P.R. Joshi et al. / Journal of Molecular Structure 1100 (2015) 80e87
81
was observed as the product. The reaction followed Arbuzov
pathway and proceeded through the lone pair on the phosphorus
as a nucleophilic centre [14]. In another study, gas phase nucleo-
philic reaction between dimethoxy methane (DMM) and HCl was
reported with HCl as a nucleophile. Though DMM possesses both
methylenic and methyl carbon atoms, only methylenic pathway
was preferred in the gas phase producing chloromethyl methyl
ether (CMME) and methanol [15].
1:1, 1:2 and 1:3 in gas phase were studied by varying the concen-
tration of CH OH to understand the reaction progress against
stoichiometry. For clarity, IR spectra are discussed in four regions
3
ꢁ1
ꢁ1
ꢁ1
(520e450 cm
,
760e720 cm
,
1050e1020 cm ,
and
ꢁ1
3050e2920 cm ).
Fig. 1 (grid A) shows infrared spectra corresponding to the PeCl
stretching vibrational mode in a N matrix. The doubly degenerate
PeCl stretching mode [ (e)] of PCl has been reported to occur at
498.9 and 494.9 cm and the non-degenerate PeCl stretching
2
n
3
3
ꢁ
1
In the present study, matrix isolation infrared technique was
used to trap and identify the product formed in the gas phase re-
ꢁ
1
mode [
n
1
(a
(e) mode of PCl
(a ) mode at 511.1 and 509.2 cm in a N
(e) mode in N
2
1
)] at 510.4 and 508.8 cm in an Ar Matrix [24]. The
ꢁ
1
action between PCl
3
and CH
3
OH. DFT computations were carried
n
n
3
3
is observed at 496.6 and 492.3 cm and the
ꢁ
1
out to support our experimental results.
1
1
2
matrix (Fig. 1a, grid A).
matrix has
Multiple features observed in the
n
3
2
. Experimental and computational methods
already been assigned to i) isotopic effect of chlorine
3
5
37
(
Cl: Cl ¼ 3:0/2:1/1:2), ii) splitting of degenerate vibrational
The prerequisite of a low temperature for matrix isolation ex-
states due to the host matrix atoms and iii) matrix site effect or
periments was achieved using a RDX-408D2 (Sumitomo Heavy
Industries Ltd.) pulsed tube, closed cycle helium compressor cooled
unification of two or more of these effects [25].
When PCl
gas in a bulb and the resultant gas mixture was deposited on to a
cryotip at 12 K, infrared spectra showed new features at 469.2,
3 3
and CH OH were premixed in a 1:1 ratio along with
ꢁ
6
cryostat. A base pressure of less than 1 ꢀ 10 mbar was obtained in
N
2
the cryostat housed to an evacuated vacuum chamber.
ꢁ
1
ꢁ1
Analytical grade PCl
3
(Merck, Purity: > 99%) and HPLC grade
466.4, 464.5, 462.1 cm and a broad feature at 511.3 cm apart
from the parent PeCl stretching features of PCl (Fig. 1c). Impor-
tantly, intensities of PeCl features of the parent PCl did drop down
at the expense of new features. When experiments were carried out
with 1:2 PCl :CH OH ratio, we observed the same new features as
in the case of 1:1 PCl :CH OH but the parent PCl peaks were
CH
3
OH (Purity: >99%) were used without any further purification.
3
However, the samples were subjected to several freeze-pump-thaw
cycles before its use. Nitrogen (INOX) with a purity of 99.9995% was
3
used as the matrix gas. PCl
bulb having 1 dm volume along with nitrogen, as a matrix gas, at
room temperature. Three experiments were performed by varying
3
and CH
3
OH were premixed in a glass
3
3
3
3
3
3
completely disappeared. The infrared intensities of the product
features were relatively high when compared to the 1:1
the concentration of CH
stant such as a) PCl :CH
1:2:1000) and c) PCl :CH
effect of stoichiometry. A typical deposition lasted for about
3
OH and keeping PCl
OH:N (1:1:1000), b) PCl
OH:N (1:3:1000) to understand the
3
concentration con-
3
3
2
3
:CH OH:N
3
2
PCl
above new features were observed in the PeCl stretching region,
when 1:3 PCl :CH OH experiment was performed (Fig. 1e, grid A).
