Quantitative evaluation of weak nonbonded Se···F interactions and their remarkable nature as orbital interactions

Article abstract of DOI:10.1021/ja016348r

To evaluate weak intramolecular nonbonded Se·"·F interactions recently characterized for a series of o-selenobenzyl fluoride derivatives (Iwaoka et al., Chem. Lett. 1998, 969-970), the temperature dependence of the nuclear spin coupling between Se and F (J_Se···F) was investigated for 2-(fluoromethyl)phenylselenenyl cyanate (1a) and bis[2-(fluoromethyl)phenyl] diselenide (1e) in CD₂Cl₂ and CD₃CN. A significant increase in the magnitude of J_Se···F was observed for both 1a and 1e upon lowering temperature, whereas the values of J_Se···F for the corresponding trifluoromethyl compounds slightly reduced or remained unchanged at low temperatures. Application of the rapid equilibrium model between two possible conformers revealed that conformer A with an intramolecular Se···F interaction is more stable in enthalpy (ΔH) by 1.23 kcal/mol for 1a (in CD₂Cl₂) and by 0.85 and 0.83 kcal/mol for 1e (in CD₂Cl₂ and CD₃CN, respectively) than conformer B, which does not have close Se···F contact. The negligible solvent effects for 1e suggested marginal electrostatic nature of the Se···F interactions. Instead, importance of the n_F → σ^*_Se-x orbital interaction was suggested by quantum chemical (QC) calculations and the natural bond orbital (NBO) analysis.

Full text of DOI:10.1021/ja016348r

Published on Web 02/08/2002
Quantitative Evaluation of Weak Nonbonded Se‚‚‚F
Interactions and Their Remarkable Nature as Orbital
Interactions
Michio Iwaoka, Hiroto Komatsu, Takayuki Katsuda, and Shuji Tomoda*
Contribution from the Department of Life Sciences, Graduate School of Arts and Sciences,
The UniVersity of Tokyo, Komaba, Meguro-ku, Tokyo 153-8902, Japan
Received June 6, 2001. Revised Manuscript Received December 11, 2001
Abstract: To evaluate weak intramolecular nonbonded Se‚‚‚F interactions recently characterized for a series
of o-selenobenzyl fluoride derivatives (Iwaoka et al., Chem. Lett. 1998, 969-970), the temperature
dependence of the nuclear spin coupling between Se and F (JSe‚‚‚F) was investigated for 2-(fluoromethyl)-
phenylselenenyl cyanate (1a) and bis[2-(fluoromethyl)phenyl] diselenide (1e) in CD2Cl2 and CD3CN. A
significant increase in the magnitude of JSe‚‚‚F was observed for both 1a and 1e upon lowering temperature,
whereas the values of JSe‚‚‚F for the corresponding trifluoromethyl compounds slightly reduced or remained
unchanged at low temperatures. Application of the rapid equilibrium model between two possible conformers
revealed that conformer A with an intramolecular Se‚‚‚F interaction is more stable in enthalpy (∆H) by
1.23 kcal/mol for 1a (in CD2Cl2) and by 0.85 and 0.83 kcal/mol for 1e (in CD2Cl2 and CD3CN, respectively)
than conformer B, which does not have close Se‚‚‚F contact. The negligible solvent effects for 1e suggested
marginal electrostatic nature of the Se‚‚‚F interactions. Instead, importance of the nF f σ*Se-X orbital
interaction was suggested by quantum chemical (QC) calculations and the natural bond orbital (NBO)
analysis.
applied to asymmetric synthesis4 and enzyme-mimetic catalytic
reactions5 but also because previous experiments have indicated
that a pseudo-multivalent state of Se might be related to
biological activities of Se compounds.6 Moreover, nonbonded
interactions involving Se may serve as potential analogues to
those involving divalent sulfur, which exists commonly in
biomolecules.7
Introduction
Weak nonbonded interactions1 play important roles in many
chemical phenomena, such as molecular recognition, confor-
mational transformation, and molecular packing in crystals.
Although recent progress in experimental methodologies and
computational technologies has made studies on weak interac-
tions significantly easier than before, it remains a difficult task
in basic chemistry to evaluate them quantitatively.2 In this paper,
we present the first quantitative estimation of weak nonbonded
interactions between selenium and fluorine (Se‚‚‚F interactions),
which have recently been characterized for a series of o-
selenobenzyl fluoride derivatives (1).3 Despite the highest
electronegativity of F among main-group elements, it has been
found that the Se‚‚‚F interaction energy for 1e (X ) SeAr) is
insensitive to solvent polarity, suggesting that the electrostatic
nature may not be important for stability.
We have recently succeeded in evaluation of the weak
nonbonded interactions between divalent Se and tertiary amino
nitrogen atoms (Se‚‚‚N interactions) by using a series of
o-selenobenzylamine derivatives as the model system.8 On the
1
basis of variable-temperature H NMR analysis and molecular
orbital (MO) calculations, the stabilization energies were
estimated as ca. ∼5-20 kcal/mol depending on the property of
the substituent at Se. It was concluded that the Se‚‚‚N interac-
tions are mostly stabilized by the orbital interaction between
the nitrogen lone pair (nN) and the low-lying antibonding orbital
(2) Mu¨ller-Dethlefs, K.; Hobza, P. Chem. ReV. 2000, 100, 143-167.
(3) Iwaoka, M.; Komatsu, H.; Tomoda, S. Chem. Lett. 1998, 969-970.
(4) (a) Fujita, K.; Iwaoka, M.; Tomoda, S. Chem. Lett. 1994, 923-926. (b)
Fujita, K.; Murata, K.; Iwaoka, M.; Tomoda, S. Tetrahedron 1997, 53,
2029-2048. (c) Wirth, T. Liebigs Ann./Recl. 1997, 2189-2196. (d) Spichty,
M.; Fragale, G.; Wirth, T. J. Am. Chem. Soc. 2000, 122, 10914-10916.
(5) (a) Iwaoka, M.; Tomoda, S. J. Chem. Soc., Chem. Commun. 1992, 1165-
1167. (b) Wirth, T.; Ha¨uptli, S.; Leuenberger, M. Tetrahedron: Asymm.
1998, 9, 547-550.
Nonbonded interactions involving divalent selenium (Se) are
intriguing subjects, not only because they have been successfully
(6) (a) Iwaoka, M.; Tomoda, S. J. Am. Chem. Soc. 1994, 116, 2557-2561.
(b) Mugesh, G.; Singh, H. B. Chem. Soc. ReV. 2000, 29, 347-357. (c)
Mugesh, G.; Panda, A.; Singh, H. B.; Punekar, N. S.; Butcher, R. J. J. Am.
Chem. Soc. 2001, 123, 839-850.
* To whom correspondence should be addressed: Phone +81 (3) 5454-
6575; fax +81 (3) 5454-6998; e-mail tomoda@selen.c.u-tokyo.ac.jp.
(1) (a) Israelachvili, J. Intermolecular and Surface Forces; Academic Press:
London, 1991. (b) Molecular Interactions; Scheiner, S., Ed.; John Wiley
and Sons: Chichester, U.K., 1997.
(7) (a) Nagao, Y.; Hirata, T.; Goto, S.; Sano, S.; Kakehi, A.; Iizuka, K.; Shiro,
M. J. Am. Chem. Soc. 1998, 120, 3104-3110. (b) Wu, S.; Greer, A. J.
Org. Chem. 2000, 65, 4883-4887. (c) Iwaoka, M.; Takemoto, S.; Okada,
M.; Tomoda, S. Chem. Lett. 2001, 132-133.
