Microwave spectroscopy of the twist C(β)-exo/C(γ)-endo conformation of prolinamide

Article abstract of DOI:10.1021/ja981775z

The rotational transitions of three isotopomers of prolinamide were measured with a Fourier-transform microwave spectrometer. A twist pyrrolidine ring conformation with C(β)-exo and C(γ)-endo reproduces the experimental moments of inertia. Stark-effect measurements of five [M] components were used to determine that the dipole moment of this conformation of prolinamide is 3.83(4) D. Kraitchman's method of isotopic substitution was used to determine an N-N distance of 2.684(2) A?.

Full text of DOI:10.1021/ja981775z

10194
J. Am. Chem. Soc. 1998, 120, 10194-10198
Microwave Spectroscopy of the Twist Câ-Exo/Cγ-Endo Conformation
of Prolinamide
Kimberly A. Kuhls, Charla A. Centrone, and Michael J. Tubergen*
Contribution from the Department of Chemistry, Kent State UniVersity, Kent, Ohio 44242-0001
ReceiVed May 21, 1998
Abstract: The rotational transitions of three isotopomers of prolinamide were measured with a Fourier-transform
microwave spectrometer. A twist pyrrolidine ring conformation with Câ-exo and Cγ-endo reproduces the
experimental moments of inertia. Stark-effect measurements of five |M| components were used to determine
that the dipole moment of this conformation of prolinamide is 3.83(4) D. Kraitchman’s method of isotopic
substitution was used to determine an N-N distance of 2.684(2) Å.
Introduction
by comparison of experimental and theoretical spectra.5 None-
theless, microwave spectroscopy of the amino acids has been
hampered by low vapor pressures and narrow working temper-
ature ranges. No glycine signal could be observed below 150
°C in the FTMW studies, but temperatures greater than 170 °C
caused rapid decomposition.3
The conformations of amino acids have been investigated by
a variety of experimental and theoretical techniques, including
molecular mechanics, NMR spectroscopy, and X-ray crystal-
lography. Structural information about these small biological
molecules reveals their conformational preferences and provides
insight about the intramolecular forces which stabilize larger
peptides and proteins. Conformational studies of gas-phase
amino acids are particularly interesting because these conforma-
tions can be directly compared with computational models
without accounting for the intermolecular effects of condensed
media.
Microwave spectra have been recorded for two different
conformers of glycine.1-3 The initial studies1 compared the
experimental rotational constants and dipole moment to values
derived from a series of structural models; the reported spectrum
was assigned to a conformation containing an intramolecular
hydrogen bond from the carboxylic acid to the amino nitrogen.
However, Hartree-Fock calculations at the 4-31G level found
that another conformer, with hydrogen bonds from the amine
to the carbonyl oxygen, is 9.2 kJ mol-1 more stable than the
initially assigned structure.4 The rotational spectrum of the more
stable conformer was later found, but its spectrum was much
weaker due to its smaller dipole moment.2 Lovas and co-
workers have recently used a Fourier-transform microwave
(FTMW) spectrometer to resolve the hyperfine structure arising
from the 14N nuclear quadrupole moment.3 The higher resolu-
tion also enabled more precise measurements of the Stark shifts
of the |MF| components and consequently a more accurate
determination of the µa and µb dipole components.
Among the amino acids, only proline (Figure 1A) has a five-
membered-ring system that joins the side chain to the backbone
at the amino nitrogen. The five-membered ring restricts torsions
about the bond between CR and the amino nitrogen, so proline
reduces the flexibility of peptides and is often found in the bends
and kinks of folded protein chains. Despite proline’s unique
connectivity, HF/4-21G6 and HF/6-31G7 calculations find that
the most stable conformation contains an intramolecular hy-
drogen bond from the amine to the carbonyl oxygen and is
similar to the lowest energy glycine and alanine conformers.
The conformations of proline and proline derivatives are
complicated by puckering of the pyrrolidine ring, which relieves
torsional strain caused by eclipsing methylene groups on the
ring edges. Microwave spectra of pyrrolidine8 (Figure 1B) and
d1-pyrrolidine9 indicate that the ring adopts an envelope structure
with the nitrogen out of plane and the amino hydrogen axial.
