The primary kinetic hydrogen isotope effect in the deprotonation of a nitroalkane by an intramolecular carboxylate group

Article abstract of DOI:10.1002/poc.1330

The rates of racemization of optically active nitropentanoic acid, and 4-deuteronitropentanoic have been compared. The rate ratio (kie) is k ^H/k^D = 5.68(±0.17) at 31°C, in good agreement with that determined by Lewis et al. for base-catalysed deprotonations using iodine-trapping methods. In a more detailed study, optically active 4-nitro-4-phenylbutanoic acid (NPBA) has also been prepared and rates of racemization measured in dimethox-yethane:water. With less than a full equivalent of triethylamine, rates are proportional to [Et₃N:]/[NPBA] . For 1 < [Et₃N:]/[NPBA] < 2, rates are independent of the ratio, consistent with racemization being dominated by deprotonation of the nitroalkane by the intramolecular carboxylate group. The solvent isotope effect is k^H2O/k^D2O = 0.73(±0.04) and rates of exchange with D₂O are equal to rates of racemization. Comparison with rates of racemization by acetate of the methyl ester yielded an effective molarity (EM = 13.7) for the intramolecular carboxylate. The kie for racemizations of NPBA and 4-deutero-NPBA is k^H/k^D = 5.78 at 25°C, and for 20 < T < 50°C, E_a^D - E_a^H = 5.5(±0.1) and A^H/A^D = 0.63(±1.03). For the acetate catalysed racemizations of the methyl ester, 25°C, k ^H/k^D = 7.43 with E_a^D - E _a^H = 5.2 kJ mol^-1 and A^H/A ^D = 1.08. In neither case is there any indication of a major tunnelling contribution on the isotopic rate ratio. A hitherto unrecognised mode of decomposition of nitronic acids, involving direct reaction with dissolved oxygen, has been identified. Copyright

Full text of DOI:10.1002/poc.1330

Special Issue
Received: 25 October 2007,
Revised: 21 December 2007,
Accepted: 21 December 2007,
Published online in Wiley InterScience: 15 May 2008
(www.interscience.wiley.com) DOI 10.1002/poc.1330
The primary kinetic hydrogen isotope effect in
the deprotonation of a nitroalkane by an
intramolecular carboxylate group
Nicholas Backstroma, Neil A. Burtona, Simon Turegaa and C. Ian F. Watta
*
The rates of racemization of optically active nitropentanoic acid, and 4-deuteronitropentanoic have been compared.
The rate ratio (kie) is kH/kD ¼ 5.68(W0.17) at 31-C, in good agreement with that determined by Lewis et al. for
base-catalysed deprotonations using iodine-trapping methods. In a more detailed study, optically active
4-nitro-4-phenylbutanoic acid (NPBA) has also been prepared and rates of racemization measured in dimethox-
yethane:water. With less than a full equivalent of triethylamine, rates are proportional to [Et3N:]/[NPBA]. For
1 < [Et3N:]/[NPBA] < 2, rates are independent of the ratio, consistent with racemization being dominated by
deprotonation of the nitroalkane by the intramolecular carboxylate group. The solvent isotope effect is
kH O=kD O ¼ 0.73(W0.04) and rates of exchange with D2O are equal to rates of racemization. Comparison with
rates of racemization by acetate of the methyl ester yielded an effective molarity (EM ¼ 13.7) for the
intramolecular carboxylate. The kie for racemizations of NPBA and 4-deutero-NPBA is kH/kD ¼ 5.78 at 25-C, and
for 20 < T < 50-C, EDa S EHa ¼ 5.5(W0.1) and AH/AD ¼ 0.63(W1.03). For the acetate catalysed racemizations of the methyl
ester, 25-C, kH/kD ¼ 7.43 with EaD S EHa ¼ 5.2 kJ molS1 and AH/AD ¼ 1.08. In neither case is there any indication of a major
tunnelling contribution on the isotopic rate ratio. A hitherto unrecognised mode of decomposition of nitronic acids,
involving direct reaction with dissolved oxygen, has been identified. Copyright ß 2008 John Wiley & Sons, Ltd.
2
2
Keywords: primary kinetic hydrogen isotope effect; racemizations; nitroalkane; intramolecular deprotonation; temperature
dependences; effective molarity
anticipated from complete loss of ground state zero-point energy
INTRODUCTION
differentials.[6–8] The temperature dependences of such rate
ratios often yield Arrhenius parameters which are incompatible
with explanations involving only ground state zero-poinꢀt 1energy
With the availability of three isotopes, and large isotopic mass
ratios, kinetic hydrogen isotope effects (kies, kH/kD, kH/kT and kD/
kT), are especially valuable in the characterisation of mechanisms
of deprotonation of carbon acids, firstly to determine if a proton
transfer is rate-limiting in a complex sequence and, when such a
situation has been discovered, then to probe the nature of the
proton transfer transition state. The usefulness of the method
follows from an understanding of the origins of the kinetic effects
at the molecular level. An excellent qualitative understanding has
been available for many years, with kinetic isotope effects
associated with changes in force constant of the vibrations of
isotopically substituted bonds as reactants proceed from ground
to transition state.[1–4] Vibrational analysis, coupled with
bond-energy bond-order relationships, has provided a useful
empirical treatment to relate the experimentally determined
isotope effects to transition state models.[5] If, however, isotope
effects are to be linked in a detailed way to changes in molecular
structure, ab initio methods must be applied and high-level
computational studies have been undertaken in a few cases.
Since rate ratios for H- and D-isotopomers at 258C are usually less
than 8, quantitatively useful models must be capable of
calculating activation energy differences much smaller than
5.7 kJ molꢀ1 presenting a tough challenge, even to modern
computational methods.
differences (for 20 < T < 2000 K, EDa ꢀ EaH ꢁ 5 kJ mol
and
0.7 < AH/AD < 1.32).[9] These effects (which are rare in ’normal’
organic reactions) have been taken as evidence for mechanisms
involving quantum mechanical tunnelling of the transferring
proton through the barrier (from the Born–Oppenheimer
potential energy surface).[10,11] An emerging explanation[12,13]
of the occurrence of such contributions in enzyme-catalysed
reactions is ‘environmentally coupled or vibrationally assisted
hydrogen tunnelling’, focusing on the role of protein dynamics
and qualitatively understood in terms of a strong coupling
between certain vibrations of the reacting protein acting to
narrow the barrier.[14] Our own computational studies on
methylamine dehydrogenase (MADH)[15,16] showed that the
structural configuration of an enzyme during transfer of a proton
from the carbon of a Schiff base to the carboxylate of an aspartate
(Fig. 1) can strongly affect the shape of the barrier to proton
transfer and that large kies might then arise, in line with
experiment. The dominating effect in these studies has been the
*
The School of Chemistry, University of Manchester, Manchester M13 9PL, UK.
E-mail: ian.watt@manchester.ac.uk
The challenge to theory has become more pronounced with
the accumulation of reports of enzyme-catalysed processes
showing kinetic isotope effects considerably larger than those
a
N. Backstrom, N. A. Burton, S. Turega, C. I. F. Watt
The School of Chemistry, University of Manchester, Manchester M13 9PL, UK
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.

N. BACKSTROM ET AL.
catalysing groups (pKa ca 10 for the phenol and ca 5 for the
carboxylic acid) and that of the carbon acid (ca 20) shows that
intramolecular carbon deprotonations would be strongly
endothermic in both cases, a situation believed to disfavour
tunnelling. The final example, in which there can be no
complicating secondary effect, is 4-nitropentanoic acid, 3, from
the work of Lewis.[21,22] In this case, the carbanion is stabilised by
a nitro group, the best carbanion stabilising group in organic
chemistry, with the anion best depicted as a nitronate with the
negative charge largely on the oxygen atoms of the group.
A pKa ¼ 7.7 is expected[23,24] for the carbon acid, so that the
imbalance with the catalysing carboxylic acid group is much
reduced (DpKa ꢁ 3) compared to that in 1 or 2. This, combined
with the high kinetic barriers associated with deprotonation of
nitroalkanes[25–27] makes 3 the most promising of this group for
tunnelling. In the event, the reported kie (kH/kD ¼ 5.5 at 258C),
again from halogen trapping experiments, is rather less than
those reported for intermolecular deprotonations of 2-nitro-
propane by other oxyanionic bases.[28] No temperature depen-
dence was reported.
In view of the sparsity of such data, we have begun a study of
isotope effects associated with intramolecular proton transfers in
synthetically accessible carbon acids, with the aim of providing
measurements for simple reactions in which ground and
transition states have been characterised as fully as experiment
allows. Experimental and computational studies are to be
compared, and results used to validate the theoretical models,
leading eventually to a computational model with predictive
power.
Figure 1. Carbon deprotonation in the active site of MADH
exo- or endothermicity of the process, which indirectly affects
the shape of the barrier. In their current state, these studies
rationalise rather than predict the results of experiment, and
because of the size and complexity of proteins, difficulties arise in
separation of the interacting contributions and inclusion of
dynamical effects to experimental accuracy.
