Diastereomers of N-α-phenylethyl-t-butylsulfinamide: Absolute configurations and predominant conformations

Article abstract of DOI:10.1021/jo702544g

(Chemical Equation Presented) N-α-Phenylethyl-t-butylsulfinamide is a complicated system for determining molecular stereochemistry because of numerous possibilities for assigning the absolute configuration and a predominant conformation. Two diastereomers

Full text of DOI:10.1021/jo702544g

Diastereomers of N-r-Phenylethyl-t-butylsulfinamide: Absolute
Configurations and Predominant Conformations
Ana G. Petrovic, Prasad L. Polavarapu,* Jozef Drabowicz, Piotr Ły z˘ wa,‡
†
,†
‡,|
‡
§
|
Marian Mikołajczyk, Wanda Wieczorek, and Agnieszka Bali n´ ska
Department of Chemistry, Vanderbilt UniVersity, NashVille, Tennessee 37235, Center of Molecular and
Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 112, 93-236 Lodz, Poland, Institute of
General and Ecological Chemistry, Technical UniVersity of Lodz, 90-924, Lodz, Zeromskiego 116, Poland,
Jan Długosz UniVersity, Institute of Chemistry and EnVironmental Protection, al. Armii Krajowej 13/15,
4
2-201 Cz e¸ stochowa, Poland
Prasad.L.PolaVarapu@Vanderbilt.edu
ReceiVed NoVember 27, 2007
N-R-Phenylethyl-t-butylsulfinamide is a complicated system for determining molecular stereochemistry
because of numerous possibilities for assigning the absolute configuration and a predominant conformation.
Two diastereomers of N-R-phenylethyl-t-butylsulfinamide derived from (-)-(S)-R-phenylethyl amine, a
(+)-diastereomer and a (-)-diastereomer, have been synthesized and their experimental chiroptical
spectroscopic properties have been measured. These properties include vibrational circular dichroism,
electronic circular dichroism and optical rotatory dispersion. Using these experimental data, in conjunction
with corresponding density functional theoretical predictions, the absolute configuration and predominant
conformations of these two diastereomers have been determined. Also, the absolute configuration of
(
-)-diastereomer has been independently confirmed by determining its structure from X-ray diffraction
data.
1
,4,9,10
Introduction
in the synthesis of enantiopure sulfinimines
as precursors
4
,10-12
for R- and â-amino acids,
R- and â-aminophospho-
Sulfonamides, obtained from
oxidation of sulfinamides, have been suggested to have
Within the family of sulfinyl derivatives,1,2 chiral sulfinamides
serve as useful chiral building blocks. In the enantioselective
asymmetric synthesis of many organic compounds, which are
important for the development of active pharmaceutical ingre-
dients and intermediates, sulfinamides were used as efficient
1
3,14
15,16
nates,
and chiral amines.
1
7
18
(
3) Senanayake, C. H.; Krishnamurthy, D.; Lu, Z.; Han, Z.; Gallou, I.
Aldrichim. Acta 2005, 38(3), 93-104.
chiral auxiliaries3 and catalysts. They found applications
-6
7,8
(4) Davis, F. A.; Zhou, P.; Chen, B. Chem. Soc. ReV. 1998, 27, 13-18.
5) Evans, D. A.; Faul, M. M.; Colombo, L.; Bisaha, J. J.; Clardy, J.;
(
Cherry D. J. Am. Chem. Soc. 1992, 114, 5977-5985.
(6) Davis, F. A.; Liu, H.; Zhou, P.; Fang, T.; Reddy, G. V.; Zhang, Y.
J. Org. Chem. 1999, 64, 7559-7567.
*
Corresponding author. Phone: (615)322-2836. Fax: (615)322-4936.
Vanderbilt University.
Polish Academy of Sciences.
Technical University of Lodz.
Jan Długosz University.
†
‡
§
|
(7) Pei, D.; Wang, Z.; Wei, S.; Zhang, Y.; Sun, J. Org. Lett. 2006, 8
(25), 5913-5915.
(1) Mikolajczyk, M.; Drabowicz J.; Kielbasinski, P. Chiral Sulfur
(8) Ellman, J. A. Pure Appl. Chem. 2003, 75, 39-46.
(9) Davis, F. A.; Reddy, R. E.; Szewczyk, J. M.; Reddy, C. V.; Portonovo,
P. S.; Zhang, H.; Fanelli, D.; Reddy, R. T.; Zhou, P.; Carroll, Patrick. J.
Org. Chem. 1997, 62, 2555-2563.
Reagents: Applications and Aymmetric and StereoselectiVe Synthesis; CRC
Press: New York, 1997.
