Journal of Molecular Spectroscopy 331 (2017) 44–52
Contents lists available at ScienceDirect
Journal of Molecular Spectroscopy
journal homepage: www.elsevier.com/locate/jms
Phosphorous bonding in PCl3:H2O adducts: A matrix isolation infrared
and ab initio computational studies
Prasad Ramesh Joshi, N. Ramanathan, K. Sundararajan, K. Sankaran ⇑
Materials Chemistry Division, Indira Gandhi Center for Atomic Research, Kalpakkam 603102, Tamil Nadu, India
a r t i c l e
i n f o
Article history:
Received 30 September 2016
In revised form 3 November 2016
Accepted 10 November 2016
Available online 12 November 2016
Keywords:
Phosphorus bonding
Ab initio
Matrix isolation
AIM analysis
NBO analysis
a b s t r a c t
Non-covalent interaction between PCl3 and H2O was studied using matrix isolation infrared spectroscopy
and ab initio computations. Computations indicated that the adducts are stabilized through novel P O
type phosphorus bonding and conventional PACl H type hydrogen bonding interactions, where the former adduct is the global minimum. Experimentally, the P O phosphorus bonded adduct was identified
in N2 matrix, which was evidenced from the shifts in the vibrational wavenumbers of the modes involving PCl3 and H2O sub-molecules. Atoms in Molecules and Natural Bond Orbital analyses have been performed to understand the nature of interactions in the phosphorus and hydrogen bonded adducts.
Interestingly, experimental evidence for the formation of higher PCl3AH2O adduct was also observed
in N2 matrix.
Ó 2016 Elsevier Inc. All rights reserved.
1. Introduction
Phosphorus is the fifth most important biogenic element after
carbon, hydrogen, oxygen, and nitrogen in terms of mass [1]. The
PAO bond (especially in phosphate esters) is the fundamental unit
present in DNA, RNA, and ATP systems and thus have significant
role in the structure, metabolism and replication in living systems
[2]. In the nuclear industry in particular, the PCl3 molecule
assumes significance as a precursor for the synthesis of phosphonates. The study of phosphorous containing molecules, involving
non-covalent interactions would be helpful to gain more insight
into the nature of weak bonding existing between phosphorus centered molecules with foreign reagents.
Studies on non-covalent interaction have been carried out since
last century and in this regard hydrogen bonding interaction was
well established experimentally as well as theoretically [3–6].
After conventional H-bonding interaction, halogen bonding was
the most widely explored interaction wherein high anisotropic
electrostatic potential seems to be responsible for interaction
between two electronegative atoms [7–10]. In the recent past,
non-covalent interactions involving oxygen and sulfur atoms
referred as chalcogen bonding have been reported [11–13].
Recently, computational studies revealed a new type of bonding
known as pnicogen bonding [14] in which interaction between a
pnicogen atom such as N, P, As, Sb and Bi atoms of Lewis acid is
⇑ Corresponding author.
E-mail address: [email protected] (K. Sankaran).
http://dx.doi.org/10.1016/j.jms.2016.11.005
0022-2852/Ó 2016 Elsevier Inc. All rights reserved.
present [15]. Solimannejad et al. observed this kind of interactions
computationally while studying the HSN and PH3 system [16]. In
their study HAPN interaction was found to be more stable compared to the generally expected PAH N interaction. Scheiner
et al. further elaborated this unusual evidence by examining simple PH3ANH3 system and noticed that HAPN interaction was
twice more stable than the PAH N interaction [17]. The stability
of pnicogen bond was probed further by replacing one of the H
atom of PH3 by more electron withdrawing groups such as Cl or
carbon chains [18–20]. Furthermore, the abilities of different donor
atoms on the strength of pnicogen bonding were studied and
revealed the following trend PN > P O > PS > Pp [21,22].
Del Bene et al. performed quantum chemical studies to understand
various parameters such as structure, binding energy, spin-spin
coupling constant and NMR properties of pnicogen bonding [23–
26]. Although, initial studies were focused on PN and PP interactions, extension of pnicogen bond to EE0 bonding where interactions between other elements of group Va such as NN, PAs or
AsAAs were available in various literatures [27–29]. The study of
pnicogen-bonded anionic adducts, bonds involving sp2 hybridized
phosphorous atom (in (H2C = PX2)), single pnicogen bonded
adducts, intramolecular pnicogen interactions such as in PHFA
(CH2)nAPHF system, and pnicogen-hydride interaction in complexes such as XH2P HBeY have been reported [30–34]. Detailed
comparison between pnicogen bonding and hydrogen or halogen
bonding was also the topic of several studies. Additionally, latest
computational studies in this aspect claimed that vibrational spec-
45
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
3. Computational methods
494.2
496.6
Ab initio calculations were performed on the PCl3AH2O system
using a Gaussian09 package [50]. Geometries of the monomer were
first optimized at B3LYP and MP2 levels of theory using 6-311++G
(d,p) and aug-cc-pVDZ basis sets. Starting from the optimized
monomer geometries, the geometry of the 1:1, 1:2 and 2:1 adducts
was then optimized without imposing any structural constraints.
Interaction energies were computed for the adducts, corrected separately for the basis set superimposition errors (BSSE) using the
method outlined by Boys and Bernadi [51] and the zero point energies (ZPE). Turi and Dannenberg showed that simultaneous application of both BSSE and ZPE corrections tend to underestimate
the interaction energies, thus both these corrections were not
applied together [52]. Vibrational wavenumber calculations were
performed on the optimized geometries to enable us to characterize the nature of stationary points and also to assign the experimentally observed vibrational wavenumbers. Indeed, all the
structures discussed in this work correspond to minima on the
potential energy surface.
To understand the nature of interaction in the 1:1 adducts, the
theory of Atoms In Molecules (AIM) proposed by Bader was applied
[53]. A (3, 1) bond critical point (BCP) that could be associated
with the intra and intermolecular interactions was searched
between the PCl3 and H2O sub-molecules of the adducts. The
488.5
479.6
511.1
509.2
507.2
A Leybold AG pulsed tube closed cycle helium compressor
cooled cryostat was used to achieve the low temperature for
matrix isolation experiments. A base pressure <1 106 mbar
was obtained in the cryostat housed in an evacuated vacuum
chamber.
