Structure and spectroscopy of dihaloformaldoximes: He I photoelectron, photoionization mass spectroscopy, mid-IR, Raman and ab initio study

Article abstract of DOI:10.1039/a605743h

The electronic and geometric structures of dichloroformaldoxime, Cl₂C=NOH, dibromoformaldoxime, Br₂C-NOH and bromochloroformaldoxime, BrClC=NOH, are investigated for the first time in the gas phase by He I photoelectron, He I and H L_{α, β, γ} photoionisation mass, and mid-IR spectroscopies, and in the solid state by Raman spectroscopy. Ab initio calculations at the MP2(fc)/6-31G** level provide an assessment of the geometric structures and, together with the IR and Raman data, permit an almost complete vibrational analysis for each molecule, which appear, in each case, to exist in one isomeric form. The calculations indicate that anti structures for the hydrogen atom position are the most stable in all cases. For BrClC=NOH, the calculations just prefer the Z-anti-structure as the thermodynamically more stable, whereas the vibrational spectroscopy, although not unambiguous, suggests the E-anti-isomer as the stereoselectively synthesised conformer. This conclusion is strengthened further by the observation that BrClC=NOH (with an exclusive E-structure), favours HCl elimination upon thermolysis, thus making it a good source of the transient BrCNO molecule. Photoelectron spectroscopy, supported by ab initio calculations and trends in the dihalo series, provides an analysis of the effect of halogen substitution on the orbitals of the parent oxime species, H₂C=NOH.

Full text of DOI:10.1039/a605743h

View Article Online / Journal Homepage / Table of Contents for this issue
Structure and spectroscopy of dihaloformaldoximes
He I photoelectron, photoionization mass spectroscopy, mid-IR, Raman and
ab initio study
Tibor Pasinszki¤ and Nicholas P. C. Westwood*
Guelph-W aterloo Centre for Graduate W ork in Chemistry, Department of Chemistry and
Biochemistry, University of Guelph, Guelph, Ontario, Canada N1G 2W 1
The electronic and geometric structures of dichloroformaldoxime, Cl CxNOH, dibromoformaldoxime, Br CxNOH and
2
2
bromochloroformaldoxime, BrClCxNOH, are investigated for the Ðrst time in the gas phase by He I photoelectron, He I and
H L
photoionisation mass, and mid-IR spectroscopies, and in the solid state by Raman spectroscopy. Ab initio calculations at
a, b, c
the MP2(fc)/6-31G** level provide an assessment of the geometric structures and, together with the IR and Raman data, permit
an almost complete vibrational analysis for each molecule, which appear, in each case, to exist in one isomeric form. The
calculations indicate that anti structures for the hydrogen atom position are the most stable in all cases. For BrClCxNOH, the
calculations just prefer the Z-anti-structure as the thermodynamically more stable, whereas the vibrational spectroscopy,
although not unambiguous, suggests the E-anti-isomer as the stereoselectively synthesised conformer. This conclusion is strength-
ened further by the observation that BrClCxNOH (with an exclusive E-structure), favours HCl elimination upon thermolysis,
thus making it a good source of the transient BrCNO molecule. Photoelectron spectroscopy, supported by ab initio calculations
and trends in the dihalo series, provides an analysis of the e†ect of halogen substitution on the orbitals of the parent oxime
species, H CxNOH.
2
We have a long-standing interest in the generation, spectros-
copy, and structure of small stable and unstable pseudoha-
lides,1 and recently have initiated investigation into the
potential of dihaloformaldoximes, XYCxNOH (where X,
Y \ halogen), and oximes in general, as sources of transient
nitrile oxides in gas-phase thermolysis or gas/solid reac-
tions.2,3 During the course of this work it became evident that
nothing is known, either theoretically or experimentally,
about the structure and spectroscopy of dihaloformaldoximes,
despite the fact that they are used in organic chemistry as pre-
cursors for the in situ generation of nitrile oxides, important
transient intermediates in 1,3-dipolar cycloadditions.
The structures of dihaloformaldoximes are particularly
interesting. With substitution by two identical halogens, the
hydrogen can adopt a syn- or anti-conformation (Scheme 1).
For the case of two di†erent halogen atoms, there is the
potential for four planar isomers due to the additional possi-
bility of a Z- or E-conformation with respect to the CxN
double bond.¤
The parent formaldoxime H CNOH, which tends to tri-
2
merise at room temperature, has been characterised structur-
ally in the gas phase by microwave spectroscopy15 and is
shown to be planar with an anti-conformation. The syn-
structure, calculated some 23(6-31G**)16 to 25(6-31G``)17 kJ
mol~1 higher in energy, has not been observed experimen-
Notwithstanding the lack of spectroscopic and structural
information, both Cl CNOH 1, and Br CNOH 2, have been
tally. Further spectroscopic investigations on H CNOH, rele-
2
2
2
known for a long time,4h7 giving ClCNO and BrCNO, respec-
tively, in high yields by HX elimination over a wide range of
pH. Since both serve as XCNO generators for [3 ] 2] cyclo-
addition reactions,8h11 they have become essential in the syn-
thesis of biologically active compounds such as pesticides and
insecticides,10 and transglutaminose inhibitors and anti-
tumour agents of use in medicinal chemistry.11 They are, we
should add, highly toxic,12 as illustrated by the historical
name of 1, phosgene oxime.
vant to the present work, include microwave,15,18 IR,19,20 and
photoelectron21h23 studies, together with
a variety of
quantum chemical calculations.16,17,24 Experiment and theory
both agree on the planar anti-ground-state structure.
In this work we present the Ðrst geometric, electronic and
vibrational investigation of the dihaloformaldoximes,
Cl CNOH 1, Br CNOH 2 and BrClCNOH 3, using both ab
2
2
initio [MP2(fc)/6-31G**] and spectroscopic methods. The gas-
phase spectroscopy includes UV photoelectron (PES), He I
Other dihaloformaldoxime derivatives are more rare. The
and H L
photoionisation mass (PIMS), and mid-IR. The
a, b, c
diiodo species, I CNOH has been synthesised,5,6 but owing to
gas-phase vibrational data is complemented by a Raman
study of the solid. Of especial interest is the linking of experi-
ment and theory to assess the molecular conformations.
2
its thermal instability at room temperature has not received
further attention. The formation of diÑuoroformaldoxime,
F CNOH, has been demonstrated in the mercury photo-
2
sensitisation of chlorodiÑuoromethane in the presence of
added NO,13 but its attempted direct synthesis from
CF ClNO and H S was not successful.14 The preparation of
2
2
the mixed halogen derivatives BrClCNOH 3, and FClCNOH
has been brieÑy described,6,14 while the other dihalo- deriv-
atives, FBrCNOH and XICNOH (where X \ F, Cl, Br),
remain unknown.
Scheme 1
¤
By convention, isomers with respect to a CxN double bond are
generally, although not exclusively, labelled as syn and anti. Since we
are using the syn/anti notation for isomers about the NwO single
bond, we prefer the use of Z and E for the double bond isomers.
¤
Permanent address: Department of Inorganic Chemistry Techni-
cal University, H-1521 Budapest, Hungary.
