Structure dependence in the solvolysis kinetics of amino acid esters

Article abstract of DOI:10.1016/j.tetlet.2010.04.063

To better understand acyl transfer reactions of oligopeptides, seventeen N-acyl amino acid esters were solvolyzed in mildly basic methanol-d₄. All show pseudo-first-order kinetics by ¹H NMR. The rate constant varies up to 400-fold with the identity of the amino acid and up to 6200-fold with the identity of the N-acyl group. The impact of the N-acyl group on the rate constant is discussed in terms of crowding, amide conformation, and amide C{double bond, long}O bond character.

Full text of DOI:10.1016/j.tetlet.2010.04.063

Tetrahedron Letters 51 (2010) 3280–3283
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Structure dependence in the solvolysis kinetics of amino acid esters
*
John Haseltine , Jason W. Runyon
Department of Chemistry and Biochemistry, Kennesaw State University, Kennesaw, GA 30144, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
To better understand acyl transfer reactions of oligopeptides, seventeen N-acyl amino acid esters were
solvolyzed in mildly basic methanol-d4. All show pseudo-first-order kinetics by 1H NMR. The rate con-
stant varies up to 400-fold with the identity of the amino acid and up to 6200-fold with the identity
of the N-acyl group. The impact of the N-acyl group on the rate constant is discussed in terms of crowd-
ing, amide conformation, and amide C@O bond character.
Received 19 March 2010
Accepted 14 April 2010
Available online 18 April 2010
Keywords:
Amino acid esters
Acyl transfer
Ó 2010 Elsevier Ltd. All rights reserved.
Crowded amides
Amide polarity
X-Pro-OEt ! X-Pro-OMe
ð1Þ
1. Introduction
As a next step in our project, we proceeded to inspect the sim-
plest esters more closely. This Letter compares N-acyl amino acid
esters undergoing acyl transfer. Even in such small compounds,
the variation of reactivity with structure is pronounced and
intriguing.
We reported previously that under mildly basic solvolysis con-
ditions, some closely related amino acid and oligopeptide esters
have strikingly varied half-lives (Eqn. 1 and Table 1).1 Each com-
pound suffers a clean acyl transfer, yielding the corresponding
methyl ester plus ethanol. The hypothesis under consideration
was that oligopeptides of different lengths would show different
degrees of acyl transfer reactivity. The data show an impact of
structural differences that are four to seven bonds distant from
the ester group. One would like to understand this impact well gi-
ven that it is similar to rate/length dependences seen in the prote-
olysis of oligopeptides. Rate/length dependences in proteolysis are
usually rationalized in terms of enzyme/substrate interactions. We
have suggested, however, that substrate structure itself plays a lar-
ger and more consequential role than formerly suspected.1
Table 1
Half-life values for ethyl esters undergoing solvolysis to their respective methyl esters
in 2.06 M i-Pr2NEt/MeOH at ambient temperature (from Ref. 1)
Compound
Piv-Pro-OEt
Piv-Pro-Pro-OEt
Ac-Pro-Pro-OEt
Piv-Sar-Pro-Pro-OEt
Ac-Pro-OEt
t1/2 (d)
1700
220
30
15
2.4
Piv = pivaloyl, Pro = prolyl, Ac = acetyl, Sar = sarcosyl.
* Corresponding author. Tel.: +1 770 499 3426; fax +1 770 423 6744.
E-mail address: jhaselti@kennesaw.edu (J. Haseltine).
Present address: Department of Chemistry, University of Alabama, Tuscaloosa, AL
35487, USA.
Figure 1. N-Acyl amino acid esters subjected to base-promoted solvolysis.
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.04.063