(2) Metzner, P.; Thuillier, A. Sulfur Reagents in Organic Synthesis;
Academic Press: London, 1994.
(10) Davis, F. A.; Fanelli, D. L. J. Org. Chem. 1998, 63, 1981-1985.
10.1021/jo702544g CCC: $40.75 © 2008 American Chemical Society
3
120
J. Org. Chem. 2008, 73, 3120-3129
Published on Web 03/15/2008

Diastereomers of N-R-Phenylethyl-t-butylsulfinamide
existence of a stereogenic nitrogen atom, three stereogenic
3
centers (C, N, S) of 2 lead to 2 ) 8 possible diastereomers
(
Figure 4). As a result, one has to consider 81 × 8 ) 648
possibilities for predicting the chiroptical properties and as-
signing the absolute configuration and predominant conforma-
tions of 2. In practice, however, this number of possibilities is
reduced, as three of the orientations each for benzene and t-butyl
groups can be reduced to two, each due to the planarity of the
benzene ring and the symmetry of the t-butyl group (vide infra).
Furthermore, four diastereomers are mirror images of the other
four. Therefore it is sufficient to do calculations on four
diastereomers, each with 36 possible conformations.
FIGURE 1. t-Butanesulfinamide (1).
In the present work, two diastereomers of 2, a (+)-diastere-
omer (pSNA-2) and a (-)-diastereomer (mSNA-2) have been
synthesized in the reaction of racemic t-butanesulfinyl chloride
and (-)-(S)-R-phenylethyl amine, and their experimental chi-
roptical spectroscopic properties have been measured. Using
these experimental data, in conjunction with corresponding
density functional predictions, the absolute configuration and
predominant conformations of these two diastereomers having
the opposite absolute configurations at the stereogenic sulfinyl
sulfur atoms have been determined. Also the absolute configura-
tions at the stereogenic carbon and sulfur atoms of mSNA-2 are
independently confirmed here by determining its structure from
X-ray diffraction data.
FIGURE 2. The chemical structure of N-R-phenylethyl-t-butylsulfi-
namide (2).
applications in medicinal chemistry. Peptidosulfinamide, a
sulfinamide derivative, has found application19 for the develop-
ment of HIV protease inhibitors.
The selective induction of chirality through asymmetric
reactions necessitates not only the synthesis of enantiopure
sulfinamides, but also the reliable determination of their absolute
stereochemistry. In recent years, chiroptical spectroscopic
2
0
methods are becoming invaluable tools for reliable determi-
nation of absolute configuration and conformations of different
types of chiral molecules. This accomplishment is made possible
by the advances in instrumentation for experimental measure-
ments, and in quantum theoretical methods for theoretical
prediction, of chiroptical spectroscopic properties. The chirop-
tical spectroscopic methods used in the present study are
vibrational circular dichroism (VCD), electronic circular dichro-
ism (ECD), and optical rotatory dispersion (ORD).
It should be noted here that two diastereomers of the
sulfinamide 2 having the (R) absolute configuration at the
stereogenic sulfinyl sulfur atom have very recently been
prepared by the reduction of optically active N-t-butylsulfi-
nylimine derived from the enantiomerically pure sulfinamide
2
2a
(
R)-1 and acetophenone. They were isolated also earlier as
the products of an asymmetric addition of methylmagnesium
bromide to optically active N-t-butylsulfinylimine derived from
the enantiomerically pure sulfinamide (R)-1 and benzaldehyde.2
It should be mentioned that the structural conclusions obtained
here from chiroptical spectroscopic data and X-ray data were
not disclosed to individual research groups working on these
analyses until this manuscript was prepared. Therefore the
structural conclusions obtained from chiroptical spectroscopic
data and X-ray data are truly independent, and such independent
establishment of stereochemistry from chiroptical spectroscopic
data reflects the reliability and importance of modern day
chiroptical spectroscopies in molecular stereochemistry.
2b
Recently, we have explored the applicability of chiroptical
spectroscopic methods for structural elucidation of chiral
sulfinamides by investigating one of the simplest sulfinamides,
t-butanesulfinamide, 1 (Figure 1). Consistent and unambiguous
stereochemical assignment was obtained21 for 1 using all three
chiroptical spectroscopic methods.
