Gas-phase dynamic NMR study of the internal rotation in N-trifluoroacetylpyrrolidine

Article abstract of DOI:10.1021/jp027696y

The barrier to the internal rotation around the C-N bond in N-trifluoroacetylpyrrolidine (TFAPYR) was characterized using dynamic NMR spectroscopy. Kinetic parameters, ΔG^?₂₉₈ = 71.4(1.2) kJ mol^-1, ΔH^? = 67.9(1.4) kJ mol^-1, and ΔS^? = -11.8(4.0) J mol^-1), were found to be lower in the gas phase than in solution. The pyrrolidine ring substituent in TFAPYR reduces the bulkiness around the C-N amide bond, stabilizing the ground state and increasing the value of the internal rotation barrier, ΔG^?₂₉₈, compared to N,N-dimethyland N,N-diethyltrifluoroacetamide. The magnitude of ΔG^?₂₉₈ in TFAPYR is also comparable to that found in primary amide bonds in peptides, emphasizing the importance of having a relatively small energy difference between cis and trans conformations in proline's role as a switch in protein signaling.

Full text of DOI:10.1021/jp027696y

3024
J. Phys. Chem. A 2003, 107, 3024-3029
Gas-Phase Dynamic NMR Study of the Internal Rotation in N-Trifluoroacetylpyrrolidine
Cristina Suarez,* Elisabeth J. Nicholas, and Molly R. Bowman
Department of Chemistry, Smith College, Northampton, Massachusetts 01063
ReceiVed: December 11, 2002
The barrier to the internal rotation around the C-N bond in N-trifluoroacetylpyrrolidine (TFAPYR) was
q
-1
q
characterized using dynamic NMR spectroscopy. Kinetic parameters, ∆G 298 ) 71.4(1.2) kJ mol , ∆H )
-
1
q
-1
6
7.9(1.4) kJ mol , and ∆S ) -11.8(4.0) J mol ), were found to be lower in the gas phase than in solution.
The pyrrolidine ring substituent in TFAPYR reduces the bulkiness around the C-N amide bond, stabilizing
q
the ground state and increasing the value of the internal rotation barrier, ∆G 298, compared to N,N-dimethyl-
q
and N,N-diethyltrifluoroacetamide. The magnitude of ∆G 298 in TFAPYR is also comparable to that found in
primary amide bonds in peptides, emphasizing the importance of having a relatively small energy difference
between cis and trans conformations in proline’s role as a switch in protein signaling.
Introduction
In an effort to understand how substitution affects the
rotational barrier around the C-N amide bond, variable-
temperature dynamic NMR (DNMR) spectroscopy has been
1-3
used extensively in the study of both solvated and gas-phase
amides.4 The amide bond is a key feature in peptide dynamics
and its restricted rotational behavior is a critical determinant of
protein conformation in general.
,5
Figure 1. A Xaa-Pro-Xaa peptide. The proline residue is in bold and
the peptide bond of interest is the one occurring between the first
carbonyl and the nitrogen on the pyrrolidine ring.
Substantial effort has been made to study experimentally and
theoretically the electronic and steric effects of substitution on
the rotational barrier of amides, particularly formamides and
acetamides.1 These systems were chosen because they serve
as amide templates with substituents akin to the variety of side
chains present in the 20 gene-coded amino acids. Proline (Pro),
-5
6
hydroxyproline, and the recently discovered pyrrolysine differ
Figure 2. N-Trifluoroacetylpyrrolidine (TFAPYR) showing carbon
numbering used throughout this work.
from the other common amino acids in having a secondary
amino group, a result of having the amine nitrogen cyclized
with the side chain. This creates a unique type of peptide bond
suited for this study because its relatively high volatility allows
both solution and gas-phase measurements. This study reports
(Figure 1).
This type of peptide bond seems to be even more rotationally
q
q
the Gibbs free energy (∆G 298), activation enthalpy (∆H ), and
restricted than the more typical primary amide peptide bond.
The five-membered ring not only seems to restrict the confor-
mational freedom available to the peptide but also causes the
loss of H bonding capability of the amino nitrogen, giving
q
activation entropy (∆S ) differences describing the rotational
barrier in TFAPYR in the gas phase and in different solutions
(
toluene-d8, 1,1,2,2-tetrachloroethane-d2, and neat). A theoretical
ground-state geometry for TFAPYR calculated at the MP2/6-
11+G* computational level is also included to understand
7
proline a unique role in the secondary structure of its proteins.