Analytical grade PCl3 (Merck, Purity: >99%) and Milli-Q ultra
pure water were used in the experiment. The samples were subjected to freeze-pump-thaw cycles before use. Nitrogen (INOX)
with a purity of 99.9995% was used as the matrix gas. PCl3 and
H2O were deposited onto a KBr-substrate maintained at 12 K by
streaming them separately through a effusive twin-jet-nozzle system. PCl3 gas was mixed with nitrogen gas in a mixing chamber
and the resultant mixture was allowed to stream through one nozzle with the flow being adjusted by a dosing valve. A second nozzle
was utilized for the deposition of H2O. Here, a bulb containing H2O
was maintained at different temperatures ranging from 55 to
80 °C to control the required vapor pressure and thereby the concentrations in the matrix. Typical sample to matrix ratio ranging
from 1 to 3:1000 for PCl3:N2 and 0.5–1.25:1000 for H2O:N2 were
used. A typical deposition lasted for about 75 min at a rate of
3 mmol/h.
Infrared spectra of the matrix-isolated samples were recorded
in transmission mode between 4000 and 400 cm1 using a BOMEM
MB 100 FTIR spectrometer with 1 cm1 resolution. After deposition, the matrix was slowly warmed to 32 K, maintained at this
temperature for 15 min and then re-cooled to 12 K. The spectra
of the matrix thus annealed were recorded again. All the spectra
reported here refer to samples annealed at 32 K unless otherwise
specified.
Absorbance
2. Experimental methods
492.3
troscopy would be a promising tool for investigation of this kind of
new bonding [35,36].
In spite of numerous computational studies, experimental
investigations are very much limited on this kind of interactions.
Hill et al. reported for the first time the possibility of stabilization
of PP type interaction using NMR spectroscopy during the characterization of carborane-phosphino derivative [37,38]. X-ray
diffraction was used to provide evidence for existence of PP interactions in anion-neutral molecule interactions, pentafluorophenyl
substituted diphosphine, and aminotetra phospines [39–41] and
PN interaction in aminoalkylferrocenyldichlorophosphanes [42],
AsAs in cyclopentadienyl arsenic compounds and EE interactions in dipnicogen dimer and their dichalcogen dimer [43,44]. Single crystal X-ray diffraction method was used by Sundberg et al. for
analyzing PP interaction in 1,2-dicarba-closo-dodecaboranes [45].
The indications of nonbonding PP interaction was observed during the study of Bis(phosphanyl)carbaborane(12) derivatives
where 13C{1H} NMR technique has been used by Hey Hawkins
et al. [46,47]. However, the specific experimental study is still
scanty and no attempt has been made in this regard. Very recently,
we have experimentally confirmed the existence of the pnicogen
bonded P O and Pp interactions involving phosphorus acceptor
at low temperatures in the PCl3ACH3OH and PCl3AC6H6 adducts
respectively using matrix isolation infrared spectroscopy [48,49].
We have introduced the specific terminology of ‘phosphorus bonding’ in place of pnicogen bonding to experimentally highlight the
‘phosphorus centric’ interactions among the pnicogen group of
atoms.
In the present study, the phosphorus bonding interaction in
PCl3AH2O adducts is reported in N2 matrix using matrix isolation
infrared spectroscopy. Experimental results are supported with
ab initio computational studies.
e
d
c
b
a
520
510
500
490
480
470
460
Wavenumber (cm-1 )
Fig. 1. Spectra of PCl3AH2O adducts in a N2 matrix, spanning the region 520–
460 cm1; matrix isolation spectra for various concentrations of PCl3/H2O/N2, (a) 1/
0/1000; (b) 1/1/1000; (c) 2/1/1000; (d) 3/1/1000 and (e) 3/1.25/1000. All spectra are
recorded at 12 K after annealing at 32 K.
46
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
electron density q(rc), Laplacian of electron density r2q(rc), three
Hessian eigenvalues and the ratio of the eigenvalues |k1|/k3 were
examined at BCP to understand the nature of the interaction. Weak
interactions are characterized by small values of q(rc) and r2q(rc)
> 0 [54]. Identification and analysis of the critical points were done
using AIM2000 package [55]. Natural Bond Orbital (NBO; version
3.1) analysis invoked through Gaussian09 was used to understand
the nature of hyper conjugative charge–transfer interactions in the
PCl3-H2O adducts [56].
Even though, computations were performed at B3LYP and MP2
levels of theory with two different basis sets, for the spectral
assignments and for the AIM and NBO analyses, only the results
of B3LYP level of theory with 6-311++G(d,p) basis set was used.
4. Result and discussion
4.1. Experimental results
3488.4
3727.2
Fig. 1 shows the infrared spectrum in the region 520–460 cm1
where non-degenerate m1(a1) and doubly degenerate m3(e) PACl
stretching modes of PCl3 appear [57]. The multiple features
observed at 511.1, 509.2, 507.2 cm1 and 496.6, 494.2, 492.3,
490.8 cm1 were assigned to the isotopic combination of chlorine
atoms in m1(a1) and m3(e) modes of PCl3 in a N2 matrix [48]. When
PCl3 and H2O were co-deposited and then annealed, new features
appeared at 488.5 and 479.6 cm1 as shown in Fig. 1b–e.