(8) Iwaoka, M.; Tomoda, S. J. Am. Chem. Soc. 1996, 118, 8077-8084.
9
1902 VOL. 124, NO. 9, 2002 J. AM. CHEM. SOC.
10.1021/ja016348r CCC: $22.00 © 2002 American Chemical Society

Weak Nonbonded Se‚‚‚F Interactions
A R T I C L E S
of Se (σ*Se-X). On the other hand, two mutually different
stabilization mechanisms (i.e., electrostatic and orbital interaction
mechanisms) were reported for intramolecular nonbonded
Se‚‚‚O interactions by Goldstein et al.9 and Barton et al.,10
respectively, depending on the model systems. Our systematic
examination using o-selenobenzaldehyde and o-selenobenzyl
alcohol derivatives suggested that the Se‚‚‚O interactions are
weaker than the corresponding Se‚‚‚N interactions and that an
orbital interaction mechanism is still a dominant factor for the
stability.11 The strong interaction between a cationic Se and an
alcoholic O atom (the Se+‚‚‚O interaction) was also extensively
analyzed in relation to selenium-based asymmetric syntheses.4d
Nonbonded Se‚‚‚F interaction is an extreme case to test the
electrostatic character of divalent Se, since F is the most
electronegative main-group element in the periodic table.
Although several compounds were found to have close Se‚‚‚F
contacts in the solid state,12 mechanistic studies of the Se‚‚‚F
interactions appeared only recently. Borisenko et al.13 suggested
the electrostatic nature of the Se‚‚‚F interaction observed for 2
in the gas phase. On the other hand, Nakanishi et al.14 discussed
orbital interactions of the close Se‚‚‚F contact in 3. These
compounds, however, possess strong steric restrictions, hence
the intrinsic nature of weak Se‚‚‚F interaction is not yet
thoroughly elucidated.
and electrostatic characters, and (3) to show usefulness of the
combined method of quantum chemical (QC) calculations and
natural bond orbital (NBO) analysis to understand these weak
interactions. We took advantage of a series of o-selenobenzyl
fluoride derivatives (1) as model compounds, which have the
freedom of internal rotation to adopt or to avoid intramolecular
Se‚‚‚F contact. In addition, the corresponding trifluoromethyl
(CF3) compounds (4) were employed as references because they
always adopt intramolecular Se‚‚‚F contact independent of the
internal rotation of the C-CF3 bond. Our preliminary investiga-
tions with this model system have already shown the presence
of nonbonded Se‚‚‚F interactions for 1 in CDCl3.3 The proposed
nF f σ*Se-X orbital interaction mechanism was theoretically
supported by Jeong and Kwon recently.16 Here, we have
developed a novel methodology to evaluate weak Se‚‚‚F
interactions by using the temperature dependence of JSe‚‚‚F. The
stabilization mechanisms are discussed on the basis of the
observed solvent effects on the interaction energies as well as
on the basis of the results from QC and NBO analyses.
Results and Discussion
Model Compounds. A series of o-selenobenzyl fluorides
(1a-g) were synthesized as model compounds to investigate
intramolecular nonbonded Se‚‚‚F interactions. Diselenide 1e was
synthesized from bis[2-(chloromethyl)phenyl] diselenide17 by
applying modified Ichihara’s fluorination method,18 in which
butyronitrile was used as a solvent instead of acetonitrile in order
to raise the reaction temperature. Compound 1e was then
converted into the other model compounds (1a-d and 1f,g) by
using the common procedures.3,8 Compounds were identified
Se‚‚‚F interaction can also serve as a useful probe for
estimating the electron-donating character of organic fluorides,
since it was previously shown by the studies of nonbonded
Se‚‚‚N and Se‚‚‚O interactions that Se is a good electron
acceptor.8,11 Although F is a strong σ electron-withdrawing
element, an electron-donating character of F lone pairs has often
been pointed out.15 Evaluation and mechanistic characterization
of Se‚‚‚F interaction is therefore of significant interest from the
viewpoint of organic fluorine chemistry.
The main purposes of this paper are 3-fold: (1) to estimate
the stabilization energies of weak nonbonded Se‚‚‚F interactions
in various solvents, (2) to delineate the stabilization mechanism
of Se‚‚‚F interactions, i.e., to assess relative importance of orbital
1
with H, 13C, 19F, and 77Se NMR spectra, although 1b was
contaminated by small amounts of 1e and an unknown byprod-
uct, and 1d slowly disproportionated into 1e and diphenyl
disulfide at room temperature.
Similar procedures were employed for the synthesis of the
corresponding CF3 compounds (4). Selenocyanate 4a was
synthesized from o-trifluoromethylaniline by diazotization,
followed by treatment with potassium selenocyanate. Compound
4a was then converted into 4e and 4f. Selenolate 4g was
prepared in situ by treatment of 4e with sodium borohydride in
CD3OD.19 All reference compounds were also characterized
spectrally.
NMR Chemical Shifts and JSe‚‚‚F Coupling Constants.
77Se NMR chemical shifts (δSe), 19F NMR chemical shifts (δF),
and nuclear spin coupling constants between 77Se and 19F
(JSe‚‚‚F) observed for 1a-g, 4a, 4e, and 4g at 298 K were
tabulated previously (Table 1 of ref 3). Nonbonded Se‚‚‚F
interactions of 1a-f have been characterized as follows.
Unexpectedly large JSe‚‚‚F coupling constants were observed
for 1a-f (JSe‚‚‚F ) ∼22.7-84.2 Hz in CDCl3).3 Because 4JSe-F
(9) (a) Goldstein, B. M.; Kennedy, S. D.; Hennen, W. J. J. Am. Chem. Soc.
1990, 112, 8265-8268. (b) Burling, F. T.; Goldstein, B. M. J. Am. Chem.
Soc. 1992, 114, 2313-2320.
(10) Barton, D. H. R.; Hall, M. B.; Lin, Z. Y.; Parekh, S. I.; Reibenspies, J. J.
Am. Chem. Soc. 1993, 115, 5056-5059.
(11) Komatsu, H.; Iwaoka, M.; Tomoda, S. Chem. Commun. 1999, 205-206.
(12) (a) Gallois, B.; Gaultier, J.; Hauw, C.; Lamcharfi, T.-D.; Filhol, A. Acta
Crystallogr. 1986, B42, 564-575. (b) Emge, T. J.; Wang, H. H.; Beno,
M. A.; Williams, J. M.; Whangbo, M.-H.; Evain, M. J. Am. Chem. Soc.
1986, 108, 8215-8223.
(13) Borisenko, K. B.; Broschag, M.; Hargittai, I.; Klapo¨tke, T. M.; Schro¨der,
D.; Schulz, A.; Schwarz, H.; Tornieporth-Oetting, I. C.; White, P. S. J.
Chem. Soc., Dalton Trans. 1994, 2705-2712.
(14) Nakanishi, W.; Hayashi, S.; Sakaue, A.; Ono, G.; Kawada, Y. J. Am. Chem.
Soc. 1998, 120, 3635-3640.
(15) (a) Tomoda, S.; Takamatsu, K.; Iwaoka, M. Chem. Lett. 1998, 581-582.
(b) Jefford, C. W.; Hill, D. T.; Ghosez, L.; Toppet, S.; Ramey, K. C. J.
Am. Chem. Soc. 1969, 91, 1532-1534. (c) Yamamoto, G.; Oki, M. J. Org.