3
The local environment surrounding a residue and the height of
the barrier to internal rotation influence the relative size of the
populations of cis and trans peptide conformations. Yet, while
less than 1% of peptide bonds overall are found to have a cis
configuration, approximately 5% of proline residues have the
better the effect of a cyclic substitution on amide geometry.
This is the first time a cyclic-substituted amide has been studied
in the gas phase.
Experimental Section
8
same cis preference. Recent findings by Andreotti et al. have
shown that the cis-trans isomerization of certain proline
residues may play an important regulatory role as switches of
catalytic activity.9
In the present study, we are interested in the gas-phase and
solution kinetic characterization of the hindered internal rotation
around the C-N bond in N-trifluoroacetylpyrrolidine (TFAPYR,
Figure 2) as a model for a Xaa-Pro peptide. TFAPYR is well-
Synthesis of TFAPYR. A solution of 0.1 mol (8.4 mL) of
pyrrolidine in 10 mL of anhydrous ether was cooled to 273 K
in a salt-ice bath, and 10 mL of a solution of anhydrous ether
containing 0.05 mol of trifluoroacetic anhydride was added
slowly with stirring. After the addition was complete, the
reaction mixture was allowed to warm to room temperature and
left to equilibrate for 30 min before distilling under reduced
pressure. After a small forerun (∼2 mL), a single fraction (6.5
mL) distilling at 350 K (10 Torr) was collected. The distillate
contained pyrrolidine, trifluoroacetamide, and the pyrrolidinium
*
To whom correspondence should be addressed. Phone: (413) 585-3838.
Fax: + 1 (413) 585-3786. E-mail: csuarez@smith.edu.
1
0.1021/jp027696y CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/05/2003

Internal Rotation in N-Trifluoroacetylpyrrolidine
J. Phys. Chem. A, Vol. 107, No. 17, 2003 3025
trifluoroacetate salt. Upon cooling, the pyrrolidinium trifluoro-
acetate salt crystallized, and the remaining liquid, composed of
primarily N-trifluoroacetylpyrrolidine, was decanted and passed
temperature and stored in 16K of memory. Digital resolutions
better than 0.37 Hz/point were obtained. The resulting solution
spectra had typical S/N ratios greater than 1500:1. A common
exponential-line-broadening factor of 0.2 Hz was applied to all
collected spectra prior to Fourier transform to enhance S/N.
2
through a 1 × 5 cm column of dry alumina to remove the
remaining salt. The column eluant (3.5 mL, 54% overall yield)
1
13
13C NMR measurements for the neat and the 10% solutions
of TFAPYR were performed on a Bruker Avance 400 with
carbon observation at 75.47 MHz. All measurements were made
on spinning samples in the locked mode (except the neat
sample). Acquisition times were between 0.54 and 2.16 s/scan
with a pulse length of 7.50 µs and a delay time of 2 s. Typically,
1000 scans were collected at each temperature (except for the
neat TFAPYR sample for which only 400 scans were necessary)
and stored in 16K of memory. Digital resolutions better than
0.92 Hz/point were obtained. The resulting solution spectra had
typical S/N ratios greater than 100:1. A common exponential-
line-broadening factor of 2 Hz was applied to all collected
spectra prior to Fourier transform to enhance S/N.
was a clear, colorless liquid. H and C NMR confirmed the
identity of the eluant to be N-trifluoroacetylpyrrolidine.
NMR Sample Preparation. Deuterated solvents were ob-
tained from Aldrich. Toluene-d8 and 1,1,2,2-tetrachloroethane-
d2 were both 99.5+ atom % D; chloroform-d1 was 99.8 atom
%
D. All solvents were used without further purification. It is
well established that solvent-amide and amide-amide associa-
tion, as well as other solvent properties, affect reaction rates;
therefore, TFAPYR concentrations for NMR solution samples
10
were kept below 15 mol %. Solution samples were prepared
1
13
by placing either a 5% ( H NMR study) or 10% ( C DNMR
study) solution of TFAPYR in the appropriate solvent and a
drop of tetramethylsilane (TMS) as a resolution and frequency
standard in 5-mm NMR tubes. For samples subjected to high
temperatures, a plug was also inserted in the tube to minimize
refluxing and evaporation effects within the active volume.
Samples were run from low to high temperatures to avoid
condensation outside this active volume.
Calibration of solution temperatures in both spectrometers
was performed using a vacuum-sealed 100% ethylene glycol
sample. The difference in chemical shifts for the two ethylene
glycol peaks was measured between 273 and 423 K and fitted
to a published chemical shift difference/sample temperature
1
3
correlation. Samples were allowed to equilibrate for 20 min
once a stable temperature was achieved. All temperatures used
in this study are calibrated temperatures and are accurate to (0.1
K.
