroleum
Research Fund administrated by the American Chemical
Society (ACS-PRF Grant #: 55140-UR4).
REFERENCES
(1) Bigeleisen, J. J. Chem. Phys. 1955, 23, 2264.
■
6660
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661

Journal of the American Chemical Society
Article
(2) Sen, A.; Kohen, A. In Quantum Effects in Enzyme Kinetics;
Allemann, R., Scrutton, N., Eds.; Royal Society of Chemistry: London,
U.K., 2009; p 164.
(3) Blanchard, J. S.; Cleland, W. W. Biochemistry 1980, 19, 3543.
(4) Cook, P. F.; Oppenheimer, N. J.; Cleland, W. W. Biochemistry
1981, 20, 1817.
(5) Cha, Y.; Murray, C. J.; Klinman, J. P. Science 1989, 243, 1325.
(6) Bahnson, B. J.; Park, D. H.; Kim, K.; Plapp, B. V.; Klinman, J. P.
Biochemistry 1993, 32, 5503.
(40) Kohen, A. Acc. Chem. Res. 2015, 48, 466.
(41) Roston, D.; Kohen, A. Proc. Natl. Acad. Sci. U. S. A. 2010, 107,
9572.
(42) Melander, L.; Saunders, W. H. In Reaction Rates of Isotopic
Molecules, 4th ed.; Krieger, R. E.: Malabar, FL, 1987; p 170.
(43) Streitwieser, A.; Jagow, R. H.; Fahey, R. C.; Suzuki, F. J. Am.
Chem. Soc. 1958, 80, 2326.
(44) Sunko, D. E.; Humski, K.; Malojcic, R.; Borcic, S. J. Am. Chem.
Soc. 1970, 92, 6534.
(45) Kuchitsu, K.; Bartell, L. S. J. Chem. Phys. 1962, 36, 2470.
(46) Allinger, N. L.; Flanagan, H. L. J. Comput. Chem. 1983, 4, 399.
(47) Mugridge, J. S.; B, G.; Raymond, K. N. Angew. Chem., Int. Ed.
Engl. 2010, 49, 3635.
(48) Mislow, K.; Graeve, R.; Gordon, A. J.; Wahl, G. H. J. Am. Chem.
Soc. 1963, 85, 1199.
(7) Rucker, J.; Cha, Y.; Jonsson, T.; Grant, K. L.; Klinman, J. P.
Biochemistry 1992, 31, 11489.
(8) Bahnson, B. J.; Colby, T. D.; Chin, J. K.; Goldstein, B. M.;
Klinman, J. P. Proc. Natl. Acad. Sci. U. S. A. 1997, 94, 12797.
(9) Wilde, T. C.; Blotny, G.; Pollack, R. M. J. Am. Chem. Soc. 2008,
130, 6577.
(49) Brown, H. C.; McDonald, G. J. J. Am. Chem. Soc. 1966, 88, 2514.
(50) Saunders, M.; Wolfsberg, M.; Anet, F. A. L.; Kronja, O. J. Am.
Chem. Soc. 2007, 129, 10276.
(51) Hayama, T.; Baldridge, K. K.; Wu, Y.-T.; Linden, A.; Siegel, J. S.
J. Am. Chem. Soc. 2008, 130, 1583.
(52) Derakhshani-Molayousefi, M.; Eilers, J. E.; Lu, Y., unpublished
results.
(53) Cheng, J.-P.; Lu, Y.; Zhu, X. Q.; Mu, L. J. Org. Chem. 1998, 63,
6108.
(54) Melander, L.; Carter, R. E. J. Am. Chem. Soc. 1964, 86, 295.
(55) Huskey, W. P. J. Phys. Org. Chem. 1991, 4, 361.
(56) Bell, R. P. The Tunnel Effect in Chemistry; Chapman & Hall:
London, 1980.
(57) Kim, Y.; Kreevoy, M. M. J. Am. Chem. Soc. 1992, 114, 7116.
(58) Dauben, H. J.; Honnen, L. R.; Harmon, K. M. J. Org. Chem.
1960, 25, 1442.
(10) Huskey, W. P.; Schowen, R. L. J. Am. Chem. Soc. 1983, 105,
5704.
(11) Klinman, J. P. In Quantum Tunnelling in Enzyme-Catalyzed
Reactions; Scrutton, N. S., Allemann, R. K., Eds.; RSC Publishing:
Cambridge, U.K., 2009; p 132.
(12) Nagel, Z. D.; Klinman, J. P. Chem. Rev. 2010, 110, PR41.
(13) Roston, D.; Kohen, A. J. Am. Chem. Soc. 2013, 135, 13624.
(14) Knapp, M. J.; Klinman, J. P. Eur. J. Biochem. 2002, 269, 3113.