Electron diffraction experiments and ab initio calculations10 (HF/
4-21 N*) confirmed that this is the lowest energy conformation
and found that the barrier to pseudorotation is 6.95 kJ mol-1
.
N-Methylpyrrolidine11 and N-cyanopyrrolidine12 also have N-
envelope structures, but with the methyl and cyano groups
equatorial.
(5) Godfrey, P. D.; Firth, S.; Hatherley, L. D.; Brown, R. D.; Pierlot, A.
P. J. Am. Chem. Soc. 1993, 115, 9687-9691.
(6) Tarakeshwar, P.; Manogaran, S. J. Mol. Struct. 1996, 365, 167-
181.
(7) Sapse, A.-M.; Mallah-Levy, L.; Daniels, S. B.; Erickson, B. W. J.
Am. Chem. Soc. 1987, 109, 3526-3529.
The microwave spectrum of a second amino acid, alanine,
was recorded and assigned to gas-phase conformations similar
to those identified for glycine; assignments were again made
(1) Brown, R. D.; Godfrey, P. D.; Storey, J. W. V.; Bassez, M.-P. J.
Chem. Soc., Chem. Commun. 1978, 547-548. Suenram, R. D.; Lovas, F.
J. J. Mol. Spectrosc. 1978, 72, 372-382.
(8) Caminati, W.; Oberhammer, H.; Pfafferott, G.; Filgueira, R. R.;
Gomez, C. H. J. Mol. Spectrosc. 1984, 106, 217-226.
(9) Ehrlichmann, H.; Grabow, J.-U.; Dreizler, H. Z. Naturforsch. 1989,
44a, 837-840.
(2) Suenram, R. D.; Lovas, F. J. J. Am. Chem. Soc. 1980, 102, 7180-
7184.
(10) Pfafferott, G.; Oberhammer, H.; Boggs, J. E.; Caminati, W. J. Am.
Chem. Soc. 1985, 107, 2305-2309.
(3) Lovas, F. J.; Kawashima, Y.; Grabow, J.-U.; Suenram, R. D.; Fraser,
G. T.; Hirota, E. Astrophys. J. 1995, 455, L201-L204.
(4) Vishveshwara, S.; Pople, J. A. J. Am. Chem. Soc. 1977, 99, 2422-
2426. Sellers, H. L.; Scha¨fer, L. J. Am. Chem. Soc. 1978, 100, 7728-7729.
Scha¨fer, L.; Sellers, H. L.; Lovas, F. J.; Suenram, R. D. J. Am. Chem. Soc.
1980, 102, 6566-6568.
(11) Pfafferott, G.; Oberhammer, H.; Boggs, J. E. J. Am. Chem. Soc.
1985, 107, 2309-2313. Caminati, W.; Scappini, F. J. Mol. Spectrosc. 1986,
117, 184-194.
(12) Su, C.-H.; Harmony, M. D. J. Mol. Spectrosc. 1985, 112, 328-
339.
S0002-7863(98)01775-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/19/1998

Twist Câ-Exo/Cγ-Endo Conformation of Prolinamide
J. Am. Chem. Soc., Vol. 120, No. 39, 1998 10195
Figure 1. Molecular structures of (A) proline, (B) pyrrolidine, and
(C) prolinamide. The molecular structure of prolinamide includes the
approximate projections of the a and b inertial axes. Ψ represents the
dihedral angle NC′CRN.
Figure 3. Ab initio model conformers. Conformer 10, which repro-
duces the experimental data best, is enlarged to show greater detail.