J. Chem. Soc., Faraday T rans., 1997, 93(1), 43È51
43

View Article Online
also occur, and so we label the four isomers, 3Z-s, 3E-s, 3Z-a
and 3E-a. They also have stable minima. All equilibrium
structures are shown in Fig. 1: the calculated total energies,
dipole moments and rotational constants are given in Table 1.
The experimental structure of the parent molecule
Experimental
The dihaloformaldoximes were synthesised using literature
methods. Cl CNOH was prepared by reduction of trichloro-
2
nitromethane (Aldrich) with tin metal.25 Br CNOH was
2
synthesised9 from oximinoacetic acid [prepared from gly-
H CNOH,15 together with a calculated structure performed
2
oxylic acid (Aldrich) and hydroxylamine hydrochloride
at the same level as that used in this work, is also shown in
(
Aldrich)]26 and bromine, and BrClCNOH from Br CNOH
Fig. 1 for comparison. This serves to indicate that the
MP2(fc)/6-31G** method does a reasonable job for the
known parent structure, providing some conÐdence in the cal-
culated structures of 1, 2 and 3, which are unknown. The CN
bond is calculated marginally too long, a typical result for
MP2. Interestingly, halogen substitution does not appear to
a†ect the length of the nominally double CN bond, a result
comparable to that observed experimentally for Cl or Br sub-
stitution on ethylene.30 Halogen substitution slightly shortens
the NwO bond, and increases the angle at nitrogen from 110¡
2
and mercury(II) chloride.6
He I PE spectra were recorded on a home-built fast-
pumping spectrometer; resolution during measurements was
ca. 45 meV. Spectra were calibrated with the known ionisation
potentials (E ) of O and N . The spectrometer, a modiÐ-
i
2
2
cation of an earlier version,27 also mass analyses ions produced
in the photoionisation process using a quadrupole mass
analyser (Hiden Analytical, 320 u), with its electron impact
source removed, mounted directly above the photoionisation
point. Single-wavelength PIMS spectra are obtained with He I
in H CNOH to ca. 112¡ and ca. 118¡ for anti and syn struc-
2
(
21.22 eV) or unÐltered H L
(10.2È12.7 eV) radiation.
tures, respectively. In general, the calculated angles are consis-
a, b, c
Although not done in coincidence, PE and PIMS spectra can
be recorded within seconds of each other, of much beneÐt in
the identiÐcation of transient and new gas-phase species.2,3
IR spectra, recorded at 0.5 cm~1 resolution, were collected
on a Nicolet 20SXC interferometer using a 20 cm single-pass
gas cell with KBr windows, which gave a spectral range from
4000 to 400 cm~1. Spectra were also recorded in the 400È200
cm~1 region using polyethylene windows and a CsI beam-
splitter. All compounds had sufficient vapour pressure for gas-
phase IR although, in practice, the vapour was pumped
continuously through the cell using a rotary pump while
maintaining the pressure constant between 250 and 300
mTorr.
Raman spectra of the crystalline compounds were recorded,
using Nd : YAG excitation (1.064 lm), on a Bruker RFS 100
FT-Raman spectrometer at 2 cm~1 resolution, and a laser
power of 100 mW.
Computational methods
Ab initio calculations were carried out at the MP2 level using
a standard 6-31G** basis set on the lighter atoms and one of
[6s4p2d] quality on bromine.28 The core electrons were not
included in the correlation energy calculations (i.e. “frozen-
coreÏ). Equilibrium molecular geometries were fully optimised
and harmonic vibrational frequencies were then calculated at
the minima using numeric second derivatives. All calculations
were performed with the GAUSSIAN-92 quantum chemistry
package.29
Results and Discussion
Calculated structures
Fig. 1 Calculated [MP2(fc)/6-31G**] geometries of the two planar
isomers of Cl CNOH, 1 and Br CNOH, 2, and the four planar
All calculated structures are planar and, in all cases, both syn
and anti conformers have stable minima. For 1 and 2, these
structures are denoted as 1s or 1a and 2s or 2a. In the case of
2
2
isomers of BrClCNOH, 3. For the notation used, see text. Also
included is the experimental (microwave) structure for the parent
H CNOH molecules,15 together with a structure calculated at the
2
3, Z- and E-isomers with respect to the CN double bond can
same level of theory as the dihaloformaldoxime structures.
Table 1 Calculateda total energies, dipole moments and rotational constants of XYCNOH (X, Y \ halogen)
total energy
kb
Ac
B
C
Cl CNOH, 1a
[1087.376 481
[5308.177 806
[3197.776 317
[3197.777 234
[1087.370 753
[5308.174 498
[3197.770 777
[3197.773 765
0.940
0.846
0.877
0.955
2.396
2.466
2.307
2.553
4.3679
2.8178
4.2926
2.9625
4.2805
2.7792
4.2113
2.9301
2.5701
1.1617
1.5170
2.0596
2.5717
1.1572
1.5173
2.0452
1.6180
0.8226
1.1209
1.2149
1.6065
0.8170
1.1154
1.2045
2
Br CNOH, 2a
2
BrClCNOH, 3E-a
BrClCNOH, 3Z-a
Cl CNOH, 1s
2
Br CNOH, 2s
2
BrClCNOH, 3E-s
BrClCNOH, 3Z-s
a Calculated at the MP2(fc)/6-31G** level. b Dipole moment in D B 3.336 ] 10~30 C m. c Rotational constants in GHz; isotopes 35Cl, 79Br, 12C,
4N, 16O, 1H.
1
44
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
tent from molecule to molecule, including the pronounced tilt
of the substituted methylene group, and the strongly bent
CNOH frame with fairly small bond angles of ca. 101¡ (anti)
and ca. 108¡ (syn) at oxygen.
Before discussing the calculated relative stabilities of con-
formers, we should, at the outset, appraise possible isomer-
isation processes, both for the syn/anti hydrogen atom, and for
the Z/E CxN double bond isomers. For the syn/anti ques-
tion, a lateral shift at oxygen (inversion) or rotation around
the NwO bond are potential processes. For the Z/E-isomers,
a lateral shift at nitrogen (inversion) or rotation around the
CxN bond, must also be considered. For simplicity we con-
sider that any inversion process goes through a planar tran-
sition state, with no involvement of combined
inversion/rotation processes.¤
The anti form is thermodynamically more stable, by 15.0
and 8.7 kJ mol~1 for 1 and 2, respectively. In-plane inversion
at oxygen is a very high energy process; for 1 the barrier from
the anti side is 219.2 kJ mol~1. Rotation around the NwO
single bond is the preferred isomerisation process, although
from the anti side it is still relatively high, 30.9 kJ mol~1 for 1,
and so rotation about NwO is not entirely free. The corre-
sponding torsional barrier for H CNOH has been calculated
2
to be even lower, 11.0 kJ mol~1 (6-31G``),17 although only
the anti-isomer is seen experimentally.