Gas-phase NMR samples were prepared in 3-cm-long re-
stricted volume tubes constructed from Wilmad high-precision
glass. The outside diameter of these tubes was 8.89 mm to allow
the tubes to fit inside a standard 10-mm Wilmad NMR tube.
Neat TFAPYR was degassed through several freeze-thaw-
pump cycles before allowing the vapor pressure at ambient
temperature to fill the restricted volume tube. Sample-handling
techniques and the vacuum system used in the preparation are
similar to those described in Ross et al.11 The gas TFAPYR
sample contained 0.7 Torr of TFAPYR, 0.5 Torr of TMS, and
Rate Analysis. Rate constants were calculated using the
1
4
computer program DNMR5. This program uses the complete
band shape (CBS) method and an iterative, nonlinear least-
squares regression analysis to obtain the best fit to the
experimental NMR spectrum. Gas-phase rate constants were
1
extracted from the H NMR exchanging spectra using the A2
5
00 Torr of argon as a bath gas. Gas-phase amide internal
a B2 portion (protons attached to C-2 and C-5) of the ring
A2C2 a B2D2 spin system. No JHF coupling was observed,
5
rotation rate constants do not exhibit pressure dependence at
these pressures.12
probably because of the inherent natural line broadening of
gaseous resonances. Solution rate constants were extracted from
1
NMR Measurements. H NMR measurements for gaseous
1
3
C NMR exchanging spectra using a ACX3 a BDX3 spin
TFAPYR were performed on a Bruker AMX500 using a 10-
mm variable temperature (VT) probe with proton observation
at 500.13 MHz. All measurements were made on spinning
samples in the unlocked mode. Acquisition times were 1.61
s/scan with a pulse length of 35 µs and a delay time of 1 s.
Short delay times are typical of gas-phase NMR because of short
relaxation times. Typically, 2000 scans were collected at each
temperature and stored in 16 K of memory. A sweep width of
4
system, where the X3 spins account for observable JCF coupling
between CF3 and C-2.
The CBS method of analysis first requires the measurement
of chemical shifts, coupling constants, and natural transverse
relaxation times (T2nat) in the absence of exchange to separate
spectral changes due to natural temperature dependence from
those due to exchange. The systems were all studied in
temperature ranges in which exchange was apparently absent
and the parameters’ natural temperature dependence could be
determined. Coupling constants showed no temperature depen-
dence in any of the systems studied. On the other hand, chemical
shifts and T2nat exhibited a dependence on temperature in most
situations. Whenever a temperature dependence of either chemi-
cal shifts or T2nat was found, a linear relationship was assumed
and adjusted parameters were estimated for the exchanging
temperatures.
The effective T2 provided to the DNMR5 program was
constructed from the temperature-adjusted T2nat, a magnetic field
inhomogeneity contribution estimated from the line width of
the reference, TMS, and the exponential-line-broadening factor.
Rate constants in the exchange region were obtained by iterating
on spectral origin, baseline height, and baseline tilt. All other
parameters were adjusted to the appropriate temperature and
held constant.
(
5102 Hz was used for all spectra, giving a digital resolution
of 0.311 Hz/point. The resulting gas-phase spectra had typical
signal-to-noise (S/N) ratios of 200:1. A common exponential-
line-broadening factor of 1 Hz was applied to all collected
spectra after Fourier transform to enhance S/N. Probe temper-
ature calibration was performed using two 36 gauge copper-
constantan type “T” thermocouples placed inside a nonspinning
dummy gas sample tube. The temperature gradient between the
top and bottom of the sample never exceeded 0.4 K. Samples
were allowed to equilibrate for 20 min once a stable temperature
was achieved. All temperatures used in this study are calibrated
temperatures.
1H NMR measurements for the neat and the 5% solutions of
TFAPYR were performed on a JEOL Eclipse 400 with a 5-mm
VT probe with proton observation at 399.78 MHz. All measure-
ments were made on spinning samples in the locked mode
(except the neat sample). Acquisition times varied from 3 to 5
s/scan with pulse angles between 15° and 60° and a common
delay time of 4 s. Typically, eight scans were collected at each
In the calculation of the kinetic activation parameters, the
Eyring equation was used to obtain ∆H and ∆S . A value of
q
q

3026 J. Phys. Chem. A, Vol. 107, No. 17, 2003
Suarez et al.