Fig. 2 shows infrared spectra in the regions 3740–3700 cm1
and 3540–3440 cm1, corresponding to the m3 and m1 vibrational
modes of H2O. In a N2 matrix, m3 and m1 modes of H2O monomer
appear at 3727.2 and 3634.6 cm1 respectively [58]. Fig. 3 (grid
‘A’) shows the matrix isolation infrared spectra of H2O in N2 matrix
at different annealing temperatures. The effect of annealing on codeposition experiments of PCl3 with H2O in N2 matrix is compared
in Fig. 3, grid ‘B’. When PCl3 and H2O were co-deposited and then
3710.2
3704.7
d
Absorbance
3719.6
e
c
b
a
3740
3720
3700 3540
3490
3440
Wavenumber (cm -1 )
Fig. 2. Spectra of PCl3AH2O adducts in a N2 matrix, spanning the region 3740–3700
and 3550–3400 cm1; matrix isolation spectra for various concentrations of PCl3/
H2O/N2 (a) 0/1/1000; (b) 1/1/1000; (c) 2/1/1000; (d) 3/1/1000 and (e) 3/1.25/1000.
All spectra are recorded at 12 K after annealing at 32 K.
annealed, new features were observed at 3719.6, 3710.2, 3704.7
and 3488.4 cm1 (Figs. 2b–d and 3, grid ‘B’).
These new features were observed only when both PCl3 and H2O
were co-deposited and annealed at 32 K. The observed new features gained in intensity as the concentration of either of the precursors were increased, indicating that these features are only due
to PCl3 and H2O adducts.
4.2. Computational results
Computations were performed for PCl3AH2O adducts at B3LYP
and MP2 levels of theory using 6-311++G(d,p) and aug-cc-pVDZ
basis sets. Fig. 4 shows the structure of 1:1 PCl3-H2O adducts A
and B computed at B3LYP/6-311++G(d,p) level of theory. Selected
structural parameters such as bond length, bond angle and dihedral angle for adducts A and B are given in Table 1. ZPE and BSSE
corrected interaction energies of adducts A and B are also listed
in the table. Interestingly, adduct A, wherein interaction between
the phosphorus atom of PCl3 and oxygen atom of H2O, referred
as P O phosphorus bonding was found to be the global minimum
at all levels of theory whereas adduct B, stabilized by hydrogen
bonded Cl H interaction was the local minimum and this adduct
occurs at 2.3 kcal/mol higher in energy than the phosphorous
bonded adduct A. The ZPE corrected energy for the formation of
adduct A is indicated to be exothermic by about 2.4 kcal/mol at
B3LYP/6-311++G(d,p) level of theory. Adduct B is barely stabilized
at this level.
4.3. Vibrational assignments
4.3.1. PCl3:H2O 1:1 adducts
Table 3 shows the comparison of experimental and computational vibrational frequencies at B3LYP/6-311++G(d,p) level of
theory.
4.3.1.1. PACl stretching region. The m3 doubly degenerate PACl
stretching mode for the isotopic combination 3:0 (35Cl:37Cl) splits
and computed to occur at 452.0 and 443.0 cm1 for the adduct A
and at 457.5 and 449.7 cm1 for the adduct B. These features are
red shifted by 2.9 and 11.9 cm1 for the adduct A and blueshifted by 2.6 cm1 and red-shifted by 5.2 cm1 for the adduct B
from the monomeric PCl3. While the experimental feature
observed at 488.5 cm1, a red shift of 8.1 cm1 can be correlated
with the computed feature for the adduct A at 443.0 cm1, the second feature, which was computed to occur at 452.0 cm1 could not
be observed as this feature is likely overlapped with the feature of
the uncomplexed PCl3. As clear feature which is red-shifted with
respect to monomeric PCl3 alone was observed at low temperature
N2 matrix, it is confirmed that adduct A alone is generated. It is
important to point out that the energetics of the adducts also favor
the production of adduct A.
It should be mentioned that the vibrational frequency shifts in
the m3 and m1 modes for other isotopic combinations of 35Cl:37Cl
are negligible even for the monomeric PCl3. Hence, we could not
observe any new features for the other isotopic combinations for
the adduct A and adduct B in our experiment.
4.3.1.2. OAH asymmetric stretching region. DFT computations
revealed a red shift of 10.9 and 8.8 cm1 for the adduct A and B
respectively in the m3 OAH asymmetric stretching mode with
respect to H2O monomer. Similarly, a red shift of 8.6 and
5.2 cm1 was computed in the m1 symmetric OAH stretching mode
for the adduct A and B. Experimentally, a new feature was
observed at 3719.6 cm1 in the m3 asymmetric stretching, a red
shift of 7.6 cm1 from the uncomplexed H2O absorption. The new
feature could be assigned to either of the adduct A or B as the
47
3700
3650
3726.9
3600
3550
3500
3450
3400
B
12 K
25 K - 12 K
30 K - 12 K
32 K - 12 K
3750
3700
3650
Wavenumber, cm-1
3600
3550
3488.4
3549.7
3633.9
3686.8
Absorbance
3549.7
3633.9
3715.4
3686.8
Absorbance
3750
A
12 K
25 K - 12 K
30 K - 12 K
32 K - 12 K
3719.6 3715.4
3710.2
3704.7
3726.9
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
3500
3450
3400
3726.9
Wavenumber, cm-1
3700
3650
3488.4
3633.9
3719.6
3710.2
3704.7
3686.8
Absorbance
3750
C
H2O/N2 (1/1000)
PCl3 /H2O/N2 (3/1/1000)
3600
3550
3500
3450
3400
Wavenumber, cm -1
Fig. 3. Matrix isolation infrared spectra spanning the region 3750–3400 cm1 at different annealing temperatures. Grid ‘A’: H2O/N2 (1/1000); Grid ‘B’: PCl3/H2O/N2 (3/1/
1000); Grid ‘C’: H2O/N2 (1/1000) and PCl3/H2O/N2 (3/1/1000) experiments annealed at 32 K.
vibrational shift of adducts with respect to H2O monomer is almost
identical. Since, the experimental feature observed in the m3 mode
of PCl3 sub-molecule was assigned to the adduct A based on our
computational considerations, the feature at 3719.6 cm1 should
be due to the m3 OAH asymmetric stretching mode of adduct A.
Moreover, as mentioned earlier, energy profile of the adducts also
support the formation of adduct A at low temperatures.
Apart from the features due to monomer, new features were
also observed in the m3 mode at 3710.2 and 3704.7 cm1 and in
the m1 mode of H2O sub-molecule, it occurred at 3488.4 cm1.