For the four isomers of 3, the anti isomers 3Z-a and 3E-a
are again the most stable, with 3Z-a being the lower (Table 1)
by only 2.4 kJ mol~1. Nevertheless, the NwO rotational
barrier for antiÈsyn isomerization is expected to be much as
for 1, ca. 31 kJ mol~1, thereby hindering isomerisation to the
higher energy syn structures. Of more signiÐcance for the
heavy atom structure of 3 is the potential for EÈZ-isomerism
resulting from hindered CN rotation or inversion at nitrogen.
From calculations performed on simple imines and for-
maldoxime,31 it would appear that inversion is a much lower
energy process, although more recently16 it has been shown
that CxN rotation via a triplet state could be even lower in
energy. For 3, inversion at nitrogen (Z ] E) is very high, 175.4
kJ mol~1, precluding isomerisation of 3Z-a to 3E-a, the two
lowest-energy structures. Since these conformers cannot easily
interchange, their relative amounts in a synthesised sample
will reÑect the selectivity of the preparative method. An issue
to be discussed below concerns which isomer is obtained, and
whether or not the present spectroscopic data can distinguish
between the Z- and E-isomers of 3.
Fig. 2 (a) He I and (b) H L
and 2
photoionization mass spectra of 1, 3
a, b, c
Photoionisation mass spectra (PIMS)
The single wavelength PIMS (either He I, 21.22 eV or
also observed weakly in the He I PIMS of 3, where
ClCNOH` features strongly. We have shown recently that a
strong NO` peak dominates the PIMS of simple nitrile
oxides,2 and so the main source of NO` in the PIMS of 1
could be the ClCNO` ion. The formation of CNOH` can
also be observed, although it appears much more strongly in
the He I PIMS of 2. This suggests that 2 is another suitable
H L
, 10.2È12.7 eV) of 1, 2 and 3 are shown in Fig. 2(a)
a, b, c
and 2(b). Fragmentation is reduced with H L
radiation as
a, b, c
shown by the increase in parent ion signal and disappearance
of the weak fragment peaks seen with He I. Fragmentation is
simplest in the case of 2 since, although some weak fragments,
Br` and BrCN`, are observed in the He I PIMS, they are not
detected with H L
radiation, the major fragments indicat-
source of CNOH` [other than HONxC(Cl)C(O)OCH ]32
a, b, c
3
ing elimination of Ðrst one, and then the second, Br atom. Br
atom loss is also the principle fragmentation step for 3, but in
both 1 and 3 fragmentation of ClCNOH` can take di†erent
routes. For 1, neutral HCl and Cl loss from the parent ion are
the two major competing processes as shown by the strong
ClCNO` and ClCNOH` fragments with H La, b, c . Another,
less important, fragmentation channel, OH and NOH elimi-
nation from M` producing Cl CN` and Cl C` ions also
for neutralisationÈreionisation experiments designed to inves-
tigate the potential interstellar species HONC. Apart from the
mass spectrometry study, HONC has only been observed by
matrix IR, interestingly from photolysis of 2.33
In all cases, the parent ion is accompanied by an (M ] 1)`
ion with the M`/(M ] 1)` relative intensity strongly depen-
dent on the pressure, as is the ClCNO`/ClCNOH` ratio in 1.
On reducing the pressure, the (M ] 1)` peak becomes negligi-
bly small, with the ClCNO` peak becoming slightly more
intense than ClCNOH`. Increasing the pressure increases the
relative intensity of the protonated forms. Such pressure
dependences clearly indicate secondary ion reactions. Given
the propensity of oximes to protonate, often as the Ðrst step of
a Beckmann rearrangement, this is not surprising.
2
2
seems to be available at He I energies, these presumably being
precursors to ClCN` and ClC`. These latter fragments are
¤
Inversion barriers have been calculated for in-plane inversion, i.e.
planar structures. It is feasible that the transition state is not planar,
resulting in a slightly lower energy barrier.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
45

View Article Online
IR and Raman spectra
The gas-phase IR and solid-phase Raman spectra are shown
in Fig. 3 and 4, respectively, with experimental and calculated
frequencies in Tables 2 and 3. In what follows below, we will
consider all oximes, 1È3, to be planar with anti-hydrogen
atoms. The justiÐcation for this is based on the precedence of
the parent anti-H CNOH molecule, and the fact that, at the
2
level of spectroscopic resolution applied herein, certainty on
the syn/anti question is not feasible.
Fig. 4 FT-Raman spectra of 1, 3 and 2
Both 1 and 2 are prolate asymmetric rotors of C symmetry;
s
i
for 1a and 2a is [ 0.31 and [ 0.66, the position of the H
calc
atom (anti/syn) having only a limited e†ect. The calculated
rotational constants of 3 depend strongly on whether the mol-
ecule is the E- or Z-isomer, and this leads to i
values of
calc
[
0.75 for 3E-a (prolate), and [0.03 for 3Z-a, maximally
asymmetrical and close to neither prolate or oblate. The
larger question here is whether or not Z- or E-conformers can
be distinguished by IR. Of the 12 fundamentals, which trans-
Fig. 3 IR spectra of 1, 3 and 2. The top portion of the Ðgure shows
the calculated [MP2(fc)/6-31G**, unscaled] band positions and rela-
tive intensities for 2.
form in C as 9a@ (in-plane) and 3aA (out-of-plane), all are both
s
IR and Raman active. The aA modes give rise to C-type bands,
Table 2 Experimental [IR (gas), Raman (solid)] and calculated vibrational frequencies (cm~1) of 1 and 2
Cl CNOH
Br CNOH
2
2
assignment
IR(g)
Raman(s)
calculateda
IR(g)
Raman(s)
calculateda
and description
3
3
3
2
2
638 R
632 Q
626 P
944
3630 R
3622 P
H
H
ca. 3216b
3860 (117)
ca. 3232b
3847 (121)
l
1 OH str.
È
È
2927
2674
1
È
È
l ] l
2
2l
3
3
687
597 R
590 P
H
1
603
1607
1642 (37)
1594, 1582c
1623 (27)
l
CxN str.
1
2
1
1
1
1
9
9
9
9
8
8
8
359 R
353 Q
346 P
000
1350 R
1346 Q
1341 P
911
H
8
H
ca. 1442b
1389 (42)
ca. 1398b
1380 (51)
917 (229)
l3 OH bend
1030, 1019c
1042 (230)
902
l (1), l (2)CNO def. ] CwX str.
4
5
79 R
74 Q
71 Q
66 P
98 R
92 Q
86 P
957 R
H
9
96, 976c
1013 (167)
987
È
996 (127)
l (1), l (2)NwO str.
949 P
5
4
852 R
843 P
H
H
È
l
] l
11
10
7
7
7
7
6
6
5
51 R
H
735 R
726 P
H
È
È
È
È
2l
40 P
11
18 R
06 P
H
704 R
697 P
H
?
56 R
44 P
17 Q
È
H
l
ClwC asym. str.
BrwC asym. str.
H
6
67, 645c
662 (9)
534
545
542 (4)
6
] dCNO
512
484
È
300
È
534 (18)
495 (2)
403 (117)
307 (0.1)
270 (3)
482 Q
È
372 Q
471
375
È
217
È
186
89
498 (31)
374 (1)
393 (103)
203 (0.5)
248 (0.7)
186 (0.1)
l
CNOH out-of plane def.