J. Haseltine, J. W. Runyon / Tetrahedron Letters 51 (2010) 3280–3283
3281
2. Results and discussion
Each of the esters in Figure 1 converts cleanly to its respective
methyl ester plus ethanol under mild conditions (0.1 M ester and
1.03 M i-Pr2NEt in methanol-d4, 21.5 1 °C).2–6 The reactions were
monitored directly by 1H NMR. The pseudo-first-order rate con-
stant was calculated from successive integrations of the diminish-
ing –CO2CH2CH3 signal of the reactant or the growing DOCH2CH3
quartet.7 The results are listed in Table 2.
Series 1–3 differ in their absolute reactivity and in their rate
constant patterns. Comparisons with control compounds show
that the amide group can boost ester reactivity significantly. The
rate constants for 1a and 2a are about 100-fold and 85-fold greater
than those for the similar simple esters ethyl cyclopentanecar-
boxylate (6) and ethyl butanoate (7). The rate constant varies sig-
nificantly with the identity of the N-acyl group, and most strongly
in series 1.
The acyl transfer mechanism probably involves a base-assisted
pyramidalization of the ester group by the solvent.8,9 Neighboring-
group participations are unlikely. Deuteration next to the amide
carbonyl was not evident in any of the kinetics trials. Also, when
the base was omitted from the reaction, the rate constants were
greatly reduced. This control experiment was run for all amide-es-
ters except 1e and 3e. Conventional neighboring-group participa-
tions should show atleast one of two consequences. They should
show deuteration if amide enolization is occurring (not expected
under our mild conditions). Alternatively, if the neutral amide
group were the nucleophile, k should be largely unaffected by
omitting the base. For most compounds in the study, deuteration
next to the ester carbonyl was also either not evident or its rate
constant was less than one-half that for acyl transfer. Enolization
of the ester group and/or ketene formation is therefore not likely
a part of the acyl transfer mechanism for those compounds.10 As
for the remaining compounds, kinetic isotope data suggest that
the ester group is not deprotonated as a part of acyl transfer events.
One possible explanation for the variation of k within each ester
series is that the ester group is directly crowded to different ex-
Figure 2. Amide conformation in series 1–3 according to NMR data and ab initio
calculations.
tents by the different N-acyl groups. This explanation is most cred-
ible for series 3 since the bulky N-t-butyl group should generally
favor conformations that have the N-acyl R group and the ester
group close to each other (Fig. 2). The largest R groups might then
interfere the most with ester pyramidalization and sponsor the
lowest rate constants.
The explanation does not apply as neatly to series 1 and 2, how-
ever, since those series do not favor an amide conformation in
which the N-acyl R group and the ester group are close to each
other. Proton NMR integrations and carbon chemical shifts indicate
the alternative amide conformation to be slightly favored for most
of series 1 and 2 (syn-1 and syn-2 in Fig. 2). Therefore, for mutual
crowding of the N-acyl and ester groups to be the main source of
variation in k, the anti conformation would need to be the more
reactive conformation for most of each series, and substantially
more reactive than the syn conformation. The possibility is not
ruled out. Note that ester 3a has a smaller rate constant than
3b–d. This is interesting because esters 1a and 2a, also bearing
the smallest N-acyl group, each show the largest rate constants
in their series. The predominant amide conformation in 3a is
evidently syn-3a, with NOE measurements supporting a prediction
Table 2
Rate constants for ester solvolysis, IR frequencies of the amide carbonyl, and NMR
chemical shifts of the amide carbonyl
a
b,c
Compound
k (10ꢀ5 sꢀ1
)
mC@O (cmꢀ1
)
dC@O (ppm)
1a
1b
1c
1d
1e
3.3
1663
1644
1644
1645
1622
160.8, 161.7
169.5, 169.7
172.5, 172.7
176.0, 176.5
176.9
0.099
0.044
0.034
0.00053
2a
2b
2c
2d
2e
8.1
2.6
1.7
1.3
0.22
1668
1646
1648
1646
1631
163.0, 163.2
171.2, 171.5
174.3, 174.6
177.5, 177.6
178.1
*
from ab initio modeling (gas phase; 6-31G level). Therefore, the
anti conformation may indeed have more of an activating effect
on ester reactivity.
3a
3b
3c
3d
3e
0.10
0.32
0.22
0.22
1654
1651
1652
1640
1628
161.8
171.8
174.6
178.6
178.7
To test the importance of amide conformation more plainly, es-
ters 4 and 5 were prepared and solvolyzed. The lactam rings in 4
and 5 lock the amide linkage into syn and anti conformations,
respectively. The rate constants are shown in Table 2. Ester 5 is
slightly more reactive than 4, despite having more branching next
to its ester carbonyl. The impact of this branching should be mild.
The rate constants for esters 6–8 imply that each alkyl substituent
next to the ester carbonyl in that control series lowers k by a factor
of about 3. Thus the data are consistent with the anti conformation
being more activating than syn, if only slightly so.
0.0045
4
5
6d
7e
8f
2.1
3.6
1682
1688
175.7
175.2
0.031
0.094
0.27
—
—
—
—
—
—
a
b
c
d
e
f
Obtained for neat compounds.
Obtained for CDCl3 solutions.
Values are given for both amide conformations, if observed.
Ethyl cyclopentanecarboxylate.
Ethyl butanoate.
A different explanation for the variation of k in each ester series
involves the activating effect of the amide group. Electron-with-
drawing groups in esters are known to facilitate acyl transfer.
Ethyl acetate.