N-R-Phenylethyl-t-butylsulfinamide, 2 (Figure 2) is a com-
plicated system for determining molecular stereochemistry. This
is because conformational flexibility resulting from four rotatable
4
bonds (Figure 3) leads, at least in principle, to 3 ) 81
conformers. Furthermore, if one assumes a possibility of the
Results and Discussion
(
11) Davis, F. A.; Szewczyk, J. M. Tetrahedron Lett. 1998, 39, 5951-
As mentioned in the introduction, the presence of three
stereogenic atoms C, N, and S in 2 leads to eight possible
diastereomers. However, since four of the diastereomers are
mirror images of the other four, only one set of non-enantiomeric
diastereomers need to be investigated. This is because chiroptical
spectra of the second set of diasteromers are obtained by
multiplying those of the first set by -1. The four diastereomers
selected as theoretical models for geometry optimization and
chiroptical predictions are SSS, SRS, SSR, and SRR. The three-
letter designations represent configurational assignment at the
stereogenic atoms C, N, and S, respectively.
5
6
1, 2222-2225.
(
13) Lefebvre, I. M.; Evans, S. A. J. Org. Chem. 1997, 62, 7532-7533.
(14) Mikolajczyk, M.; Lyzwa, P.; Drabowicz, J. Tetrahedron: Asymmetry
1
997, 8 (24), 3991-3994.
15) Han, Z.; Krishnamurthy, D.; Pflum, D.; Grover, P.; Wald, S. A.;
Senanayake, C. H. Org. Lett. 2002, 4 (23), 4025-4028.
16) Moreau, P.; Essiz, M.; Merour, J.; Bouzard, D. Tetrahedron:
Asymmetry. 1997, 8 (4), 591-598.
17) Clennan, E. L.; Chen, M.; Greer, A.; Jensen, F. J. Org. Chem. 1998,
3, 3397-3402.
18) Bharatam, P. V.; Kaur, A.; Kaur, D. J. Phys. Org. Chem. 2002, 15,
(
(
(
6
1
1
(
97-203.
2
1
(19) Moree, W. J.; Van der Marel, G. A.; Liskamp, R. J. J. Org. Chem.
Investigations on a smaller sulfinamide 1 revealed that
neither the geometries nor relative electronic energies are
995, 60, 5157-5169.
(20) For recent reviews, see (a) Polavarapu, P. L. Chem. Rec. 2007, 7,
1
1
25-136. (b) Polavarapu, P. L. Int. J. Quantum Chem. 2006, 106, 1809-
814.
(22) (a) Tanuwidjaja, J.; Peltier, H. M.; Ellman, J. A. J. Org. Chem.
2007, 72, 626-629. (b) Cogan, D. A.; Liu, G.; Ellman, J. A. Tetrahedron,
1999, 55, 8883-8904.
(21) Petrovic, A. G.; Polavarapu, P. L. J. Phys. Chem. A 2007, 111,
1
0938-10943.
J. Org. Chem, Vol. 73, No. 8, 2008 3121

Petrovic et al.
+
-
FIGURE 3. Newman projections around B and C bonds of 2 showing trans (T), plus gauche (G ), and minus gauche (G ) conformations.
FIGURE 4. Four diastereomers, SSS, SRR, SRS, and SSR, of 2. The three-letter designations represent configurational assignment at the stereogenic
atoms C, N, and S, respectively. The remaining diastereomers are mirror images of these four. The conformation of each diastereomer shown is of
lowest energy at the B3LYP/6-31G* level of theory.
influenced significantly in the presence of solvent. Therefore
all calculations were done for 2, as isolated molecule and solvent
influence was not investigated. Exploration of the conforma-
tional space for each diastereomer was achieved by varying the
dihedral angles corresponding to four rotatable bonds (A, B,
C, and D in Figure 2). The conformational search was initiated
with incremental 120° rotations around B and C bonds in order
is associated with a planar phenyl ring, a 90° rotation was
sufficient for conformational exploration. On the other hand,
the t-butyl group associated with the bond D is C3-symmetric,
hence requiring only one 60° rotation for the conformational
search.
Special attention was given to monitoring the configurational
preference of nitrogen during the geometry optimization process
+
-
23,24
to obtain trans (T), plus gauche (G ), and minus gauche (G )
because this chiral center is prone
to configurational
conformations. Newman projections of these conformations are
given in Figure 3. The combination of T, G , and G confor-
mations associated with the two dihedral angles has resulted
inversion. As a result of nitrogen inversion during the optimiza-
tion, some of the initial structures of SRS-2 have converged
into one of the stable conformations of SSS-2, and likewise some
of the initial structures of SSR-2 have converged into the stable
conformations of SRR-2.
+
-
+
-
+
+
+
in nine distinct conformations (TT, TG , TG , G T, G G ,
+
-
-
-
+
-
-
G G , G T, G G , G G ). For each diastereomer, these con-
formers have been subjected to further geometry optimization.
