Kinetic isotope and thermodynamic analysis of the nornicotine-catalyzed aqueous aldol reaction

Article abstract of DOI:10.1016/j.tet.2005.09.094

A series of kinetic isotope effects and thermodynamic studies were performed to test key predictions of a computationally derived model for a nornicotine-catalyzed aqueous aldol reaction. The relative energies of the two computationally-derived transition

Full text of DOI:10.1016/j.tet.2005.09.094

Tetrahedron 62 (2006) 352–356
Kinetic isotope and thermodynamic analysis of the
nornicotine-catalyzed aqueous aldol reaction
*
Claude J. Rogers, Tobin J. Dickerson and Kim D. Janda
Departments of Chemistry and Immunology and The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla CA 92037, USA
Received 5 July 2005; revised 17 August 2005; accepted 9 September 2005
Available online 11 October 2005
Abstract—A series of kinetic isotope effects and thermodynamic studies were performed to test key predictions of a computationally derived
model for a nornicotine-catalyzed aqueous aldol reaction. The relative energies of the two computationally-derived transition states were
challenged using the proton inventory, which demonstrated that a single water molecule from the solvent is involved in, or before, the rate-
limiting step. These results suggest the importance of proton transfer in the aqueous aldol reaction and may assist the development of other
aqueous organocatalytic processes.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
Recently, organocatalysis has become an increasingly
useful tool in the construction of complex molecular
skeletons.1 The intense interest in the field has led to the
development of a wide variety of catalysts capable of
accelerating a diverse set of reactions including aldol
additions,2 Mannich reactions,3 [4C2]4 and [3C2]5
cycloadditions, a-aminations,6 epoxidations,7 and cyclo-
propanations.8 Despite this flurry of activity, there are few
examples of truly aqueous organocatalysts.9 The develop-
ment of synthetically viable aqueous organocatalysts
would be a boon to the field of green chemistry, assisting
the development of environmentally benign chemical
processes. In our laboratory, we have shown that
nornicotine 1, a metabolite of nicotine and constituent of
tobacco, can catalyze aqueous aldol reactions under
physiologically relevant conditions (Scheme 1).10 Aside
from the pharmacological implications,11 this reaction is
noteworthy because it was the first, and still one of the few,
examples of non-enzymatic aqueous enamine-based
chemistry.9 While the reaction proceeds in good yield
with sufficiently activated substrates, the rate and substrate
compatibility of the reaction must be improved before it can
be considered synthetically useful. With these thoughts in
mind, we recently initiated a research program aimed at
explicating the mechanism of this reaction to better
Scheme 1. An aqueous aldol reaction catalyzed by 30 mol % nornicotine 1
in 200 mM phosphate buffer at 37 8C.
understand the molecular basis for rate enhancement, and
thus develop improved aqueous organocatalysts.
To elucidate the structural requirements for effective
catalysis, a series of nornicotine analogs were used to
determine the linear free energy relationship between small
changes in the structure of the catalyst and the rate of the
reaction.12 By replacing nornicotine with meta and para
substituted 2-arylpyrrolidines, the rate of the reaction
increased with a corresponding increase in the electron-
withdrawing nature of the substituents on the aryl ring of the
catalyst. The positive value of the slope of the Hammett
plot, r, may be due to the lower pKa of the pyrrolidine
nitrogen of analogs containing electron-withdrawing groups
on the aryl ring. Thus, the perturbed pKa of these analogs
effectively increases the amount of available catalyst under
the reaction conditions. While these results explained why
proline (predominantly zwitterionic at pH 8.0) and
pyrrolidine (pKaZ11.4) are poor aqueous organocatalysts,
it did not give insight into why the nornicotine-catalyzed
aldol reaction similarly fails in organic solvent.9,10
Keywords: Nornicotine; Organocatalysis; Kinetic isotope effects.
*
Corresponding author. Tel.: C1 858 784 2515; fax: C1 858 784 2595;
e-mail: kdjanda@scripps.edu
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.09.094

C. J. Rogers et al. / Tetrahedron 62 (2006) 352–356
353
Scheme 2. A computationally derived mechanism for a nornicotine-catalyzed aldol reaction in water.