In the 760e720 cm region, for 1:1 ratio of PCl
along with N gas (Fig.1c, grid B), a strong feature at 753.1 cm and
weak features at 732.5 and 726.9 cm were observed. In the case
of 1:2 PCl :CH OH experiment (Fig. 1d, grid B), alike 1:1
PCl :CH OH experiment, all three features were observed. The in-
tensity of the features at 732.5 and 726.9 cm were higher and no
3 3
:CH OH experiment (Fig. 1d, grid A). Interestingly, none of the
(
3
3
2
3
3
ꢁ
1
9
0 min at ~3 mmol/h rate.
3 3
and CH OH
ꢁ
1
Infrared spectra of the matrix-isolated samples were recorded in
2
ꢁ1
ꢁ1
transmission mode between 4000 and 400 cm using a BOMEM
ꢁ
1
MB 100 FTIR spectrometer with 1 cm resolution. All the spectra
reported here are those obtained after deposition of samples onto a
KBr substrate maintained at 12 K.
3
3
3
3
ꢁ
1
ꢁ
1
Computations were performed using a Gaussian 94W package
significant change in the intensity of 753.1 cm
observed. In the case of 1:3 PCl :CH OH experiment (Fig. 1e, grid B),
the 753.1 cm feature was completely absent, while the features at
feature was
[
16] at B3LYP/6-311þþG(d,p) level of theory without imposing any
3
3
ꢁ
1
geometrical constraints. Geometries of reactants (PCl and CH OH)
3
3
ꢁ1
and possible products were optimized by keeping all geometrical
parameters free during optimization process. The optimized ge-
ometries were then used to obtain vibrational frequencies which
enable us to characterize the nature of stationary points and to
assign the experimentally obtained frequencies. The computed
vibrational frequencies were scaled on a mode-by-mode basis for
the different modes for assigning the experimental features. It is
recognized that matrix perturbs vibrational frequencies of the
trapped species, with the magnitude of perturbation depending
upon vibrational frequency of the mode. We therefore consider that
a mode-by-mode scaling is an appropriate method to account for
the varying degrees of matrix influence and helps in bringing the
computations in better agreement with experimental values over
the entire range of the vibrational wavenumbers [17e22]. The
conformational isomers of the reaction products were also studied.
The computed intensities and scaled frequencies were used before
plotting a simulated vibrational spectrum using SYNSPEC program
732.5 and 726.9 cm were observed with highest intensity.
ꢁ
1
In the 1050e1020 cm
region, Fig. 1b (grid C) the feature
observed at 1034.0 cm corresponds to OeC stretch of monomeric
CH OH [26e28]. When 1:1 PCl :CH OH experiment (Fig. 1c, grid C)
in N matrix was performed, the feature due to OeC stretching of
CH OH was absent and new strong features at 1044.4 and
1042.1 cm and a weak feature at 1026.8 cm were observed.
Similar observation was noticed, in the 1:2 PCl :CH OH experiment
ꢁ1
3
3
3
2
3
ꢁ
1
ꢁ1
3
3
ꢁ
1
(Fig. 1d, grid C) but the intensity of the feature at 1026.8 cm was
found to be increased. No parent signature of OeC stretch at
ꢁ1
1034.0 cm was seen in this experiment, confirming the complete
reaction of CH OH with PCl . In the case of 1:3 PCl :CH OH
experiment (Fig. 1e, grid C), apart from the feature due to OeC
3
3
3
3
ꢁ
1
stretch of monomeric CH
3
OH, the 1026.8 cm
feature was
observed with increased intensity.
ꢁ
1
ꢁ1
In the 3000e2800 cm region, a new feature at 2853.7 cm
ꢁ
1
along with a low intense feature at 2963.3 cm were observed in
1:1 PCl :CH OH experiment (Fig. 1c, grid D) which did grow sub-
stantially in the 1:2 PCl :CH OH experiment (Fig. 1d, grid D).
Whereas for the 1:3 PCl :CH OH (Fig. 1e, grid D) experiment, the
intensity of the feature at 2853.7 cm decreased considerably and
[
23] where a Lorentzian line profile with a full-width-at-half-
3
3
ꢁ1
maximum of 1 cm was assumed.
3
3
3
3
ꢁ
1
3
. Results and discussion
ꢁ1
the intensity of the 2963.3 cm feature increased.
3.1. Experimental section
New features observed in all the aforementioned spectra in the
different regions indicate that the reactants PCl
3 3
and CH OH un-
3 3
Three experiments with different ratios of PCl :CH OH such as
dergo reaction to yield new products. Importantly, these new