3282
J. Haseltine, J. W. Runyon / Tetrahedron Letters 51 (2010) 3280–3283
The relative power of such groups depends on their specific elec-
tronic character. If the amide group increases ester reactivity by
acting as an electron-withdrawing group or an electron sink, two
different amides need not be alike in this respect. Different N-acyl
R groups, by sponsoring differences in mean and dynamic amide
geometry, orbital structure, and charge distribution, may afford
differences in electron-withdrawing power and charge
accommodation.
Modeling of the syn conformation in series 1 was done to pre-
dict the impact of R on amide geometry. Some details are shown
in Table 3. As the size of R increases, the R–C–N angle (opposite
the carbonyl oxygen) and the adjacent C–N–C angle are each pre-
dicted to widen. The O@C and C–N bonds of the O@C–N substruc-
ture are each predicted to lengthen slightly.
That the R group does affect amide electronic character is indi-
cated empirically by the amide carbonyl stretching frequencies,
listed in Table 2. The value of mC@O drops significantly from the
smallest to the largest R group within each series. However, the
value of mC@O is almost constant when R = Me, Et, and i-Pr within
each series, so R is probably not affecting C@O stretching by its
Figure 3. Comparison of hybrid states for the amide carbonyl carbon as a function
of R–C–N angle. Only sigma-bonding orbitals of the carbonyl carbon are shown.
gen), thereby favoring an incremental shift of the bond’s electron
density slightly closer to carbon. The carbon might therefore gain
‘extra’ shielding by 13C NMR, appearing further upfield than ex-
pected (as seen for 1e, 2e, and 3e). Each loss in the volume of the
major lobe would also remove density from the C/O internuclear
space, thereby tending to weaken the C@O sigma bond and possi-
bly lowering mC@O as we observe.
Are the data consistent with the hypothesis that differences in
amide electronic character affect ester reactivity? Certainly the
largest decreases in rate constant correspond to the largest de-
creases in mC@O in each series. The hypothesis seems reasonable
for now if we suppose that k and mC@O might each be affected by
multiple factors.
specific mass or by a through-bond effect. The jumps in mC@O corre-
spond to the expected increases in crowding between the R group
and nitrogen’s other substituents. In series 1, for example, while
the size of the R group increases from 1a (R = H) to 1b (R = CH3),
the value of mC@O decreases by about 20 cmꢀ1. From 1b to 1c and
1d, the size of R increases again, but the R groups in 1c and 1d
can be rotated to avoid additional crowding relative to 1b. This
may explain why mC@O varies little over this sub-series. From 1d
(R = i-Pr) to 1e (R = t-Bu), however, an increase in crowding cannot
be avoided through facile bond rotations, and mC@O drops again by
about 20 cmꢀ1. Low values of mC@O are not unusual for crowded
amides and ketones.11–13
The 13C NMR chemical shifts of the amide carbonyl also indicate
an interesting impact by the R group (Table 2). For each series of
compounds, the signal is near 160 ppm when R = H. When
R = Me, it is in the range of 169–172 ppm. When R = Et, it is 3–
4 ppm further downfield, and when R = i-Pr, it is 3–4 ppm still fur-
ther downfield. When R = t-Bu, however, the C@O signal appears at
a frequency similar to that seen when R = i-Pr. A linear trend of
C@O signals toward lower field with increasing alkyl substitution
at the alpha carbon is known for several functional groups.14,15
The chemical shifts of 1e, 2e, and 3e represent deviations from this
trend.
Such deviations have been reported previously for crowded ke-
tones.13,16 They make sense in terms of the expected impact of
crowding on the C@O carbon hybridization. Consider that if the
R–C–N angle of the amide group widens for any reason, carbon’s
atomic orbital in the sigma bond to oxygen would gain in p charac-
ter. This is shown in simple form in Figure 3, comparing three hy-
brid states for the C@O carbon based on benchmark R–C–N angles
of 109°, 120°, and 180°. Only carbon’s sigma-bonding orbitals are
shown. From left to right in this series, carbon uses an sp, sp2, or
p orbital, respectively, to make its sigma bond to oxygen. As the
R–C–N angle widens, each increase in p character increases the rel-
ative volume of the orbital’s minor lobe (pointed away from oxy-
3. Conclusion
The nature of the amide group in N-acyl amino acid esters influ-
ences the acyl transfer reactivity of the ester group. The rate con-
stant varies positively with the infrared stretching frequency of
the amide carbonyl and negatively with the size of the N-acyl
group. A high importance of direct crowding between the N-acyl
and ester groups is not generally indicated by the present data.
Conformation within the amide group is judged to be at least
somewhat important. Crowding within the amide group is proba-
bly most responsible for variation in the amide’s electronic charac-
ter, evident by IR and NMR, which may in turn affect ester
reactivity. Further tests of these ideas with dipeptide and tripep-
tide esters are in progress.
Acknowledgments
We are grateful to the College of Science and Mathematics of
Kennesaw State University for support of this work. We are also
grateful to Mr. Alex M. Morrison and Prof. Kevin P. Gwaltney of
our Department for performing NOE and low-temperature NMR
experiments.
References and notes
Table 3
1. Fan, Y.-H.; Grégoire, C.-A.; Haseltine, J. Bioorg. Med. Chem. 2004, 12, 3097.
2. Amide-esters 1a–e, 2a–e, and 3a–e were prepared by N-acylation of proline
ethyl esterꢁHCl,3 sarcosine ethyl esterꢁHCl (Aldrich), and N-(t-butyl)glycine
ethyl ester,4 respectively. Formylations were performed in ethyl formate using
excess triethylamine and/or heat and pressure. Lactam-ester 4 was prepared by
alkylating the sodium salt of 2-pyrrolidinone (NaH/DMF) with ethyl
bromoacetate. Lactam-ester 5 was prepared from its corresponding methyl
ester by treatment with K2CO3/ethanol.5
Calculated bond angles and bond lengths at the amide group in compounds 1a–e, syn
conformation (Hartree–Fock level, 6-31G* basis set)
Ester
R–C–N
C–N–C
O@C (Å)
C–N (Å)
1a
1b
1c
1d
1e
113.49°
117.13°
117.01°
118.59°
122.31°
125.33°
126.88°
127.06°
127.69°
130.75°
1.197
1.203
1.203
1.205
1.206
1.343
1.355
1.356
1.355
1.362
3. Proline ethyl esterꢁHCl was prepared by applying the method of Garbers, C. F.;
Schmid, H.; Karrer, P. Helv. Chim. Acta 1955, 38, 1490.
4. Gribble, G. W.; Hirth, B. H. J. Heterocycl. Chem. 1996, 33, 719.