The theory thus remains unclear in its application in enzymes,
and it has been pointed out[17] that, while the chemistry
occurring in the active site of enzyme-substrate complexes has
often been modelled by intramolecular reactions in small
molecules,[18] measurements of kinetic isotope effects in
intramolecular proton, hydride and hydrogen atom transfers in
such models are rare. Indeed, a search for such measurements
in intramolecular deprotonations of carbon acids revealed only
the three examples shown in Fig. 2. All reactions were carried in
aqueous medium, and in no case were temperature depen-
dences reported.
Acetophenones 1 and 2 are from the work of Bell.[19,20] In
these, the acidic carbon site is adjacent to the ketonic carbonyl
and phenolic or carboxylic acid groups provide the catalysing
intramolecular base. In both cases, rates of enolization were
determined by halogen trapping. An elevated isotope effect was
reported for 1 but close reading of the original paper suggests
that this result should be viewed with caution. Additionally, the
acidic methyl groups were fully labelled, so observed ratios were
combinations of primary and secondary effects which were not
disentangled. The large difference between the acidities of the
As noted earlier, values of kH/kD significantly larger than those
which might be expected from complete loss of isotopically
induced ground state zero-point energy differentials are rare in
‘small molecule’ chemistry, occurring mainly (but not exclusively)
in deprotonations of nitroalkanes. Some of the very largest of
these[29–31] (kH/kD ¼ 50 at 258C) have, on reinvestigation, turned
out to be artefacts arising from unrecognised loss of isotopic label
during the rate measurements,[32,33] but values between 14 < kH/
kD < 20 remain to be explained. These are often associated with
deprotonations in aprotic media by hindered bases[34,35] but
deprotonations of 2-nitropropane by hindered pyridine bases in
protic medium are particularly well characterised,[36,37] with kH/
kD ¼ 19.5 at 258C in EtOH, and temperature dependence (a 128
range only!) giving AH/AD ¼ 7.1 and EaD ꢀ EHa ¼ 12.5 kJ molꢀ1
,
compatible with the intervention of tunnelling. Pressure
dependences also are compatible with a large tunnelling
contribution in the 1H-isotopomer,[38] but the tritium isotope
effect, kH/kT, has also been measured[39] and the value of the
exponent a ¼ 1.42 in the Swain–Schaad relationship,[40] kH/
kT ¼ (kH/kD)a, does not support a high-tunnelling contribution.
It is perhaps worth repeating that the large kinetic isotope
effects and unusual Arrhenius behaviour taken as indicative of
proton tunnelling may also arise from mechanistic complexities.
Quite reasonable combinations of microscopic rate constants and
conventional primary isotope effects may produce such
results,[41,42] for example, in reactions in which products are
formed from common intermediates. Even where kinetic isotope
effects fall within the ‘normal’ range, anomalous Arrhenius
behaviour may arise from a mechanism involving internal
return.[43] Checking for such complexities in enzyme-catalysed
reactions is far from trivial but an essential part of any study of
small molecule reactions intended to provide data for the
construction and optimisation of theoretical models.
Figure 2. Available kinetic isotope effects in intramolecular deprotona-
tions of carbon acids
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 603–613

PRIMARY KINETIC HYDROGEN ISOTOPE EFFECT
RESULTS AND DISCUSSION
Choice of substrate and experimental design
Nitroalkanes offer a combination of particularly high carbanion
stability with high kinetic barrier to deprotonation (the
nitroalkane anomaly[25–27]), a combination thought to favour
tunnelling pathways, and numerous detailed studies of inter-
molecular deprotonation exist for comparison. With the excep-
tion of Lewis’ study of 4-nitropentanoic acid 3, intramolecular
variants have not been examined, but an oxidative conversion of
nitroalkanes to corresponding aldehydes or ketones catalysed by
the flavoenzyme, nitroalkane oxidase, has been shown to involve
deprotonation of an enzyme-bound nitroalkane by the carbox-
ylate group of an aspartate residue,[44] with a remarkable 109-fold
rate enhancement over the acetate induced reaction. A kinetic
isotope effect, kH/kD ¼ 7.9, has been reported for the reaction of
1,1-dideuterionitroethane.[45,46]
Scheme 1. Preparation and labelling of 4-nitropentanoic acid, 3, and
4-nitro-4-phenylbutanoic acid, 4
In this work, we have firstly re-examined the behaviour of
3 to establish techniques and methods, and then undertaken a
more detailed examination of the behaviour of 4-nitro-
4-phenylbutanoic acid 4 in which the methyl group of 3, is
replaced by phenyl. In both nitroalkanes, formation a nitronate
anion eliminates the stereogenic centre adjacent to the nitro
group, so that loss of optical activity from resolved compounds
offers a useful handle on extent of the deprotonation reaction.
Replacement of methyl by phenyl induces a favourable small shift
of the carbon acidity (7.39[47] as opposed to 7.7[23,24]) and offers
the valuable technical advantages of enhanced UV activity and
ease of preparation of optically active material, but is not
expected to make major changes to the chemistry.
regarded as a complication in studies of deprotonations of
simple nitroalkanes or in reversions of nitronic acid tautomers to
nitroalkanes in near neutral aqueous medium. Lewis, however,
noted that the anion of 3 was relatively unstable, decomposing to
4-oxopentanoic acid 5 (R ¼ Me) with evolution of a gas presumed
to be nitrous oxide.[21,22] To account for the apparent occurrence
of this Nef chemistry under the mildly basic conditions of his
observations, he suggested that the carboxylate group might
participate in forming the putative tetrahedral intermediate of
the hydrolytic Nef reaction by nucleophilic attack on the carbon
of the nitronate or nitronic acid[54] to give a g-lactone 6 as shown
in Scheme 2.
Our first observations of the behaviour of aqueous solutions of
4-nitropentanoate confirmed the ready formation of 5. Since the
chemistry leading to 5 was not well understood, and might have
complicated measurements of deprotonation rates, this aspect of
the reactivity was briefly investigated further using 4 (rather
than 3) because of its more convenient UV–Vis properties.
1-Phenylnitropropane then serves as a reference compound
lacking the catalysing intramolecular carboxylate group.
Solutions of phenylnitropropane (ca 10ꢀ5 M) in distilled but
undegassed water exhibited the expected shoulder at 240 nm
Preparative considerations and preliminary
reactivity studies.
Methyl esters of both 3 and 4 are readily accessible by
conjugated addition of nitroethane or phenyl nitromethane to
methyl acrylate,[48] and pyridine-catalysed exchange with
OD-methanol affords the required deuteriated materials. Extents
of deuteriation could be readily determined from 1H-NMR spectra
by integration since the signal from the exchangeable hydrogen
adjacent the nitro group was well resolved (a doublet of doublets
at d 4.68 in 3 and d 5.60 in 4) and readily compared with signals
from non-exchanging sites. As anticipated from the established
behaviour of nitroalkanes,[49,50] hydrolysis of the esters by
heating in aqueous TFA occurred without exchange of hydron at
the position a-to the nitro group.
In the case of 3, we could not find conditions for resolution of
either the ester or the acid by chiral chromatography, and
resorted to an established method[51] involving crystallisations of
the quinine salt of the 3, in both isotopomeric forms. Again,
extents of deuteriation after resolution were readily measured
from 1H-NMR spectra by integration, but, in our hands, the
resolution by this method was accomplished only with some
wash-out of deuterium, and decomposition of the nitroalkane by
processes discussed later. For 4, chiral HPLC of the ester, either
before or after deuteriation, yielded optically active material,
which was then hydrolysed with no detectable loss of either label
or of optical activity (Scheme 1).
Hydrolyses of nitronic acid tautomers of nitroalkanes to
ketones or aldehydes (the Nef reaction[52]) usually occurs at
relatively high acidities[53] (ca 3 M HCl) and is not normally
Scheme 2. Nef chemistry on nitronates or nitronic acids carrying car-
boxylic acid residues
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

N. BACKSTROM ET AL.
characteristic of the nitro group.[47] Addition of aqueous
sodium hydroxide then yielded a new peak, lmax ¼ 275 nm
(emax ¼ 6300 Mꢀ1 cmꢀ1), with a shoulder to 225 nm, consistent
with formation of the nitronate anion, and this basic solution was
stable. Acidification to pH 1 immediately shifted the peak to
262 nm with an increased extinction, changes associated with
formation of the nitronic acid tautomer, expected to
have pKa ꢁ 4.[54] This spectrum, however, did not, as expected,
revert cleanly to that of the original phenylnitropropane, instead
slowly developing a new peak at 244 nm (emax 10 300 Mꢀ1 cmꢀ1).
When changes were complete, GLC analysis of the solution and
comparison with authentic samples confirmed recovery of
phenylnitropropane and of propiophenone, the Nef product.