Chem. 1984, 49, 1913-1917. (d) Yamamoto, G.; Oki, M. Tetrahedron
Lett. 1985, 26, 457-460. (e) Gribble, G. W.; Kelly, W. J. Tetrahedron
Lett. 1985, 26, 3779-3782. (f) Dunitz, J. D.; Taylor, R. Chem. Eur. J.
1997, 3, 89-98. (g) Plenio, H. Chem. ReV. 1997, 97, 3363-3384. (h)
Thalladi, V. R.; Weiss, H.-C.; Bla¨ser, D.; Boese, R.; Nangia, A.; Desiraju,
G. R. J. Am. Chem. Soc. 1998, 120, 8702-8710.
(16) Jeong, M.; Kwon, Y. Chem. Phys. Lett. 2000, 328, 509-515.
(17) Iwaoka, M.; Tomoda, S. Phosphorus Sulfur Silicon Relat. Elem. 1992, 67,
125-130.
(18) Ichihara, J.; Matsuo, T.; Hanafusa, T.; Ando, T. J. Chem. Soc., Chem.
Commun. 1986, 793-794.
(19) Miyashita, M.; Hoshino, M.; Yoshikoshi, A. Tetrahedron Lett. 1988, 29,
347-350.
9
J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002 1903

A R T I C L E S
Iwaoka et al.
Table 1. JSe‚‚‚F Coupling Constants Observed for 1a, 1e, and 1f in
Various NMR Solvents at 298 K
JSe‚‚‚Fb (Hz)
1a
(X ) CN)
1e
1f
solvent
ꢀa
(X ) SeAr)
(X ) Me)
(CD3)2SdO
CD3CN
(CD3)2CdO
CD2Cl2
CDCl3
C6D6
46.45
35.94
20.56
8.93
4.81
2.27
39.0
62.9
68.7
76.9
84.2
81.8
32.6
30.3
30.0
24.7
23.6
23.0
29.4
27.1
26.2
23.9
22.7
22.1
Figure 1. Equilibrium between two stable conformers (A and B) of 1.
By use of the rapid equilibrium model (Figure 1), it is possible
to evaluate the Se‚‚‚F interaction energies (HSe‚‚‚F) experimen-
tally as follows. First, the observed value of JSe‚‚‚F represents
weighted average of those for conformers A and B:
a Dielectric constants of the solvents at 298 K quoted from ref 21. bThe
1
values of JSe‚‚‚F coupling constants were measured from the expanded H-
decoupled 77Se NMR spectra.
JSe‚‚‚F ) JSe‚‚‚F(A)øA + JSe‚‚‚F(B)(1 - øA)
(1)
coupling constants are usually small [4JSe-F ) ∼10-28 Hz for
2,4,6-tris(trifluoromethyl)phenylseleno derivatives],20 the larger
values of JSe‚‚‚F for 1a-f suggested the presence of direct
through-space nonbonded interactions between Se and F. The
magnitude of JSe‚‚‚F correlated positively with the downfield
shifts of δF, except for 1c (X ) Br).3 The propensity suggested
that the F atom loses more electrons as the Se‚‚‚F interaction
becomes stronger. Therefore, the electrons may delocalize from
F to Se in the Se‚‚‚F interactions. The JSe‚‚‚F coupling was not
detected for selenolate 1g (JSe‚‚‚F ) 0 Hz in CD3OD), in which
the excessive negative charge at Se may cause repulsion with
the F lone pair electrons to avoid close Se‚‚‚F contact.
Consequently, the JSe‚‚‚F coupling could be observed only when
nonbonded Se‚‚‚F interaction existed in solution.
Table 1 shows solvent effects on JSe‚‚‚F observed for 1a
(X ) CN), 1e (X ) SeAr), and 1f (X ) Me) at room
temperature. The value of JSe‚‚‚F for 1a decreased significantly
with an increasing dielectric constant of the solvents,21 while
completely opposite trends were observed for 1e and 1f. Since
the cyano group is a strong electron-withdrawing substituent,
the Se‚‚‚F interaction of 1a would possess electrostatic character
to some extent. Hence the interaction might become weaker in
polar solvents. Similar electrostatic contribution was previously
found for the nonbonded Se‚‚‚N interaction of N,N-dialkyl-2-
(aminomethyl)phenylselenenyl cyanate.8 On the other hand,
remarkable enhancement of JSe‚‚‚F in polar solvents observed
for 1e and 1f suggested that the electrostatic character of the
Se‚‚‚F interactions in this system may be negligible. The result
is quite surprising considering the high electronegativity of F.
Rapid Equilibrium between Conformers A and B. 77Se
NMR signals of reference compounds appeared as quartets
(JSe‚‚‚F ) 63.2 Hz for 4a, 53.0 Hz for 4e, 48.4 Hz for 4f, and
15.6 Hz for 4g), indicating free rotation of the CF3 groups at
room temperature. In addition, low-temperature 1H NMR spectra
of 1a-f showed a sharp singlet in the region of benzylic protons
even at -90 °C at 500 MHz. To explain the dynamic NMR
spectra, rapid equilibrium between two stable conformers has
been proposed (Figure 1):3 Conformer A has close atomic
contact between Se and F with almost linear X-Se‚‚‚F atomic
alignment, and conformer B has the C-F bond away from the
Se-X bond, which lies out of the aromatic plane.
where JSe‚‚‚F(A) and JSe‚‚‚F(B) are the Se‚‚‚F coupling constants
for conformers A and B, respectively. øA is a temperature-
dependent variable representing the relative population of
conformer A. It should be noted that the value of JSe‚‚‚F(B) for
1a-f would be approximately zero [JSe‚‚‚F(B) ≈ 0 Hz; namely,
JSe‚‚‚F ≈ JSe‚‚‚F(A)øA] because the nuclear spin coupling between
Se and F was not observed for 1g, for which conformer B is
predicted to be the only stable conformer in solution. Second,
øA is related to enthalpy and entropy changes (∆H and ∆S) of
the equilibrium (Figure 1) by the free energy relationship:
∆G ) ∆H - T∆S ) -RT ln {(1 - øA)/øA}
(2)
where T and R represent temperature and the gas constant,
respectively. Third, since rotational barriers of the CH2F group
and the Se-X moiety must be low as indicated by the low-
temperature NMR spectra, the enthalpy for the interaction
between Se and F (HSe‚‚‚F) is approximately equal to ∆H:
HSe‚‚‚F ≈ ∆H
(3)
Thus, in principle a quantitative estimate of HSe‚‚‚F can be
experimentally determined by measuring JSe‚‚‚F values at various
temperatures if JSe‚‚‚F(A) remains constant. This critical condition
was practically examined by using the JSe‚‚‚F values for the
corresponding CF3 compounds (4) at various temperatures.
Temperature Dependence of JSe‚‚‚F for 1a and 1e. Com-
pounds 1a (X ) CN) and 1e (X ) SeAr) were selected to
investigate the temperature dependence of JSe‚‚‚F because they
are remarkably more stable than the other model compounds.
CD3CN and CD2Cl2 were selected as polar and less polar
solvents, respectively. The values of JSe‚‚‚F were measured from
the expanded 77Se NMR spectra, in which the signals are split
into a doublet due to coupling with the nuclear spin of 19F. As
seen in Figure 2, the splitting of 77Se NMR signals was enhanced
at low temperatures in CD2Cl2 for both 1a and 1e. Similar
propensities were also observed in CD3CN. The JSe‚‚‚F vs T plots
for 1a and 1e are shown in Figures 3 and 4, respectively, where
the data for the reference compounds (4a and 4e) are also
plotted.