a
TABLE 1: Chemical Shifts (ppm) and Coupling Constants
(
Hz) of Gaseous and Solution TFAPYRb
H
2
-C-2
H
2
-C-3c
H
2
-C-4c
2
H -C-5
C-6
C-8
Gas
ν
H
3.619
6.2
1.933
1.845
3.530
6.9
3
J
HH
Toluene
1.112
23.24
ν
H
2.982
46.10
6.9
1.181
3.076
ν
C
26.09
1.5
47.19
6.7
d
155.00 117.30
35.4 288.2
3
J
HH
J
CF
3.7
d
1,1,2,2-Tetrachloroethane
ν
H
3.609
46.53
6.9
1.988
26.19
1.895
23.43
3.530
47.43
6.9
ν
C
155.24 116.31
36.9 287.5
3
J
HH
1
J
CF
3.7
1.5
d
d
Figure 3. Temperature-dependent gas-phase 500 MHz H NMR spectra
of TFAPYR protons attached to C-2 and C-5.
Neat
1.913
24.15
ν
H
3.637
47.34
7.0
2.021
26.96
3.495
48.17
7.0
ν
C
155.80 117.80
36.1 287.5
3
J
HH
J
CF
3.7
1.5
d
d
a
All chemical shifts are referenced to TMS signals set to 0.000 ppm
(
gaseous TMS resonates ∼3 ppm downfield from solution TMS).
b
1
Uncertainties in chemical shift values are ca. (0.001 ppm for H and
1
3
ca. (0.01 ppm for C signals; coupling constants carry uncertainties
3
c 3
on the order of ca. (0.4 Hz for JHH and ca. (0.9 Hz for JCF
for C-3 and C-4 were not extracted from the pseudoquintets; the value
of ν was obtained from the frequency of the strongest line within the
multiplet. Not observable.
.
J
HH
H
d
0
.5 was assigned to the transmission coefficient (κ). The Gibbs
q
q
free energy of activation, ∆G 298, was obtained from ∆G )
∆
of temperature.
q
q
q
q
H - (298 K)∆S assuming ∆H and ∆S to be independent
Computational Methods. Ground-state geometries were first
optimized at the B3LYP/G3-11+G* level followed by an MP2
calculation with the 6-311+G* basis set using the program
Gaussian 98.15
Figure 4. Eyring plots of gas-phase and solution TFAPYR exchange
rate constants: (9) 1,1,2,2-tetrachloroethane; ([) toluene; (2) neat;
(b) gas.
1
Figure 3 shows temperature-dependent gas-phase H NMR
spectra of TFAPYR exhibiting exchange broadening of the C-2
and C-5 methylene resonances. Upper traces correspond to
experimental spectra, while lower traces correspond to DNMR-
simulated line shapes at their respective temperatures. Below
294 K, two separate signals can be observed for C-2 and C-5,
while beyond 393 K, only one signal represents the fast
exchange between the two distinct magnetic environments.
The Eyring plots, obtained from the rate constants extracted
using total-line-shape analysis of all four systems, are presented
in Figure 4. Table 2 lists the experimental rate constants (k)
characterizing the internal rotation around the C-N bond in
gaseous, neat, and 10% (v/v) solution TFAPYR. Table 3 lists
Results
Comparison of the gas-phase and solution results indicates
considerable medium effects. At room temperature, both gaseous
and solution TFAPYR H NMR spectra exhibit four distinct
1
sets of signals: two triplets (downfield) and two pseudoquintets
(upfield). The triplets correspond to protons attached to carbons
C-2 and C-5, and the pseudoquintets are actually multiplets
resulting from non-first-order behavior of the protons attached
to carbons C-3 and C-4. In all solutions at room temperature,
13
the C NMR spectra of TFAPYR exhibit two quartets downfield
C-6 and C-7 quartets due to C-F coupling) and four apparent
singlets upfield corresponding to C-5, C-2, C-3, and C-4. Table
(
q
q
q
the values of the kinetic parameters, ∆G 298, ∆H , and ∆S ,
obtained from those rate constants. Finally, Table 4 summarizes
and Figure 5 illustrates the calculated MP2 geometry for the
ground state of TFAPYR.
1
13
1
lists the corresponding H and C chemical shifts for gaseous
and solution TFAPYR.
Assignment of the 1H and 13C NMR signals to the ground-
state structure of TFAPYR in the different solvents was made
using JHF values, aromatic solvent shifts, and 2D COSY and
HETCOR sequences. Coupling between fluorine and hydrogen,
JHF, was observed in a degassed sample of TFAPYR in CDCl3.