A rational explanation of this phenomenon is that an explicit
proton must transfer at the transition state for the reaction to
occur. This explanation was supported by the results of a
computational study into the mechanism of the nornicotine-
catalyzed aqueous aldol reaction.13 In this study, the
calculations were simplified by assuming that the enamine
was preformed, and the energy of the enamine was the zero-
point for the reaction. Based on this assumption, the
predicted mechanism revealed that, unlike an aldol reaction
in organic solvent, the reaction proceeds through a two-step
mechanism, with each step requiring an explicit water
molecule from the solvent (Scheme 2). Carbon–carbon bond
formation proceeds through a trimolecular six-membered
transition state TS1 and forms a stable, albeit short-lived,
hemiaminal intermediate. The rate-limiting step is
hydrolysis of the hemiaminal via TS2 to give the product
with concomitant regeneration of the catalyst.
two explicit water molecules from the solvent are involved
in, or before, the rate-limiting step. Therefore, the rate of the
reaction in deuterium oxide should be significantly reduced
relative to the rate of the reaction in water, as two O–(D)
bonds should be broken based on this prediction. The
isotope effect was determined in buffered water and buffered
deuterium oxide.15 Rate constants were measured under
pseudo-first-order conditions by using acetone as the donor
and 4-nitrobenzaldehyde as the acceptor. Experiments were
performed at 37 8C, under conditions in which the reaction
rate was maximized relative to the uncatalyzed reaction
(pH 8.0). While a large isotope effect was predicted based
upon the proposed computational mechanism, the measured
value (kH/kDz3) was surprisingly small. To provide an
explanation for the magnitude of the kinetic isotope effect,
the number of water molecules involved in the reaction, the
proton inventory, was determined by observing the effect of
increasing the atom fraction of deuterium (n) on the
corresponding observed rate constant (kn). In addition to
the number of protons from the solvent involved in the
transition state, the transition state fractionation factors can
be calculated for the reaction.16 The fractionation factor (f)
is a measure of the preference of the solute (SH) for
combination with deuterium, relative to water (ROH),17 that
is,
Although the trimolecular TS1 is unusual and would seem
to be entropically unfavorable, there is empirical evidence
that the concerted transfer of a proton in the transition state
is important. For example, the calculations for the
uncatalyzed reaction required proton transfer in the
transition state for both organic and aqueous media.
Considering that a hydrogen-bound water could activate
the aldehyde for nucleophilic attack, and the large excess of
water in the reaction (w55 M), the assumption of an
explicitly involved water molecule is reasonable. Further-
more, there is evidence for a trimolecular transition state
involving water in the aqueous Diels–Alder reaction of
methyl vinyl ketone and cyclopentadiene.14 Analogous to
the proposed mechanism for the nornicotine-catalyzed
aqueous aldol reaction, water accelerates the reaction by
acting as a Lewis acid by providing a hydrogen bond to
methyl vinyl ketone. We report herein upon a series of
kinetic isotope and thermodynamic experiments designed to
help determine the validity of the computationally derived
mechanism for the nornicotine-catalyzed aqueous aldol
reaction.
½SDŠ=½SHŠ
f Z
½RODŠ=½ROHŠ
The proton inventory requires application of the Gross–
Butler Eq. (1),
Transition stateð1Kn CnfiÞ
Q
i
kn Z kH
(1)
Reaction stateð1Kn CnfjÞ
Q
j
where fi and fj are the fractionation factors for all
exchangeable hydrogens in the transition state and reactant
state, respectively. According to the proposed mechanism,
the two protons from water in TS1 and TS2 show a change
between reagent and transition state, thus Eq. (1) simplifies
to give,
2. Results and discussion
kn Z kHð1Kn Cnf‡TS1Þð1Kn CnfT‡S2
Þ
(2)
A key prediction of the computational mechanism is that

354
C. J. Rogers et al. / Tetrahedron 62 (2006) 352–356
Figure 1. Proton inventory for the nornicotine-catalyzed aqueous aldol
reaction. The magnitude of the error bars are obscured by the corresponding
point.
Figure 2. Eyring analysis for the nornicotine-catalyzed aqueous aldol
reaction evaluated over a range of temperatures between 15–55 8C.
where f‡ is the transition-state fractionation factor. There-
fore, the theoretical mechanism predicts a non-linear
dependence between kn and n.
experiments. Provided that DGTS1ODGTS2 upon inclusion
of the entropic penalties in the calculation, the results of the
proton inventory are entirely consistent with the proposed
mechanism. Since TS1 requires ordering three molecules,
the entropic penalty is expected to be larger than the penalty
for TS2, thus it is possible that their inclusion could make
the first step rate limiting.
In contrast to the predicted dependence, a plot of the
observed rate constant versus increasing atom fraction of
deuterium was linear (R2Z0.999, Fig. 1), indicating only
one water molecule from the solvent is involved in the
transition state. Therefore, a key prediction of the
computational mechanism appears to be incorrect. When
applied to the data, the Gross–Butler equation simplifies to
give,
To further characterize the mechanism, the activation
parameters DS‡ and DH‡ were calculated. Measuring the
rate of the reaction at different temperatures allowed for the
application of the Erying equation,
kn Z kHð1Kn Cnf‡Þ
(3)
ꢀ
ꢁ
DS‡
kB
h
DH‡
lnðkÞ Z
Cln
K
(4)
consistent with the involvement of a single water molecule,
with f‡Z0.32. In this case, the value of the transition state
fractionation factor is also the inverse of the isotope effect.
R
RT
where kB and h are the Boltzman and Planck’s constants,
respectively. The plot of ln(k/T) versus 1/T was found to
be linear over the range of temperatures examined (R2Z
0.996, Fig. 2). The calculated value of DS‡ was K9.8G
Having established that a single water molecule participates
in the reaction at or prior to the rate-limiting step, we
considered the implications of the magnitude of the primary
kinetic isotope effect, as the value is significantly lower than
the theoretical maximum. An explanation for this reduction
is that isotope effects are lower for proton transfer between
two oxygen atoms than for a transfer from oxygen to
carbon.16 However, the magnitude of the isotope effect
also depends on the geometry of the transition state.
Isotope effects are maximized when the geometry of the
proton-in-flight during the transition state approaches
linearity. Interestingly, the reduced isotope effect is
consistent with the predicted geometry of both six-
membered transition states TS1 and TS2, as the bond
angles required (w1098) would lower the magnitude of the
isotope effect.18
1.2 cal KK1 molK1 and DH‡ was 10.9G0.4 kcal molK1
.
While the value of DS‡ is negative, as predicted, it is not
possible to make any mechanistic determinations based on
the sign or the magnitude of the value without a reference
reaction of the same standard state.19
Further investigation into the mechanistic assumptions was
provided by isotopic substitution of the hydrogens at
reactive centers of the aldehyde and acetone. For example,
an inverse a-secondary kinetic isotope effect was expected
upon isotopic substitution of the aldehydic proton due to the
sp2/sp3 transition at the reactive carbon. The effect of this
substitution was measured with 4-nitro-(a-d1)benzaldehyde,
synthesized according to a known procedure.20 Indeed, the
deuterated aldehyde perturbed the rate of the reaction by a
magnitude consistent with an inverse secondary isotope
effect (Table 1).
While the results of proton inventory experiment are
contradictory with the predicted involvement of two
molecules of water involved in or before the rate-limiting
step, these results provide evidence that one molecule of
water is involved in the reaction, and are consistent with the
proposed six-membered transition state. Furthermore, these
results are in accord with the computational mechanism,
provided the relative energies are inaccurate as a result of
the lack of entropic corrections in the computational
Upon substitution of acetone with d6-acetone, a primary
isotope effect was observed due to C–(D) bond cleavage
when forming the enamine. Additionally, the hybridization
change at the enamine methylene after carbon–carbon bond
formation should result in an inverse secondary kinetic
isotope effect. However, the magnitude of the effect was

C. J. Rogers et al. / Tetrahedron 62 (2006) 352–356
355
Table 1. Summary of kinetic and thermodynamic parameters for the
nornicotine-catalyzed aqueous aldol reaction
was followed by monitoring the height and area of the peak
corresponding to 4-hydroxy-4-(4-nitrophenyl)butan-2-one
(tRZ6.2 min) and extrapolating concentrations from a
standard curve obtained using a synthetic standard.21
Entry
Parameter
Valuea
1
2
3
4
5
6
kH/kD (Water)
kH/kD (aldehyde)
kH/kD (acetone)
f‡
3.05G0.10
0.89G0.08
4.76G0.11
0.32G0.01
K9.8G1.2b
10.9G0.4c
4.2. General procedure for kinetic experiments
DS‡
Nornicotine was added as a 300 mM solution in DMSO to a
solution of 200 mM phosphate buffer (pH 8.0) and acetone.