Figure 3 suggests two important features as to the Se‚‚‚F
interaction of 1a (the NC-Se‚‚‚F interaction). First, the Se‚‚‚F
interaction is attractive in both solvents (HSe‚‚‚F > 0). This can
be predicted as follows. According to the rapid equilibrium
model (Figure 1 and eq 1), the increase of the magnitude of
(20) (a) Bertel, N.; Roesky, H. W.; Edelmann, F. T.; Noltemeyer, M.; Schmidt,
H. G. Z. Anorg. Allg. Chem. 1990, 586, 7-18. (b) Labahn, D.; Bohnen, F.
M.; Herbst-Irmer, R.; Pohl, E.; Stalke, D.; Roesky, H. W. Z. Anorg. Allg.
Chem. 1994, 620, 41-47.
(21) Techniques of Chemistry; Riddick, J. A., Bunger, W. B., Sakano, T. K.,
Eds.; John Wiley and Sons: New York, 1986; Vol. 2.
J
Se‚‚‚F observed for 1a at low temperatures should be due to the
enhancement of øA or JSe‚‚‚F(A). However, the JSe‚‚‚F(A) value
9
1904 J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002

Weak Nonbonded Se‚‚‚F Interactions
A R T I C L E S
Figure 2. 77Se NMR spectra (95.35 MHz) observed for 1a (left panel)
and 1e (right panel) in CD2Cl2 at various temperatures.
Figure 4. Plots of JSe‚‚‚F observed for 1e (O) and 4e (9) in CD2Cl2 and
CD3CN as a function of temperature. The solid (in CD2Cl2) and dashed
(CD3CN) curves for 1e were depicted with the parameters obtained by data
fitting.
and 4e were unexpectedly not much different from those in CD3-
CN. The observation again suggested that the Se‚‚‚F interaction
should not have much electrostatic character.
The solvent effect on JSe‚‚‚F observed for 1a was different
from that for 1e as discussed above. The discrepancy may be
due to the preferential solvation at the SeCN moiety of 1a, unlike
the CH2F moiety of 1a, since only a marginal solvent effect
was observed for 1e. This is consonant with the hydrophobicity
of organic fluorides.22
Evaluation of Se‚‚‚F Interaction Energies (HSe‚‚‚F). By
fitting the temperature-dependent JSe‚‚‚F data obtained for 1a
and 1e to eqs 1-3, stabilization energies of the nonbonded
Se‚‚‚F interactions (HSe‚‚‚F) were evaluated.
For the case of 1e (the Se-Se‚‚‚F interaction), the fitting
was straightforward since the value of JSe‚‚‚F(A) should be
constant at various temperatures as suggested by the constant
value of JSe‚‚‚F for 4e. The values of HSe‚‚‚F for 1e were estimated
as 0.85 ( 0.20 kcal/mol in CD2Cl2 and 0.83 ( 0.22 kcal/mol
in CD3CN. Thus, negligible solvent effects were quantitatively
confirmed. The interaction energies determined were signifi-
cantly smaller than those previously estimated for Se‚‚‚N
interactions (∼5-20 kcal/mol)8 due to the lower nucleophilicity
of F than N. The values of ∆S and JSe‚‚‚F(A), which were
simultaneously converged, were 4.2 ( 0.2 eu and 73 ( 21 Hz,
respectively, in CD2Cl2 and 3.9 ( 0.6 eu and 81 ( 33 Hz,
respectively, in CD3CN. The positive values of ∆S were
consistent with the larger conformational flexibility of conformer
B.
On the other hand, the data fitting was difficult for the case
of 1a because the value of JSe‚‚‚F(A) should not be constant at
low temperatures, as suggested by the decrease of JSe‚‚‚F
observed for 4a. However, it may be a reasonable assumption
in CD2Cl2 that the value of JSe‚‚‚F(A) for 1a is approximately
constant because the JSe‚‚‚F values for 4a decreased only slightly
and the JSe‚‚‚F value for 1a increased greatly with lowering
temperature. On the basis of this hypothesis, the NC-Se‚‚‚F
interaction energy (HSe‚‚‚F) of 1a was estimated as 1.23 ( 0.18
Figure 3. Plots of JSe‚‚‚F observed for 1a (O) and 4a (9) in CD2Cl2 and
CD3CN as a function of temperature. The solid curve (in CD2Cl2) for 1a
was depicted with the parameters obtained by data fitting. The other curves
were drawn for convenience with no physical meaning.
for 1a must be reduced slightly at low temperatures because
the JSe‚‚‚F value for the reference compound (4a) decreased with
lowering the temperature. Therefore, the equilibrium would
proceed toward conformer A at low temperatures. Second, the
NC-Se‚‚‚F interaction is affected by polarity of the solvent
employed. The values of JSe‚‚‚F for both 1a and 4a were smaller
in CD3CN than those in CD2Cl2. This is probably due to stronger
solvation by CD3CN than that by CD2Cl2. The observed
decrease in the JSe‚‚‚F value for 4a at low temperatures can also
be ascribed to the effects of solvation because the solvation
would slightly weaken the Se‚‚‚F interaction by reducing the
net charge of the electrophilic Se atom.
On the other hand, different behaviors were observed for 1e.
As seen in Figure 4, the value of JSe‚‚‚F for 1e significantly
increased with lowering the temperature in both CD3CN and
CD2Cl2, whereas that for 4e remained constant. The constant
value of JSe‚‚‚F for 4e suggested that the value of JSe‚‚‚F(A) for
1e is independent of temperature in both solvents. Therefore,
the observed increase of the JSe‚‚‚F value for 1e should be directly
due to the enhancement of the relative population of conformer
A (øA) at low temperatures. This implies that the nonbonded
Se‚‚‚F interaction of 1e (the Se-Se‚‚‚F interaction) may be
attractive. Furthermore, the values of JSe‚‚‚F in CD2Cl2 for 1e
(22) Organofluorine Compounds in Medicinal Chemistry and Biomedical
Applications; Filler, R., Kobayashi, Y., Yagupolskii, L. M., Eds.; Elsevi-
er: Amsterdam, 1993.
9
J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002 1905

A R T I C L E S
Iwaoka et al.
Table 2. Summary of Quantum Chemical (QC) Calculations on 5a-f and the Natural Bond Orbital (NBO) Analysis
b
c
c
c
d
e
∆EAB
(kcal/mol)
rSe‚‚‚F
(Å)
F‚‚‚Se−X
(deg)
ωAr-C
∆E
∆qSe
∆qFf
(e-)
del
compd
methoda
(deg)
(kcal/mol)
(e-)
5a
I
II
I
II
I
II
I
II
I
II
I
-1.1
2.92
2.79
2.97
2.69
3.25
2.76
3.08
2.91
3.10
2.94
3.14
2.99
160.6
168.8
155.6
170.7
133.4
168.6
159.4
171.6
158.0
171.4
157.6
169.6
54.2
49.4
56.2
47.0
67.0
51.2
60.8
54.2
62.7
55.9
62.9
58.1
3.5
0.0048
-0.0076
-0.0073
-0.0025
-0.0050
-0.0046
-0.0046
-1.2
5b
5c
5d
5e
5f
+0.1
3.0
1.0
2.1
2.0
1.6
0.0045
0.0010
0.0027
0.0024
0.0024
-0.5
+1.3
-1.1
+0.1
-0.0
+0.7
+0.2
+0.3 (+0.2)g
-0.2 (+0.0)g
II
a QC calculations were carried out at the B3LYP/6-31G(d,p)//HF/6-31G(d,p) level (method I) and the B3LYP/6-31G(d,p)//B3LYP/6-31G(d,p) level (method
II) by use of Huzinaga’s 43321/4321/311 basis sets for Se and Br. NBO analysis was performed at the HF/6-31G(d,p) level. b Energy of conformer A
relative to that of conformer B, corrected with the unscaled zero-point energies. c Selected structural parameters obtained for conformer A. ωAr-C is the
dihedral angle along the Ar-CH2F bond. d Total stabilization energy due to the orbital interactions between SeX and F for conformer A revealed by the
NBO deletion analysis. e Change of the electron population for the σ*Se-X orbital due to the Se‚‚‚F orbital interaction of conformer A. f Change of the
electron population for the nF orbitals due to the Se‚‚‚F orbital interaction of conformer A. g Energy of conformer A relative to that of conformer C.