ClwC, BrwC sym. str.
OH torsion
1
7
1
8
1
9
0
l
l
l
l
l
3
80 Q
È
1
È
È
È
ClCCl sciss., BrCBr rock
CNO out-of-plane def.
È
È
2
ClCCl rock, BrCBr sciss.
261
244 (0.4)
9
6
0
5
lattice mode
lattice mode
a Calculated at the MP2(fc)/6-31G** level; IR intensities (km mol~1) in parentheses. b Broad bands (associated OH, see text). c Splitting due to
crystal Ðeld e†ects.
46
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Table 3 Experimental [IR (gas), Raman (solid)] and calculated vibrational frequencies (cm~1) of 3
calculateda
assignment
and description
IR(g)
Raman(s)
Z-a
E-a
3
3
3
2
2
1
1
1
1
1
9
9
9
9
9
8
8
8
635 R
631 Q
626 P
940
H
ca. 3195b
3852 (117)
3857 (121)
l1 OH str.
È
È
l ] l
2
2l
3
3
682
607 R
597 P
354 R
349 Q
344 P
H
1
602
D1416b
99
1633 (33)
1387 (42)
1633 (30)
1383 (52)
l
CxN str.
2
H
l3 OH bend
68 R
H
9
1016 (139)
992 (255)
1005 (116)
968 (250)
l
NwO str.
58 P
49 R
45 Q
40 P
88 R
83 Q
78 P
4
5
H
952
È
l
CNO def. ] CwX str.
H
l
] l
1
0
11
7
7
7
7
6
6
5
49 R
40 P
H
È
È
2l
11
18 R
08 P
44 R
35 P
09 Q
È
H
?
H
668, 643c
545 (5)
656 (7)
l
Cl(Br)C str. ] dCNO
6
502, 493
È
501 (27)
395 (106)
486 (4)
268 (3)
249 (0.3)
208 (0.3)
523 (21)
402 (114)
381 (0.6)
248 (0.7)
248 (0.2)
230 (0.5)
l
CNOH out-of plane def.
0
OH torsion
1
1
7
1
8
9
l
l
l
l
l
1
È
È
È
È
383
305
253
242
Br(Cl)C str.
CNO out-of-plane def.
BrCCl scissors
BrCCl rock
2
1
8
03
6
lattice mode
lattice mode
a Calculated at the MP2(fc)/6-31G** level; IR intensities (km mol~1) in parentheses. b Broad bands (associated OH, see text). c Splitting due to
crystal-Ðeld e†ects.
whereas, depending on the direction of the a-principal axis,
the a@ fundamentals give rise to A-, B- or A/B-type bands. The
molecules are shown in their calculated principal axis system
in Fig. 5. Of note is the fact that in 1, the a-principal axis runs
essentially down the (zigzag) CNO backbone, whereas in 2
the a-axis is essentially normal to the CN bond. Such orienta-
tional di†erences also hold true for 3Z-a and 3E-a (syn-hydro-
gen atoms do not incur any signiÐcant di†erence) implying
that band shapes may provide a clue to the dominant isomer.
Using calculate rotational constants (Tables 2 and 3), and the
equations of Seth-Paul,34 the calculated PR separations of A-,
B- and C-type bands are 14.8, 10.5 and 22.1 cm~1 for 1; 10.3,
7
.3 and 15.4 cm~1 for 2 and ca. 12, 9 and 18 cm~1 for 3
(average of 3E-a and 3Z-a), although for hybrid bands it is to
be expected that the separations would be slightly smaller.
The assignment of the fundamentals is based on the com-
plementary nature of the IR and Raman data, the ab initio
calculations, inter-comparisons between molecules 1, 2 and 3,
and comparison with similar molecules, notably H CNOH,20
2
CH (H)CNOH,35 (CH ) CNOH,36 and 1,1-dihalo-substituted
3
3 2
ethylenes.37,38 The ideal systems for comparison would be the
dihaloimine species but, regrettably, apart from F CxNH,39
2
these species are unknown. However, Cl CxNCl is known;
2
the IR has been obtained,40,41 and will provide some guid-
ance. We shall Ðrst discuss the assignments and then possible
conformers based on the calculations and band shapes (see
below).
There are two important points. The Ðrst is that the ab
initio harmonic frequencies, also included in Tables 2 and 3,
are in surprisingly good agreement with experiment. The top
of Fig. 3 shows the calculated (unscaled) vibrational fre-
quencies and intensities for 2, directly above the experimental
spectrum. A similar, excellent correspondence obtains also for
1
, and implies that, in the case of 3, the calculated vibrational
frequencies could be used as a guide for distinguishing Z/E-
conformers. To this end, calculated values for both the Z- and
E-isomers of 3 are given in Table 3. The second point is that
both the gas-phase IR and the solid-state Raman spectra for
molecules, 1, 2 and 3 (Fig. 3 and 4) show only sufficient bands
Fig. 5 Molecules 1a, 2a, 3E-a and 3Z-a shown in their calculated
[MP2(fc)/6-31G**] principal axis system
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
47

View Article Online
for one isomer or, at least, a very dominant isomer, and so
they are assigned to one species.
appears around 550È600 cm~1. However, the potential-energy
distribution (PED) calculated using unscaled ab initio force
constants as input to the BMAT program,44 indicates that
Assignments: 4000–1250 cm~1 region. Details of some of
the bands are shown in Fig. 6. The strong band in the IR
spectra between 3640 and 3620 cm~1 [Fig. 6(a)] can be
this mode is strongly coupled with l , the CwCl anti-
6
symmetric stretch, and so we assign it to the present position
on the basis of the calculated position and intensity. The NO
stretch and CNO deformation are distinguished on the basis
of their intensities and rotational proÐles [Fig. 6(d)], and they
switch position on going from 1 ] 3 ] 2 (Tables 2 and 3) such
that, in 2, the higher wavenumber band belongs to the NO
stretch, whilst in 1, this fundamental appears, as a more pro-
nounced A-type band, on the low-wavenumber side. This
switch is reproduced by the ab initio calculations.
assigned unambiguously to l , the OH stretch. Such a vibra-
1
tion in free oximes generally occurs in the 3650È3500 cm~1
region.42,43 The next two IR bands around 1610È1590 and
1
360È1340 cm~1 originate from the CxN stretch, l , and the
2
OH in-plane bend, l , respectively. Details are shown in Fig.
3
6(b) and 6(c), illustrating the change in position and (in the
case of l ) the change in rotational proÐle. The OH bend
2
shows an A-type proÐle, the trend in PR separation consistent
Weak bands seen in the IR spectra of both chloro deriv-
atives at 660È640 cm~1 appear with weakÈmedium intensity
in the Raman, and are split due to crystal-Ðeld e†ects. These
with the change in rotational constants. Values for these fun-
damentals in the simple alkyl oximes, CH (H)CNOH35 and
3
(
CH ) CNOH36 are 1655/1675 and 1268/1344 cm~1, respec-
bands are assigned to the CwCl stretch, l ; the correspond-
3
2
6
tively. Two very weak bands observed in the IR of all three
molecules in the ranges 2670È2690 and 2920È2950 cm~1 are
assigned to 2l and l ] l , respectively (Tables 2 and 3).