In successive experiments, the ratio of propiophenone to
phenylnitropropane was enormously variable, but it was clear
that the production of the ketone was reduced by use of
degassed materials and solvents, and enhanced by use of
oxygenated water.
plex.[59,60] In contrast, oxidative denitrifications of nitroalkanes,
catalysed by flavin-dependent enzymes of bacterial, fungal and
plant origin[60–63] consume molecular oxygen, and as we have
noted earlier,[44] at least one enzyme-catalysed denitrification
involves a deprotonation of the nitroalkane by the carboxylate of
an active site aspartate residue.[44] There are interesting parallels
here, and perhaps also some warnings for the interpretation of
data on deprotonations of nitroalkanes when these have been
monitored by UV–Vis spectroscopy. These observations will be
pursued elsewhere; for the present, the important conclusion is
that this oxidative Nef chemistry occurs after deprotonation of
the nitroalkane and will not complicate measurements of
deprotonation rates by racemization of optically active nitroalk-
anes.
Reaction kinetics
For direct comparability with the results of Lewis, the solvent
chosen for the study of 3 was that used by Lewis, 54:46 (wt:wt)
t-BuOH:water. To confirm conditions under which intramolecular
deprotonations might be observed we initially examined
racemizations in the presence of varying amounts of tetra-
A solution of 4 (ca 10ꢀ4 M) in undegassed water similarly
exhibited only a shoulder at 240 nm, and addition of dilute
sodium hydroxide produced a new peak at 275 nm, associated
formation of the nitronate, again stable for prolonged periods. On
acidification, however, the solution did not yield any indication of
a peak corresponding to the nitronic acid, but immediately
methylguanidine (TMG), a strong base (for TMGHþ, pKa ¼ 13.6[64]
)
capable of deprotonating quantitatively both the carboxylic acid
and the nitroalkane. Under these conditions, with undegassed
solvents, material recovered from solutions at completion of the
racemizations contained up to 20% of 4-oxopentanoic acid, 5
(R ¼ Me).
showed a new absorption at 244 nm (emax 11 000 Mꢀ1 cmꢀ1
)
corresponding closely to that of 4-oxo-4-phenylbutanoic acid,
5(R ¼ Ph), the Nef product, whose presence was confirmed by
subsequent isolation and comparison with authentic material.
When solvents and reagents were degassed, acidification of the
nitronate produced a peak at 266 nm consistent with formation
of the nitronic acid. Ketone formation was suppressed, and 4 as
well as 5 could then be recovered after the cycle of basification
and re-acidification.
With less than stoichiometric amounts of TMG, loss of optical
activity was first order. With excess TMG, behaviour was complex,
with a fast initial phase followed by a prolonged slow phase. In
view of the complexity across the range, initial rates, (df/dtt ¼ 0),
were determined for all reactions and divided by initial rotations
(fo) to yield comparable first order rate constants which are
presented in Table 1 and graphically in Fig. 3.
These observations are consistent with formation of ketonic
products, not from a hydrolytic Nef reaction, but from a process
involving reaction between dissolved oxygen and the nitronic
acids, with the carboxylic acid in 4 facilitating the reaction.
Observations on more concentrated solutions by 1H-NMR
spectroscopy support this conclusion. Solutions of the
nitronate anion of 4 (10ꢀ2 M) were prepared by addition to an
excess of sodium deuteroxide in undegassed 30:70 vol:vol:
CD3CN:D2O mixture. Besides the phenyl signals, the spectrum
showed 2H-triplets[55] at d 2.99 and at d 2.04 and in this basic
solution there was no indication of decay either to nitroalkane or
to Nef product. Addition of 2 molar equivalents of CD3COOD
(relative to the sodium deuteroxide) induced an immediate small
change in chemical shifts with distinct broadening of the signal at
d 2.99, and then evolution of the spectrum to that of a mixture of
4 and 5 (R ¼ Ph) over a period of less than 5 min at 208C. The
amount 5 formed was compatible with consumption of the
available oxygen in the NMR tube[56] and, if air was bubbled
through the tube immediately after acidification, 5 was the
exclusive product. Repetition of the experiment with thoroughly
degassed solutions, with addition of acetic acid, under argon,
completely suppressed formation of 5 (R ¼ Ph). When phenylni-
tropropane, was used similar results were obtained but the
conversions were much slower, taking place over 45 min.
Oxidative cleavage of nitronates is a synthetically preferred
method of converting nitroalkanes to carbonyl compounds,[57]
but with one exception,[58] molecular oxygen has not been used
as the oxidant, and the only in vitro reactions of dioxygen with
nitronates reported are those of a nitronate–copper com-
The plot shows a break at base equivalence point linking two
linear sections, and is consistent with the reaction sequence
shown in Scheme 3, in which the chiral compound which we
*
depict as a dibasic acid, HO ꢂ CH, is first deprotonated
ꢀ
*
stoichiometrically to yield the carboxylate, O ꢂ CH. The
kinetically significant processes are then deprotonations at the
carbon of the nitroalkane. The intramolecular process (governed
by kintra) results in racemization, but, because the nitroalkane is at
least 103 times less basic than the carboxylic acid, does not
consume additional base. With excess base, deprotonations
Table 1. Initial rates for racemizations of 0.018 M
4-nitropentanoic acid, 3, in 54:46 (wt:wt) t-BuOH:water at
30.98C in the presence of tetramethylguanidine
[TMG]/[3]
105 k/sꢀ1
0.251
0.522
0.996
1.136
1.445
1.512
1.990
2.590
3.85
7.22
16.9
88.2
215
277
530
827
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 603–613

PRIMARY KINETIC HYDROGEN ISOTOPE EFFECT
which describes the increase in rate proportional to the fraction
of the 3 in its carboxylate form, and confirms the absence of
significant competing water-catalysed reaction. The slope of
the line, 16.9 ꢃ 10ꢀ5 sꢀ1, is then the rate constant for the
intramolecular reaction and in reasonable agreement with the
value (18.1 ꢃ 10ꢀ5 sꢀ1) obtained by Lewis for the same reaction
using pyridine as base.[21,22] Lewis also obtained a value for the
second order rate constant for acetate catalysed iodination of the
methyl ester of 4-nitropentanoic acid[65] in the same medium
(4.0 ꢃ 10ꢀ5 Mꢀ1 sꢀ1), and the ratio of first and second order rate
constants gives the effective molarity, EM ꢁ 4.3, of the catalysing
carboxylate in 4-nitropentanoic acid, a low value but comparable
with those tabulated for other such intramolecular carbon acid
deprotonations.[66–68]
The much steeper line after equivalence point has the form:
105 ꢃ kinit ¼ 517ðꢄ14Þ ꢃ ½base ratioꢅ ꢀ 511ðꢄ25Þ
(2)
Figure 3. Effect of added tetramethylguanidine on rates of racemization
of optically active 4-nitropentanoic acid 3 at 30.98C in 54:46 (wt:wt)
t-BuOH:water
and represents the contribution from proton abstractions by the
excess strong base from the nitroalkane moiety of the
carboxylate. Correction for initial NPA concentration gives a
second order rate constant, kinter ¼ 2.9 ꢃ 10ꢀ2 Mꢀ1 sꢀ1. Since
TMG is sufficiently basic to generate some hydroxide in this
medium, this rate constant will itself be a complex quantity
reflecting contributions from both active bases.
For the reactions of 4 and its ester, the solvent mixture was
changed to 2:1 wt:wt dimethoxyethane:water (28.6:71.4 mole
ratio), to permit later use of D2O in the experiments.
Racemizations of 4 in this mixture were similarly examined,
but with the TMG replaced by the less basic triethylamine
(pKa ¼ 11.01 for Et3NHþ,) which, when in excess, was shown by
separate UV–Vis observations to be capable of fully deprotonat-
ing both carboxyl group and the CH of the nitroalkane in this
medium. For these reactions, losses of optical activity were
first order for base ratios <1, and showed little deviation from
first order behaviour with 1 < base ratio < 2. For full compar-
ability with the behaviour of 3 and TMG however, initial rates
were determined as before and are presented in Table 2 and
graphically in Fig. 4. Recovery of material from solutions at
completion of the racemizations now yielded 4-nitro-4-
phenylbutanoic acid 4, with less than 2% of 4-oxo-4-
phenylbutanoic acid 5(R ¼ Ph).
Scheme 3. Reaction scheme for loss of optical activity in reaction of
strong bases with 3 or 4
governed by kinter yield a carboxylate-nitronate dianion, ꢀO ꢂ Cꢀ,
with consumption of an additional equivalent of added base.