Under these conditions, small quartets due to the spin-spin
interaction between fluorines (CF3) and protons (CH2) were
observed within each triplet. Observable H-F quartets yield
Discussion
1
In both the gas phase and solution, the H NMR spectra of
TFAPYR seem to follow traditional amide assignment patterns.
Protons trans to the oxygen resonate at higher frequencies than
those cis (except in toluene-d8 in which as in all aromatic
solvents the resonance absorption belonging to the protons trans
to the carbonyl is shifted farther upfield than the cis proton
5
larger JHF values for the hydrogens assigned to C-2 (0.89 Hz)
1
than for the hydrogens assigned to C-5 (0.64 Hz) because of
the proximity of the C-2 to the CF3 group. This is consistent
with previous studies of other trifluoroacetamides.1
absorption ). Coupling constants, both JHH and JHF, appear to
be of the same order and relative magnitude as those observed
in other trifluoracetamides.
6,17

Internal Rotation in N-Trifluoroacetylpyrrolidine
J. Phys. Chem. A, Vol. 107, No. 17, 2003 3027
TABLE 2: Exchange Rate Constants (k, s-1) for Gas-Phase
state (TS). When the nitrogen is part of a strained ring system,
the TS is destabilized relative to the GS and a higher nitrogen
inversion barrier is expected. Experimental values obtained for
and Solution TFAPYR
1,1,2,2-
-1 18
toluene
10% (v/v)
tetrachloroethane
10% (v/v)
N-methylaziridine in C6H12 (79.5 kJ mol ) and N-methylpyr-
T (K)
gas
neat
-1 19
rolidine in Freon solvents (33 kJ mol ) confirm that less-
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
38.0
44.0
49.9
55.9
61.9
71.1
73.8
82.3
85.7
89.0
92.4
93.5
95.7
99.1
04.7
28.9(1.8)
44.7(2.5)
63.7(3.7)
91.1(5.3)
137.6(7.7)
strained rings systems (pyrrolidine vs aziridine rings) yield lower
barriers to nitrogen inversion. Acetylation of the amine further
lowers the barrier. The pure p lone pair orbital of the TS has
better symmetry for pπ-pπ overlap than the lone pair in the
GS with its mix of s and p hybridization. The barrier to nitrogen
inversion of N-acetylaziridine in CH2CHCl has an experimental
16.8(0.9)
32.2(1.6)
27.5(1.4)
12.9(0.6)
25.3(1.3)
300.6(25.6)
603.8(49.6)
-
1 20
47.0(2.4)
60.4(3.0)
79.5(4.0)
98.9(5.0)
value of <29 kJ mol
compared to the previously cited
-1
N-methylaziridine (79.5 kJ mol ), so a large drop in barrier
height is expected for N-acetylpyrrolidines and in particular
TFAPYR in terms of lower ring strain and acetylation.
69.9(3.5)
57.2(3.0)
137.6(7.4)
Similarly, barriers to ring pseudorotation in five-membered
102.8(5.1)
136.1(6.8)
85.4(4.3)
104.6(5.3)
-1
rings start at an order of magnitude lower (0-17 kJ mol ) than
21
those to nitrogen inversion. Therefore, ring pseudorotation is
09.5 2556.5(240.6)
10.3
expected to be too fast to be observable in this TFAPYR DNMR
study.
191.6(9.6)
155.0(8.0)
q
q
q
TABLE 3: Internal Rotation ∆G 298, ∆H , and ∆S in
Gas-Phase and Solution TFAPYR
On the other hand, the barriers for internal rotation around
the amide bond are known to fall within DNMR limits (20-85
1,1,2,2-
-1 22
kJ mol ). In our study, the splitting patterns observed in the
toluene tetrachloroethane
10% (v/v)
1
H exchanging NMR signals of TFAPYR (two triplets, hydro-
gas
10% (v/v)
neat
gens attached to C-2 and to C-5, and two pseudoquintets,
G 298 (kJ mol-1) 71.4(1.2) 78.6(2.5)
H (kJ mol
S (J mol
q
80.9(1.5)
77.9(2.0)
-10.0(5.0)
80.4(0.1)
78.0(0.1)
-7.8(0.1)
∆
∆
∆
hydrogens attached to C-3 and to C-4) and the number of 13
C
q
-1
)
67.9(1.4) 75.0(3.2)
q
-1
exchanging NMR signals match those expected if fast nitrogen
inversion, fast ring pseudorotation, and slow internal rotation
)
-11.8(4.0) -12.0(8.3)
TABLE 4: Computational Molecular Parameters (MP2/
-311+G*) for the Ground-State Geometry of TFAPYRa
q
q
q
all take place. Therefore, the ∆G 298, ∆H , and ∆S values
measured for TFAPYR in this gas-phase/solution study can be
unambiguously assigned to the internal rotation around the C-N
amide bond.
6
r
r
r
r
r
r
C6N
C6O
NC5
NC2
C7C6
C7F
1.350
1.226
1.473
1.475
1.545
1.349
∠ OC6N
124.65
115.73
111.86
118.33
129.18
112.33
τ
τ
τ
τ
OC6C7F
OC6NC5
OC6NC2
C2C3C4C5
109.38
-0.90
-176.06
∠ C7C6N
∠ FC7C6
∠ C6NC5
∠ C6NC2
∠ C5NC2
Several magnetic field strengths were used in the acquisition
of the temperature-dependent NMR data. Gas-phase data were
1
collected at 500 MHz to maximize the separation of the H
a
TFAPYR signals and, therefore, increase the temperature
interval over which rates could be extracted. It is well-known
that the exchange temperature interval depends on the frequency
difference, δν, between the sites involved in the exchange. On
the other hand, C spectra of 10% (v/v) and neat TFAPYR
solutions were collected at 300 MHz to lower their exchange
temperature intervals and enable the inclusion of fast-exchange
data in the final analysis.
Bond lengths are in Å, and angles and dihedral angles are in deg.
1
3
The 500-MHz 1H NMR spectrum of gaseous TFAPYR
exhibits slow exchange below 294 K, coalescence at 348 K,
and fast exchange above 393 K. The resulting internal rotational
q
Gibbs free energy of activation barrier, ∆G 298, is found to be
-
1
lower in the gas phase (71.4 kJ mol ) than in any of the
q
solutions examined. The ∆G 298 values for TFAPYR in toluene,
1
,1,2,2-tetrachloroethane, and neat (78.6, 80.9, and 80.4 kJ
-
1
mol , respectively) also exhibit solvent dependence. Toluene-
d8 and 1,1,2,2-tetrachloroethane-d2 were chosen for their
solubility properties, high boiling points (toluene bp ) 111 °C;
Figure 5. Computed geometry (MP2/6-311+G*) for N-trifluoro-
1,1,2,2-tetrachloroethane bp ) 145 °C), and dielectric constants.
acetylpyrrolidine.
Toluene has a relatively small dielectric constant (ꢀ ) 2.4)
compared to 1,1,2,2-tetrachloroethane (ꢀ ) 8.2), and neat
TFAPYR is expected to have an even higher ꢀ value, based on
its relatively high dipole moment (µ ) 5.33 D; calculated using
TFAPYR’s ground-state geometry optimized with Gaussian 98.)
Free-energy activation barriers in amides are known to increase
with increasing solvent polarity, which is directly related to
There are three main conformational processes that dominate
the stereodynamics of N-acetyl-subtituted pyrrolidines: nitrogen
inversion, ring pseudorotation, and internal rotation around the
C-N bond. In this TFAPYR study, the internal rotation around
the C-N bond was the only process observed.
The nitrogen inversion barrier in TFAPYR is expected to be
too small to be observed by DNMR. In trialkylamines, the
process of nitrogen inversion changes the geometry of the
pyramidal ground state (GS) conformation to a planar transition
4
,23
dielectric constant. In this study, the dielectric constants are
in the order toluene < 1,1,2,2-tetrachloroethane < neat, and
q
the solution ∆G 298 values are in the order toluene < neat <

3028 J. Phys. Chem. A, Vol. 107, No. 17, 2003
Suarez et al.
q
TABLE 5: Gas-Phase ∆G 298 for Selected NR
2
-Substituted
there is no gas-phase value available for the internal rotation
Trifluoroacetamides
q
∆G 298 of N-methyltrifluororacetamide to compare because only
the strongly favored cis configuration is experimentally observed
2
7
q
at different temperatures. Nevertheless, internal rotation ∆G 298
values for N-methylacetamides are known to be relatively large
(
-
1
a value of 87.0 kJ mol for N-methylacetamide in 1,2-
dichloroethane has been reported). The dielectric constant for
1
,2-dichloroethane is known to be ∼10 comparable to 1,1,2,2-
q
tetrachloroethane’s dielectric value of 8.2. The ∆G 298 value
for TFAPYR in 1,1,2,2-tetrachloroethane is 80.9 kJ mol .
Therefore, the peptide bond in proline is expected to have an
-
1
q
internal rotation ∆G 298 of similar magnitude to TFAPYR and
to primary amide bonds in peptides.