In the Raman spectra of all three molecules, weak broad
bands ascribable to solid-phase OH stretching and bending
vibrations are observed in the regions 3200 and 1400 cm~1,
respectively. Such gasÈsolid shifts are typical for
oximes,36,42,43 and represent the di†erence between free and
associated OH.
ing antisymmetric stretches in the dichloro species Cl CCH37
2
2
and Cl CNCl41 occur at 796 and 959 cm~1. For all CwX
2
modes there is, from the PED, a strong coupling with dCNO,
3
2
3
leading to frequencies that are generally to lower wavenum-
ber. The corresponding CBr antisymmetric stretch in 2 is
2
seen weakly in the IR at 534 cm~1 (696 cm~1 in
Br CxCH ),38 but is more intense in the Raman at 545
2
2
cm~1. A band in the IR spectra of all three molecules (Fig. 3)
at 520È480 cm~1 shows a classic C-type rotational envelope,
and since the corresponding band in the Raman is very weak,
it is assigned to l , the CNOH out-of-plane deformation. An
1
250–480 cm~1 region. The two close, strong, IR bands just
1
0
below 1000 cm~1 are assigned to l and l ; they originate
additional band, which appears strongly in the Raman of 1 at
484 cm~1, but not in the IR, can be ascribed, on the basis of
4
5
from the NO stretch, and the CNO in-plane deformation
Tables 2 and 3). The NO stretch in oximes typically appears
(
its position and Raman intensity as the CCl symmetric
2
around 930 cm~1,42 and is weak in the Raman,36 as seen here.
The assignment of the more intense band of the two to the
CNO deformation is atypical since this fundamental generally
stretch, l .
7
In addition to the above-noted fundamentals, there are two
very weak IR bands in the 900È700 cm~1 region, not seen in
the Raman, which are assigned to the l ] l combination
1
0
11
band, and to the Ðrst overtone of the OH torsion, l
1 (Tables
1
2
and 3).
4
80–150 cm~1 region. There is a strong IR band at 380È370
cm~1 (recorded using polyethylene windows for 1 and 2,
only), which cannot be found in the Raman spectra of 1, and
is potentially masked in the Raman of 2 and 3 by strong
CwBr-based fundamentals (see below). It has a pronounced
Q-branch and, as a C-type band, is assigned to the OH
torsion, l . We note that OwHÉÉÉN interactions in solid
1
1
(
CH ) CNOH36 are seen at 118 cm~1, in the location of what
3
2
we have described as lattice modes.
As noted above, strong bands in the Raman spectra of 2
and 3 around 370È380 cm~1 correspond to the CBr sym-
2
metric stretch (in 2) and the CBr stretch (in 3). Below 310
cm~1 three additional bands are expected, corresponding to
the CNO out-of-plane deformation, and the CX (or CXY)
2
rock and scissors modes. The CNO out-of-plane deformation
l
is expected to be weak, and is partially masked by the
1
2
CCl vibrations in 1 at ca. 300 cm~1, but can be seen weakly
2
in the mixed compound 3 at 305 cm~1. The rock l and scis-
8
sors l modes are then ascribed to the remaining pairs of
9
strong Raman bands at 300/261 cm~1 in 1, 217/186 cm~1 in 2
and 248/230 cm~1 in 3 (Tables 2 and 3).
Conformers. Apart from the calculations, which indicate
thermodynamic stability and barriers, the IR spectra provide
the best opportunity to decide which isomers are observed.
This is still somewhat difficult, given that any band contour
analysis can be fraught with problems, especially at low
resolution. However, some general comments may be made,
using both calculated frequencies and band envelopes, to
provide some guidance for future high-resolution work.
Fig. 6 Details of, (a) l , the OH stretch, (b) l , the CN stretch, (c) l ,
1
2
3
the OH bend and (d) l /l , the NO stretch/CNO deformation, in 1, 2
First, with regard to the hydrogen atom, we believe that the
anti-isomer exists in all cases; it is the thermodynamically
4
5
and 3 (note; the l /l bands switch position, see text for assignment)
4
5
48
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
most stable and as mentioned, the anti structure has been
found, both experimentally and theoretically, for the parent
formaldoxime H CNOH15h18,24 and other oximes. Some evi-
2
dence that this holds also for the dihaloformaldoximes comes
from the rotational band envelopes of l which are of type A,
1
B/(A) and B in molecules 1, 3 and 2, respectively [Fig. 6(a)], in
agreement with a CNOH anti structure and the principal axis
systems of Fig. 5.
The calculations for the E- and Z-isomers of 3, i.e. 3E-a and
Z-a suggest 3Z-a as the thermodynamically more favourable
3
isomer, albeit by only 2.4 kJ mol~1. As noted above, the IR
and Raman spectra can be assigned to essentially one isomer;
this accords with the high (175 kJ mol~1) barrier for nitrogen
inversion. From the calculations (Table 3), the largest, and
possibly most signiÐcant, di†erence between the frequencies of
the two isomers occurs in the case of the CwX stretches
(
XxCl or Br; l and l ). Given the magnitude of the di†er-
6
7
ence for the two isomers (ca. 110 cm~1), and the aforemen-
tioned accuracy of the computated vibrational frequencies, it
would appear that the E-isomer is favoured experimentally.
Although the calculated Z/E thermodynamic preference is too
close to call, with the Z-isomer predicted to be more slightly
more stable, it is clear that the preparative method produces,
stereoselectively, only one isomer. The synthetic method
Fig. 7 He I photoelectron spectra of (a) 1, (b) 2 and (c) 3
tion techniques,47 with the valence orbital sequence,
.......(5a@)2(6a@)2(1aA)2(7a@)2(2aA)2
.
and so this provides a suitable starting point.
involves an apparently simple reaction between 2 and HgCl
2
The PE spectrum of H CNOH consists of a pair of bands
in aqueous solution. Possible mechanistic interpretations
2
at 10.61/11.14 eV assigned to n (2aA) and n (7a@) and, fol-
could involve mercury(II) ion mediated selectivity, or aqueous
equilibria between the oxime and the nitrile oxide followed by
selective nucleophilic attack of the chloride anion on the
nitrile oxide. Some evidence for the latter mechanism comes
from NMR studies,45 which indicate that the nitrogen lone
pair develops trans to the attacking Cl~. Such a mechanism
would account for the exclusivity of the E-isomer, noted here.
The formation of the E-conformer of 3 has important chemi-
cal consequences, especially with regard to the products of
thermolysis (see below).
~
~
lowing a gap, another pair at 14.25/14.86 eV assigned to n
1aA) and n (6a@), where ] and [ refer to bonding and anti-
`
(
`
bonding combinations, respectively, of N and O p orbitals. In
the outer-valence region, replacement of the two hydrogen
atoms by two halogen atoms will add four characteristic
halogen p-type orbitals (two a@ and two aA) into the gap region
(
with some encroachment onto the Ðrst two bands), together
with two p orbitals arising from CwX bonding and anti-
bonding combinations.