The initial linear portion of the plot in Fig. 4 has the equation
105 ꢃ kinit ¼ 16:9 ðꢄ1:3Þ ꢃ ½base ratioꢅ ꢀ 0:48 ðꢄ0:75Þ
(1)
The plot of rate constant against base ratio shows that, with
less than the full equivalent of base, rates are again proportional
to the fraction of acid in its salt form with no indication of a
Table 2. Initial rates for racemizations of 0.023 M
4-nitro-4-phenylbutanoic acid 4, in 2:1 (wt::wt) DME:water at
30.048C in presence of triethylamine
[Et3N]/[4]
104 k/sꢀ1
0.148
0.493
0.642
0.790
0.888
0.987
1.480
1.628
1.974
2.50
13.6
16.8
20.2
23.0
27.2
26.3
27.2
27.9
Figure 4. Effect of added triethylamine on rates of racemization of
optically active 4-nitro-4-phenylbutanoic acid 4 at 30.048C in 2:1 (wt:wt)
DME:water
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

N. BACKSTROM ET AL.
significant water-catalysed process
involvement in deprotonation of the nitroalkane. The effect is not
compatible with involvement of solvent derived hydron in
covalent bonding changes in the rate-limiting step.
104 ꢃ kinit ¼ 26:6ðꢄ0:9Þ ꢃ ½base ratioꢅ ꢀ 0:42ðꢄ0:58Þ (3)
The second was that some exchange of nitroalkane proton
with solvent might occur independently of the racemization (i.e.
The slope of the line gives the rate constant for the
,
spontaneous reaction of the salt of 4, k ¼ 26.6(ꢄ0.9)ꢃ 10ꢀ4 sꢀ1
k
exchange > kracemisation),[70] possibly by
a discrete transient
some 15-fold faster than the reaction of 3, reflecting both the
change in medium, and the slightly enhanced acidity of the
carbon acid associated with replacement of methyl by phenyl in
the nitroalkane. In contrast to the behaviour of 3 with TMG, rates
are little affected by excess of the weaker triethylamine showing,
in terms of the reaction scheme (Scheme 3), that this base neither
competes efficiently with the intramolecular carboxylate, nor
generates sufficient hydroxide in this medium at the concen-
trations used for its effects to be obvious at the base
concentrations used.
pyramidal hydrogen-bonded carbanion of the type suggested
by Bordwell et al.[71] Such exchange has been shown to lead to
erroneously low values for kD in reactions of deuteriated
isotopomers.[72] The excellent first order behaviour observed in
racemizations in the D2O/DME mixture in the determinations
solvent isotope effect militates against this, but exchange
reaction of 4 in D2O/d10–DME was also monitored directly by
1H-NMR spectroscopy to follow loss of the signal associated the
nitroalkane hydrogen (d 5.6). With one equivalent of triethyla-
mine, the first order rate constant at 30.18C was 33.2
(ꢄ2.8) ꢃ 10ꢀ4 sꢀ1, giving an apparent kexchange/kracemisation ¼ 1.25.
1.25. However, allowance for the solvent isotope effect associated
with replacement of H2O in the racemization experiments
To estimate the effective molarity of the intramolecular
carboxylate, racemizations of optically active methyl ester of
4
were measured also using buffered acetic acid
([HOAc]:[NaOAc] ¼ 1.0). These were cleanly first order and the
dependence of rates on concentration of sodium acetate
(Table 3) is given by the relationship:
by D2O in the NMR experiment (kH O=kD O ¼ 0.73) reduces the
ratio to 0.91(ꢄ0.14), and we conclude that, within the combined
experimental uncertainties, kracemisation ¼ kexchange in these reac-
tions.
2
2
105 ꢃ kobs ¼ 19:3ðꢄ3:0Þ ꢃ ½NaOAcꢅ þ 1:5ðꢄ0:6Þ
(4)
Comparing the second order rate constant for catalysis by
acetate with that for spontaneous reaction of the carboxylate of 4
yields EM ¼ 13.7, marginally larger only than that found for 3.
Generally, the behaviours of 3 and 4 are closely comparable,
but before moving to the primary kinetic isotope effects in 4, we
consider two possibly related complications to both measure-
ment and modelling.
The first is that water might be involved as a proton shuttle in
the intramolecular deprotonation, and this has been probed by
measurement of the solvent isotope effect. Rates for racemiza-
tions of optically active 4 were measured in at 30.048C
in a D2O–DME mixtures, using triethylamine with [Et3N]/
[4] ¼ 0.800. Loss of optical activity was first order, and duplicate
runs gave k ¼ 28.6 (ꢄ0.5) ꢃ 10ꢀ4 sꢀ1. From Equation (3), the rate
constant for reaction in an H2O/DME mixture at the same
Primary kinetic isotope effects
The practical difficulties in measuring kinetic isotope effects in
these racemizations are associated with ensuring that the
isotopomeric reactants experience identical conditions of
solvent, temperature and degree of neutralisation. To achieve
this, we have resolved partially deuteriated mixtures of
isotopomers and then monitored the course of racemization
of these isotopically mixed samples, catalysed by less than a full
equivalent of TMG, in the case of 3, and Et3N in the case of 4. This
procedure assumes firstly that g-isotopic substitution has no
significant effect on the acidity of the carboxylic acid, and
secondly, that the substitution does not significantly affect the
optical activity. For the first, we note that the effects at the
a-deuteriation on acidity of carboxylic acids are small
CD
(KaCH3 =Ka ¼ 0.03 in acetic acid)
and suggest that the
[73–75]
3
base ratio is 20.8 ꢃ 10ꢀ4 sꢀ1, giving kH O=kD O ¼ 0.73(ꢄ0.04), for
the intramolecularly catalysed process. Interpretation here is
complicated a little by the use of mixed solvent, but this ratio is
close to that reported for the deprotonations of 2-nitropropane
2
2
effects of isotopic substitution even more remote from the
carboxylate can safely be ignored. For the second, the resolution
and measurement of optical activity in compounds such
[76]
PhCHDCH3
shows there must indeed be an effect from
by hydroxide in water[28] (kH O=kD O ¼ 0.68), and to the
fractionation factor for isotopic distribution between bulk water
and water hydrogen-bonded to oxy-anionic base (f ¼ 0.74).[69]
The inverse effect is interpreted in terms the stronger
DOD. . ..ꢀOR bonds compared to HOH. . ..ꢀOR, and a requirement
for solvation of the oxy-anionic base to be reduced prior to its
2
2
isotopic substitution, but the associated changes in specific
rotations are very small compared to those for the compounds
used, and we believe that this effect can also be safely ignored.
With these simplifying assumptions, the optical rotation, ft, of
the partially neutralised isotopic mixture is expected to decay
following relationship (5), where finf is the optical rotation at
infinite time:
H:t
D:t
ft ¼ fðinfÞ ꢀ Fr Df eꢀk ꢀ ð1 ꢀ FrðHÞÞDf eꢀk
(5)
ðHÞ
Table 3. Rate constants for racemizations of methyl
4-nitro-4-phenylbutanoate in 1:1 NaOAC:HOAc buffered 2:1
DME: water at 30.048C
The fraction here (Fr(H)) is the fraction undeuterated within the
optically active component, and not in the total mixture which
may contain racemic material, both labelled and unlabelled, but
which is ‘transparent’ to polarimetry. The relationships then sums
the contributions from each optically active isotopomer to the
total optical rotation of the mixture, and rates, extracted
from time dependences of the rotations by non-linear
regression,[77,78] are presented in Table 4, along with the
associated activation parameters.
[NaOAc]
105 k/sꢀ1
0.140
0.210
0.282
4.32
5.30
7.05
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 603–613

PRIMARY KINETIC HYDROGEN ISOTOPE EFFECT
Table 4. Primary kinetic isotope effects for base-catalysed racemizations of carboxylic acids of 3 and 4, and of the methyl ester of 4
T (8C)
104 kH/sꢀ1
kH/kD
Activation parameters
Acid 3 in t-BuOH: H2O. Salt formation with TMG.zH
DG303 ¼ 88.6(W3.8) kJ molS1
30.9
1.69 (ꢄ1.3)
5.68 (W0.17)
Acid 4 in DME:H2O. Salt formation with Et3N.
5.99 (W0.28)a
EDa S EHa ¼ 5.5(W0.1) kJ molS1, AH/AD ¼ 0.63(W1.03)
20.16
32.04
50.02
25.0
11.6 (W0.3)
35.5 (W0.8)
151 (W2.5)
5.51 (W0.06)a
4.87 (W0.12)a
EHa ¼ 67.6(W1.2) kJ molS1, AH ¼ 1.68(W1.2) T 109
DHHz ¼ 65.1(W1.2) kJ molS1, DSHz ¼ S78.8(W3.8) J KS1 molS1
zH
DG ¼ 88.6(W3.8) kJ molS1
*
*
18.5
5.78
298
Methyl ester of 4 in DME:H2O in presence of 0.25 M NaOAc
7.18 (W0.09)
6.33 (W0.17)
30.40
50.00
25.0
4.18 (W0.01)
21.1 (W0.02)
EDa S EHa ¼ 5.2 kJ molS1, AH/AD ¼ 1.08
*
*
2.58
7.43
a Mean of two experiments.