These new vibrational features could possibly be due to higher
adducts of PCl3 and H2O and is discussed in the subsequent section.
4.3.2. Computations on the higher multimeric adducts of PCl3:H2O
To find out the next possible site of attack by PCl3 or H2O in the
adduct A, DFT computations were carried to optimize the higher
adducts of PCl3AH2O. Two structures corresponding to the
(PCl3)2-H2O [adducts C and D] and one structure to PCl3A(H2O)2
[adduct E] were optimized using B3LYP/6-311++G(d,p) level of theory as shown in Fig. 5.
Adduct C is stabilized by two P Cl and one P O interaction.
The bond distances between the P1Cl8, P1Cl6 and P5 O10 in
adduct C are 3.933, 4.405 and 2.927 Å, respectively. It should be
mentioned that the P O bond distance in the adduct C is shorter
than the adduct A, could be due to the presence of P Cl interaction in the adduct. The selected structural parameters for the
adducts C, D and E are given in Table 2. In the case of adduct D,
the H2O molecule is sandwiched between the two PCl3 molecules
and the structure is stabilized by two P O interaction. The bond
distances between P1 O5 and P8 O5 are 3.044 Å. The counterpoise corrected interaction energy for the adduct C and D are
2.73 and 4.25 kcal/mol, respectively, clearly showing that the
adduct C is less stable than adduct D.
In cyclic adduct E, H2O dimer is bonded to PCl3 molecule. This
adduct is stabilized by P O phosphorus bonding, O H and
C H conventional hydrogen bonding interactions. The bond
lengths P1 O9, O6 H10 and H5Cl4 are 2.819, 1.862 and
2.737 Å, respectively. The BSSE corrected interaction energy for
this adduct is found to be 9.41 kcal/mol. Computations revealed
that adduct E is energetically more favorable than adduct C and
D and hence there is a higher probability to observe the adduct E
in the PCl3AH2O experiments.
4.3.3. Vibrational assignments of higher multimeric PCl3:H2O adducts
The experimental and computed vibrational wavenumbers of
PCl3AH2O multimeric adducts are given in Table 4.
In the OH stretching region of H2O sub-molecule, new features
were observed at 3710.2 and 3704.7 cm1 in the m3 mode in addition to the feature at 3719.6 cm1 which was assigned to the 1:1
PCl3AH2O adduct A. Similarly, a new feature was observed at
3488.4 cm1 in the m1 mode of H2O sub-molecule. The intensity
of these new features increased as the concentration of either of
PCl3 or H2O was varied and they are red shifted by 17.0,
22.5 cm1 (m3 modes) and 146.2 cm1 (m1 mode) from the H2O
monomer absorption respectively. It should be mentioned that
for the various concentrations of H2O used in our experiments,
when deposited at 12 K, vibrational features characteristic of H2O
dimer were observed at 3715.4 (m3 mode) and 3549.7 cm1 (m1
mode) [59]. When the matrix was annealed at higher temperatures
from 12 to 32 K, gradually these features decreased in intensity
and got completely disappeared at 32 K (Fig. 3, grid ‘A’). As it is evident from Fig. 3, grid ‘A’ that at higher annealing temperatures, the
H2O dimer is utilized in the formation of H2O multimer as the multimer grows substantially at these higher annealing temperatures.
For experiments when PCl3 and H2O were co-deposited, features
of H2O dimer were initially observed and at higher annealing
48
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
2.954 Å
BCP
0.0 kcal/mol
1:1 adduct A
2.821 Å
BCP
2.3 kcal/mol
1:1 adduct B
Fig. 4. Computed structure of 1:1 PCl3AH2O adducts optimized at B3LYP/6-311++G(d,p) level of theory.
Table 1
Geometrical parameters and interaction energies of 1:1 adducts A and B calculated at
B3LYP/6-311++G(d,p) of theory.
Parametersa
a
b
Adducts
Adduct A
Adduct B
P1AO5
P1ACl4
P1ACl3
P1ACl2
Cl2AH5
O5AH6/O6AH5
\Cl4P1O5
\P1Cl2H5
\H6O5P1
Cl2H5O6
\H6O5H7
Tor\H6O5P1Cl4
Tor Cl2H5O6H7
Dipole momentb
2.954
2.110
2.109
2.106
–
0.963
176.9
–
116.2
–
105.4
119.1
–
2.06
–
2.089
2.097
2.103
2.821
0.966
–
112.1
–
176.8
–
–
123.8
3.53
Interaction energies (kcal/mol)
ZEP/BSSE corrected
2.43/2.67
0.14/0.44
Bond length in Å and bond angle and torsion angle in °.
Dipole moment in Debye.
of H2O multimers. The intensity of the features due to H2O
multimers was also found to be less in these experiments in
comparison to the one where H2O alone with the same concentration was deposited (Fig. 3, grid, ‘C’).
It is important to point out that among the higher adducts,
adduct E wherein H2O dimer interacts with PCl3 is energetically
more favorable and hence most likely to be formed in the matrix.
DFT computations showed that the m3 and m1 modes of adduct E
are red-shifted by 18.7, 44.9 and 190.1 cm1, which agrees well
with the experimental features observed in the m3 and m1 mode
of H2O sub-molecule at 3710.2/3704.7 and 3488.4 cm1, respectively, a red shift of 17.0, 22.5 and 146.2 cm1 (Fig. 2).
In the PACl stretching region, a new feature was observed at
479.6 cm1 (Fig. 1), which picks up in intensity with a red shift
of 17.0 cm1 from the PCl3 monomer absorption. This experimental
feature correlates well with the computed red shift of 21.2 cm1
for the adduct E.
Experimentally, no new features could be discerned in the m3
and m1 modes of PCl3 and in the H2O vibrational regions for the
adducts C and D.
4.4. Nature of interaction
temperatures, the new vibrational features at 3719.6, 3710.2,
3704.7 and 3488.4 cm1 with the concomitant reduction in the
features of H2O dimer (Fig. 3, grid ‘B’) were noticed. Hence, the
decrease in the intensity of H2O dimer could also be due to the formation PCl3A(H2O)2 adduct (adduct E) in addition to the formation
4.4.1. AIM analysis
The nature of the interaction in the phosphorus bonding adduct
A and hydrogen bonding adduct B was studied using AIM analysis.
A (3, 1) bond critical point BCP was located between phosphorus
49
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
BCP
BCP
3.933 Å
2.927 Å
4.405 Å
RCP
BCP
2:1 Terminal-H2O shaped adduct C: -3.59/-2.73
3.044 Å
BCP
2:1 Sandwich-H2O shaped adduct D: -4.41/-4.25
BCP
BCP
2.819 Å
RCP
1.862 Å
2.737 Å
BCP
1:2 Cyclic adduct E: -7.51/-9.41
Fig. 5. Computed structure of (a) terminal-H2O shaped adduct C; (b) sandwich-H2O shaped adduct D and (c) cyclic adduct E optimized at B3LYP/6-311++G(d,p) level of
theory. ZPE/BSSE corrected interaction energies (kcal/mol) are given alongside.
Table 2
Geometrical parametersa and interaction energies of adduct C, D and E calculated at B3LYP/6-311++G(d,p) level of theory.
a
b
Parameters
Adduct C
Parameters
Adduct D
Parameters
Adduct E
P5AO10
P5ACl8
P5ACl6
P5ACl7
P1ACl8
P1ACl2
P1ACl3
P1ACl4
2.927
2.116
2.106
2.104
3.933
2.100
2.098
2.096
P1AO5
P7AO5
P1ACl4
P1ACl2/P1ACl3
P8ACl9
P8ACl8
P8ACl10
\O5P1Cl4
3.044
3.044
2.103
2.105
2.103
2.104
2.105
178.1
\O10P5Cl8
\Cl8P1Cl2
\Cl6P1Cl2
176.5
171.9
129.9
178.1
108.1
112.3/110.2
2.819
2.117
2.100
2.131
2.737
0.965
0.962
0.962
0.974
173.4
163.8
137.2
Dipole momentb
4.03
\O5P8Cl10
\P1O5P8
\P1O5H6/
\P1O5H6
1.28
P1AO9
P1ACl2
P1ACl3
P1ACl4
H5ACl4
O6AH5
O6AH7
O9AH8
O9AH10
\O9P1Cl2
\O6H10O9
\Cl4H5O6
Bond length in Å and bond angle in °.
Dipole moment in Debye.
3.24
50
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
Table 3
Computed and experimental vibrational frequencies, shifts in the frequencies and mode assignments for PCl3AH2O adducts. Computations were performed using B3LYP/6-311+
+G(d,p) level of theory.
Computational vibrational frequencies
H2O
m (cm1)
a
b
c
d
e
Experimental vibrational frequencies
(in N2 matrix)
Vibrational mode assignments
3921.7 (57)a
3816.7 (9)
1602.9 (67)
3910.8 (67)
3912.9 (75)
3808.1 (23)
3811.5 (34)
1615 (74)
1611.4 (64)
D mb
–
–
–
10.9
8.8
8.6
5.2
+12.1
+8.5
m (cm1)
3727.2
3634.6
1597.6
3719.6
–c
–c
–c
–c
–c
Dm
–
–
–
7.6
–c
–c
–c
–c
–c
m3 (OAH) in H2O
m1 (OAH) in H2O
m2 (HAOAH) in H2O
m3 (OAH) in Adduct A
m3 (OAH) in Adduct B
m1 (OAH) in Adduct A
m1 (OAH) in Adduct B
m2 (HAOAH) in Adduct A
m2 (HAOAH) in Adduct B
PCl3
479.7 (37)
–
511.1
–
m1 (PACl, a1)d in PCl3
478.4 (38)
–
509.2
–
476.5 (38)
–
507.2
–
474.3 (36)
–
–c
–c
454.9 (166)
–
496.6 (vs)
–
456.6
453.2
454.7
451.8
451.8
(166)
(166)
(163)
(163)
(163)
–
–
–
3.1
494.2
492.3 (s)
494.2
490.8 (sh)
–c
–
–
–
–
–c
474.8
480.0
452.0
443.0
457.5
449.7
(57)
(43)
(178)
(182)
(181)
(163)
4.9
+0.3
2.9
11.9
+2.6
5.2
–c
–c
488.5
486.2
–c
–c
–c
8.1
10.4
–c
(35Cl:37Cl = 3:0)
m1 (PACl, a1) in PCl3
(35Cl:37Cl = 2:1)
m1 (PACl, a1) in PCl3
(35Cl:37Cl = 1:2)
m1 (PACl, a1) in PCl3
(35Cl:37Cl = 0:3)
m3 (PACl, e)e in PCl3
(35Cl:37Cl = 3:0)
m3 (PACl) in PCl3
(35Cl:37Cl = 2:1)
m3 (PA-Cl) in PCl3
(35Cl:37Cl = 1:2)
m3 (PACl, e) in PCl3
(35Cl:37Cl = 0:3)
m1 (PACl, a1)d in adduct A (35Cl:37Cl = 3:0)
m1 (PACl, a1)d in adduct B (35Cl:37Cl = 3:0)
m3 (PACl, e)e in adduct A
(35Cl:37Cl = 3:0)
m3(PACl, e)e in adduct B
(35Cl:37Cl = 3:0)
Intensities in km/mol given in parentheses.
Shift = Dm = madduct mmonomer.
Features are not observed experimentally.
PACl nondegenerate vibrational stretching mode (m1).
PACl doubly degenerate vibrational stretching mode (m3).
(P1) of PCl3 and oxygen (O6) of H2O in the adduct A and between
chlorine of PCl3 and hydrogen of H2O for the adduct B (Fig. 4).