This further optimization has involved varying dihedral angles
corresponding to bonds A and D (Figure 2). Since the bond A
(23) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
John Wiley & Sons, Inc: New York, 1994.
(24) Tanaka, M.; Aida, M. Chem. Phys. Lett. 2006, 417, 316-319.
3122 J. Org. Chem., Vol. 73, No. 8, 2008

Diastereomers of N-R-Phenylethyl-t-butylsulfinamide
TABLE 1. Fractional Populations for Stable Conformations of the
four dominant conformations (C1-C4). Figure 6A shows the
correlation between the experimental spectra for pSNA-2 and
the B3LYP/6-31G* theoretical spectra for SSS-2 (see Supporting
Information for more details). Figure 6B displays corresponding
correlation between the experimental VCD spectra of mSNA-2
and the predicted VCD spectra for SRR configuration. Differ-
ences in the VCD band intensities among population weighted
and experimental spectra could result from different sources:
a
Four Diastereomers of 2
configuration
conformation
SSS
SRR
SRS
SSR
C1
C2
C3
C4
C5
C6
0.769
0.094
0.077
0.040
0.012
0.008
0.671
0.145
0.141
0.043
0.984
0.016
0.892
0.074
0.034
(a) the predicted relative intensities need not be accurate enough
at the theoretical level used; (b) relative populations of conform-
ers in solution environment used for experimental measurement
can be different from that for an isolated molecule considered
by the theoretical model.
a
The populations are based on Gibbs free energies obtained with the
B3LYP functional and the 6-31G* basis set.
The configurational assignment derived from VCD spectral
analysis can be confirmed from a comparison between experi-
mental and predicted ECD spectra. The negative ECD band at
209 nm for pSNA-2 corresponds to the B3LYP/6-31G*-predicted
negative ECD band at 213 nm (Figure 7A). On the other hand,
the positive ECD of mSNA-2 at 221 nm corresponds to the
B3LYP/6-31G*-predicted positive ECD band at 242 nm (Figure
7
3
B). Finally, a consideration of experimental and B3LYP/6-
1G*-predicted ORD spectra also validates the VCD-based
conclusion that pSNA-2 corresponds to SSS configuration (Figure
A), and mSNA-2 corresponds to SRR configuration (Figure 8B);
8
pSNA-2 has positive optical rotation at all wavelengths mea-
sured, as does SSS-2. Similarly, mSNA-2 has negative optical
rotation at all wavelengths measured, as does SRR-2. It can be
seen from Figures 6-8 that all three chiroptical methods exhibit
agreement in the above-stated stereochemical assignments for
the two experimentally considered diastereomers.
In the above analysis it should be noted that VCD spectral
analysis of 2 is critical for the assignment of absolute config-
uration of diastereomers. Because of the presence of numerous
vibrational transitions, comparison between experimental and
predicted VCD spectra (Figure 6) allowed for distinguishing
between different diastereomers, as only one diastereomer
provides satisfactory match with the experimental VCD spectra.
ECD spectral analysis (Figure 7), on the other hand, does not
offer as much discrimination between different diastereomers
because there is only one experimental ECD band in the
accessible experimental region, and the sign of that one ECD
band can be predicted for more than one diastereomer. In the
case of pSNA-2, the one negative ECD band observed could be
seen in the predicted spectra for SSS, SRS, and SSR diastereo-
mers. In the case of mSNA-2, the one positive ECD band
observed could be seen in the predicted spectra for SRR, RRR,
RSR, and RSR diastereomers. Similarly, ORD spectral analysis
FIGURE 5. Experimental VA (bottom panel), VCD, and noise spectra
for two diastereomers, pSNA-2 and mSNA-2. VCD and noise traces
are not shown in the region with excessive absorption intensity. The
spectra of mSNA are shifted up for clarity.
The geometry optimization using the B3LYP functional and
6-31G* basis set has resulted in six stable conformations for
SSS-2, four for SRR-2, two for SRS-2, and three for SSR-2
diastereomers. The relative populations of the conformers of
each diastereomer, determined from Gibbs free energies, are
given in Table 1.
(Figure 8) also does not offer as much discrimination between
different diastereomers because the experimental optical rota-
tions at all accessible wavelengths have a single sign, and that
sign can be predicted for more than one diastereomer. In the
case of pSNA-2, positive ORD has been predicted for SSS, RSS,
and RSR diastereomers. In the case of mSNA-2, negative ORD
has been predicted for SRR, RRR, SSR, and SRS diastereomers.
Thus, while VCD spectral analysis of 2 provided the critical
determination role, ECD and ORD spectral analyses of 2
provided a confirmatory role. In the chiroptical spectroscopic
analysis of several diastereomers of a chiral molecule, VCD
spectroscopy generally provides a better discriminatory role than
ECD or ORD.