*
Extrapolated values.
than that in the a staggered-to-gauche change in butane
(3.8 kJ molꢀ1). If the effects on the pKa of carboxylic acids of
moving from water to DMSO (4.8 in water to 12.3 in DMSO[80]) are
taken to reflect the effects of desolvation on the stability of
the anionic base in a polar non-aqueous environment, then the
energetic cost of the desolvation may be as much as
42.8 kJ molꢀ1, so that half of the measured free energy barrier
might be accounted for by processes not linked directly to the
proton transfer which occurs in step B, with no water molecule
having more than ‘interested-spectator’ status. This may lead to a
carbanion in which charge remains largely localised on carbon,[71]
and the process is then completed by charge and solvent
re-organisation to yield the planar, optically inactive nitronate in
step C. The extent of linkage between the proton motion and the
secondary re-organisation remains to be established, but the fact
of a substantial kinetic isotope effect and the equality of
exchange and racemization rates demands that step C is
substantially faster than the reverse of step B.
Table 4 includes Arrhenius and Eyring parameters for the
H-isotopomer of 4, and we note that the reaction exhibits a large
negative entropy of activation. The kie for the intramolecular
reaction in 4 does not differ significantly for that found for 3, and
the temperature dependence yields Arrhenius parameters
indicating that the rate ratios are associated with differences
in activation energies (EaD ꢀ EHa ) rather than in pre-exponential
factors (AH/AD), and thus not consistent with major tunnelling
contributions.
Conclusions
Taken together, the combined results of the studies described
above are consistent with the racemizations of salts 3 and 4
involving a direct transfer of proton from the nitroalkane carbon
to an anionic oxygen of the intramolecular carboxylate via a
six-centre cyclic array, and that rates of racemization are a reliable
measure of deprotonation rates. The kie found in this work for the
racemization 4-nitropentanoic acid, 3 (kH/kD ¼ 5.68 ꢄ 0.17 at
318C) is satisfyingly close to that reported by Lewis for
deprotonations of 4-nitropentanoic acid, using iodine-trapping
methods (kH/kD ¼ 5.5 ꢄ 0.5). Replacement of methyl by phenyl,
and change of solvent from aqueous t-BuOH to aqueous
dimethoxyethane makes little difference with kH/kD ¼ 5.51 ꢄ 0.06
0.06 at 328C for racemizations of 4 in aqueous dimethoxyethane.
In considering the detailed course of the reaction (Scheme 4),
we assume that ground state conformation of the reactant in the
medium is the extended one shown, minimizing torsional
interactions along the chain, and expect that the anionic oxygens
of the carboxylate are solvated, with hydrogen bonding with a
number of water molecules,[79] although we show only one here.
Before the intramolecular proton transfer can occur, there must
be conformational adjustment to bring the basic carboxylate
group within reacting distance of the nitroalkane hydrogen, and
desolvation of the oxyanion (Step A). Since the transfer occurs
through a six-centre cyclic array, the energy increase associated
with the conformational change is likely to be only a little greater
Scheme 4. Separation of steps in the racemizations of the salts of 4 (L ¼ H and D)
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

N. BACKSTROM ET AL.
Notably, the kie for the intramolecular reactions is significantly
smaller than that catalysed intermolecularly by acetate (this work,
7.2 at 258C), but again, the intermolecular kinetic isotope effect is
associated with differences in activation energy (EDa ꢀ EaH) rather
than in pre-exponential factors. The smaller intramolecular effect
may be associated with some stiffening of isotopically sensitive
modes in the cyclic transition state, and indeed, may be signalling
that the six-membered ring is not an ideal size for accommodat-
ing a proton transfer. Studies on intramolecular hydrogen
bonding using diamides[81] have suggest that rings as large as
nine may be necessary to accommodate fully relaxed linear
hydrogen bonds.
These measurements on a well-defined intramolecular process
will be used to test and parametrise computational methods
which can then be applied to much more complex reacting
systems, and the results of a parallel computational study will be
reported separately.
pressure to yield 8.9 g of yellow oil which was purified by
bulb-to-bulb distillation (1008C at 1 mm Hg). dH (300 MHz, CDCl3)
4.68 (1H, m, 14 lines, W ¼ 35 Hz), 3.72 (3H, s), 2.43 (2H, m, 13 lines,
width ¼ 57 Hz), 2.29 (1H, m, 14 lines, width ¼ 37 Hz), 2.12, (1H, m,
12 lines, W ¼ 34 Hz), 1.61 (3H, d, J ¼ 8 Hz); dC (75 MHz, CDCl3)
172.5, 89.9, 82.5, 51.9, 34.0, 29.9, 28.728, 21.6, 19.16; nmax/cmꢀ1
,
2995, 2955, 1743, 1543, 1554, 1438.
4-Nitropentanoic acid, 3
Methyl 4-nitropentanoate (4.1 g) and 2:1 vol:vol TFA:water
(100 ml) were heated at 808C overnight in an RB flask before
removal of solvents under reduced pressure by rotary evapor-
ation. Toluene (30 ml) was added to the residue in the flask, and
then also evaporated under reduced pressure to leave a white
crystalline solid, recrystallised for ether:petroleum (3.54 g). mp
338C (lit., 34.532[51]). dH (300 MHz, CDCl3) 11.5 (1H, s), 4.85 (1H, m,
14 lines, W ¼ 35 Hz), 2.44 (2H, m, 13 lines, width ¼ 57 Hz), 2.24 (1H,
m, 14 lines, width ¼ 37 Hz), 2.07, (1H, m, 12 lines, W ¼ 34 Hz), 1.77
(3H, d, J ¼ 8 Hz); dC (75 MHz, CDCl3) 177.9, 89.9, 82.6, 30.3, 29.0,
19.05; nmax/cmꢀ1, 2583 (b), 2994, 1712, 1548, 1450, 1416. m/z
(CIþ) 179 (51%, M þ NHþ4 ) 162 (100%, M þ Hþ), 146,(64%).
EXPERIMENTAL SECTION
General methods
NMR spectra were recorded on Varian Unity INOVA-300 or Bru¨ker
500 AMX spectrometers. Chemical shifts are parts per million
(ppm) downfield from tetramethylsilane. Signal splittings are
quoted as: singlet (s), doublet (d), triplet (t), quartet (q) and
multiplet (m). Infrared spectra were recorded on an ATI Mattson
Genesis Series FTIR and were run as liquid films or as films
evaporated from deuteriated chloroform. UV–Vis spectra were
recorded on a Cary 50 UV–Visible spectrophotometer. Mass
spectra were recorded on a Kratos MS25, Fisons VG Trio 2000 or
on a Micromass platform. Modes of ionisation are electron impact
(EI), positive chemical ionisation (þCI) using ammonia and fast
atom bombardment (FAB), positive electrospray (þES) and
negative electrospray (-ES). Melting points were recorded on a
Kofler heated stage microscope and are uncorrected. Thin layer
chromatography (TLC) was carried out on Polygram Sil G/UV254
0.25 mm silica gel plates with solvent systems as indicated.
Analytical gas chromatography was carried out using: Perki-
n-Elmer capillary gas chromatography model 8310 with flame
ionisation detection (FID). Analytical HPLC was carried out using:
Waters 510 HPLC pump, Bondclone C18 100 ꢃ 8 mm column,
Perkin-Elmer LC480 Diode Array System with detection at 255
and 285 nm pre-mixed degassed solvent systems 75:25
Methyl 4-deuterio- 4-nitropentanoate, 3(L ¼ D)
Methyl 4-nitropentanoate (2.0 g) was dissolved in methanol-OD
(10 ml) and pyridine (1 ml) added.. The reaction mixture was
heated to 608C and the exchange was followed by 1H-NMR
spectroscopy until appropriate deuterium incorporation, by
integration of the signal at d4.68. The solvent was then removed
under reduced pressure, the residue taken up in ether, and the
pyridine removed with successive washes with dil. hydrochloric
acid. The ethereal solution was then dried over sodium sulphate,
before evaporation and distillation as before.
4-Deutero-4-nitropentanoic acid, 3(L ¼ D)
Methyl-4-deutero–4-nitropentanoate (2.0 g) was hydrolysed by
overnight treatment with 2:1 water:TFA mixture at 808C
and isolation as described above to give crystals, (1.75 g).
dH (300 MHz, CDCl3) 11.5 (1H, s), 2.44 (2H, m, width ¼ 50 Hz), 2.24
(1H, m, 6 lines, width ¼ 31 Hz), 2.07, (1H, m, 6 lines, W ¼ 30 Hz),
1.70 (3H, s); dC (75 MHz, CDCl3) 177.9, 89.9, 82.6, 30.3, 29.0, 19.05;
nmax/cmꢀ1, 2583 (b), 2994, 1712, 1548, 1450, 1416.
MeOH:H2O or as stated in the text, flow rate 2 ml minꢀ1
.
Resolution of 4-nitropentanoic acid and
4-deutero-4-nitropentanoic acid
The compounds
4-Nitropentanoic acid was resolved by crystallisation of its
quinine salt as described by Theilacker and Wendtland.[51] The
acid (1.0 g) was dissolved in acetone (10 cm3) with gentle
warming. Quinine (2.1 g) was also dissolved in acetone (10 cm3),
and then the two warm solutions mixed and allowed to stand
until no more crystals formed. The mother liquor was then
pipetted away and the crystals washed with a little cold acetone.