exo indicates that Câ and C′ are on the opposite sides of the
CR-N-Cδ planessee conformer 10 in Figure 3). DeTar and
Luthra modeled the ring-puckering conformations of prolyl
systems using molecular mechanics and found two broad
potential energy minima approximately corresponding to a Cγ-
exo envelope structure and a Câ-exo/Cγ-endo twist structure.14
DeTar and Luthra also investigated proline ring conformations
1
in solution using H and 13C chemical shifts of complexes
between N-acylproline esters and europium or ytterbium shift
reagents; they found that the observed shifts were reproduced
by a 60:40 mixture of twist and envelope conformations.16
Proline ring conformations, determined from the X-ray structures
of 40 different proline derivatives and peptides, cluster about
these two minima as well.14
The ring-puckering conformations of prolyl systems are often
described with use of the pseudorotation equation:15,17 øi )
ømax cos(P + 4πi/5) in which øi represents a bond torsion angle
of the ring, ømax is the maximum amplitude of puckering, and
P is the phase angle of puckering. The amplitude and phase of
ring puckering have been determined from many X-ray crystal-
lographic and NMR experiments; these experiments indicate
that the most frequent conformations have P ≈ 0° or P ≈ 180°
and are approximately twist structures with Câ-endo and Cγ-
exo (Haasnoot et al. refer to this as “N-domain”) or the closely
Figure 2. Semiempirical model conformers.
Ring puckering in proline and prolinamide may differ from
that in pyrrolidine because of the attachment of the carbonyl at
CR; prolinamide and its principal axes are shown in Figure 1C.
The X-ray crystal structure of L-proline,13 originally interpreted
as an envelope conformation, is consistent with a “twist”
pyrrolidine ring conformation with Câ-exo and Cγ-endo14,15 (Câ-
(15) Haasnoot, C. A. G.; De Leeuw, F. A. A. M.; De Leeuw, H. P. M.;
Altona, C. Biopolymers 1981, 20, 1211-1245.
(16) DeTar, D. F.; Luthra, N. P. J. Org. Chem. 1979, 44, 3299-3305.
(17) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94, 8205-
8212.
(13) Kayushina, R. L.; Vainshtein, B. K. SoV. Phys.-Crystallogr. 1966,
10, 698-706.
(14) DeTar, D. F.; Luthra, N. P. J. Am. Chem. Soc. 1977, 99, 1232-
1244.

10196 J. Am. Chem. Soc., Vol. 120, No. 39, 1998
Kuhls et al.
related Câ-exo/Cγ-endo structure (“S-domain”).15 The confor-
mational domains are quite broad, however, and include
conformations better described as envelopes (including the Cγ-
exo envelope structure).
Table 1. Center Frequencies for the Rotational Transitions of
14N2-Prolinamide
transition
obsd freq, MHz
obsd - calcd, MHz
303-202
313-212
312-211
322-221
321-220
404-303
414-313
413-312
212-101
303-212
313-202
404-313
9025.684
8707.280
9513.585
9126.676
9227.307
11925.940
11583.590
12651.430
7799.835
7284.940
10448.656
10503.316
-0.163
-0.242
-0.072
0.034
-0.131
0.028
0.054
0.116
-0.010
0.083
Experimental Section
The rotational spectra were recorded with a Balle-Flygare18
spectrometer consisting of a microwave cavity established by two 36-
cm diameter aluminum mirrors. The mirrors slide on four 1.27-cm
diameter stainless steel rails and can be positioned with a motorized
micrometer. Microwave radiation, generated by a Hewlett-Packard
83711B synthesized frequency generator, is coupled into the cavity by
an L-shaped antenna; molecular emission is detected by a similar
antenna on the opposite mirror and frequency reduced by a heterodyne
circuit described elsewhere.19 The molecular signal is then digitized
with use of a Keithley-MetraByte DAS-4101 data acquisition board in
a personal computer.
0.144
0.069
Table 2. Spectroscopic Constants for the Isotopic Species of
Prolinamide
Two solid aluminum plates, 30 cm × 30 cm × 0.5 cm, straddle the
cavity between the mirrors. The plates are separated by 27 cm and
are attached to the mirror guide rails with plastic spacers. Each plate
can be charged up to 10 000 V with opposite polarity for Stark effect
measurements. The electric field is calibrated by measuring the Stark
shift of the 1r0 transition of OCS (µ ) 0.715196 D20).
Prolinamide was heated to 150 °C and incorporated into an argon
expansion with a backing pressure of 1.5 atm. The expansion was
pulsed into the microwave cavity perpendicular to the cavity axis. This
arrangement results in Doppler splittings of 25 kHz and a line width
of 12 kHz for each component, when measured with carbonyl sulfide.
The frequency resolution of the spectrometer, determined by the number
of data points recorded for the free induction decay, was 4 kHz for
these measurements (1024 data points).