For each of the four considered diastereomers, the most stable
conformer (labeled, C1), is shown in Figure 4. All conformations
identified as stable represent minima on the potential energy
surface, as no imaginary vibrational frequencies have been
predicted. Theoretical VCD, ECD, and ORD spectra for the
stable conformations have been obtained using the same B3LYP
functional and 6-31G* basis set. The comparison of population-
weighted VCD, ECD, and ORD spectra for each of the eight
diastereomers with the corresponding experimental data for two
diastereomers (Figures 5-8) has indicated that pSNA-2 corre-
sponds to SSS configuration with six dominant conformations
The dominant conformers of SSS-2 and SRR-2 are further
(C1-C6), and mSNA-2 corresponds to SRR configuration with
examined by reoptimizing their geometries at a higher B3LYP/
J. Org. Chem, Vol. 73, No. 8, 2008 3123

Petrovic et al.
FIGURE 6. Comparison of the experimental VA and VCD spectra of mSNA-2 and pSNA-2, with population-weighted spectra predicted for four
diastereomers at the B3LYP/6-31G* level.
aug-cc-pVDZ level. These reoptimizations have resulted in the
convergence of four conformations (C1′-C4′) each for SSS-2
and SRR-2. The relative populations determined from Gibbs
energies of the conformers determined at the B3LYP/aug-cc-
pVDZ level are given in Table 2.
and mSNA-2 has the SRR configuration. The vibrational assign-
ments derived from B3LYP/aug-cc-pVDZ calculations are
provided in the Supporting Information.
To confirm the configurational assignments derived from
VCD spectral analysis, ECD and ORD spectra predicted at the
higher B3LYP/aug-cc-pVDZ level are also considered. For this
purpose the conformers with relative populations greater than
10% have been used. The spectral comparisons associated with
ECD data are displayed in Figure 11, and that associated with
ORD data are shown in Figure 12. In terms of the ECD spectra,
pSNA-2 displays an overall satisfactory qualitative agreement
with the population-weighted spectra of SSS-2.
The negative experimental ECD band at 209 nm for pSNA-2
corresponds to the B3LYP/aug-cc-pVDZ-predicted negative
ECD band at 234 nm for SSS-2 (Figure 11A). In terms of mSNA-
2, its positive experimental ECD band at 221 nm corresponds
to the B3LYP/aug-cc-pVDZ-predicted positive ECD band at
The B3YP/aug-cc-pVDZ-calculated vibrational absorption
(VA) and VCD spectra of individual conformers along with
the population-weighted spectra for SSS-2 have been compared
to the experimental spectra of pSNA-2 in Figure 9. An analogous
spectral comparison between mSNA-2 and SRR-2 is given in
Figure 10. The double-headed arrows in these figures point to
the correlated bands between the population-weighted VCD
spectra of SSS-2 and SRR-2 and the experimental spectra of
pSNA-2 and mSNA-2, respectively (see Supporting Information
for more details). The overall satisfactory correlation among
multiple VCD bands at both theoretical levels (Figures 6 and
9) leads to the conclusion that pSNA-2 has the SSS configuration,
3124 J. Org. Chem., Vol. 73, No. 8, 2008

Diastereomers of N-R-Phenylethyl-t-butylsulfinamide
TABLE 2. Fractional Populations for Stable Conformations of the
SSS and SRR Diastereomers of 2a
configuration
conformation
SSS
SRR
C1′
C2′
C3′
C4′
0.908
0.043
0.036
0.014
0.650
0.173
0.168
0.009
a
The populations are based on Gibbs free energies obtained with the
B3LYP functional and the aug-cc-pVDZ basis set.
positive ECD band SRR-2, at both basis set levels, can be
justified. Similarly, the correlation between the negative ECD
band of pSNA-2 and the negative ECD band of SSS-2 at B3LYP/
aug-cc-pVDZ level can be justified. The close positioning of
the 6-31G*-predicted ECD band at 213 nm for SSS-2 (Figure
7
) with that of pSNA-2 is most likely fortuitous.
The absolute configurations derived from VCD and ECD
spectral analyses are supported by a good qualitative agreement
of ORD between experimental and B3LYP/aug-cc-pVDZ-
predicted ORD spectra (Figure 12). The specific rotation signs
for the population-weighted SSS-2 remain positive at each of
the six wavelengths considered, which is in agreement with that
of pSNA-2. Likewise, specific rotation signs at all wavelengths
considered for the population-weighted SRR-2 remain negative
as is the case for mSNA-2 (see Supporting Information).