kcal/mol in CD2Cl2. The simultaneously converged values of
∆S and JSe‚‚‚F(A) were 2.9 ( 0.4 eu and 120 ( 5 Hz,
respectively. The estimated Se‚‚‚F interaction energy for 1a
(1.23 kcal/mol) was 0.38 kcal/mol larger than that for 1e (0.85
kcal/mol) in the same solvent, being consistent with the higher
electrophilicity of selenocyanates than diselenides. For the data
obtained in CD3CN, such a fitting could not be applied because
of larger solvent effects: the Se‚‚‚F interaction should be weaker
than 1.23 kcal/mol because solvation around the SeCN moiety
may attenuate the Se‚‚‚F atomic contact.
QC Calculations and the NBO Analysis. To rationalize the
experimental results, QC calculations were carried out on
5a-f, two of which have a simplified substituent at Se to save
computational time. Gaussian 94 and 98 programs23,24 were used
for all calculations. Structures were fully optimized at the HF/
6-31G(d,p) or B3LYP/6-31G(d,p) level by use of Huzinaga’s
43321/4321/311 basis sets25 for Se and Br. Single-point energies
were subsequently calculated at the level of the density
functional theory (B3LYP).26 To analyze the mechanism of the
observed nonbonded Se‚‚‚F interactions, NBO analysis of
Weinhold and co-workers27 was performed.
two stable conformers in the following discussion. For the case
of 5f (X ) Me), the most stable conformer was found to be
conformer C at the B3LYP/6-31G(d,p)//B3LYP/6-31G(d,p)
level. Hence, three conformers are considered for 5f. The results
of QC calculations and the NBO analysis are summarized in
Table 2.
For all compounds, energies of conformers A and B were
close to each other (∆EAB ) -1.2 to +1.3 kcal/mol), suggesting
that both conformers may exist in solution. Thus, the fitting
analysis of the NMR data based on a rapid equilibrium model
(Figure 1) could be rationalized. The values of ∆EAB for 5a
(∆EAB ) -1.1 and -1.2 kcal/mol in methods I and II,
respectively) were in excellent agreement with the Se‚‚‚F
interaction energy experimentally estimated for 1a in CD2Cl2
(HSe‚‚‚F ) 1.23 ( 0.18 kcal/mol). Therefore, it was strongly
suggested that solvent effect on the Se‚‚‚F interaction, and hence
the electrostatic character, is insignificant. On the other hand,
the values of ∆EAB for 5e (∆EAB ) +0.7 and +0.2 kcal/mol in
methods I and II, respectively) were a little different from those
estimated for 1e (HSe‚‚‚F ) 0.85 ( 0.20 and 0.83 ( 0.22 kcal/
mol in CD2Cl2 and CD3CN, respectively). This is probably due
to the simplified molecular structure of 5e compared with 1e.
The effect of electron correlation on the Se‚‚‚F interactions
is obviously seen in Table 2. Comparing the Se‚‚‚F nonbonded
distance (rSe‚‚‚F) of conformer A optimized at the HF/6-31G-
(d,p) level (method I) with those optimized at the B3LYP/6-
31G(d,p) level (method II), it was revealed that more accurate
treatment of electron correlation causes strengthening of the
Se‚‚‚F interaction for 5a-f. The most remarkable change was
Three to six possible conformers were located for 5a-f: two
corresponded to conformers A and B (Figure 1), and the others
included an unstable conformer with both the C-F and Se-X
bonds tilted out of the aromatic plane to the same direction
(conformer C), one with both the C-F and Se-X bonds lying
on the aromatic plane, etc. Except for 5f, the energies of
conformers A and B were found to be obviously lower than
those of the other conformers. Therefore, we consider only the
(24) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.; 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.; Baboul, A. G.; 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.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian
98; Gaussian, Inc.: Pittsburgh, PA, 1998.
(23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B.
G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94; Gaussian, Inc.:
Pittsburgh, PA, 1995.
(25) Gaussian Basis Sets for Molecular Calculations; Huzinaga, S., Ed.;
Elsevier: Amsterdam, 1984.
(26) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785-789. (b)
Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(27) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899-926.
9
1906 J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002

Weak Nonbonded Se‚‚‚F Interactions
A R T I C L E S
Figure 5. Molecular structure of conformer C obtained for 5f (X ) Me)
at the B3LYP/6-31G(d,p) level by use of Huzinaga’s 43321/4321/311 basis
sets for Se.
Figure 6. Contour map of LUMO+3 for conformer A of 5a (X ) CN) at
the HF/6-31G(d,p) level by use of Huzinaga’s 43321/4321/311 basis sets
for Se. The map was drawn at the plane on the Se, F, and N atoms with a
line interval of 0.05. Thin solid lines mean intersection of the nodal surface.
observed for 5c (X ) Br): The value of rSe‚‚‚F for 5c was 0.49
Å smaller in method II than in method I. For the other
compounds, the decrease in rSe‚‚‚F was ∼0.13-0.28 Å. Although
the effect of electron correlation is significant, it is notable that
the values of rSe‚‚‚F obtained at the Hartree-Fock level (∼2.92-
3.25 Å in method I) were already smaller than the sum of the
van der Waals radii (3.37 Å).28 Therefore, electron correlation
may not be an essential factor for formation of the nonbonded
Se‚‚‚F interactions.
For the case of 5f (X ) Me), the relative energies of
conformer A to conformers B and C were -0.2 and +0.0 kcal/
mol, respectively, in method II, suggesting that the three
conformers exist in solution for 1f. The most stable was found
to be conformer C. This unique character of 5f () 1f) may be
due to dual weak intramolecular interactions, i.e., C-H‚‚‚Se29
and C-H‚‚‚F hydrogen bonds, as shown in Figure 5. The
obtained H‚‚‚Se and H‚‚‚F nonbonded distances were 2.89 and
2.24 Å, respectively, which were significantly shorter than the
corresponding sum of the van der Waal radii (3.10 and 2.67 Å,
respectively).28
interactions between the SeX moiety and the F atom (∆Edel in
Table 2) showed a significant magnitude of Se‚‚‚F orbital
interactions (1.0∼3.5 kcal/mol). The values of ∆qSe and ∆qF,
which indicate electron flow in the formation of the nonbonded
Se‚‚‚F interaction, suggested that the F atom slightly donates
electrons to the SeX moiety. Thus, the character of the nF f
σ*Se-X orbital interaction was supported theoretically.30
Mechanisms of Se‚‚‚F Interaction. Stabilization mechanisms
for weak nonbonded interactions generally involve several
physical factors, such as electrostatic, orbital (charge-transfer),
and dispersion interactions. Their relative contributions depend
on the interaction systems. For the case of the Se‚‚‚F interaction
observed for 1e (the Se-Se‚‚‚F interaction), the electrostatic
character is presumably not important as revealed by the
negligible solvent effects, while the orbital character may be
significant as suggested by the large JSe‚‚‚F coupling constant
estimated for conformer A of 1e [JSe‚‚‚F(A) ) 73 Hz in CD2Cl2
and 81 Hz in CD3CN] and by the QC and NBO analyses on 5e.
On the other hand, an electrostatic character may contribute
to some extent for the Se‚‚‚F interaction of 1a (the NC-Se‚‚‚F
interaction) as suggested by the solvent effect on the JSe‚‚‚F
coupling constant. However, the contribution from orbital
interactions should be more important, based on the large value
of JSe‚‚‚F estimated for conformer A of 1a [JSe‚‚‚F(A) ) 120 Hz
in CD2Cl2] and the results of the QC and NBO analyses of 5a
(Table 2 and Figure 6). Since the QC calculation result (i.e.,
∆EAB) in vacuo almost quantitatively reproduced the fitting
result (i.e., HSe‚‚‚F) that was based on experimental observations
in solution, the solvent effect may be rather small. Therefore,
the orbital character would be predominant for stabilization of
the NC-Se‚‚‚F interaction.
Structural parameters of conformer A (rSe‚‚‚F
and ωAr-C) obtained for 5a-f indicated the distinct tendency
that the shorter the nonbonded Se‚‚‚F atomic distance (rSe‚‚‚F
,
F‚‚‚Se-X,
)
is, the more linear the nonbonded F‚‚‚Se-X angle ( F‚‚‚Se-
X) is and the smaller the dihedral angle between the aromatic
plane and the C-F bond (ωAr-C) is. The linear propensity of
the Se‚‚‚F interactions suggested a major contribution from the
orbital interaction between the antibonding orbital of the Se-X
bond (σ*Se-X) and the lone pair of F (nF) on the stability. Indeed,
the mixing of σ*Se-X and nF was apparent for the case of 5a as
graphically shown in Figure 6. Such n f σ*Se-X orbital
interactions have also been suggested for the corresponding
Se‚‚‚N8 and Se‚‚‚O11 interactions as the major stabilization
factor.
To evaluate the magnitude of the nF f σ*Se-X orbital
interaction, NBO deletion analysis,27 which should allow
quantitative estimation of the total energy due to the orbital
interactions between two specific atoms or groups in a molecule
by deleting the Fock matrix elements, was applied to conformers
A of 5a-f. The obtained stabilization energies due to the orbital
Table 2 shows that the relative electronic energy (∆EAB) and
Se‚‚‚F atomic distance (rSe‚‚‚F) of conformer A significantly
change depending on the calculation methods (method I vs
method II). Since the B3LYP method incorporates the effect
of electron correlation more efficiently, the results suggested
that the contribution from dispersion forces may also be involved
for the attractive nature of nonbonded Se‚‚‚F interactions.
Conclusions
(28) Bondi, A. J. Phys. Chem. 1964, 68, 441-451.
By means of data fitting of the nuclear spin coupling constant
between Se and F (JSe‚‚‚F) observed for 1a and 1e at various
(29) (a) Iwaoka, M.; Tomoda, S. J. Am. Chem. Soc. 1994, 116, 4463-4464.
(b) Iwaoka, M.; Komatsu, H.; Tomoda, S. Bull. Chem. Soc. Jpn. 1996, 69,
1825-1826. (c) Narayanan, S. J.; Sridevi, B.; Chandrashekar, T. K.; Vij,
A.; Roy, R. Angew. Chem., Int. Ed. Engl. 1998, 37, 3394-3397. (d)
Michalczyk, R.; Schmidt, J. G.; Moody, E.; Li, Z.; Wu, R.; Dunlap, R. B.;
Odom, J. D.; Silks, L. A. Angew. Chem., Int. Ed. Engl. 2000, 39, 3067-
3070.
(30) According to ref 16, the value of ∆Edel was 0.0 kcal/mol for 5f. The
discrepancy between the results of ref 16 and this work (∆Edel ) 1.6 kcal/
mol for 5f) may be ascribed to the difference in the basis sets employed.
9
J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002 1907

A R T I C L E S
Iwaoka et al.
and sulfuryl chloride (4.8 µL, 0.06 mmol) was added to the solution.
After having stirred for 1.5 h at 40 °C, the reaction mixture was
concentrated under reduced pressure. The resulting crude product
(brown oil) contained 1b (79%), 1e (16%), and an unknown byproduct
(5%) according to the integration of H NMR. Further purification of
1b could not be performed because of the high reactivity of 1b. Spectral
data for 1b: 1H NMR δ 7.90 (d, J ) 7.9 Hz, 1H), 7.51 (d, J ) 7.9 Hz,
temperatures to the rapid equilibrium model (Figure 1), stabi-
lization energies of the nonbonded Se‚‚‚F interactions (HSe‚‚‚F
)
were estimated as 1.23 kcal/mol for 1a (in CD2Cl2) and 0.85
and 0.83 kcal/mol for 1e (in CD2Cl2 and CD3CN, respectively).
The stabilization energies were significantly smaller than the
corresponding Se‚‚‚N8 and Se‚‚‚O11 interactions, being consis-
tent with the lower electron-donating ability of F lone pairs.
The negligible solvent effect observed for 1e suggested little
electrostatic nature of the Se‚‚‚F interaction despite the high
electronegativity of F. On the other hand, a small solvent effect
was observed for 1a, suggesting the Se‚‚‚F interaction has a
slight electrostatic character. The solvation site, however, may
be the SeCN moiety rather than the CH2F moiety.
As for the major mechanism of the nonbonded Se‚‚‚F
interactions, the nF f σ*Se-X orbital interaction was strongly
suggested on the basis of the observed large JSe‚‚‚F coupling
constants as well as the results obtained from QC calculations
and the NBO analysis. The stabilization mechanism is essentially
the same as the previously investigated Se‚‚‚N8 and Se‚‚‚O11
interactions, although the nucleophilicity of F is much lower
than those of N and O. Therefore, the strong electrophilicity of
a divalent organic selenium as well as the weak, but distinct,
electron-donating character of the lone pairs of organic fluo-
rides15 can be concluded.
1
1H), 7.45 (t, J ) 7.9 Hz, 1H), 7.42 (t, J ) 7.9 Hz, 1H), 5.65 (d, JHF
)
49.2 Hz, 2H); 13C NMR δ 138.0 (d, JCF ) 17.7 Hz), 134.8, 131.5 (d,
JCF ) 2.6 Hz), 130.6, 129.7 (d, JCF ) 0.8 Hz), 128.0 (d, JCF ) 8.6 Hz),
84.3 (d, JCF ) 167.6 Hz); 19F NMR δ -205.2 (t, JHF ) 49.2 Hz); 77Se
NMR δ 978.5 (d, JSeF ) 80.1 Hz).
2-(Fluoromethyl)phenylselenenyl Cyanate (1a). Compound 1b,
prepared from 1e (193 mg, 0.51 mmol), was dissolved in dry THF (20
mL) under nitrogen. Cyanotrimethylsilane (175 µL, 1.32 mmol) was
added to the solution. After 1 h, the reaction mixture was concentrated
under reduced pressure. Product 1a was obtained in 80% yield (176
mg) as a colorless oil from the residue by silica-gel column chroma-
tography (hexane-CH2Cl2) followed by gel-permeation chromatogra-
phy. Spectral data for 1a: 1H NMR δ 7.86 (d, J ) 7.0 Hz, 1H), 7.47
(d, J ) 4.2 Hz, 1H), 7.45-7.41 (m, 2H), 5.52 (d, JHF ) 46.9 Hz, 2H);
13C NMR δ 137.0 (d, JCF ) 17.9 Hz), 134.0, 130.9 (d, JCF ) 1.9 Hz),
130.0, 129.6 (d, JCF ) 8.3 Hz), 123.1, 101.4 (d, JCF ) 6.4 Hz), 84.1
(d, JCF ) 168.6 Hz); 19F NMR δ -204.3 (t, JHF ) 46.9 Hz); 77Se NMR
δ 288.6 (d, JSeF ) 84.2 Hz). Anal. Calcd for C8H6FNSe: C, 44.88; H,
2.82; N, 6.54. Found: C, 44.59; H, 2.89; N, 6.52.
2-(Fluoromethyl)benzeneselenenyl Bromide (1c). Compound 1e
(19.0 mg, 0.05 mmol) was dissolved in CH2Cl2 (2 mL) under nitrogen,
and bromine (2.59 µL, 0.05 mmol) was added to the solution. After 15
min, the reaction mixture was evaporated. Product 1c was almost
quantitatively obtained as a dark brown oil. Because of the labile nature
of 1c, purification was not performed. Spectral data for 1c: 1H NMR
δ 7.93 (d, J ) 7.6 Hz, 1H), 7.52 (d, J ) 7.6 Hz, 1H), 7.47 (t, J ) 7.6
Hz, 1H), 7.37 (t, J ) 7.6 Hz, 1H), 5.68 (d, JHF ) 47.4 Hz, 2H); 13C
NMR δ 139.3 (d, JCF ) 16.9 Hz), 137.0, 131.1, 129.6 (d, JCF ) 3.0
Experimental Section
General Procedures. Commercially available organic and inorganic
reagents were used without further purification. Tetrahydrofuran (THF)
was dried over sodium wire and was distilled under nitrogen. Dichlo-
romethane (CH2Cl2) was dried over calcium hydride and was distilled
under nitrogen before use. Methanol (MeOH) was distilled under
nitrogen. Other organic solvents were used without purification. Gel-
permeation chromatography was carried out by using a JAI LC-908
1
Hz), 128.5 (d, JCF ) 4.3 Hz), 128.0 (d, JCF ) 9.5 Hz), 84.2 (d, JCF )
system equipped with JAIGEL-H columns (chloroform as eluent). H
168.6 Hz); 19F NMR δ -208.0 (t, JHF ) 47.4 Hz); 77Se NMR δ 801.0
(d, JSeF ) 43.1 Hz).
(500 MHz), 13C (125.65 MHz), 19F (470.40 MHz), and 77Se (95.35
MHz) NMR spectra were measured on a Jeol R500 spectrometer in
CDCl3 containing tetramethylsilane as an internal standard for 1H and
13C NMR. For 19F NMR, fluorobenzene (δ -113.0 ppm) was added as
an internal standard. For 77Se NMR, dimethyl selenide (δ 0 ppm) in
CDCl3 was used as an external standard. JSe‚‚‚F coupling constants of 1
and 4 were measured from the expanded 1H-decoupled 77Se NMR
spectra in appropriate solvents at various temperatures. The temperatures
were not corrected.
2-(Fluoromethyl)benzeneselenenyl Phenyl Sulfide (1d). Compound
1e (19.2 mg, 0.05 mmol) was dissolved in CH2Cl2 (2 mL). Pyridine (2
drops) and benzenethiol (26.1 µL, 0.25 mmol) were added to the
solution. After 2.5 h, the reaction mixture was concentrated under
reduced pressure to afford a yellow oil, which contained 1d, 1e, and
1
PhSSPh. The yield of 1d was 30% according to the integration of H
NMR. Due to the disproportionation of 1d, purification could not be
performed. Spectral data for 1d: 1H NMR δ 7.78 (d, J ) 7.3 Hz, 1H),
7.51-7.21 (m, 8H), 5.48 (d, JHF ) 49.1 Hz, 2H); 13C NMR δ 136.7
(d, JCF ) 16.4 Hz), 136.3, 132.5, 130.3, 129.7 (d, JCF ) 1.3 Hz), 129.0,
128.5 (d, JCF ) 7.5 Hz), 128.4, 127.7, 127.5, 83.9 (d, JCF ) 167.6 Hz);
Bis[2-(fluoromethyl)phenyl] Diselenide (1e). According to the
literature method,18 the well-ground 2:3 (w/w) mixture of potassium
fluoride and calcium fluoride (from an artificial source) (5.20 g, 35.8
mmol of KF) was activated for 5 h at 150 °C under reduced pressure
(0.1 mmHg). The mixture was added to the solution of bis[2-
(chloromethyl)phenyl] diselenide17 (1.51 g, 3.67 mmol) in butyronitrile
(10 mL) under nitrogen. After refluxing for 85 h, the reaction mixture
was filtered. The filtration residue was washed with ether, and the ether
layer was combined with the filtrate. After evaporation, the resulting
crude product was purified by silica-gel column chromatography
(hexane-CH2Cl2). Product 1e was obtained in 47% yield (0.65 g) as
yellow crystals. Spectral data for 1e: 1H NMR δ 7.65 (d, J ) 7.6 Hz,
2H), 7.41 (d, J ) 7.6 Hz, 2H), 7.35 (t, J ) 7.6 Hz, 2H), 7.25 (t, J )
19F NMR δ -206.3 (t, JHF ) 49.1 Hz); 77Se NMR δ 499.7 (d, JSeF
)
38.7 Hz).
2-(Fluoromethyl)phenyl Methyl Selenide (1f). Compound 1e (193
mg, 0.51 mmol) and methyl iodide (0.2 mL, 3.2 mmol) were dissolved
in MeOH (2 mL) under nitrogen, and sodium borohydride was added
to the solution until the yellow color disappeared. After 2 h, the reaction
mixture was poured into aqueous NaHCO3 (10 mL) and was extracted
with ether. Product 1f was obtained as a colorless oil (162 mg, 78%)
after purification by gel-permeation chromatography. Spectral data for
1f: 1H NMR δ 7.48 (d, J ) 7.0 Hz, 1H), 7.39 (d, J ) 7.0 Hz, 1H),
7.30-7.23 (m, 2H), 5.49 (d, JHF ) 47.7 Hz, 2H), 2.32 (s, 3H); 13C
NMR δ 136.8 (d, JCF ) 17.6 Hz), 131.9 (d, JCF ) 3.8 Hz), 131.2,
129.4 (d, JCF ) 2.5 Hz), 128.4 (d, JCF ) 7.5 Hz), 126.6, 84.1 (d, JCF
) 167.6 Hz), 7.7; 19F NMR δ -208.1 (t, JHF ) 47.7 Hz); 77Se NMR
δ 161.1 (d, JSeF ) 22.7 Hz). Anal. Calcd for C8H9FSe: C, 47.31; H,
4.47. Found: C, 47.01; H, 4.41.
7.6 Hz, 2H), 5.42 (d, JHF ) 47.9 Hz, 4H); 13C NMR δ 138.0 (d, JCF
)
16.2 Hz), 135.3, 130.0 (d, JCF ) 4.1 Hz), 129.5 (d, JCF ) 1.7 Hz),
129.1, 128.1 (d, JCF ) 8.8 Hz), 84.0 (d, JCF ) 169.7 Hz); 19F NMR δ
-208.1 (t, JHF ) 47.9 Hz); 77Se NMR δ 437.3 (d, JSeF ) 23.6 Hz).
Anal. Calcd for C14H12F2Se2: C, 44.70; H, 3.22. Found: C, 44.64; H,
3.37.