The mother liquor was retained for subsequent crops. The solid
was then dissolved in the minimum amount of methanol (5 cm3)
and sulphuric acid solution (20 cm3 of 20%) added. The resolved
compound was extracted into ether (30 cm3) which was dried
over sodium sulphate before evaporation of solvent to leave a
clear oil which slowly crystallised, (0.42 g, [a]D ¼ 138 in
acetonitrile). Similar treatment of 4-deuterio-4-nitropentanoic
acid yielded crystals (0.42 g. [a]D ¼ 9.68 in acetonitrile). Integ-
Phenylnitropropane
Propiophenone oxime was oxidised with sodium perborate in
acetic acid according to the method of Olah et al.[82], to give
phenylnitropropane in 45% yield after distillation; b.p. 728C
at 2 mm Hg (lit 758C at 1 mm Hg).
Methyl 4-nitropentanoate
AmberlystTM A-21 (5.00 g), nitroethane (6.0 g) and methyl acrylate
(5.81 g) were mixed and the solution left stirring overnight at
408C. The reaction was monitored by NMR and on completion the
mixture was cooled and filtered. The resin was washed with a
little dichloromethane, and the combined liquids dried over
sodium sulphate before removal of solvent under reduced
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 603–613

PRIMARY KINETIC HYDROGEN ISOTOPE EFFECT
ration of the signals at d 4.85 and at d 1.70 yielded estimate of the
deuterium incorporation in samples used in isotope effect
measurements.
29.8, 28.5; nmax(film)/cmꢀ1, 1703(s), 1552; m/z (CI) 227, 180(100%);
(EI) 163, 117(100%); (ES-) 208(M-1, 100%), 113(40%).
For the deuteriated material the 1H-NMR spectrum was closely
similar to the all-H material, with diminution of the signal at d5.65,
consistent with 77% deuteriation; nmax(film)/cmꢀ1, 1715, 1552;
m/z (CI) 228(75%), 181(100%); (EI) 164(65%), 118(100%).
Methyl 4-nitro-4-phenylbutanoate
This ester was prepared by the method of Ballini and Bosica.[48]
Amberlyst-A21TM resin (1.2 g) was added to a stirred mixture of
methyl acrylate (1.1 g, 12.8 mmol) and nitrophenylmethane[83]
(1.42 g, 10.2 mmol), under nitrogen. After stirring overnight at
258C, the resin was separated from the liquid and extracted
3 times with 20 ml portions of ether. The liquid and combined
ether extracts were then dried over anhydrous sodium sulphate,
and the ether was removed under vacuum yielding methyl
4-nitro-4-phenylbutanoate as a clear oil (1.8 g, 78.3%), of
sufficient purity for use in subsequent hydrolysis. Distillation of
a portion yielded material, b.p. 1348C at 1.4 mm of Hg (lit., 2058C
at 0.4 mm); dH (300 MHz, CDCl3) 7.44 (5H, m), 5.65 (1H, dd, J ¼ 5.8,
J ¼ 8.5), 3.73(3H, s), 2.82(1H, m, 14 lines, width 34 Hz), 2.55–2.37
(3H, m); dC (75 MHz, CDCl3) 172.1, 133.9, 129.9, 127.9, 127.6, 89.9,
51.8, 30.0, 28.8; nmax(film)/cmꢀ1, 1737(s), 1552; m/z (CI) 241
(100%), 177; (EI) 177, 117; (ES-) 222(M-1, 20%), 208(100%).
Sensitivities to oxygen
Observations by UV-vis spectroscopy were on a Cary 50 Bio
Spectrometer. NMR experiments used a 400 MHz INOVA instru-
ment with probe held at 208C.
To a NMR tube containing 1 ml of phenylnitropropane and
0.5 ml of pivalic acid as internal reference was added CD3CN (230
ml) and standardised NaOD in D2O (460 ml of 0.0906 M). For initial
experiments no precautions for deoxygenation were taken. The
NMR spectrum and then d4-acetic acid (4.5 ml) was injected into
the tube to give a solution with pD of 5.6 at 258C. Spectra were
then collected. For reactions without oxygen, the tube was fitted
with a subbaTM seal and nitrogen was bubbled slowly for 1 h
through the solution via syringe needle before addition of the
phenylnitropropane and pivalic acid and NaOD, and then
again before acidification. Simliar experiments were carried
out with tubes containing a weighed amount (0.0011 g) of
4-nitro-4-phenylbutanoic acid in CD3CN (230 ml) and standar-
dized NaOD in D2O (460 ml of 0.0891 M). Results are described in
the text.
Deuteration of methyl 4-nitro-4-phenylbutanoate
Methyl 4-phenyl-4-nitrobutanoate, (0.5 g, 2.25 mmol) was added
to a 25 ml round bottom flask containing of methanol-D (4.0 cm3)
and of pyridine (1.0 cm3). This was left for 24 h at 508C, then the
methanol and pyridine were removed under vacuum. The
process was then repeated with fresh reagents. This yielded an
oil, which was purified by bulb-to-bulb distillation yielding a
clear oil, methyl 4-deuterio-4-nitro-4-phenylbutanoate (0.454 g,
2.03 mmol, 90.4%). Its proton NMR spectrum was closely similar
to the all-H material, with diminution of the signal at d5.65,
consistent with 77% deuteriation; nmax(film)/cmꢀ1, 1738(s), 1552;
m/z (CI) 242(100%), 178(65%); (EI) 178(15%), 118(100%),
105(80%), (ES-) 236(100%), 222(M-1, 90%).
Reaction kinetics
Polarimetry
Optical rotations were monitored using a Perkin-Elmer PE 341
polarimeter. Reactions were run in a thermostatted cell whose
temperature was monitored at the cell using a platinum
resistance thermometer. Cell temperatures were constant over
the course of a kinetic run to ꢄ0.028C. Cell volume was 5 ml and
the path length was 10 cm. The recorder voltage from the
polarimeter and the output of the platinum resistance
thermometer were taken via a Pt100 4-channel analogue-
to-digital converter (from Pico Technology Limited) directly to a
standard PC and logged continuously (1 reading per second)
using the supplied data logging software (Picolog). Reactions
were monitored for at least five halflives (of the slower
component when mixtures of isotopomers were used). Data
was transferred to Microsoft Excel for treatment, usually
non-linear least squares fitting to the appropriate function using
the Solver add-in. Uncertainties and statistics associated with
fitting were generated using the SolvStat macro of Billo.[84] For
fitting to Equation 5 in the text to extract kinetic isotope effects,
there is a strong covariance between the kH/kD and the fraction of
deuteriated material, Fr(H). Initial fitting was therefore carried out
with the value of Fr(H) fixed at that determined by NMR analysis of
the sample of 3 or 4 used. Fitting was then repeated with Fr(H)
treated as a fully adjustable parameter. Data were only deemed
acceptable if relaxing the restriction on Fr(H) did not significantly
change its value or any of the other derived parameters.
Resolution of Methyl 4-nitro-4-phenylbutanoate and Methyl
4-deuterio-4-nitro-4-phenylbutanoate
Optically active methyl 4-nitro-4-phenylbutanoate was obtained
by chromatography on a 25 ꢃ 1 cm ChiracelTM OD-H column
eluting with 1% 2-propanol in hexane at 4 ml minꢀ1. A full
separation of enantiomers was not achieved but judicious cutting
of the peaks yielded samples with [a]365 up to 758 for
dichloromethane solutions.
4-Nitro-4-phenylbutanoic acid and
4-deuterio-4-nitro-4-phenylbutanoic acid, 4 (L ¼ H and D)
Methyl 4-nitro-4-phenylbutanoate (70 mg, 0.315 mmol) and 2:1
v:v TFA:D2O solution (2 cmꢀ3) was placed in a screw-top test tube
which was heated at 558C for 12 h, with reaction monitoring by
periodic 1H-NMR spectroscopy. Removal of volatiles under
vacuum yielded 4-nitro-4-phenylbutanoic acid as a waxy solid
(48 mg), mp 78–838C and material used in measurements was
purified by crystallisation from petrol giving white needles, mp
94–988C.