L-Prolinamide was used as purchased from Aldrich. The 15N-labeled
isotopes were prepared by a four-step synthesis starting from L-proline.21
The first step was to add a carbobenzoxy protecting group to the proline
amino group by reaction with benzyl chloroformate in sodium hydroxide
solution. CBZ-proline was obtained as a viscous oil. Next, the
carboxylic acid was esterified with p-nitrophenol with use of the
coupling reagent dicyclohexylcarbodiimide; the needles of CBZ-proline-
p-nitrophenyl ester melted at 92-94 °C. The amide was formed by
reacting the activated ester with anhydrous ammonia generated in a
vacuum line from NH4Cl and excess NaOH. Subsequent hydrogenation
of the CBZ-prolinamide in cyclohexene with palladium catalyst22
removed the carbobenzoxy protecting group and resulted in the
formation of the desired isotopomer of prolinamide. The final yield
of crude product was 40% over the four steps. The intermediate
compounds were characterized with use of melting points and FTIR
spectroscopy; a sample of 14N2-labeled final product was also character-
ized by comparison of the rotational spectrum to that of the com-
mercially available material. Two 15N-labeled isotopomers were
synthesized: a single 15N at the amide position (15Na-prolinamide),
prepared with the most abundant isotope of proline and 15NH3, and the
doubly substituted species (15N2-prolinamide) in which both the amide
and amine nitrogens were 15N labeled. The doubly substituted species
was prepared from 15N-L-proline and 15NH3 (Cambridge Isotope
Laboratories).
prolinamide
15Na-prolinamide 15N2-prolinamide
A (MHz)
B (MHz)
C (MHz)
DJ (kHz)
3640.355 (89)
1655.717 (20)
1386.497 (15)
a
a
a
3593.340 (24)
1641.516 (7)
1371.998 (6)
1.1 (2)
-9.5 (11)
a
3558.635 (6)
1635.443 (3)
1365.514 (2)
0.98 (3)
-4.2 (1)
-0.24 (4)
3.7
DJK (kHz)
d1 (kHz)
∆νrms (kHz) 114.9
22.4
a Value fixed to 0.0 kHz in the fit.
other and could not be resolved. The rotational transitions,
therefore, appeared as broad clusters of partially resolved
components spread over 600 kHz. Approximate line centers
were estimated to be near the center of each cluster of hyperfine
components; we estimate the uncertainty of these line centers
to be 200 kHz. Rotational constants (Table 2) were obtained
by fitting the line centers to the Watson S-reduction Hamiltonian
using the Ir representation; ∆νrms ) (νobs - νcalc rms
) 114.9
kHz. No c-type transitions were observed, suggesting that
the projection of the dipole moment on the c-inertial axis is
small.
)
Twelve and fifteen rotational transitions were observed for
15Na-prolinamide and 15N2-prolinamide, respectively. Again, no
c-type transitions could be found, despite accurate predictions
based on fits of the measured transitions. The rotational
transitions of 15Na-prolinamide were somewhat easier to resolve
due to the presence of only one quadrupoler nucleus, but
overlapping hyperfine structure still prevented assignment of
the components. These transitions spanned approximately 300
kHz, and the estimated center frequencies have uncertainties
of 75 kHz. Fits of the transitions to the Watson Hamiltonian,
including the distortion constants DJ and DJK, resulted in ∆νrms
) 22.4 kHz. Each rotational transition of the 15N2 isotopomer
had only a single component with a line width of 40 kHz. ∆νrms
) 3.7 kHz for the spectral fit.
Results
Measurements of the Stark shifts were made for five |M|
components of the 15N2-labeled species to avoid the complicating
effects of nuclear quadrupole splittings. We measured the Stark
shifts of M ) 0 and |M| ) 1 components arising from different
rotational transitions because the higher |M| components had
very fast Stark shifts (shifting ∼500 kHz in a 75 V/cm electric
field) which could not be calibrated against OCS. The
experimental Stark shifts were then fit to the a, b, and c
components of the dipole moment by using second-order Stark
coefficients calculated from the rotational constants. The best-
fit values were µa ) 3.13 (4) D, µb ) 2.21(2) D, µc ) 0.1(4) D,
and µtot ) 3.83(4) D, where the large uncertainties reflect the
small Stark shifts: typically ∆νStark e 400 kHz. Table 3
Twelve a- and b-type rotational transitions were measured
for the most abundant isotope of prolinamide (Table 1). Each
rotational transition consists of many hyperfine components
arising from the two 14N nuclei; these components overlap each
(18) Balle, T. J.; Flygare, W. H. ReV. Sci. Instrum. 1981, 52, 33-45.