(15) Basran, J.; Masgrau, L.; Sutcliffe, M. J.; Scrutton, N. S. In Isotope
Effects in Chemistry and Biology; Kohen, A., Limbach, H. H., Eds.;
Taylor & Francis, CRC Press: Boca Raton, FL, 2006; Chapter 25, p
671.
(16) Kohen, A. In Isotope Effects in Chemistry and Biology; Kohen, A.,
Limbach, H. H., Eds.; Taylor & Francis, CRC Press: Boca Raton, FL,
2006; Chapter 28, p 743.
(17) Hay, S.; Scruton, N. S. Nat. Chem. 2012, 4, 161.
(18) Kreevoy, M. M.; Ostovic, D.; Truhlar, D. G.; Garrett, B. C. J.
Phys. Chem. 1986, 90, 3766.
(59) Zhu, X.-Q.; Dai, Z.; Yu, A.; Wu, S.; Cheng, J.-P. J. Phys. Chem. B
2008, 112, 11694.
(60) Thompson, M. S.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104,
(19) Ostovic, D.; Roberts, R. M. G.; Kreevoy, M. M. J. Am. Chem. Soc.
1983, 105, 7629.
4106.
(61) Bunting, J. W.; Sindhuatmadja, S. J. Org. Chem. 1981, 46, 4211.
(62) Fukuzumi, S.; Tokuda, Y.; Kitano, T.; Okamoto, T.; Otera, J. J.
Am. Chem. Soc. 1993, 115, 8960.
(63) Picot, A.; Milliet, P.; Cherest, M.; Lusinchi, X. Tetrahedron Lett.
1977, 43, 3811.
(20) Saunders, W. H., Jr. J. Am. Chem. Soc. 1985, 107, 164.
(21) Klinman, J. P. In Enzyme Mechanism from Isotope Effects; Cook,
P. F., Ed.; CRC Press: Boca Raton, FL, 1991; p 127.
(22) Rucker, J.; Klinman, J. P. J. Am. Chem. Soc. 1999, 121, 1997.
(23) Cheng, M.-C.; Marsh, E. N. G. Biochemistry 2007, 46, 883.
(24) Bahnson, B. J.; Anderson, V. E. Biochemistry 1991, 30, 5894.
(25) Sikorski, R. S.; Wang, L.; Markham, K. A.; Ravi Rajagopalan, P.
T.; Benkovic, S. J.; Kohen, A. J. Am. Chem. Soc. 2004, 126, 4778.
(26) Kohen, A.; Jensen, J. H. J. Am. Chem. Soc. 2002, 124, 3858.
(27) Lu, Y.; Qu, F.; Moore, B.; Endicott, D.; Kuester, W. J. Org.
Chem. 2008, 73, 4763.
(28) Lu, Y.; Qu, F.; Zhao, Y.; Small, A. M.; Bradshaw, J.; Moore, B. J.
Org. Chem. 2009, 74, 6503.
(29) Lu, Y.; Bradshaw, J.; Zhao, Y.; Kuester, W.; Kabotso, D. J. Phys.
Org. Chem. 2011, 24, 1172.
(30) Hammann, B.; Razzaghi, M.; Kashefolgheta, S.; Lu, Y. Chem.
Commun. 2012, 48, 11337.
(31) Liu, Q.; Zhao, Y.; Hammann, B.; Eilers, J.; Lu, Y.; Kohen, A. J.
Org. Chem. 2012, 77, 6825.
(32) Kashefolgheta, S.; Razzaghi, M.; Hammann, B.; Eilers, J.;
Roston, D.; Lu, Y. J. Org. Chem. 2014, 79, 1989.
(33) Liao, M.-S.; Lu, Y.; Scheiner, S. J. Comput. Chem. 2003, 24, 623.
(34) Liao, M.-S.; Lu, Y.; Parker, V. D.; Scheiner, S. J. Phys.Chem. A
2003, 107, 8939.
(35) Lu, Y.; Zhao, Y.; Handoo, K. L.; Parker, V. D. Org. Biomol.
Chem. 2003, 1, 173.
(36) Pudney, C. R. J. L.; Sutcliffe, M. J.; Hay, S.; Scrutton, N. S. J. Am.
Chem. Soc. 2010, 132, 11329.
(37) Stojkovic,
A. Chem. Commun. 2010, 46, 8974.
(38) Stojkovic, V.; Kohen, A. Isr. J. Chem. 2009, 49, 163.
́
V.; Perissinotti, L. L.; Lee, J.; Benkovic, S. J.; Kohen,
́
(39) Sam, H.; Sutcliffe, M. J.; Scrutton, N. S. In Quantum Tunnelling
in Enzyme Catalyzed Reactions; Allemann, R., Scrutton, N., Eds.; Royal
Society of Chemistry: London, U.K., 2009; Vol. 9, p 199.
6661
DOI: 10.1021/jacs.5b03085
J. Am. Chem. Soc. 2015, 137, 6653−6661