Satisfactory correlations between experimental and predicted
spectra, from each of the three chiroptical spectroscopic
methods, at both basis set levels, provide confidence to the
conclusions established about the stereochemistry of diastere-
omers of 2. They are fully supported by X-ray analysis carried
out for the mSNA-2 diastereomer. Crystal data, experimental
details, and the basic geometric parameters are provided in the
Supporting Information. The X-ray crystal structure of mSNA-2
(Figure 13), which can be compared to the SRR diastereomer
in Figure 4, unequivocally revealed that the absolute configura-
FIGURE 7. Comparison of the experimental ECD spectra of mSNA-2
and pSNA-2 with population-weighted spectra predicted for four
diastereomers at the B3LYP/6-31G* level.
2
45 nm for SRR-2 (Figure 11B). As noted in the study21 of
smaller sulfinamide 1, it is not unusual for the electronic
2
5,26
transitions predicted with the B3LYP functional to appear
at longer wavelengths than the experimentally observed transi-
tions. With some blue shift of the predicted ECD bands,
correlation between the positive ECD band of mSNA-2 and the
FIGURE 8. Comparison of the experimental ORD spectra of mSNA-2 and pSNA-2 with population-weighted spectra predicted for four diastereomers
at the B3LYP/6-31G* level.
J. Org. Chem, Vol. 73, No. 8, 2008 3125

Petrovic et al.
FIGURE 9. Comparison of the experimental VA and VCD spectra of pSNA-2 with those predicted for SSS-2 at the B3LYP/aug-cc-pVDZ
level.
tions at the stereogenic carbon and sulfur atoms in mSNA-2 are
S) and (R), respectively. The bond angles about the sulfur atom
in the crystal structure are in the range 110.6(1)-98.6(1)°, and
the assumed position of the lone pair of electrons indicates a
distorted tetrahedral environment.
Finally, the sulfinamide diastereomer designated here as
mSNA-2 with (S)-absolute configuration at the C atom and (R)-
absolute configuration at the S atom corresponds to the sulfina-
mide, designated as 5a in ref 22a and sulfinamide 4c in ref
Conclusion
(
On the basis of the experimental and density functional theory
(DFT)-predicted VCD spectra, the absolute configurations and
dominant conformations for two diastereomers of N-phenylethyl-
t-butylsulfinamide (2) have been determined. One diastereomer,
pSNA-2 has the SSS configuration with one predominant
conformation, while the second diastereomer mSNA-2 has the
SRR configuration with three predominant conformations. These
conclusions are supported by the ECD and ORD spectral
analyses and also by the X-ray analysis carried out for mSNA-
2
2b. The sulfinamide diastereomer designated here as pSNA-2,
with (S)-configurations at both C and S atoms has not been
reported before, but the one with (R)-configurations at both C
and S atoms has been reported as 4a in ref 22a.
2
.
Experimental Section
(
25) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454-
The Dextrorotatory and Levorotatory Diastereomers of N-
R-Phenylethyl-t-butylsulfinamide, pSNA-2 and mSNA-2. A solu-
tion of (S)-(-)-R-phenylethylamine (2.45 g, 20 mmol) and triethy-
4
64.
(26) Polavarapu, P. L.; He, J.; Jeanne, C.; Ruud, K. Chem. Phys. Chem.
2
005, 6, 2535-2540.
3126 J. Org. Chem., Vol. 73, No. 8, 2008

Diastereomers of N-R-Phenylethyl-t-butylsulfinamide
FIGURE 10. Comparison of the experimental VA and VCD spectra of mSNA-2 with those predicted for SRR-2 at the B3LYP/aug-cc-pVDZ basis
level.
lamine (2.2 g, 22 mmol) in Et
with vigorous stirring to a solution of t-butanesulfinyl chloride (2.8
g, 20 mmol) in Et
was stirred for 6 h at -70 °C and 72 h at room temperature, the
solution was diluted with diethyl ether (50 mL) and quenched with
2
O (10 mL) was added dropwise
11-43 gave the pure levorotatory diastereomer (0.978 g) with [R]589
) -91.0 (c ) 1.24, MeCO Et); white solid mp ) 70 °C.