In the typical primary amide peptide bonds, steric conflicts
between adjacent R-carbon substituents destabilize the cis
configuration sufficiently that less than 1% of peptide bonds
overall have a cis preference. The energy difference between
the cis and trans isomers of the peptide bond analogue
a
Reference 11. b Reference 23.
q
1
,1,2,2-tetrachloroethane. The higher ∆G 298 value for 1,1,2,2-
tetrachloroethane is not completely unexpected because halo-
genated solvents are known to enhance the solvent polarity of
their solutions.24
-1 28
N-methylacetamide has been calculated to be ∼11 J mol .
In comparison, the energy difference between the cis and trans
Only one other study has addressed the effect of cyclic
-1
configurations in proline has been found to be only 2.1 kJ mol
substituent ring size on barriers to amide rotation. 25 In the 13
C
because of the similar steric environments for the ring R-carbon
substituents in both configurations. This small difference in
energy probably accounts for the higher percentage (∼5%) of
NMR study of N-acetylpyrrolidine, Pinto et al. measured the
G of the internal rotation around the C-N bond to be 81.2-
0.8) kJ mol in (CD3)2SO, determined at the coalescence
q
∆
(
-
1
8
proline found in the cis configuration. This combination of
temperature of 403.7 K. This barrier fits within the solvent
dielectric series because the value of ꢀ for (CD3)2SO (46.7) is
higher than the value for toluene and lower than the expected
value for neat TFAPYR (the value of µ for (CD3)2SO is 3.9 D,
compared to 5.33 D calculated for TFAPYR).
relatively high barrier to internal rotation and relatively small
energy difference between cis and trans configurations seems
to be ideal properties for proline to have in its role as a switch.
We conclude that the internal rotation around the C-N amide
bond in N-trifluoroacetylpyrrolidine is much faster in the gas
phase than in solution, following similar patterns found in other
substituted amides. The five-membered ring substituent in
TFAPYR reduces the bulkiness around the C-N amide bond,
compared to N,N-dimethyl- and N,N-diethyltrifluoroacetamide,
q
It is also noteworthy to compare the gas-phase ∆G 298 value
of N-alkyl versus this N-cyclic trifluoroacetamide. True and
LeMaster and co-workers have done extensive studies on the
gas-phase barriers to rotation for trifluoroacetamides and have
4
,26
evaluated the results in terms of substituent bulkiness.
The
q
stabilizing the ground state and increasing the ∆G 298. The
current understanding is that the ground-state structures of
trifluoroacetamides are destabilized by bulky N-substituent
groups. The destabilization is attributed to steric hindrance and
the concomitant loss of the amide group planarity.23 This trend
is seen not only in the trifluoroacetamides but also in the
acetamides and formamides.26 The cyclization of the alkyl side
chains in pyrrolidine pulls the CH2 groups away from both the
carbonyl electrons and the fluorines decreasing the bulkiness
around the amide nitrogen. Table 4 lists and Figure 5 illustrates
some of the computational molecular parameters obtained at
the MP2/6-311+G* level for the ground-state geometry of
TFAPYR.
q
magnitude of the internal rotation barrier ∆G 298 in TFAPYR
is comparable to that found in primary amide peptide bonds,
emphasizing the importance of the relatively small energy
difference between cis and trans conformations in proline’s role
as a switch in protein signaling.
Acknowledgment. The authors thank the University of
Massachusetts at Amherst (500 MHz) and Wellesley College
(300 MHz) for the use of their NMR spectrometers and in
particular Charles Dikinson and Susan Kohler for their NMR
help. We thank David Bickar for his invaluable counsel, Ken
Wiberg for his computational advice, and Carole LeMaster for
her great editorial help.
This calculation predicts ground-state planarity of the amide
moiety (τOC6NC5 ) -0.90°) and ∠ C6NC5, ∠ C6NC2, and
∠
C5NC2 angle values of 118.33°, 129.18°, and 112.33°,
References and Notes
respectively. In comparison, the same MP2/6-311+G* level
ground-state calculations for dimethyl and diethyl trifluoroac-
etamide reveal similar planarity but smaller ∠ C6NC5 and
(1) Stewart, W. E.; Siddall, T. H. I. Chem. ReV. 1970, 70, 517.
(2) Wiberg, K.; Rush, D. J. J. Am. Chem. Soc. 2001, 123, 2038.
(3) Bragg, R. A.; Clayden, J.; Morris, G. A.; Pink, J. H. Chem.sEur.