The Ðrst band in the spectra of the dihaloformaldoximes
Although difficult to distinguish at 0.5 cm~1 resolution, the
A- A/B- and B-type band proÐles concur in almost all cases,
except for the NO stretch, with the E-conformation of 3.
between 9.8 and 10.2 eV originates from the 4a@ (n ) MO
~
which is CxN bonding and NO antibonding (n Èn ). In
CN
O
H CNOH this n /n orbital is observed at 10.61 eV,23 with
2
CN O
the result that there is a ca. 0.4È0.6 eV destabilisation upon
halogen substitution due to mesomeric e†ects. The second
band in the PE spectrum of 1 at 11.36 eV (Fig. 7) can be
Photoelectron (PE) spectra
The He I PE spectra are shown in Fig. 7 with experimental
and calculated E listed in Table 4. The shifts in PE bands, on
assigned to 11a@, an antibonding admixture (n ) of the in-
plane lone pair orbitals on N and O, and is clearly stabilised
is
~
going from 2, through 3, to 1 can be easily followed. On the
basis of these trends, the known PE spectra of the parent
formaldoxime21h23 and 1,1-dihaloethylenes,46 and calcu-
lations using KoopmansÏ theorem, the assignment is relatively
by inductive e†ects compared to the corresponding MO in
H CNOH. The consequences of dihalo-substitution on the
2
Ðrst two, closely spaced, energy levels of H CNOH is that
2
they are now clearly separated. The 11a@ (n ) MO is not
~
straightforward, although in C symmetry there is consider-
s
able mixing, especially amongst molecular orbitals (MOs) of a@
observed in the spectrum of the dibromo- derivative 2,
symmetry. The assignment for H CNOH has been put on a
because it is completely hidden under intense halogen lone-
pair bands at 10.68 and 11.27 eV, and it shows only as a
2
Ðrm basis by calculations involving many-body GreenÏs func-
Table 4 Experimentala and calculatedb,c ionisation potentials (eV) of the dihaloformaldoximes
Cl CNOH
BrClCNOH
Br CNOH
2
2
exp.
calc.
exp.
calc.d
exp.
calc.
orbital character
1
0.12
1.36
1.87
2.38
2.90
3.50
10.80 (4aA)
13.18 (11a@)
12.76 (10a@)
13.33 (3aA)
13.84 (9a@)
15.01 (2aA)
16.77 (8a@)
10.05
(11.4)
11.16
11.85
12.43
13.13
14.83
15.4
10.59 (4aA)
13.01 (11a@)
11.82 (10a@)
12.55 (3aA)
13.34 (9a@)
14.60 (2aA)
16.02 (8a@)
17.31 (1aA)
18.53 (7a@)
19.93 (6a@)
21.14 (5a@)
9.83
È
10.49 (4aA)
12.75 (11a@)
11.37 (10a@)
11.89 (3aA)
12.42 (9a@)
13.96 (2aA)
15.57 (8a@)
17.17 (1aA)
17.93 (7a@)
19.78 (6a@)
20.94 (5a@)
n Èn
CN
O
1
1
1
1
1
n Èn
N
O
10.69
11.27
11.75
12.64
14.35
15.2
16.3
17.6
È
n (p)
X
n (n)
X
n (p)
X
n (n) ] n
X
O
O
G
p
` ] n
1
5.6
CX
1
7.46 (1aA)
n ] n
O
CN
1
1
7.0
8.3
È
18.79 (7a@)
19.99 (6a@)
21.47 (5a@)
16.7
17.9
È
p
~ ] n
CX
O
n ] n
O
N
p
a Vertical E s. b Calculated via KoopmansÏ theorem. c Valence orbital numbering in assignment order. d E-a isomer.
i
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
49

View Article Online
shoulder (estimated at 11.4 eV) in 3. The KoopmansÏ calcu-
lations (Table 4) fail to give the correct sequence here, prefer-
ring to stabilise excessively this MO at the expense of the
strongly halogen-based MOs.
amounts of CO and CO were also identiÐed. In short, the
mixed bromochloro compound 3 selectively eliminates HCl,
producing BrCNO and not ClCNO, and actually appears to
be a better route to gas-phase BrCNO than 2, the dibromo-
species.2
2
The assignment of the four sharp bands in all three mol-
ecules between 10.5 and 13.5 eV to the halogen-based lone
pair orbitals is fairly certain. The exact sequence of halogen-
based MOs for dihalo compounds in local C symmetry has
Conclusion
2
v
H CNOH is a much studied molecule. In this work we have
a much chequered history,46,48,49 with the most likely
2
provided the Ðrst spectroscopic investigation of the dihalo-
sequence (in order of increasing E ) being b , a (n), a and
i
2
2
1
substituted
species,
namely
dichloroformaldoxime,
b (n),48 although a switch of a and a has been sug-
1
2
1
Cl CxNOH, dibromoformaldoxime, Br CxNOH and
gested.46,49 For the dihaloformaldoxime molecules these
halogen-based MOs appear to follow the Ðrst sequence, as do
the calculations, and so we assign them as 10a@, 3aA, 9a@ and
2
2
bromochloroformaldoxime, BrClCxNOH, principally in the
gas phase (PES, PIMS, IR), but also in the solid phase
(Raman). This, coupled with ab initio calculations at the
2
a@ in order of increasing E . The fourth band of this cluster
i
MP2(fc)/6-31G** level has provided us with an assessment of
their electronic and geometric structures, together with an
analysis of their vibrational and photoelectron spectra. As in
the case of the parent formaldoxime molecule, anti structures
(
2aA) around 12.6È13.5 eV shows a broader FranckÈCondon
proÐle than the others owing to mixing with orbitals on the
CNOH group.
The next band in 2 at 14.35 eV, and that at 16.3 eV ori-
ginate from the two CwBr p orbitals, assigned as 8a@ and 7a@.
They are observed at 13 ^ 0.2 eV and 15.2 eV in the spectrum
(with respect to the hydrogen position) are the thermodynami-
cally most stable in all cases. In the case of 3, a ZÈanti con-
Ðguration (with respect to the CxN double bond) is predicted
to be the most stable, but only just. The gas-phase investiga-
tions by mid-IR coupled with an analysis of band positions,
band shapes and a comparison with the ab initio calculations
suggest however that the next stable isomer, the E-anti con-
former, is the one uniquely favoured. This conclusion is sup-
ported by Raman studies of the solid. This notion is further
justiÐed by the fact that 3, by virtue of an exclusive E-
structure, is a good thermolytic source of the transient nitrile
oxide BrCNO, HCl elimination being favoured. He I photo-
electron spectroscopy shows easily followed trends from the
parent molecule through to BrClCNOH, enabling all bands to
be assigned with a considerable degree of certainty.