(19) Tubergen, M. J.; Flad, J. E.; Del Bene, J. E. J. Chem. Phys. 1997,
107, 2227-2231.
(20) Tanaka, K.; Ito, H.; Harada, K.; Tanaka, T. J. Chem. Phys. 1984,
80, 5893-5905.
(21) Bodansky, M.; Bodansky, A. The Practice of Peptide Synthesis;
Springer-Verlag: Berlin, 1994.
(22) Brieger, G.; Nestrick, T. J. Chem. ReV. 1974, 74, 567-580. Jackson,
A. E.; Johnstone, R. A. W. Synthesis 1976, 685-687.

Twist Câ-Exo/Cγ-Endo Conformation of Prolinamide
J. Am. Chem. Soc., Vol. 120, No. 39, 1998 10197
Table 3. Comparison of the Observed and Calculateda Stark
that there are many different low-energy conformations, al-
though the experimental spectrum may not correspond to the
lowest energy structure.
Effects for 15N2-Prolinamide
∆ν/E2 (MHz kV-2 cm2)
The model structures were used to predict moments of inertia
(using the program STRGN25) which were then compared with
the experimental values for the identification of the conforma-
tion. Tables 4 and 5 report the root-mean-square averages of
the differences between observed and calculated moments of
inertia: ∆Irms where ∆I ) Ix(obs) - Ix(calc) and x ) a, b, and
c for each isotopomer. The different ring conformers predict
very different moments of inertia for the three isotopic species;
∆Irms ranges from 10.3 to 44.7 amu Å2 for the semiempirical
structures and from 2.1 to 35.9 amu Å2 for the ab initio
structures. In Tables 4 and 5 the Câ-exo/Cγ-endo twist
conformations have the lowest ∆Irms values, although the
semiempirical model is more puckered (ømax ) 41.4°) than the
corresponding ab initio model (ømax ) 32.9°); see below. The
ab initio twist conformer (10) has the lowest ∆Irms and therefore
best describes the ring conformer observed experimentally.
The model conformations described in Tables 4 and 5 all
have Ψ ) 0°; the effect of this dihedral angle on the moments
of inertia was systematically considered for each of the models.
For conformer 10, Ψ ) 0° gives ∆Irms ) 2.1 amu Å2. ∆Irms
rapidly increases as the dihedral angle changes and comes to
maxima at Ψ ) 50° and 230° (∆Irms ) 24.8 and 27.8 amu Å2,
respectively). Between these maxima, at Ψ ) 102°, a second
region of low ∆Irms values occurs (∆Irms ) 2.8 amu Å2).
Minima occur at different dihedral angles for the remaining
conformers, but their ∆Irms values never fall below 9.0 amu Å2,
except for conformer 1, which has ∆Irms ) 4.1 amu Å2 at Ψ )
80° and 180°. The experimental moments of inertia eliminate
all Ψ rotamers of conformers 1-9 but do not distinguish
between the Ψ ) 0° and 102° rotamers of conformer 10.
The projections of the dipole moment were used to distinguish
the two rotamers of conformer 10. Hartree-Fock calculations
(6-31G*) predict µtot ) 4.14 D, µa ) -3.31 D, µb ) 2.44 D,
and µc ) -0.48 D for the Ψ ) 0° rotamer of conformer 10
and µtot ) 4.84 D, µa ) -3.78 D, µb ) 2.79 D, and µc ) -0.57
D for the Ψ ) 102° rotamer. These conformations (and all
conformations 1-10) have the amino hydrogen exo and the lone
pair endosconsistent with an intramolecular hydrogen bond
from the amide to the amine (Ψ ) 0° describes a structure with
the amide oriented toward the amine). The orientation of the
amide in the Ψ ) 102° rotamer is inconsistent with intramo-
lecular hydrogen bonding, so the amino hydrogen may be either
exo or endo. Thus another HF/6-31G* model of conformer 10
with Ψ ) 102°, but with the amino hydrogen endo, was
computed to have µtot ) 4.58 D, µa ) -2.34 D, µb ) 3.83 D,
transition
|M|
obsd
calcd
312-211
404-303
414-313
413-312
303-212
0
1
1
0
0
4.0(1)
3.08(7)
5.3(3)
-3.20(4)
-2.22(4)
3.9
3.21
5.2
-3.28
-2.3
a Calculated with µa ) 3.13 D, µb ) 2.21 D, µc ) 0.1 D, and the
rotational constants in Table 2.
Table 4. Semiempirical Model Conformations
∆Irms
,
rel energy,
kJ mol-1
conformer
description
amu Å2
1
2
3
4
5
6
Câ-exo/Cγ-endo twist
Cγ-exo envelope
Câ-endo envelope
amino N-endo envelope
amino N-exo envelope
planar
10.3
39.3
41.6
38.8
44.7
20.6
53.4
22.0
56.9
63.3
165.7
0.0
Table 5. Ab Initio Model Conformations
∆Irms
,
rel energy,
kJ mol-1
conformer
description
amu Å2
7
8
9
Cγ-endo envelope
Câ-endo envelope
Câ-endo/Cγ-exo twist
Câ-exo/Cγ-endo twist
9.4
35.9
34.1
2.1
0.0
5.2
1.2
2.4
10
compares the experimental Stark shifts to the shifts calculated
with the best-fit dipole moment components.
Discussion
The amide derivative of proline was chosen for these
experiments because the melting point is much lower for
prolinamide (95-97 °C) than for proline (228 °C). Also, proline
decomposes at its melting temperature, while prolinamide is
stable over a relatively large temperature range.
We used the program Hyperchem23 with the AM1 param-
etrization to model the conformations of prolinamide starting
from the default values for the bond distances and angles of
the prolyl residue. Complete optimizations, in which no
constraints were applied to the structure, converged to a planar
ring conformation with Ψ ) 0° (eclipsing nitrogens about the
NC′CRN dihedral angle, see Figure 1C). Other conformations,
analogous to the envelope and twist structures described by
Kayushina and Vainshtein,13 DeTar and Luthra,14 and Haasnoot
et al.,15 were modeled by constraining Ψ (fixed at 0°) and some
of the ring torsional angles; these structures are shown in Figure
2. The relative energies of the six conformers are given in Table
4, and the atomic coordinates for these structures are available
as Supporting Information.
The conformations of prolinamide were also modeled at the
Hartree-Fock level (6-31G*) with PCSpartan24 software (Figure
3 and Table 5). The fully optimized structure converged to a
Cγ-endo envelope ring conformation with Ψ ) 0°. Two twist
structures and an additional envelope conformation were found
within 6 kJ mol-1 of the minimum energy structure. These
structures may not represent local minima on the potential
energy surface since Ψ and the ring torsional angles were
constrained as described above. The model calculations reveal
and µc ) 0.90 D. The experimental dipole components (µtot
)
3.83(4) D, |µa| ) 3.13(4) D, |µb| ) 2.21(2) D, and |µc| ) 0.1-
(4) D) are in good agreement with the dipole components
predicted for the Ψ ) 0° conformation but are inconsistent with
either prediction for the Ψ ) 102° conformation. The final
structure of the experimental conformation is best described as
a twisted ring with Câ-exo, Cγ-endo, and Ψ ) 0°.
The Kraitchman method26 can be used to calculate the
absolute values of the atomic coordinates of the nitrogens from
the experimental moments of inertia (Table 6); the relative signs
of the coordinates, however, must be deduced from the
molecular structure. Since the center of mass is located above
(25) Schwendeman, R. H. Critical EValuation of Chemical and Structural
Information; Lide, D. R., Paul, M. A., Eds.; National Acadamy of
Sciences: Washington, DC, 1974; pp 94-115.
(26) Kraitchman, J. Am. J. Phys. 1953, 21, 17-24. Gordy, W.; Cook,
R. L. MicrowaVe Molecular Spectra; Wiley: New York, 1984.
(23) Hyperchem, release 4; Hypercube: Waterloo, ON, 1994.