Table 5 gives the value of electron density [q(rc)] and Laplacian
of electron density [r2q(rc)] at these BCPs for the adducts A and
B computed at B3LYP/6-311++G(d,p) level of theory. As can be seen
from the table that the magnitude of (q(rc)) for adduct A (PCl)
and adduct B (Cl H) at the BCPs is of the order 102 and 103 a.
u., respectively, clearly showing that the former adduct is stronger
than the latter which is in accordance with the interaction energies. The small values of (q(rc)) and small positive values of (r2q
(rc)) are indicative of weak nature of the interaction of close shell
type. The small value of |k1|/k3 also shows the weak nature of interaction in PCl3AH2O adducts.
4.4.2. NBO analysis
The charge transfer delocalization interaction (n-r⁄) generally
leads to red shifting of hydrogen bonds. NBO analysis is a useful
tool to estimate the strength of charge transfer delocalization by
looking at the magnitude of the second order perturbation energy
(E2).
The results of NBO analysis of 1:1 PCl3AH2O adducts A and B
computed at B3LYP/6-311++G(d,p) level of theory are listed in
Table 6. In H2O molecule, two lone pairs are present which have
ðrÞ
ðpÞ
been distinguished as s-rich (nO6 ) and pure-p (nO6 ) lone-pair NBOs
[60]. Similarly, the lone pairs on chlorine are also distinguished as
r and p pairs.
In 1:1 phosphorus bonded adduct A, a reduction in the electron
ðrÞ
occupancy by 0.0186e for the nO6 non-bonding orbital and an
increase in the electron occupancy by 0.0039e for the r⁄(P1ACl4)
bond of the PCl3 sub-molecule with respect to PCl3 monomer was
noticed. The E2 energy for this interaction is 3.90 kcal/mol clearly
showing that adduct A is mainly stabilized by charge transfer
ðrÞ
nO6 ? r⁄(P1ACl4) interaction. In the adduct A we also observe
ðpÞ
delocalization interactions from nO6 ? r⁄(P1ACl4) and the E2
energy for this interaction is 0.34 kcal/mol. As can be seen
from the Table 6 that the second order E2 energy was found to
ðrÞ
ðrÞ
ðrÞ
be in the order nO6 ? r⁄(P1ACl4) > nO6 ? r⁄(P1ACl3) = nO6 ?
r⁄(P1ACl2), with the reason could be due to the orientation of
chlorine atoms in the PCl3 molecule. In the adduct A, P1ACl4 bond
is almost linear to oxygen lone pair whereas P1ACl2 and P1ACl3
bonds are away from the linearity and due to this the effective
ðrÞ
delocalization occurs between the nO6 and r⁄(P1ACl4), which
results in elongation of PACl bond and a red-shift in PACl stretching mode of adduct A. Although, it is marginal, delocalization interðrÞ
action from phosphorus lone pair (nP1 ) to r⁄(O6AH5) and
r⁄(O6AH7) also contributes to the stabilization of P O interaction and E2 energies for these charge transfer interaction was found
to be 0.12 kcal/mol. Furthermore, these charge transfer results in
the increase in the electron occupancies in the antibonding orbitals
which makes the bond to lengthen with a concomitant red shift in
OAH stretching mode of adduct A.
51
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
Table 4
Computed and experimental vibrational frequencies, shifts in the frequencies and mode assignments for the PCl3 dimer, H2O-terminal shape (adduct C), sandwich H2O shape
(adduct D) and cyclic shape (adduct E). Computations were performed using B3LYP/6-311++G(d,p) level of theory.
a
b
Computational vibrational frequencies
Experimental vibrational frequencies (in
N2 matrix)
m (cm1)
m (cm1)
Dm
(PCl3)2AH2O (terminal-H2O shape - adduct C)
3913.3 (68)a
8.4
3809.9 (23)
6.8
1614.1 (76)
11.2
478.3 (42)
1.4
472.6 (48)
7.1
456.8 (224)
1.9
453.7 (148)
1.2
447.5 (199)
7.4
433.7 (118)
21.2
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
m3 (OAH)
m1 (OAH)
m2 (HAOAH)
m1 (PACl)
m1 (PACl)
m3 (PACl)
m3 (PACl)
m3 (PACl)
m3 (PACl)
PCl3AH2OAPCl3 (sandwich-H2O shape - adduct D)
3896.6 (61)
25.1
3796.7 (29)
20.0
1617.4 (68)
14.5
477.7 (12)
2.0
472.9 (91)
6.8
455.0 (317)
0.1
451.4 (250)
3.5
448.7 (65)
6.2
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
m3 (OAH)
m1 (OAH)
m2 (HAOAH)
m1 (PACl)
m1 (PACl)
m3 (PACl)
m3 (PACl)
m3 (PACl)
PCl3A(H2O)2 (cyclic adduct E)
3903.0 (137)
3876.8 (76)
3626.6 (420)
1640.4 (41)
1611.6 (82)
481.9 (148)
449.9 (79)
433.7 (143)
3710.2
3704.7
3488.4
–
–
–
–
479.6
17.0
22.5
146.2
–
–
–
–
17.0
m3 (OAH)
m3 (OAH)
m1 (OAH)
m2 (HAOAH)
m2 (HAOAH)
m1 (PACl)
m3 (PACl)
m3 (PACl)
D mb
18.7
44.9
190.1
37.5
8.7
2.2
5.0
21.2
Vibrational mode assignments
Intensities in km/mol given in parentheses.
Shift = Dm = madduct mmonomer.
Table 5
Properties of (3, 1) bond critical points of adducts A (phosphorus bonded) and B (hydrogen bonded) computed at B3LYP/6-311++G(d,p) level of theory.
a
b
c
Adducts
q(rc)a
r2q(rc)b
k1c
k2c
k3c
|k1|/k3
Adduct A
P1 O5
0.01425
0.03944
0.01136
0.01000
0.06080
0.18684
Adduct B
Cl2 H5
0.00605
0.01748
0.00499
0.00464
0.02712
0.18399
q(rc) is electron density.
r2q(rc) is Laplacian of electron density.
k1, k2, and k3 are three eigenvalues of Hessian Matrix.
Table 6
Electron occupancies, donor-acceptor delocalization interaction and second order perturbation energies (E2, kcal/mol) of various NBOs for adducts A and B computed at B3LYP/6311++G(d,p) level of theory.
Adducts
NBO
Occupancy
Donor-acceptor delocalization interaction
E2 (kcal/mol)
Adduct A
ðrÞ
nO5
⁄
1.97825 (1.99688)b
ðrÞ
nO5
⁄
? r (P1ACl4)
3.90
nO5 ? r⁄(P1ACl4)
ðpÞ
0.34
ðrÞ
0.09
ðrÞ
0.08
ðrÞ
0.12
ðrÞ
0.12
ðrÞ
0.33
ðpÞ
0.25
r (P1ACl4)
ðpÞ
0.10368 (0.09977)a
1.99706 (1.99745)b
ðrÞ
0.10368 (0.09977)a
1.97825 (1.99688)b
ðrÞ
0.09969 (0.09977)a
1.97825 (1.99688)b
nO5
r⁄(P1ACl4)
nO5
r⁄(P1ACl3)
nO5
r⁄(P1ACl2)
ðrÞ
0.09965 (0.09977)
1.99284 (1.99417)a
ðrÞ
0.00074 (0.00002)b
1.99284 (1.99417)a
nP1
r⁄(O5AH6)
nP1
r⁄(O6AH7)
Adduct B
a
0.00076 (0.00002)b
ðrÞ
1.95127 (1.95755)a
ðpÞ
0.00172 (0.00003)b
1.94012 (1.94915)a
nCl2
r⁄(H5AO6)
nCl2
r⁄(H5AO6)
b
a
Occupancy of monomeric PCl3 is given in parentheses.
Occupancy of monomeric H2O is given in parentheses.
0.00172 (0.00003)b
nO5 ? r⁄(P1ACl3)
nO5 ? r⁄(P1ACl2)
nP1 ? r⁄(O6AH5)
nP1 ? r⁄(O6AH7)
nCl2 ? r⁄(H5AO6)
nCl2 ? r⁄(H5AO6)
52
P.R. Joshi et al. / Journal of Molecular Spectroscopy 331 (2017) 44–52
ðrÞ
In the adduct B, the charge transfer occurs from ncl2 ?
r (H5AO6) and
ðpÞ
ncl2
? r (H5AO6) and the E2 energies for these
interactions are 0.33 and 0.25 kcal/mol, which is very weak when
⁄
⁄
ðrÞ
compared to the nO6 ? r⁄(P1ACl4) interaction.
5. Conclusion
In the present work, the experimental evidence for 1:1 PCl3A(H2O) adduct A and 1:2 PCl3A(H2O)2 adduct E was obtained in
N2 matrix. The formation of the adducts was evidenced from the
red-shifts in the PACl and OAH vibrational modes of PCl3 and
H2O sub-molecules. Ab initio computations preformed on 1:1 PCl3AH2O adducts revealed that adduct A is stabilized by phosphorus
bonded P O interaction while adduct B is stabilized through
hydrogen bonding interaction. Two types of 2:1 PCl3AH2O adducts
C and D were computed and these adducts were also stabilized by
phosphorus bonding. One type of 1:2 PCl3AH2O adduct was computed with a cyclic structure (adduct E), which was found to be
stabilized both by phosphorus and hydrogen bonding interactions.
Since, 1:2 PCl3AH2O adduct (adduct E) has a larger interaction
energy, this adduct was experimentally identified at low temperatures. Experimental vibrational frequencies of PCl3AH2O adducts
corroborate well with computation results. AIM and NBO analyses
together with ab initio computations unambiguously proved that
the phosphorus bonded adduct is stronger than hydrogen bonded
adduct.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
Acknowledgements
The authors thank Prof. D. Mohan, Department of Chemical
Engineering, A.C. Tech Campus, Anna University, Chennai for providing the computational support. Dr. P.R.J. is thankful to IGCAR,
Department of Atomic Energy, India for providing Research Associate fellowship.
[43]
[44]
[45]
[46]
[47]
[48]
References
[1] M.A. Pasek, D.S. Lauretta, Astrobiology 5 (2005) 515–535.
[2] E. Maciá, M.V. Hernández, J. Oró, Origins Life Evol. Biosphere 27 (1997) 459–
480.
[3] P. Goymer, Nat. Chem. 4 (2012) 863–864.
[4] M.D. Joestan, L.J. Schaad, Hydrogen Bonding, Marcel Dekker, New York, 1974.
[5] G.C. Pimentel, A.L. McClellan, The Hydrogen Bond, W.H. Freeman, San
Francisco, 1960.
[6] E. Arunan, G.R. Desiraju, R.A. Klein, J. Sadlej, S. Scheiner, I. Alkorta, Pure Appl.
Chem. 83 (2011) 1637.
[7] P. Politzer, P. Lane, M.C. Concha, Y. Ma, J.S. Murray, J. Mol. Model. 13 (2007)
305–311.
[8] G. Cavallo, P. Metrangolo, T. Pilati, G. Resnati, M. Sansotera, G. Terraneo, Chem.
Soc. Rev. 39 (2010) 3772–3783.
[9] K. Eskandari, H. Zariny, Chem. Phys. Lett. 492 (2010) 9–13.
[10] V.R. Hathwar, T.N. Guru Row, J. Phys. Chem. A 114 (2010) 13434–13441.
[11] F.T. Burling, B.M. Goldstein, J. Am. Chem. Soc. 114 (1992) 2313–2320.
[12] M. Iwaoka, S. Takemoto, S. Tomoda, J. Am. Chem. Soc. 124 (2002) 10613–
10620.
[13] M.V. Vener, A.N. Egorova, V.G. Tsirelson, Chem. Phys. Lett. 500 (2010) 272–
276.
[14] The origin and etymology of this term is suggested by Dutch chemist Anton
Eduard van Arkel in the early 1950s. Root of this term is the Greek word
‘‘pnikta” means ‘‘suffocated or strangled things” (where – gen derives from the
Greek word ‘‘genno” means ‘‘to produce”). However this term is subject of
controversy as par spelling and etymology is concerned. The recent
computational articles preferred to report pnicogen rather than pnicogen.