Racemizations of 4-nitropentanoic acid, 3
For all-H material, dH (300 MHz, CDCl3) 7.44 (5H, m), 5.65 (1H,
dd, J ¼ 5.8, J ¼ 8.5), 2.82(1H, m, 14 lines, width 36 Hz), 2.55–2.37
(3H, m); dC (75 MHz, CDCl3) 177.1, 133.7, 130.0, 129.1, 127.6, 89.7,
For racemizations of 4-nitropentanoic acid, a stock solution of
optically active 4-nitropentanoic acid (0.181 M) in 54:46 wt:wt
t-BuOH: water was prepared, having an optical rotation
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

N. BACKSTROM ET AL.
f ¼ –3.4118 (at 598 nm and 30.98C). These solutions showed slow
loss of optical activity on storage (presumably via the small
equilibrium concentration of the carboxylate) and stored
solutions were acidified ([HCl] ¼ 2.5 ꢃ 10ꢀ3 M) to suppress
ionisation of the carboxylic acid. For the rate measurements,
weighed quantities of TMG (0.103 g, 0.208 g, 0.026 g, 0.206 g,
0.1501 g, 0.157 g, 0.054 g, 0.1176 g) were added with shaking to
5 ml of the stock solution in a volumetric flask, which was
thermostatted to the same temperature as the polarimeter cell.
showed 4-nitro-4-phenylbutanoic acid and no more than 2% of
4-oxo-4-phenylbutanoic acid.
Kinetics of deuteriation of 4-nitro-4-phenylbutanoic acid,
4, by NMR
The solvent was a mixture of d10-DME (2.91 g) and D2O (1.45 g).
4-Nitro-4-phenylbutanoic acid (0.032 g) was dissolved in the
solvent mixture (0.75 ml), and transferred to an NMR tube. After
equilibration to probe temperature, an aliquot of triethylamine
was added by micrometer syringe, the tube shaken quickly and
then returned to the probe. Spectra were then recorded at
appropriate intervals, with at least 10 s between acquisitions.
The probe temperature was measured before and after the
experiment using signal separations in the spectrum of ethylene
glycol. Decrease in ratio of integrals of signal at d5.45 to
the isolated signal of a non-exchanging hydrogen at d2.28 with
time was first order. Base ratios were measured by comparing
integrals from the methyl signal of the triethylamine with that at
d2.28. In two separate experiments rate constants were
31.7(ꢄ2.8) ꢃ 10ꢀ5 s (at 28.48C) and 34.7(ꢄ2.8) ꢃ 10ꢀ5 s (at
31.78C) for fully neutralised materials. Uncertainties take into
account the standard deviation of the fit, and in the base ratio.
Primary kinetic isotope effects in racemizations of
4-nitropentanoic acid, 3
Studies were conducted at 30.9, 41.1, and 49.38C. Separate stock
solutions of weighed amounts of optically active partially
deuteriated 4-nitropentanoic acid and of TMG in 54/46 w/w
tert-Butanol/water were prepared. Concentrations varied
between 0.68 and 0.15 M for the acid and 0.6 and 0.08 M for
the TMG, with concentrations chosen so that mixing would yield
solution of partially neutralised acid with a known acid:base ratio,
adjusted to give a conveniently observable rate. Before mixing,
these were brought to the reaction temperature and then 2.5 cm3
of each solution transferred to a thermostatted flask for mixing
before transferring by cannula to the cell as quickly as possible.
Data logging zero time was taken from the time of mixing, and
changes in optical activity at l ¼ 598 nm were monitored for at
least 30 ꢃ t1/2 anticipated for the H-isotopomer.
Primary kinetic isotope effects in racemizations of
4-nitro-4-phenylbutanoic acid, 4
Separate stock solutions of optically active partially deuteriated
4-nitro-4-phenylbutanoic acid (0.221 g in 45.1 ml) and a more
concentrated solution of triethylamine (0.86 g in 15 ml) were
made up in the 2:1 DME:H2O solvent mixture, as before. The acid
solution (5.0 ml) was placed in a stirred flask and equilibrated to
the reaction temperature. Triethylamine solution (0.1 ml) was
then added by micrometer syringe as rapidly as possibly for
mixing before transfer by cannula to the cell. Data logging zero
time was taken from the time of mixing, and changes in optical
activity at l ¼ 365 nm were monitored for at least 10 ꢃ t1/2
anticipated for the H-isotopomer.
Material was recovered from these racemizations by acidifica-
tion of the solutions with dil. hydrochloric acid, and extraction
with dichloromethane. Examination by 1H-NMR spectroscopy
showed, in all cases, the presence of laevulinic acid (up to 20% of
total) as well as 4-nitropentanoic acid.
Racemizations of 4-nitro-4-phenylbutanoic acid, 4
The reaction medium was 2:1 wt:wt mixture of dimethoxyetha-
ne:water, made up using dimethoxyethane freshly distilled
from calcium hydride and distilled deoxygenated water, and
stored in the dark under nitrogen. Stock solutions of optically
active 4-nitro-4-phenylbutanoic acid (0.24 g in 50.5 ml) and a
more concentrated solution of triethylamine (0.86 g in 15.0 ml)
were made up, under nitrogen, and sonicated briefly before
mixing under nitrogen. Typically, an aliquot of the acid solution
(5.0 ml) was placed in a stirred flask thermostatted to the reac-
tion temperature (30.068C) and a volume of the triethylamine
solution added by micrometer syringe as rapidly as possibly.
The mixed solution was then transferred by cannula to the
polarimeter cell, and changes in optical rotation monitored at
l ¼ 365 nm. Data logging was initiated at time of mixing of acid
and basic solutions. For a series of measurements the acid–base
ratio was varied and initial rotations were between 0.28 and 0.268.
For the solvent isotope effects, stock solutions of optically
active (0.276 g in 20.0 ml) and of triethylamine (0.105 g in 25 ml)
were made up in a mixture of DME (102.6 g) and D2O (56.4 g).
For the reaction, these solutions were separately equilibrated to
temperature and equal volumes (2.6 ml) mixed at the reaction
temperature before transfer to the cell as before. Changes in
optical rotation monitored at l ¼ 365 nm, where initial rotations
were ca 0.168.
Racemizations of the methyl ester of 4
Three acetate buffers (1:1 HOAc:NaOAc) in 2:1 DME:water were
prepared with [NaOAc] ¼ 0.28, 0.42 and 0.56 M. Equal volumes
(2.5 ml) of each buffer and a stock solution of 4-nitro-
4-phenylbutanoic acid (0.13 g in 12 g) were equilibrated to
temperature and mixed before transfer by cannula to the cell, and
changes in optical activity at l ¼ 365 nm were monitored for at
least 5 ꢃ t1/2
.
For the kinetic isotope effects in the racemization, separate
stock solutions in 2:1 DME:water of optically active partially
deuteriated 4-nitro-4-phenylbutanoic acid (0.223 g in 50 ml) and
an acetate buffer (AcOH 0.752 g, 0.0125 mol, and NaOAc, 1.025 g,
0.0125 mol, in 25 ml) were made up. The solutions were
separately equilibrated to temperature and equal volumes
(2.6 ml) mixed at reaction temperature (308C or 508C) before
transfer by cannula to the cell. Data logging zero time was taken
from the time of mixing.
Acknowledgements
Material was recovered from these racemizations by acidifica-
tion of the solutions with dil. hydrochloric acid, and extraction
into dichloromethane. Examination by 1H-NMR spectroscopy
We thank the University of Manchester and EPSRC (Grant number
GR/580486/01) for studentships (to support Simon Turega and
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 603–613

PRIMARY KINETIC HYDROGEN ISOTOPE EFFECT
Nick Backstrom respectively). We thank Raman Sharma and
Michelle Thorley for valuable discussions throughout this work
and Hannah Rablen for valuable experimental assistance. The
polarimeter was a generous gift from Astra-Zeneca, Alderley Park.
[40] C. G. Swain, E. C. Stivers, J. F. Reuwer, L. F. Schaad, J. Am. Chem. Soc.
1958, 80, 5885–5893.
[41] A. Thibblin, J. Phys. Org. Chem. 1988, 1, 161–167.
[42] A. Thibblin, P. Alhberg, Chem Soc. Reviews 1989, 18, 209–224.
[43] H. H. Koch, D. B. Dahlberg, J. Am. Chem. Soc. 1980, 102, 6102–6017.
[44] P. F. Fitzpatrick, M. Allen, A. Nagpal, M. P. Valley, Arch. Biochem.
Biophys. 2005, 433, 157–165.
REFERENCES
[45] M. P. Valley, S. E. Tichy, P. F. Fitzpatrick, J. Am. Chem. Soc. 2005, 127,
2062–2066.
[46] D. T. Major, D. M. York, J. Gao, J. Am. Chem. Soc. 2005, 127,
16374–16375.
[47] For nitrophenylethane, F. G. Bordwell, W. J. Boyle, J. Am. Chem. Soc.
1972, 94, 3907–3911.
[48] R. Ballini, G. Bosica, Eur. J. Org. Chem. 1998, 355–357.
[49] I. Erden, J. R. Keeffe, F. P. Xu, J. B. Zheng, J. Am. Chem. Soc. 1993,
115(21), 9834–9835.
[50] N. C. Deno, R. W. Gaugler, T. Spencer, J. Am. Chem. Soc. 1968, 31,
1968–1969.
[1] J. Bigeleisen, M. Wolfsberg, Adv. Phys. Chem. 1958, 1, 15–76.
[2] M. Wolfsberg, M. J. Stern, Pure Appl. Chem. 1964, 8, 225–242.