(24) PCSpartan 1.0; Wavefunction, Inc.: Irvine, CA, 1996.

10198 J. Am. Chem. Soc., Vol. 120, No. 39, 1998
Kuhls et al.
Table 6. Nitrogen Atomic Coordinates (Angstroms) from
Kraitchman Substitution Calculations
were simultaneously fit to the pseudorotational parameters for
both N- and S-domain conformations and the relative amounts
of the two conformations, described by the mole fraction of
the N-domain conformation. The corresponding N-domain
parameters from solution NMR (PN ) -3.0° and ømax,N ) 38.2°)
amino nitrogen
amide nitrogen
|a|
|b|
|c|
0.874(2)
0.999(2)
0.628(2)
1.533(1)
1.221(1)
0.539(3)
are very similar to model structure 9 (P ) -3.6° and ømax
)
39.3°) and describe a twist structure with Câ-endo and Cγ-exo.
Since the mole fraction fit to 0.5, both conformations are equally
populated in solution. Structure 9 is inconsistent with the
experimental microwave spectra because of its large ∆Irms value.
While no spectroscopic evidence (such as unassigned transitions)
for a second conformation was observed in the 875 MHz
frequency range searched, we cannot eliminate the possibility
of additional stable conformations in the expansion.
CR (see Figure 1), the a coordinates of the nitrogens have
opposite signs. The relative signs of the b and c coordinates
depend on the conformation and orientation of the amide; for
conformer 10 with Ψ ) 0° the two nitrogens have the same
sign for their b coordinates and opposite signs for their c
coordinates. Using these Kraitchman coordinates we found that
the N-N distance is 2.684(2) Å.
We used the structure of conformer model 10 to calculate
the pseudorotational phase angle, P, and the maximum out-of-
plane pucker amplitude, ømax, using the definitions of Westhof
Conclusions
Experimental moments of inertia of three isotopomers of
prolinamide were compared to several semiempirical and ab
initio models. A twist Câ-exo/Cγ-endo structure reproduces the
experimental values best. Stark effect measurements were used
to find µtot ) 3.83(4) D; both dipole moment and moment-of-
inertia data were used to determine that the NC′CRN dihedral
angle is approximately zero. The Kraitchman analysis of the
experimental moments of inertia indicates that the N-N distance
is 2.684(2) Å for this conformation.
and Sundaralingam:27 tan P ) B/A and ømax ) (A2 + B2)1/2
.
The torsional angles and origin of the pseudorotational cycle
were defined according to Schmidt et al.;28 the constants A and
B can then be calculated by using
4
2
4π
A )
ø cos
(i - 2)
∑
i
(
)
5 i)0
5
and
Acknowledgment. This work was supported by a grant from
the National Science Foundation (CHE-9700833). Acknowledg-
ment is also made to the donors of the Petroleum Research Fund,
administered by the American Chemical Society, for the support
of this research. The authors wish to thank Dr. Thomas Janini,
Dr. James Dudones, and Professor Paul Sampson for helpful
discussions regarding the synthesis of the isotopically labeled
species.
4
2
4π
(i - 2)
B ) -
ø sin
∑
i
(
)
5 i)0
5
Structure 10 has P ) 185° and ømax ) 32.9° and is consistent
with the conformations determined from NMR and X-ray
crystallography; structure 10 is especially close to the S-domain
conformation of prolinamide derived from NMR coupling
constants (PS ) 192° and ømax,S ) 33.3°).15 In the analysis of
Supporting Information Available: Tables of principal axis
atomic coordinates of model conformations 1-10 (10 pages
print/PDF). See any current masthead page for ordering
information and Web access instructions.
3
the prolinamide NMR data, the JHH NMR coupling constants
(27) Westhof, E.; Sundaralingam, M. J. Am. Chem. Soc. 1983, 105, 970-
976.
(28) Schmidt, J. M.; Bru¨schweiler, R.; Ernst, R. R.; Dunbrack, R. L.;
Joseph, D.; Karplus, M. J. Am. Chem. Soc. 1993, 115, 8747-8756.
JA981775Z