Spectral Measurements. The VA and VCD spectra were
2
2
O (100 mL) at -70 °C. After the reaction mixture
-
1
recorded in the 2000-870 cm region using a Fourier Transform
VCD spectrometer modified27 to reduce the level of artifacts. The
spectra of pSNA and mSNA were obtained with 3 h data collection
water. The organic phase was separated and washed with 5% H
2
-
-
1
SO solution, 5% K CO solution, and water and dried over MgSO
4
2
3
4
.
time, at 4 cm resolution, in a variable path length cell with BaF
windows. Measurements were done in CCl solvent. Solvent
2
After the solvent was removed under reduced pressure, the crude
sulfinamide (4.35 g) was purified by column chromatography on
silica gel (110 g). The column was washed subsequently with
petroleum ether/diethyl ether [1:1 v/v (86 × 10 mL)] and diethyl
ether (43 × 10 mL). Evaporation of petroleum ether/diethyl ether
4
absorption has been subtracted from the presented absorption
spectra. The enantiomers or racemic mixtures of the studied
diastereomers are not available, and therefore the VCD spectra of
solvent were subtracted from those of the samples.
(
1:1) fractions 41-56 gave the pure dextrorotatory diastereomer
The experimental VA, VCD, and noise spectra for pSNA-2
measured at 0.189 M concentration with 300 µm path length are
shown in Figure 5. Those for mSNA-2, measured at 0.343 M
concentration with 200 µm path length are also shown in Figure 5.
(0.52 g) with [R]589 ) +41.3 (c ) 1.55, MeCO
2
Et). Oil: Anal.
Calcd. for C12
H19NOS: C, 63.94; H, 8.50; N, 6.22; S, 14.24.
Found: C, 63.81; H, 8.73; N, 6.53; S, 14.64. Evaporation of
petroleum ether/diethyl ether (1:1) fractions 21-40 gave another
-
1
In these spectra, the ∼1030-1103 cm region is not shown
oil sample of the dextrorotatory diastereomer (2.2 g) with [R]589
)
1
+
39.65 (c ) 1.16, MeCO
2
Et), the H NMR of which was identical
(27) Shanmugam, G.; Polavarapu, P. L. J. Am. Chem. Soc. 2004, 126,
to that of the first sample. Evaporation of diethyl ether fractions
10292-10295.
J. Org. Chem, Vol. 73, No. 8, 2008 3127

Petrovic et al.
concentration to determine28 the intrinsic rotations. The experimental
ORD data presented here correspond to these intrinsic rotations.
Calculations. Geometry optimizations and chiroptical spectro-
scopic predictions for 2 were undertaken with quantum mechanical
programs2 using DFT and the B3LYP functional. The previous
9,30
21
investigation on 1 revealed that calculations with the smaller
-31G* basis set with the B3LYP functional are not adequate for
6
obtaining reliable predictions of electronic properties for 1. On the
other hand, calculations with the larger aug-cc-pVDZ basis set with
the B3LYP functional were found21 to provide a consistent
stereochemical assignment. Guided by these earlier results, both
6-31G* and aug-cc-pVDZ basis sets were used for the calculation
of chiroptical spectroscopic properties of 2.
The theoretical absorption and VCD spectra of 2 were simulated
-
1
with Lorentzian band shapes and a 5 cm half-width at 1/2 of
peak height. Frequencies calculated at the B3LYP/ 6-31G* level
were scaled by a factor of 0.9613. Frequencies obtained at the
B3LYP/aug-cc-pVDZ level have not been scaled.
The theoretical ECD spectra of 2 were simulated from the first
30 singlet f singlet electronic transitions using Lorentzian band
shapes and a 20 nm half-width at 1/2 of peak height. Rotational
strength values, calculated with velocity representation, have been
used for the ECD spectral simulations. The predicted ECD
intensities of 2 have been scaled-up by a factor of 6 in order to
facilitate a qualitative comparison with the experimental ECD
spectra.
FIGURE 11. Comparison of experimental and B3LYP/aug-cc-pVDZ
theoretical ECD spectra of 2. The inset shows the calculated spectrum
with the location of individual electronic transitions and their ECD
intensities.
because of the excessive absorbance in this region. Separate
measurements, at 0.175 M concentration and 140 µm path length
for pSNA, and at 0.138 M concentration and 170 µm path length
for mSNA, were undertaken to measure the absorption and associ-
ated VCD signals in this missing region. These two separate
measurements have been combined to compare the experimental
spectra in the entire region of the measurement with the predicted
spectra.
Crystal and Molecular Structure of the Levorotatory Dias-
tereomer of N-R-Phenylethyl-t-butylsulfinamide, mSNA-2. (A)
Crystal data for mSNA-2. C12
H19NOS, colorless, plate 0.22 ×
0.15 × 0.07 mm, orthorhombic, space group P2
1 1 1
2 2
, a ) 5.7913
3
(19), b ) 12.976 (3), c ) 17.478 (4) Å, V ) 1313.4 (6) Å, M )
r
-
3
-1
225.35, Z ) 4, dcalcd ) 1.140 g cm , µ ) 1.99 mm , T ) 296 (2)
K, F(000) ) 488.