[3] M. J. Stern, M. Wolfsberg, J. Chem. Phys. 1966, 45, 2618–2630.
[4] For the most recent review, F. E. Romesberg, R. Schowen, Adv. Phys.
Org. Chem. 2004, 39, 27–77.
[5] L. B. Sims, D. E. Lewis, Bond order methods for calculating isotope
effects in organic reactions, Isotopes Org. Chem. 1984, 6, 161–259.
[6] B. A. Bahnson, J. P. Klinman, Methods Enzymol. 1995, 249, 373–397.
[7] J. Basran, S. Patel, M. J. Sutcliffe, N. S. Scrutton, J. Biol. Chem. 2001,
276, 6234–6242.
[51] W. Von Theilacker, G. Wendtland, Justus Leibigs Ann. Chem. 1950, 570,
33–53.
[52] J. U. Nef, Ann. 1894, 280, 263–291.
[8] J. Basran, M. J. Sutcliffe, N. S. Scrutton, Biochemistry 1999, 38,
3218–3222.
[53] H. W. Pinnock, Org. React. 1990, 38, 655–792.
[54] E. S. Lewis, in Chemistry of the Functional Groups, Suppl F, Part 2,
Chapter 16 (Ed.: S. Patai), Wiley and Sons, 1982.
[55] W. R. Bowman, B. T. Golding, W. P. Watson, J. Chem. Soc., Perkin Trans. 2
1980, 731–736.
[9] M. E. Schneider, M. J. Stern, J. Am. Chem. Soc 1972, 94, 1517–1522.
[10] The earliest consideration of nuclear tunnelling seems to have been
by F. Hund, Z. Phys. 1927, 43, 805–826; for a fuller historical view see
Reference [11].
[11] (Ed.: R. P. Bell). The Tunnel Effect in Chemistry, Chapman and Hall,
London, 1980, Chapter 1.
[56] Using the tube volume and a solubility of oxygen in water (8.2 ppm
at 258C).
[12] M. J. Knapp, J. P. Klinman, Eur. J. Biochem. 2002, 269, 3113–31121.
[13] M. J. Sutcliffe, N. S. Scrutton, Eur. J. Biochem. 2002, 269, 3096–3102.
[14] D. Antoniou, S. Nunez, S. D. Schwarz, Adv. Phys. Org. Chem. 2006, 41,
315–362, and references therein.
[57] R. Ballini, M. Petrini, Tetrahedron 2004, 60, 1017–1047.
[58] M. D. Nikalje, A. I. Sayyed, G. K. Dewkar, A. Sudalai, Tetrahedron Lett.
2000, 41, 959–961.
[59] E. Balogh-Hergovich, G. Speier, M. Reglier, L. Gioirgi, L. Parkany, Inorg
Chem. 1998, 37, 6535–6537.
[15] G. Tresadern, S. Nunez, P. Faulder, H. Wang, I. H. Hillier, N. A. Burton,
Faraday Discuss. 2002, 122, 223–242.
[60] E. Balogh-Hergovich, Z. Greczi, J. Kaiser, G. Speier, G. Huttner, L.
Zsolnai, Eur. J. Inorg. Chem. 2002, 1687–1696.
[61] C. R. Hipkin, M. A. Salem, D. Simpson, S. J. Wainwright, Biochem. J.
1999, 340, 491–495.
[62] P. F. Fitzpatrick, A. M. Orville, M. P. Nagpal, M. P. Valley, Arch. Biochem.
Biophys. 2005, 433(1), 157–165.
[16] P. F. Faulder, G. Tresadern, K. K. Chohan, N. S. Scrutton, M. J. Sutcliffe,
I. H. Hillier, N. A. Burton, J. Am. Chem. Soc. 2001, 123, 8604–8605.
[17] A. Kohen, J. P. Klinman, Acc. Chem. Res. 1988, 31, 397–404.
[18] A. J. Kirby, Angew. Chem., Int Ed. Engl. 1996, 35, 707–724.
[19] R. P. Bell, D. W. Earls, J. B. Henshall, J. Chem. Soc., Perkin Trans. 2 1976,
39–45.
[63] H. N. Little, J. Biol. Chem. 1951, 193, 347–358.
[64] S. J. Angyal, W. K. Warburton, J. Chem. Soc. 1951, 2492–2494.
[65] H. Wilson, J. D. Caldwell, E. S. Lewis, J. Org. Chem. 1973, 38, 564–566.
[66] A. J. Kirby, Adv. in Phys. Org. Chem. 1980, 17, 183–278.
[67] A. J. Kirby, Angew Chem,. Int. Ed. Engl. 1996, 35, 707–724.
[68] A. J. Kirby, Acc. Chem. Res. 1997, 30, 290–296.
[69] R. Alvarez, R. Schowen, in Isotopes in Organic Chemistry, Vol. 7,
Chapter 1., (Eds.: E. Buncel, C. C. Lee), Elsevier PL, Amsterdam, 1989.
[70] D. J. Cram, Fundamentals of Carbanion Chemistry, Chapter 3,
Academic Press, New York, 1965.
[71] F. G. Bordwell, J. E. Bartmess, J. A. Hautala, J. Org. Chem. 1978, 43,
3107–3113.
[72] O. Rogne, J. Chem. Soc., Chem. Commun. 1977, 695.
[73] S. Y. Bary, H. Gilboa, E. A. Halevi, J. Chem. Soc., Perkin Trans. 2 1979,
938–942.
[74] A. Streitwieser, H. S. Klein, J. Am. Chem. Soc. 1963, 85, 2759–2763.
[75] M. Paabo, R. G. Bates, R. A. Robinson, J. Phys. Chem. 1966, 70,
540–543.
[76] E. L. Eliel, J. Am. Chem. Soc. 1949, 71, 3970–3972.
[77] H. Strehlow, Rapid Reactions in Solution, VCH, New York, 1992.
[78] H. Strehlow, Advances in molecular relaxation and interaction pro-
cesses. 1978, 12, 29–46.
[20] R. P. Bell, D. W. Earls, J. Chem. Soc., Perkin Trans. 2 1976, 45–46.
[21] H. Wilson, E. S. Lewis, J. Am. Chem. Soc. 1972, 94, 2283–2285.
[22] H. Wilson, J. D. Caldwell, E. S. Lewis, J. Org. Chem. 1973, 38, 564–566.
[23] H. F. Gilbert, J. Am. Chem. Soc. 1980, 102, 7059–7065.
[24] D. Turnbull, S. J. Maron, J. Am. Chem. Soc. 1943, 65, 212–218.
[25] W. J. Albery, C. F. Bernasconi, A. J. Kresge, J. Phys. Org. Chem. 1988, 1,
29–31.
[26] C. F. Bernasconi, Adv. Phys. Org. Chem. 1992, 27, 119–238 and
references therein.
[27] J. P. Guthrie, J. Am. Chem. Soc. 1997, 119, 1151–1152.
[28] V. Gold, S. J. Grist, J. Chem. Soc., B 1971, 2282–2285.
[29] E. H. Caldin, S. Mateo, J. Chem. Soc., Chem. Commun. 1973, 854–855.
[30] E. H. Caldin, S. Mateo, J. Chem. Soc., Faraday Trans. 1 1975, 71,
1876–1904.
[31] E. H. Caldin, S. Mateo, J. Chem. Soc., Faraday Trans. 1 1976, 72,
112–121.
[32] A. J. Kresge, M. F. Powell, J. Phys. Org. Chem. 1990, 3, 55–61.
[33] E. F. Caldin, S. Mateo, P. Warrick, J. Am. Chem. Soc. 1981, 103, 202–204.
[34] P. Pruszynski, A. Jarczewski, J. Chem. Soc., Perkin Trans. 2 1986,
1117–1120.
[35] W. Galezowski, A. Jarczewski, J. Chem. Soc., Perkin Trans. 2 1989,
1647–1656.
[79] M. Mautner, W. L. Seick, J. Am. Chem. Soc. 1986, 108, 7525–7259.
[80] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456–463.
[81] S. H. Gellman, B. R. Adams, Tetrahedron Lett. 1989, 30, 3381–3384.
[82] G. A. Olah, P. Ramaiah, C. Lee, G. K. S. Prakash, Synlett 1992, 337–341.
[83] A. P. Black, F. H. Babers, Org. Synth. 1943, CV 2, 512–515.
[84] E. J. Billo, Excel for Chemists, 2nd ed. Wiley-VCH, New York, 2001.
[36] E. S. Lewis, L. H. Funderburk, J. Am. Chem. Soc. 1967, 89, 2322–
2327.
[37] R. P. Bell, D. M. Goodall, Proc. Roy Soc. (London) 1966, A294, 273–
297.
[38] N. Isaacs, K. Javaid, J. Chem. Soc., Perkin Trans. 2 1979, 1583–1586.
[39] E. S. Lewis, J. K. Robinson, J. Am. Chem. Soc. 1968, 90, 4337–4342.
J. Phys. Org. Chem. 2008, 21 603–613
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc