2584 J . Org. Chem., Vol. 63, No. 8, 1998
Genov et al.
LAOCOON IV program.22 The parameters were accurate to
0.1 Hz or better. The conformational preferences were evalu-
ated by means of NMR spectroscopy relating the observed
vicinal proton-proton coupling constants to the couplings in
the individual conformers:
further minimized, and used in the next run as the fixed
species. For each compound, multiple dockings with either
five molecules of CHCl3 or C5H5N or (CH3)2CO or CH3OH to
each of conformers ga, ag, or gg.
The ab initio calculations were carried out using SPARTAN
SGI Version 4.1.227 on a SGI O2 180 MHz R5000SC 64Mb RAM
IRIX 6.3 computer. A list of 27 conformers was generated with
the SYBYL MM force field28 by rotation about the C1-C2,
P-C1, and C2-O2 bonds. All conformers were further
optimized using the PM3 method.29 Then the most stable
conformers were optimized at the HF/3-21G(*) level.16 Single-
point calculations were performed with the HF method using
a 6-31G* basis set.16
J obs
)
niJ i
∑
i)1,3
n1 + n2 + n3 ) 1
The proton-proton coupling constants in the individual
conformers were calculated using the Haasnoot-Altona equa-
tion14 and the group electronegativities reported by Inamoto
et al.23 Full details about the spectral analysis and the
assignment of the diastereotopic R-methylene protons HA and
HB are given in ref 2c.
IR spectra of 2a -e in solid state (KBr disks) and in DCCl3
and tetrachloromethane solutions were recorded on a Nicolet
Impact 400 spectrometer. The solution concentrations are
given in mol % in order to have the same molar ratio between
solvent and solute in all cases. For the solutions, sodium
chloride cells with 0.5 mm path length and quartz cells with
5 cm path length were used. The solvents were obtained
commercially and used without further purification.
X-r a y Cr ysta l Str u ctu r e Deter m in a tion s. P r oced u r e.
Data were recorded on a FAST TV detector diffractometer (Mo
KR radiation) using previously described methods30 and cor-
rected for Lorentz and polarization effects. Systematically
absent data indicated the space group in each case. Most
non-H atom positions were estimated using direct methods,31
and all were located by means of sequential difference Fourier
syntheses.32 These atoms were refined32 anisotropically using
merged data. Hydrogen atoms were introduced in theoretical
positions (OH, 0.82; CarylH, 0.93; CmethineH, 0.98; CmethyleneH,
0.97; CmethylH, 0.96 Å) and assigned common, refined, isotropic
displacement parameters according to three types, namely
aromatic, methyl, and aliphatic nonmethyl. Finally, an ab-
sorption correction33 based on these converged models was
applied to the data and degrees of freedom restricted ap-
propriately for the final refinement32 cycles. The weighting
scheme was w-1 ) σ2(Fo)2 + xP2, where 3P ) (Fo)2 + 2(Fc)2
(Table 1). A listing of specific details is given in Table 1.
Syn th esis. The synthesis of phosphonate 2a is described
elsewhere.2c
Gen er a l P r oced u r e for P r ep a r a tion of (2-Hyd r oxy-
a lk yl)d ip h en ylp h osp h in e Oxid es.5 To a stirred solution of
diphenylmethylphosphine oxide in dry THF was added a
solution of 1.6 M n-butyllithium at 0 °C under N2. After 30
min, the reaction solution was cooled to -78 °C (acetone-solid
CO2), and a solution of appropriate aldehyde in THF was added
dropwise at such a rate that the solution temperature was
maintained at -78 °C. The reaction mixture was allowed to
warm to room temperature over 2 h, and water was added.
THF was removed under reduced pressure, and brine was
added to the aqueous residue before extraction with dichlo-
romethane. The extract was dried (MgSO4), and dichlo-
romethane was evaporated to give the crude product. In this
way the following compounds were prepared.