J. Haseltine, J. W. Runyon / Tetrahedron Letters 51 (2010) 3280–3283
3283
5. Williams, P. D.; Perlow, D. S.; Payne, L. S.; Holloway, M. K.; Siegl, P. K. S.; Schorn,
T. W.; Lynch, R. J.; Doyle, J. J.; Strouse, J. F.; Vlasuk, G. P.; Hoogsteen, K.;
Springer, J. P.; Bush, B. L.; Halgren, T. A.; Richards, A. D.; Kay, J.; Veber, D. F. J.
Med. Chem. 1991, 34, 887.
6. Each compound in Table 2 that is not well known (1a, 1c, 2c–e, 3a–e, and 5)
gave satisfactory 1H NMR, 13C NMR, IR, and high-resolution mass spectral
data.
7. For each kinetics trial, 1.00 mL of a 1.03 M (i-Pr)2NEt/MeOH-d4 solution was
added to a measured amount of ester so that the initial concentration of ester
was about 0.1 M. Kinetics data were collected as a series of signal integrations
with a Bruker Avance DPX-300 NMR spectrometer. Each rate constant was
calculated by the least-squares method for the simple linear regression
equation ln (%RCO2Et) = ꢀkt + C. Each value in Table 2 is the average of at
least two determinations.
9. Mitton, C. G.; Gresser, M.; Schowen, R. L. J. Am. Chem. Soc. 1969, 91, 2045.
10. If deuteration next to the ester carbonyl and acyl transfer were both occurring
by a single mechanism, for example, via ketene formation, then the rate
constant for deuteration could equal or exceed one-half k. The factor of one-
half arises from the fact that in such a mechanism, one N–CH2CO2Et proton
would be lost per event versus two CO2CH2CH3 protons.
11. Kondo, K.; Iida, T.; Fujita, H.; Suzuki, T.; Yamaguchi, K.; Murakami, Y.
Tetrahedron 2000, 56, 8883.
12. Rao, Y.; Li, X.; Danishefsky, S. J. J. Am. Chem. Soc. 2009, 131, 12924.
13. Qin, X.-r.; Ishizuka, Y.; Lomas, J. S.; Tezuka, T.; Nakanishi, H. Magn. Reson. Chem.
2002, 40, 595.
14. Levy, G. C.; Lichter, R. L.; Nelson, G. L. Carbon-13 Nuclear Magnetic Resonance
Spectroscopy, 2nd ed.; Wiley: New York, 1980. pp 137–152.
15. Breitmaier, E.; Voelter, W. Carbon-13 NMR Spectroscopy, 3rd ed.; VCH:
Weinheim, Germany, 1990. pp 216–232.
8. Mitton, C. G.; Schowen, R. L.; Gresser, M.; Shapley, J. J. Am. Chem. Soc. 1969, 91,
2036.
16. Jackman, L. M.; Kelly, D. P. J. Chem. Soc. (B) 1970, 102.