Finally, we would point out that the molecules described
herein, are ideal species for structural study by microwave
spectroscopy since they include two halogen atoms, each with
two isotopes, which have a large quadrupolar coupling,
together with small nitrogen quadrupole coupling. FT-
microwave spectroscopy is thus the perfect tool for these mol-
ecules, given the ability to see weak isotopes in natural
abundance. With the signiÐcant di†erence in the predicted
dipole moments and rotational constants for the Z- and E-
isomers of 3, they should be easy to distinguish.
of 1,1-dibromoethylene46 and are shifted toward higher E in
i
2
, as expected, owing to the negative inductive e†ect of the
NOH group, i.e. strongly mixed with n . The remaining two
0
weaker bands in the spectrum of 2 (where they show most
clearly) at 15.2 and 17.6 eV can be assigned to 1aA (n ) and 6a@
`
(
n ), the oxygen out-of-plane orbital being mixed strongly
`
with n , and the oxygen in-plane orbital being mixed with
the nitrogen lone-pair orbital. These MOs are thus stabilised
CN
with respect to those of H CNOH, with the new CwX
2
p-based MOs intervening. The shift of the last four bands by
replacing Br by Cl atoms can be easily followed (Fig. 7) and
hence the same assignment obtains for 3 and 1 in the 14È19
eV region (Table 4). In the case of 1, the broad band at 15.6
eV corresponds to both the CwCl p (8a@) and n ] n (1aA)
O
CN
orbitals.
Thermolysis of 3, BrClCNOH
We have shown previously that thermolysis, or reaction of 2
with dehydrohalogenating reagents, leads to the formation of
the transient BrCNO molecule.2 In the present work, we
believe from the IR data that reaction between 2 and HgCl
2
to produce 3 selectively favours the E-isomer. This structure,
with Cl and OH groups on the same side of the CN double
bond, is stereochemically favourable for HCl elimination. To
this end we undertook the thermolysis of 3 using a quartz
tube (8 mm id) surrounded by a furnace along a 15 cm length,
attached directly to the inlet of the PEÈPIMS spectrometer or
Ñow-through IR cell. For a more efficient pyrolysis the tube
was loosely packed with quartz chips. The major products of
the thermolysis were, indeed, HCl and BrCNO (Scheme 2),
lending strong support to our thesis of an E-isomer, even
though, on bond-strength grounds, one might expect a pre-
ferred elimination of HBr to yield ClCNO. Formation of
N.P.C.W. thanks the Natural Sciences and Engineering
Research Council of Canada (NSERC) for operating and
equipment grants in support of this work. T.P. thanks
NSERC for the award of a NATO Science Fellowship.
References
1
T. Pasinszki and N. P. C. Westwood, J. Phys. Chem., 1995, 99,
649; N. P. C. Westwood, Chem. Soc. Rev., 1989, 18, 317; T.
Veszpremi, T. Pasinszki and M. Feher, J. Chem. Soc., Faraday
1


T rans., 1991, 87, 3805; M. Alaee, E. G. Livingstone and N. P. C.
Westwood, J. Am. Chem. Soc., 1993, 115, 2871; T. Pasinszki, H.
Yamakado and K. Ohno, J. Phys. Chem., 1993, 97, 12718.
HNCO, ONCCNO and Br side products were also observed.
2
Decomposition of 3 commenced at 300 ¡C with increasing
2
3
T. Pasinszki and N. P. C. Westwood, J. Phys. Chem., 1995, 99,
temperature decreasing the amount of unreacted precursor
and increasing the HCl yield. The BrCNO yield depends
strongly on temperature. The highest BrCNO/side product
ratio was observed at ca. 350È400 ¡C, and further increases in
temperature increased the relative amounts of side products.
Formation of Br2 and ONCCNO may be explained by the
decomposition of BrCNO. At higher temperatures small
6
401.
T. Pasinszki and N. P. C. Westwood, J. Am. Chem. Soc., 1995,
117, 8425; T. Pasinszki and N. P. C. Westwood, J. Phys. Chem.,
1996, 100, 16856.
I. De Paolini, Gazz. Chim. Ital., 1930, 60, 700.
L. Birkenbach and K. Sennewald, L iebigs Ann. Chem., 1931, 489,
4
5
7.
6
7
8
L. Birkenbach and K. Sennewald, Chem. Ber., 1932, 65, 546.
G. Endres, Chem. Ber., 1932, 65, 65.
J. E. Baldwin, C. Hoskins and L. Kruse, J. Chem. Soc., Chem.
Commun., 1976, 795; G. R. Humphrey and S. H. B. Wright, J.
Heterocycl. Chem., 1989, 26, 23; K. Halling, I. Thomsen and K. B.
G. Torssell, L iebigs Ann. Chem., 1989, 985; D. Chiarino and M.
Fantucci, J. Heterocycl. Chem., 1991, 28, 1705; N. Sewald and K.
Burger, L iebigs Ann. Chem., 1992, 947; P. Pevarello, R. Amici, M.
Scheme 2
50
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Colombo and M. Varasi, J. Chem. Soc., Perkin T rans. 1., 1993,
25 E. Gryszkiewicz-Trochimowski, K. Dymowski and E. Schmidt,
Bull Soc. Chim. France, 1948, 597.
26 E. G. E. Jahngen Jr. and E. F. Rossomando, Synth. Commun.,
1982, 12, 601.
27 D. C. Frost, S. T. Lee, C. A. McDowell and N. P. C. Westwood,
J. Electron. Spectrosc. Relat. Phenom., 1977, 12, 95.
2
3
151; E. C. Boyd and R. M. Paton, T etrahedron L ett., 1993, 34,
169.
9
0
P. A. Wade, J. F. Bereznak, B. A. Palfey, P. J. Carroll, W. P.
Dailey and S. Sivasubramanian, J. Org. Chem., 1990, 55, 3045.
A. Ch. T. Hsu (Rohm and Haas Co.) Eur. Pat. Appl., EP 232,012
1
1
987; D. Munro and B. Patel (Shell Int. Res. Maatschappij B. V.)
28 R. C. Binning and L. A. Curtiss, J. Comput. Chem., 1990, 11, 1206.
29 M. J. Frisch, G. W. Trucks, M. Head-Gordon, P. M. W. Gill, M.
W. Wong, J. B. Foresman, B. G. Johnson, H. B. Schlegel, M. A.
Robb, E. S. Replogle, R. Gomperts, J. L. Andres, K. Raghava-
chari, J. S. Binkley, C. Gonzalez, R. L. Martin, D. J. Fox, D. J.
Defrees, J. Baker, J. J. P. Stewart and J. A. Pople, GAUSSIAN
92, Revision E.1, Gaussian, Inc., Pittsburgh PA, 1992.
30 M. D. Harmony, V. W. Laurie, R. L. Kuczkowski, R. H. Schwen-
deman, D. A. Ramsey, F. J. Lovas, W. J. La†erty and A. G. Maki,
J. Phys. Chem. Ref. Data, 1979, 8, 619.
Eur. Pat. Appl., EP 555,931 1993; M. Mizukai, N. Tsukaguchi, T.
Yoneda, Y. Kondo, H. Oota, Sh Kato and J. Tanaka (Sankyo
Co.) Jpn. Kokai T okkyo Koho, JP 05,105,672 [93,105,672] 1993.
A. A. Hagedorn III, B. J. Miller and J. O. Nagy, T etrahedron
L ett., 1988, 21, 229; P. A. Wade, M. K. Pillay and S. M. Singh,
T etrahedron L ett., 1982, 23, 4563; R. V. Stevens and R. P.
