Phosphorous bonding in PCl3:H2O adducts: A matrix isolation infrared and ab initio computational studies

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 HAP


N 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 HAP


N 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 P


N > P


O > P


S > P


p [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 P


N and P


P interactions, extension of pnicogen bond to E


E0 bonding where interactions between other elements of group Va such as N


N, P


As 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 P


P 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 P


P interactions in anion-neutral molecule interactions, pentafluorophenyl substituted diphosphine, and aminotetra phospines [39–41] and P


N interaction in aminoalkylferrocenyldichlorophosphanes [42], As


As in cyclopentadienyl arsenic compounds and E


E interactions in dipnicogen dimer and their dichalcogen dimer [43,44]. Single crystal X-ray diffraction method was used by Sundberg et al. for analyzing P


P interaction in 1,2-dicarba-closo-dodecaboranes [45]. The indications of nonbonding P


P 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 P


p 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 P1


Cl8, P1


Cl6 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 H5


Cl4 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 (P


Cl) 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. Macia´, M.V. Herna´ndez, J. Oro´, 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.