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
The detailed history of origin of the term can be found in G.S. Girolami, J.
Chem. Educ. 86 (2009) 1200–1201.
S. Scheiner, J. Chem. Phys. 134 (2011) 164313.
M. Solimannejad, M. Gharabaghi, S. Scheiner, J. Chem. Phys. 134 (2011)
024312.
S. Scheiner, J. Chem. Phys. 134 (2011) 094315.
S. Scheiner, J. Phys. Chem. A 115 (2011) 11202–11209.
U. Adhikari, S. Scheiner, Chem. Phys. Lett. 536 (2012) 30–33.
U. Adhikari, S. Scheiner, J. Phys. Chem. A 116 (2012) 3487–3497.
S. Scheiner, U. Adhikari, J. Phys. Chem. A 115 (2011) 11101–11110.
U. Adhikari, S. Scheiner, J. Chem. Phys. 135 (2011) 184306.
J.E. Del Bene, I. Alkorta, G.S. Sanz, J. Elguero, Chem. Phys. Lett. 512 (2011) 184–
187.
J.E. Del Bene, I. Alkorta, G.S. Sanz, J. Elguero, J. Phys. Chem. A 115 (2011)
13724–13731.
J.E. Del Bene, I. Alkorta, G.S. Sanz, J. Elguero, J. Phys. Chem. A 116 (2012) 3056–
3060.
J.E. Del Bene, I. Alkorta, J. Elguero, J. Phys. Chem. A 118 (2014) 2360–2366.
S. Scheiner, Chem. Phys. Lett. 514 (2011) 32–35.
X. An, R. Li, Q. Li, X. Liu, W. Li, J. Cheng, J. Mol. Model. 18 (2012) 4325–4332.
P. Politzer, S.J. Murray, J. Mol. Model. 14 (2008) 659–665.
J.E. Del Bene, I. Alkorta, J. Elguero, J. Phys. Chem. A 118 (2014) 3386–3392.
J.E. Del Bene, I. Alkorta, J. Elguero, J. Phys. Chem. A 117 (2013) 6893–6903.
I. Alkorta, J. Elguero, M. Solimannejad, J. Phys. Chem. A 118 (2014) 947–953.
G.S. Sanz, I. Alkorta, C. Trujillo, J. Elguero, ChemPhysChem 14 (2013) 1656–
1665.
F. Ma, A. Li, Comput. Theor. Chem. 1045 (2014) 78–85.
D. Setiawan, E. Kraka, D. Cramer, Chem. Phys. Lett. 614 (2014) 136–142.
D. Setiawan, E. Kraka, D. Cramer, J. Phys. Chem. A 119 (2014) 1642–1656.
W.E. Hill, L.M.S. Trivino, Inorg. Chem. 17 (1978) 2495–2498.
W.E. Hill, L.M.S. Trivino, Inorg. Chem. 18 (1979) 361–364.
R.S. Drago, N. Wong, D.C. Ferris, J. Am. Chem. Soc. 113 (1991) 1970–1977.
A.C. Marr, M. Nieuwenhuyzen, C.L. Pollock, G.C. Saunders, Organometallics 26
(2007) 2659–2671.
C. Ganesamoorthy, M.S. Balkrishna, J.T. Mague, H.M. Tuononen, Inorg. Chem.
47 (2008) 7035–7047.
S. Tschirschwitz, P. Lönnecke, E. Hey-Hawkins, Dalton Trans. (2007) 1377–
1382.
E.V. Avtomonov, K. Megges, S. Wocadlo, J. Lorberth, J. Organomet. Chem. 524
(1996) 253–261.
K.W. Klinkhammer, P. Pyykkö, Inorg. Chem. 34 (1995) 4134–4138.
M. Sundberg, R. Uggla, C. Vinas, F. Teixidor, S. Paavola, R. Kivekas, Inorg. Chem.
Commun. 10 (2007) 713–716.
S. Bauer, S. Tschirschwitz, P. Lönnecke, R. Franck, B. Kirchner, M.L. Clarke, E. HHawkins, Eur. J. Inorg. Chem. 2009 (2009) 2776–2788.
S. Zahn, R. Franck, E. H-Hawkins, B. Kirchner, Chem. Eur. J. 17 (2011) 6034–
6038.
P.R. Joshi, N. Ramanathan, K. Sundararajan, K. Sankaran, J. Phys. Chem. A 119
(2015) 3440–3451.
N. Ramanathan, K. Sundararajan, K. Sankaran, Phys. Chem. Chem. Phys. 18
(2016) 19350–19358.
M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery, Jr., J.E. Peralta, F. Ogliaro,
M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M.
Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C.
Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth,
P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman,
J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian, Inc., Wallingford CT, 2009.
S.F. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553–566.
L. Turi, J.J. Dannenberg, J. Phys. Chem. 97 (1993) 7899–7909.
R.F.W. Bader, Atoms in Molecules: A Quantum Theory, Clarendon Press,
Oxford, 1994.
R.G.A. Bone, R.F.W. Bader, J. Phys. Chem. 100 (1996) 10892–10911.
F.B. Koning, R.F.W. Bader, W.H. Tang, J. Comput. Chem. 96 (2000) 6796.
E.D. Glendening, J.E. Carpenter, F. Weinhold, NBO Version 3.1.
M. Hawkins, M.J. Almond, A.J. Downs, J. Phys. Chem. 89 (1985) 3326–3334.
S. Coussan, A. Loutellier, J.P. Perchard, S. Racine, Y. Bouteiller, J. Mol. Struct. 471
(1998) 37–47.
R.M. Bentwood, A.J. Barnes, W.J. O-Thomas, J. Mol. Spec. 84 (1980) 391–404.
A.D. Clauss, S.F. Nelsen, M. Ayoub, J.W. Moore, C.R. Landis, F. Weinhold, Chem.
Educ. Res. Pract. 15 (2014) 417–434.