(2-Hyd r oxy-2-p h en yleth yl)d ip h en ylp h osp h in e Oxid e
(2b). Diphenylmethylphosphine oxide (4 g, 18.6 mmol), n-
BuLi (1.6 M in hexane, 12.5 mL, 18.6 mmol), and benzaldehyde
(2.1 g, 20.4 mmol) were reacted as above to produce a viscous
oil, which after washing with hexane crystallized. It was
purified by recrystallization from hexane/acetone to give white
crystals (4.8 g, 80%): mp 143-145 °C; IR (KBr)/cm-1 3197
(OH), 1179 (PdO). Anal. Calcd for C20H19O2P: C, 74.52; H,
5.94. Found: C, 74.78; H, 6.00%.
(2-Hyd r oxy-3,3-d im eth ylbu tyl)d ip h en ylp h osp h in e Ox-
id e (2c). Diphenylmethylphosphine oxide (2 g, 9.3 mmol),
n-BuLi (1.6 M in hexane, 6.3 mL, 9.6 mmol), and trimethyl-
acetaldehyde (0.88 g, 10.2 mmol) reacted as above giving a
crystalline product, which was purified by recrystallization
Determination of molecular weight for phosphonate 2a was
carried out as previously described.11
Com p u ter Sim u la tion s. MM modeling was carried out
using the COSMIC-90 program.24 Partial charges were cal-
culated by the LIVERPOOL-2 method.25
The calculations were carried out using bond lengths (PdO,
P-C, P-O) and bond angles (OPC, CPC, OPO, PCC, etc.)
obtained from our X-ray crystallographic studies. The reason
for this was that the PdO bond length included in the original
COSMIC force field is 1.428 Å, whereas it has been reported
that the PdO length is usually in the range of 1.475-1.490 Å
and it is almost insensitive to molecular environment, hydro-
gen bonding, or the nature of any substituents.26 Similarly,
bond angles involving phosphorus are not included in the force
field, and a default value of 109.5° is normally used; X-ray
data for phosphine oxides have shown that there is a consistent
deviation of the bond angles from the tetrahedral values with
CPO higher (112-114°) and CPC lower (104-107°).26
Each structure was then optimized by well-established
methods such as CONJ UGATE GRADIENT and TORMIN.
The different conformers, produced upon rotation of all rotat-
able bonds in 5° steps, were further minimized and stored.
In the molecular-docking program the stationary molecule,
in this case one of the conformers ga, ag, and gg, is placed at
the center of 1 nm diameter sphere, and the mobile molecule
(the single solvent molecule) is placed in turn at points on the
sphere, in all a total of 204 of such starting points. At each
starting point the mobile molecule is rotated about its axis in
60° steps about all three Cartesian axes, and the lowest energy
obtained is used as the starting orientation for a subsequent
energy-minimization procedure involving 500 iterations or
until the change in energy between succesive iterations is
<0.0001 kcal mol-1. In this procedure the bimolecular species
of lowest energy from the first docking process was stored,
(21) A good overview of phosphoryl infrared spectroscopy can be
found in Thomas, L. C. Interpretation of the Infrared Spectra of
Organophosphorus Compounds; Heyden: New York, 1994.
(22) Castellano, S. M.; A. A. Bothner-By, J . Chem. Phys. 1967, 47,
5443.
(23) Inamoto, N.; Masuda, S. Chem. Lett. 1982, 1003-1006.
(24) (a) Vinter, J . G.; Davies, A.; Saunders: M. R J . Comput.-Aided
Mol. Des. 1987, 1, 31-51. (b) Moley, S. D.; Abraham, R. J .; Hawarth,
I. S., J ackson, D. E.; Saunders: M. R.; Vinter, J . G. J . Comput.-Aided
Mol. Des. 1991, 5, 475-504.
(27) SPARTAN Version 4.1.2. SGI IRIX 6.3.; Wavefunction, Inc.,
18401 Von Karman Ave., no. 370, Irvine, CA 92715.
(28) Clark, M.; Cramer III, R. D.; Opdensch, N. J . Comput. Chem.
1989, 10, 982-1012.
(29) Stewart, J . J . P. J . Comput. Chem. 1989, 10, 209-220.
(30) Darr, J . A.; Drake, S. R.; Hursthouse, M. B.; Malik, K. M. A.
Inorg. Chem., 1993, 32, 5704-5708.
(25) (a) Abraham, R. J .; Griffiths, L.; Loftus, P. J . J . Comput. Chem.
1982, 3, 407-416. (b) Abraham, R. J .; Hudson, B. D. J . Comput. Chem.
1984, 5, 562-570. (c) Abraham, R. J .; Hudson, B. D. J . Comput. Chem.
1985, 6, 173-181.
(26) Gilheany, D. In The Chemistry of Organophosphorus Com-
pounds; Hartley, F., Ed., Wiley & Sons Ltd.: New York, 1992; Vol. 2,
Chapter 1.
(31) Sheldrick, G. M. SHELXS-86, Program for the Solution of
Crystal Structures, University of Goettingen, 1987.
(32) Sheldrick, G. M. SHELXL-93, Program for the Refinement of
Crystal Structures, University of Goettingen, 1993.
(33) Walker, N.; Stewart, D. Acta Crystallogr. 1983, A39, 158-166.
(34) Davies, E. K. SNOOPI, Chemical Crystallography Laboratory,
University of Oxford, 1982.