2-(Fluoromethyl)benzeneselenenyl Chloride (1b). Compound 1e
(19.1 mg, 0.05 mmol) was dissolved in dry THF (2 mL) under nitrogen,
Sodium [2-(Fluoromethyl)phenyl]seleno(trimethoxy)borate (1g).
9
1908 J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002

Weak Nonbonded Se‚‚‚F Interactions
A R T I C L E S
Compound 1g was generated in an NMR sample tube by the reaction
of 1e with excess amounts of sodium borohydride in CD3OD.19
Although 1H and 13C NMR spectra of the reaction mixture were
significantly complicated, the formation of 1g was characterized by
19F and 77Se NMR spectra. 6H,12H-Diseleno[b,f][1,5]diselenocin29a,b
was detected as a major byproduct during the prolonged measurement
of the NMR spectra. Spectral data for 1g: 19F NMR δ -217.6 (t, JHF
) 49.3 Hz); 77Se NMR δ -24.1 (s).
2-(Trifluoromethyl)phenylselenenyl Cyanate (4a). 2-(Trifluorom-
ethyl)aniline (3.32 g, 20.6 mmol) was suspended in 6 M HCl (40 mL)
at 0 °C, and 3 M NaNO2 (8 mL, 24 mmol) was added to the mixture
dropwise. After 1 h, a saturated aqueous solution of sodium acetate
was added dropwise until the pH of the reaction solution reached about
6. Then the mixture was poured into aqueous potassium selenocyanate
(3.70 g, 20.5 mmol). The resulting mixture was extracted with ether.
After purification by silica-gel column chromatography (hexane-CH2-
Cl2), 4a was obtained as a brown oil (4.40 g, 85%). Spectral data for
4a: 1H NMR δ 7.97 (d, J ) 7.7 Hz, 1H), 7.74 (d, J ) 7.7 Hz, 1H),
7.59 (t, J ) 7.7 Hz, 1H), 7.52 (t, J ) 7.7 Hz, 1H); 13C NMR δ 133.9,
133.6, 129.8 (q, JCF ) 31.6 Hz), 129.2, 127.6 (q, JCF ) 5.4 Hz), 123.3
(q, JCF ) 274.2 Hz), 121.0, 100.7 (q, JCF ) 4.7 Hz); 19F NMR δ -60.6
(s); 77Se NMR δ 341.7 (q, JSeF ) 63.2 Hz).
Bis[2-(trifluoromethyl)phenyl] Diselenide (4e). Compound 4a (1.05
g, 4.2 mmol) and NaOH (7.12 g, 0.18 mol) were dissolved in methanol
(50 mL), and the mixture was stirred at room temperature for 24 h.
The reaction mixture was added to water and was extracted with CH2-
Cl2. Product 4e was obtained as yellow crystals (0.64 g, 68%) after
purification by silica-gel column chromatography (hexane). Compound
4e was known in the literature.31 Spectral data for 4e: 1H NMR δ 7.88
(d, J ) 7.7 Hz, 2H), 7.60 (d, J ) 7.7 Hz, 2H), 7.39 (t, J ) 7.7 Hz,
2H), 7.31 (t, J ) 7.7 Hz, 2H); 13C NMR δ 133.6, 132.7, 129.5 (q, JCF
data for 4f: 1H NMR δ 7.59 (d, J ) 7.8 Hz, 1H), 7.47 (d, J ) 7.8 Hz,
1H), 7.39 (t, J ) 7.8 Hz, 1H), 7.24 (t, J ) 7.8 Hz, 1H), 2.34 (s, 3H);
13C NMR δ 132.0, 131.5, 131.0, 130.1 (q, JCF ) 30.4 Hz), 126.7 (q,
JCF ) 5.2 Hz), 125.7, 124.1 (q, JCF ) 272.0 Hz), 7.5; 19F NMR δ
-61.7 (s); 77Se NMR δ 217.7 (q, JSeF ) 48.4 Hz).
Sodium [2-(Trifluoromethyl)phenyl]seleno(trimethoxy)borate
(4g). Compound 4g was generated in an NMR sample tube by the
reaction of 4e with excess amounts of sodium borohydride in CD3-
OD.19 Formation of 4g was unambiguously characterized by NMR
spectra. Spectral data for 4g: 1H NMR δ 7.82 (t, J ) 8.3 Hz, 1H),
7.28 (t, J ) 8.3 Hz, 1H), 6.88 (t, J ) 8.3 Hz, 2H); 13C NMR δ 142.1,
140.6, 134.2 (q, JCF ) 26.4 Hz), 130.3, 126.5 (q, JCF ) 5.3 Hz), 126.5
(q, JCF ) 273.5 Hz), 122.2; 19F NMR δ -63.6 (s); 77Se NMR δ 69.5
(q, JSeF ) 15.6 Hz).
Data Fitting. The values of JSe‚‚‚F for 1a and 1e were measured in
1
CD2Cl2 and CD3CN at various temperatures from the expanded H-
decoupled 95.35 MHz 77Se NMR spectra with a digital resolution of
0.15 Hz. The data obtained were fitted to eqs 1-3 by using
KaleidaGraph version 3.08 (Synergy Software, Reading, PA), in which
the Levenberg-Marquardt algorithm32 was employed for the nonlinear
least-squares method.
QC Calculations and the NBO Analysis. All computational
calculations were carried out by using Gaussian 94 and 98 programs23,24
installed on DEC Alpha21164A Unix workstations. Stable conformers
of 5a-f were first sought by changing the two dihedral angles along
the Ar-SeX and Ar-CH2F bonds independently with an interval of
30° at the HF/3-21G(d) level. The obtained conformers (conformers
A, B, C, etc.) were fully optimized at the HF/6-31G(d,p) (method I)
and B3LYP/6-31G(d,p) (method II) levels by use of Huzinaga’s 43321/
4321/311 basis sets25 without any restrictions. The single-point energies
were then calculated by applying the density functional theory at the
B3LYP/6-31G(d,p) level.26 Zero-point energies (ZPE) calculated at the
appropriate level were not scaled. NBO deletion analysis27 was
performed with standard parameters at the HF/6-31G(d,p) level. The
contour map of LUMO+3 for 5a (Figure 6) was drawn by using a
homemade program.
) 30.3 Hz), 128.5, 127.5, 126.6 (q, JCF ) 7.6 Hz), 123.9 (q, JCF
)
274.0 Hz); 19F NMR δ -60.1 (s); 77Se NMR δ 454.0 (q, JSeF ) 53.0
Hz).
2-(Trifluoromethyl)phenyl Methyl Selenide (4f). Compound 4a
(527 mg, 2.1 mmol) and methyl iodide (0.33 mL, 5.3 mmol) were
dissolved in MeOH (6 mL) under nitrogen, and sodium borohydride
was added to the solution until the yellow color disappeared. After 2
h, the reaction mixture was added to aqueous NaHCO3 and was
extracted with ether. Product 4f was obtained as a colorless oil (417
mg, 83%) after purification by gel-permeation chromatography. Spectral
Acknowledgment. This work was supported by Grants in
Aid for Scientific Research 11120210 and 10133207 from the
Ministry of Education, Science and Culture, Japan.
JA016348R
(31) McKillop, A.; Koyunc¸u, D.; Krief, A.; Dumont, W.; Renier, P.; Trabelsi,
M. Tetrahedron Lett. 1990, 31, 5007-5010.
(32) Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431-441.
9
J. AM. CHEM. SOC. VOL. 124, NO. 9, 2002 1909