The solution was briefly vortexed, and then incubated at
37 8C. The reaction was initiated by the addition of
4-nitrobenzaldehyde as a 100 mM solution in DMSO. The
final concentrations of reactants were 2.4 mM nornicotine,
240 mM acetone, and 1–8 mM 4-nitrobenzaldehyde in 10%
DMSO. The progress of the reaction was monitored by
removing 10 mL aliquots of the solution at various times
during the reaction and diluting them to a total volume of
500 mL. Then, 20 mL of these samples were injected onto an
analytical RP-C18 HPLC column for analysis. Pseudo-first-
order rate constants were determined by linear regression
analysis.
DH‡
a Error for all parameters determined by propagation of error analysis.
b Value given in cal KK1 molK1
c Value given in kcal molK1
.
.
obscured by the primary isotope effect. Considering
multiple isotope effects are assumed to be additive, we
attempted to deconvolute the secondary kinetic isotope
effect by directly measuring the observed Keq of enamine
formation via 2D 1H–1H ROESY and NOESY NMR
spectroscopy. Unfortunately, the error associated with the
measurement was too large to calculate a reasonable
equilibrium constant as the diagonal peak of the enamine
methyl overlapped substantially with adjacent nornicotine
peaks of much greater intensity (data not shown). None-
theless, the results of the acetone and aldehyde kinetic
isotope effects indicate that carbon–carbon bond and
enamine formation occur in or before the rate-limiting
step, consistent with the proposed mechanism.
4.3. General procedure for kinetic isotope experiments
Kinetic isotope effects were measured by substituting the
appropriate amount of deuterated substrate into the reaction.
The observed rate constant (kD) was then compared to the
rate of the reaction with only protiated substrates (kH). The
proton inventory was determined by measuring the observed
rate constant at increasing percent of phosphate buffered
D2O (200 mM potassium phosphate, pD 8.0)16 in phosphate
buffered H2O (200 mM potassium phosphate, pH 8.0).
3. Conclusion
Based on the data obtained in this study, we are able to reject
the relative energies of the computationally derived
mechanism for the nornicotine-catalyzed aqueous aldol
reaction. Provided TS1 is the rate-limiting step, the data is
entirely in accord with the computational mechanism.
However, this does not preclude other mechanistic
possibilities that are also consistent with the data. While
we were not able to provide evidence to directly support the
proposed trimolecular transition state, these results under-
score the importance of proton transfer in aldol organo-
catalysis, as the nornicotine-catalyzed reaction fails in
organic solvent most likely because there are few available
protons to participate in the transition state. Combined with
our previous results,12 we are able to provide a clearer
picture into the mechanistic demands of the aqueous aldol
reaction, which should assist in the development of future
green organocatalysts.
Acknowledgements
We thank the reviewer for valuable suggestions and
gratefully acknowledge financial support from The Skaggs
Institute for Chemical Biology.
References and notes
1. Dalko, P. I.; Moisan, L. Angew. Chem., Int. Ed. 2004, 43,
5138–5175.
2. For examples, see: (a) Pidathala, C.; Hoang, L.; Vignola, N.;
List, B. Angew. Chem., Int. Ed. 2003, 42, 2785–2788. (b) Peng,
Y.-Y.; Ding, Q.-P.; Li, Z.; Wang, P.-G.; Cheng, J.-P.
4. Experimental
´
Tetrahedron Lett. 2003, 44, 3871–3875. (c) Cordova, A.;
Notz, W.; Barbas, C. F., III J. Org. Chem. 2002, 67, 301–303.
´
3. Cordova, A. Acc. Chem. Res. 2004, 37, 102–112.
4.1. General methods
4. (a) Juhl, K.; Jørgensen, K. A. Angew. Chem., Int. Ed. 2003, 42,
1498–1501. (b) Thayumanavan, R.; Dhevalapally, B.;
Sakthivel, K.; Tanaka, F.; Barbas, C. F., III Tetrahedron
Lett. 2002, 43, 3817–3820. (c) Ramchary, D. B.; Chowdari,
N. S.; Barbas, C. F., III Angew. Chem., Int. Ed. 2003, 42,
4233–4237.
All chemicals were obtained from commercial suppliers.
Nornicotine was distilled prior to use. HPLC solvents were
filtered and degassed prior to use. All isotopic solvents and
reagents were above 99.9% enrichment, if available,
otherwise the highest commercially available enrichment
was used. All HPLC experiments were preformed using a
C18 reverse-phase column and an isocratic mobile phase of
25% acetonitrile in water with 0.1% TFA, with a flow rate of
1.0 mL minK1, monitoring at 254 nm. Product formation
5. (a) Jen, W. S.; Wiener, J. J. M.; MacMillan, D. W. C. J. Am.
¨
Chem. Soc. 2000, 122, 9874–9875. (b) Karlsson, S.; Hogberg,
H. E. Tertahedron: Asymmetry 2002, 13, 941–945.
6. Duthaler, R. O. Angew. Chem., Int. Ed. 2003, 42, 975–978.

356
C. J. Rogers et al. / Tetrahedron 62 (2006) 352–356
´
7. (a) Marigo, M.; Franzen, J.; Plousen, T. B.; Zhuang, W.;
Jørgensen, K. A. J. Am. Chem. Soc. 2005, 127, 6964–6965. (b)
Aggarwal, V. K.; Lopin, C.; Sandrinelli, F. J. Am. Chem. Soc.
2003, 125, 7596–7601.
13. Dickerson, T. J.; Lovell, T.; Meijler, M. M.; Noodleman, L.;
Janda, K. D. J. Org. Chem. 2004, 69, 6603–6609.
14. (a) Blake, J. F.; Jorgensen, W. L. J. Am. Chem. Soc. 1991, 113,
7430–7432. (b) Blake, J. F.; Lim, D.; Jorgensen, W. L. J. Org.
Chem. 1994, 59, 803–805.
8. Kunz, R. K.; MacMillan, D. W. C. J. Am. Chem. Soc. 2005,
127, 3240–3241.
15. The relationship, pDZpHobsC0.6, was used to determine the
pD of buffered D2O solutions.
9. For other examples, see: (a) Reymond, J.-L.; Chen, Y. J. Org.
Chem. 1995, 60, 6970–6979. (b) Cordova, A.; Notz, W.;
Barbas, C. F., III Chem. Commun. 2002, 3024–3025. (c)
Hatikka, A.; Arvidsson, P. I. Tetrahedron: Asymmetry 2004,
15, 1831–1834. (d) Chimni, S. S.; Mahajan, D. Tetrahedron
2005, 61, 5019–5025.
16. Schowen, R. L. Prog. Phys. Org. Chem. 1972, 9, 275–332.
17. Gold, V. Adv. Phys. Org. Chem. 1969, 7, 295–331.
18. Chiao, W. B.; Saunders, W. H. J. Am. Chem. Soc. 1978, 100,
2802–2805.
19. Carpenter, B. K. Determination of organic reaction
mechanisms; Wiley: New York, 1984.
10. Dickerson, T. J.; Janda, K. D. J. Am. Chem. Soc. 2002, 124,
3220–3221.
20. Spletstoser, J. T.; White, J. M.; Georg, G. I. Tetrahedron Lett.
2004, 45, 2787–2789.
11. (a) Dickerson, T. J.; Janda, K. D. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 15084–15088. (b) Dickerson, T. J.; Janda, K. D.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8182–8187.
12. Rogers, C. J.; Dickerson, T. J.; Brogan, A. P.; Janda, K. D.
J. Org. Chem. 2005, 70, 3705–3708.
21. List, B.; Lerner, R. A.; Barbas, C. F.; III J. Am. Chem. Soc.
2000, 122, 2395–2396.