(B) Data Collection. Enraf-Nonius CAD-4 diffractometer with
graphite monochromatized Cu KR radiation. Measured reflections
3051 (θmax 75.0°), 2699 independent (Rint 0.074).
(C) Structure Solution. direct method, anisotropic refinement
The concentration-dependent VA spectra of 2 have also been
measured (not shown) in CCl solvent, in the concentration range
4
of ∼77-15 mg/mL (0.343-0.067 M), to examine the presence of
dimers. However, no evidence of dimer formation was apparent
for either diastereomer.
The ECD spectra were recorded using a 0.01 cm path length
quartz cell. The concentrations used were 0.0195 M for pSNA and
2
on F for all non H atoms, while H atoms were generally fixed in
idealized positions (with the exception of some N-H protons,
whose positions were determined from a difference map) and
refined with their displacement parameters riding on those of their
parent atoms. The structure was refined over 2654 reflections with
I > 2σ (I) (145 refined parameters with 18.3 reflections on
parameter). The correct absolute structure was proved by Flack
0.0213 M for mSNA in hexane. The optical rotations at six discrete
wavelengths (633, 589, 546, 436, 405, and 365 nm) were measured
with a 1 dm cell. The concentration-dependent ORD measurements
-
3
31
for 2 were obtained in the concentration ranges 38.6 × 10 to 3.0
parameter x ) 0.00(2). For all data, the final wR2 was 0.131, R1
-
4
-3
-3
×
10 g/mL (0.0865-0.0006 M) for pSNA-2, and 19.5 × 10 to
) 0.048, S ) 1.08, max ∆F ) 0.56 e Å . Data processing was
-4
32
33
1
.5 × 10 g/mL (0.1713-0.0013 M) for mSNA-2. The concentra-
carried out with the CAD-4 Manual and SDP; structure solution
SHELXS;34 structure refinement SHELXL. Full details (excluding
35
tion-dependent specific rotations have been extrapolated to zero
FIGURE 12. Comparison of the experimental and B3LYP/aug-cc-pVDZ theoretical ORD spectra of 2.
128 J. Org. Chem., Vol. 73, No. 8, 2008
3

Diastereomers of N-R-Phenylethyl-t-butylsulfinamide
Acknowledgment. Computational work was partially sup-
ported by the National Center for Supercomputing Applications
under TG-CHE060063T and utilized the NCSA IBM P690 and
SGI Altix systems. Teaching assistantship from Vanderbilt
University (to A.G.P.) is greatly appreciated. The studies in Lodz
were in part financially supported (as Grant PBZ-KBN-126/
TO9/02 for J.D.) in the frame of a scientific project by the
government fund created to support science in the 2006-2009
period.
Supporting Information Available: General experimental
information, correlation between experimental and predicted VCD
bands, vibrational assignments, correlation between experimental
and predicted ORD data, tables of Cartesian coordinates and
energies of different conformers of 2 at B3LYP/aug-cc-pVDZ level,
tables of X-ray data, a figure showing intermolecular hydrogen
bonding in the crystal structure, and proton NMR spectra. This
material is available free of charge via the Internet at http://pubs.
acs.org.
FIGURE 13. The molecular structure of mSNA-2 determined from
X-ray diffraction data, showing 50% probability ellipsoids.
the structure factors) for the structure have been deposited at the
Cambridge Crystallographic Data Center, as supplementary publica-
tion no. CCDC 669314 and can be obtained free of charge upon
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, U.K.
JO702544G
[
e-mail: deposit@ccdc.cam.ac.uk].
(32) Schagen, J. D.; Straver, L.; van Meurs, F.; Williams, G. CAD-4
Manual; Data collection program; Enraf-Nonius: Delft, The Netherlands,
1989.
(33) Frenz, B. A. SDP; Data reduction program; Enraf-Nonius: Delft,
The Netherlands, 1986.
(28) Polavarapu, P. L.; Petrovic, A. G.; Wang, F. Chirality 2003, 15,
S143-S149.
(29) Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998 (full citation is
given in the Supporting Information).
(34) Sheldrick, G. M. SHELXS97; University of G o¨ ttingen: G o¨ ttingen,
Germany, 1997.
(35) Sheldrick, G. M. SHELXL97; University of G o¨ ttingen: Gottingen,
Germany, 1997.
(
30) Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004 (full citation
is given in the Supporting Information).
31) Flack, H. D. Acta Crystallogr., Sect. A 1983, 39, 876-881.
(
J. Org. Chem, Vol. 73, No. 8, 2008 3129