∠
C6NC2 angles. Table 5 lists the values of these angles and
J. 2002, 8, 1279.
the corresponding gas-phase free energy of activation barriers
for all three compounds.
(
4) True, N. S.; Suarez, C. AdV. Mol. Struct. Res. 1995, 1, 115.
(5) Taha, A. N.; Neugebauer-Crawford, S. M.; True, N. S. J. Phys.
Chem. A 2000, 104, 7957.
The height of the free energy of activation barrier does not
determine the relative populations of the cis and trans rotational
conformers, but it definitely restricts the interconversion between
them. The role of proline peptide bonds as conformational
switches in proteins is based on their relatively large barrier to
internal rotation and their relatively small energy difference
between cis and trans conformers. We expect TFAPYR to have
(
6) Hao, B.; Gong, W.; Ferguson, T. K.; James, C. M.; Krzycki, J. A.;
Chan, M. K. Science 2002, 296, 1462.
7) Marks, D. B.; Marks, A. D.; Smith C. M. Basical Medical
Biochemistry. A Clinical Approach; Williams and Wilkins: Baltimore, MD,
996.
(
1
(8) MacArthur, M. W.; Thornton, J. M. J. Mol. Biol. 1991, 218, 397.
(9) Brazin, K.; Mallis, R. J.; Fulton, D. B.; Andreotii, A. H. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 1899.
10) Duffy, E. M.; Severance, D. L.; Jorgensen, W. L. J. Am. Chem.
Soc. 1992, 114, 7535.
q
an internal rotation ∆G 298 comparable to that of the standard
(
peptide analogue N-methyl trifluroacetamide. Unfortunately,

Internal Rotation in N-Trifluoroacetylpyrrolidine
J. Phys. Chem. A, Vol. 107, No. 17, 2003 3029
(
11) Ross, B. D.; True, N. S.; Decker, D. L. J. Phys. Chem. 1983, 87,
(18) Drakenberg, t.; Lehn, J. M. J. Chem. Soc., Perkin Trans. 2 1972,
8
8
3
9.
(
532.
(19) Lambert, J. B.; Oliver, W. L. J. J. Am. Chem. Soc. 1969, 91, 7774.
20) Anet, F. A. L.; Osyany, J. M. J. Am. Chem. Soc. 1967, 89, 352.
12) Ross, B. D.; Wong, L. T.; True, N. S. J. Phys. Chem. 1985, 89,
36.
(
19.
(
(
13) Ammann, C.; Meier, P.; Merbach, A. E. J. Magn. Res. 1982, 46,
The ∆G values from this paper were recalculated and reported by Jennings,
W. B.; Boyd, D. R. Strained Rings. In Cyclic Organonitrogen Sterody-
namics; Lambert, J. B., Takeuchi, Y., Eds.; VCH Publishers: New York,
14) (a) Stephenson, D. S.; Binsch, G. Program No. 365. (b) LeMaster,
C. B.; LeMaster, C. L.; True, N. S. DNMR5; Program Nos. 569 and
QCMP059; QCPE, Indiana University: Bloomington, IN.
1
991.
(21) Riddell, F. G. The Stereodynamics of Five-Membered Nitrogen-
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
Containing Rings. In Cyclic Organonitrogen Stereodynamics; Lambert, J.
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.6; Gaussian,
Inc.: Pittsburgh, PA, 1998.
B., Takeuchi, Y., Eds.; VCH Publishers Inc.: New York, 1991; p 159.
(22) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New
York, 1982.
(23) Suarez, C.; LeMaster, C. B.; LeMaster, C. L.; Tafazzoli, M.; True,
N. S. J. Phys. Chem. 1990, 94, 6679.
(
24) Perng, B. C.; Newton, M. D.; Raineri, F. F.; Friedman, H. L. J.
Chem. Phys. 1996, 104, 7153.
25) Pinto, M. B.; Grindley, T. B.; Szarek, W. A. Magn. Reson. Chem.
986, 24, 323.
26) LeMaster, C. L.; LeMaster, C. B.; True, N. S. J. Am. Chem. Soc.
(
1
1
(
999, 121, 4478.
(
16) Neugebauer Crawford, S. M.; Taha, A. N.; True, N. S.; LeMaster,
C. B. J. Phys. Chem. A 1997, 101, 4699.
17) Ng, S. J. Chem. Soc. A 1971, 1586.
(27) True, N. S.; Ross, B. D. J Phys. Chem. 1984, 88, 3216.
(28) Langley, C. H.; Allinger, N. L. J. Phys. Chem. A 2002, 106,
5638.
(