Polniaszek, T etrahedron, 1983, 39, 743; A. L. Castelhano,
D. H. Pliura and M. C. Venuti [Syntex (U.S.A.), Inc.] US Pat.
1
1
4
,970,297, 1990; M. De Amici, C. De Micheli and V. Misani,
T etrahedron, 1990, 46, 1975; P. Pevarello and M. Varasi, Synth.
Commun., 1992, 22, 1939; J. C. Rohlo†, J. Robinson III and J. O.
Gardner, T etrahedron L ett., 1992, 33, 3113; G. D. Diana, D. Cut-
li†e, D. L. Volkots, J. P. Mallamo, T. R. Bailey, N. Vescio, R. C.
Oglesby, T. J. Nitz, J. Wetzel, V. Giranda, D. C. Pevear and F. J.
Dutko, J. Med. Chem., 1993, 36, 3240.
31 R. D. Bach and G. Wolber, J. Org. Chem., 1982, 47, 239; R. D.
Bach and G. Wolber, J. Org. Chem., 1982, 47, 245.
32 C. E. C. A. Hop, K.-J. van den Berg, J. L. Holmes and J. K.
Terlouw, J. Am. Chem. Soc., 1989, 111, 72.
33 G. Maier, J. H. Teles, B. A. Hess Jr. and L. J. Schaad, Angew.
Chem. Intl. Ed. Engl., 1988, 27, 938.
1
2
S. Chanachev, Nauch. T rudove V isshiya Med. Inst., 1958, 5, 173
34 W. A. Seth-Paul, J. Mol. Struct., 1969, 3, 403.
(Chem. Abstr., 1962, 56, 665); A. J. McAdams and M. H. Jo†e,
35 G. Geiseler, H. B oŽ hlig and J. Fruwert, J. Mol. Struct., 1973, 18,
U.S. Dep. Commun., PB 168119, 1965 (Chem. Abstr., 1967, 66,
3548k).
M. G. Bellas, O. P. Strausz and H. E. Gunning, Can. J. Chem.,
965, 43, 1022.
N. N. Tarovenko and S. P. Motornyi, Zh. Obshch. Khim., 1960,
43.
9
36 W. C. Harris and S. F. Bush, J. Chem. Phys., 1972, 56, 6147.
37 J. C. Evans, J. Chem. Phys., 1959, 30, 934; F. Winther and D. O.
Hummel, Spectrochim. Acta, Part A, 1967, 23, 1839; Y. Yamaoka
and K. Machida, J. Mol. Spectrosc., 1980, 83, 21.
38 J. R. Scherer and J. Overend, J. Chem. Phys., 1960, 32, 1720.
39 H. B uŽ rger and G. Pawelke, J. Chem. Soc., Chem. Commun., 1988,
105; P. Groner, H. Nanaie, J. R. Durig and D. D. DesMarteau, J.
Chem. Phys., 1988, 89, 3983.
40 J. M. Burke and R. W. Mitchell, Spectrochim. Acta, Part A, 1972,
28, 1649.
1
3
4
1
1
3
0, 4066.
1
1
5
6
I. N. Levine, J. Chem. Phys., 1963, 38, 2326.
D. R. Armstrong and G. T. Walker, J. Mol. Struct. (T heochem),
1
987, 149, 369.
1
7
8
P. D. Adeney, W. J. Bouma, L. Radom and W. R. Rodwell, J.
Am. Chem. Soc., 1980, 102, 4069.
1
I. N. Levine, J. Mol. Spectrosc., 1962, 8, 276; M. G. K. Pillai, J.
Phys. Chem., 1962, 66, 179; V. K. Kaushik and K. Takagi, J.
Phys. Soc. Jpn., 1978, 45, 1975; A. Klesing and D. H. Sutter, Z.
Naturforsch., T eil A, 1990, 45, 817.
41 K. Kalcher and W. Kosmus, Z. Naturforsch., T eil A, 1981, 36,
391; K. Kalcher, E. Nachbauer and A. Popitsch, J. Mol. Spectro-
sc., 1983, 102, 174.
42 G. Socrates, Infrared Characteristic Group Frequencies, Wiley,
Chichester, 1980.
1
2
9
0
S. Califano and W. Lu
L. R. Zumwalt and R. M. Bader, J. Chem. Phys., 1939, 7, 235; A.
Azman, D. Hadzi, J. Kidric, B. Orel and C. Trampuz, Spectro-
Ž the, Z. Phys. Chem., 1956, 6, 83.
43 D. Hadz— i and L. Premu, Spectrochim. Acta, Part A, 1967, 23, 35.
—
—
—
—
44 D. F. McIntosh and M. R. Peterson, QCPE Program No.
QCMP067 (1989) available from the Quantum Chemistry
Program Exchange, Indiana University, Bloomington, Indiana.
chim. Acta, Part A, 1971, 27, 2499; G. Duxbury, R. M. Percival,
D. Devoy and M. R. M. Mahmoud, J. Mol. Spectrosc., 1988, 132,
3
80; G. Duxbury, J. Mol. Spectrosc., 1988, 132, 393; R. A. Bannai
45 Ch. Grundmann and P. GruŽ nanger, T he Nitrile Oxides, V ersatile
and G. Duxbury, J. Opt. Soc. Am. B, 1994, 11, 170.
A. Dargelos and C. Sandorfy, J. Chem. Phys., 1977, 67, 3011.
D. C. Frost, W. M. Lau, C. A. McDowell and N. P. C. West-
wood, J. Phys. Chem., 1982, 86, 3577.
T ools of T heoretical and Preparative Chemistry, Springer-Verlag,
2
2
1
2
New York, 1971, p. 154.
46 K. Wittel and H. Bock, Chem. Ber., 1974, 107, 317.
47 R. Cambi and W. von Niessen, J. Electron Spectrosc. Relat.
Phenom., 1985, 37, 69.
48 J. C. Bunzli, D. C. Frost, F. G. Herring and C.A. McDowell, J.
Electron Spectrosc. Relat. Phenom., 1976, 9, 289.
49 K. Wittel and H. Bock, in T he Chemistry of Functional Groups,
Supplement D, ed. S. Patai and Z. Rappoport, Wiley, Chichester,
1983, ch. 28.
2
2
3
4
J. P. Dognon, C. Pouchan, A. Dargelos and J. P. Flament, Chem.
Phys. L ett., 1984, 109, 492.
W. F. Hwang and H. A. Kuska, J. Mol. Struct., 1978, 48, 239; M.
T. Nguyen and T. K. Ha, J. Mol. Struct. (T heochem), 1982, 88,
1
1
27; K. Akagi, Y. Tanabe and T. Yamabe, J. Mol. Struct., 1983,
02, 103; C. J. M. Huige, A. M. F. Hezemans and K. Rasmussen,
J. Comput. Chem., 1987, 8, 204; R. Glaser and A. Streitwieser, J.
Am. Chem. Soc., 1989, 111, 7340; A. H. Otto, J. Mol. Struct.
(
T heochem), 1991, 235, 489; M. T. Nguyen and L. G. Vanquick-
enborne, J. Chem. Soc., Perkin T rans 2, 1993, 1969.
Paper 6/05743H; Received 16th August, 1996
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
51