Spectroscopic and ab initio investigation of 2,6-difluorophenylacetylene–amine complexes: coexistence of C–H⋯N and lone-pair⋯π complexes and intermolecular coulombic decay

Abstract

Binary complexes of 2,6-difluorophenylacetylene with methylamine, dimethylamine, trimethylamine and triethylamine were investigated using one colour resonant two photon ionization and infrared-optical double resonance spectroscopic techniques combined with high level ab initio calculations. All four amines form C_Ac–H⋯N hydrogen-bonded complexes. Additionally trimethylamine and triethylamine form complexes characterized by Lp⋯π interactions, due to the electron deficient nature of the phenyl ring of 2,6-difluorophenylacetylene. The Lp⋯π interacting structure of the 2,6-difluorophenylacetylene–trimethylamine complex is about 1.5 kJ mol⁻¹ higher in energy than the C_Ac–H⋯N hydrogen-bonded structure, which is the global minimum. Energy decomposition analysis indicates that the electrostatics and dispersion interactions favour the formation of C_Ac–H⋯N and Lp⋯π complexes, respectively. Interestingly the C_Ac–H⋯N hydrogen-bonded complex of 2,6-difluorophenylacetylene–triethylamine showed a smaller shift in the acetylenic C–H stretching frequency than the 2,6-difluorophenylacetylene–trimethylamine complex. The observed fragmentation of the binary complexes of 2,6-difluorophenylacetylene with the four amines following resonant two-photon ionization can be explained on the basis of the intermolecular coulombic decay process.

---

1. Introduction

One of the major challenges in hydrogen bonding is to predict, a priori, how the individual functional groups in a multifunctional molecule behave when they are made to interact with suitable hydrogen bonding partners. In the case of phenylacetylene (PHA) it has been shown that the intermolecular structures are highly dependent on the nature of the interacting molecule and even subtle changes, such as the addition of a methyl group, can change the intermolecular structure drastically.¹ Therefore PHA has been termed as a hydrogen bonding chameleon.² Interaction of ammonia and amines with PHA led to a plethora of structures. Two isomers of PHA–ammonia and PHA–methylamine complexes were observed and include a C_Ac–H⋯N hydrogen-bonded complex in each case.^3,4 On the other hand C_Ac–H⋯N hydrogen-bonded complexes were not observed for the binary complexes of PHA with several tertiary amines, and they largely interact via dispersion dominated C_Me–H⋯π interactions.⁵ Substitution of fluorine on the benzene ring of PHA can lead to the following changes: (a) increase in the acidity of both the acetylenic and aromatic C–H groups (especially in the ortho position), (b) lowering of the π electron density of the benzene ring, and (c) participation of fluorine in hydrogen bonding interactions.

An interesting question that arises is whether the C_Me–H⋯π complexes between PHA and tertiary amines persist with the substitution of fluorine on the benzene ring of PHA? Or does it lead to the formation of lone-pair⋯π (Lp⋯π) complexes? Several examples of Lp⋯π interaction have been reported in the literature, both in chemistry and biology.⁶ This interaction mostly pertains to the aromatic rings with reduced π electron density, such as in fluorine substituted aromatic rings.⁷ In an effort to understand the Lp⋯π interaction vis-à-vis the conventional hydrogen bonding, investigations on binary complexes of 2,6-difluorophenylacetylene (DFPHA) with four alkyl amines viz. methylamine (MA), dimethylamine (DMA), trimethylamine (TMA) and triethylamine (TEA) were carried out. The present investigation uses a double-barrel approach with the infrared-optical double-resonance spectroscopic technique along with ab initio calculations to understand the nature of the interaction of the amines with DFPHA.

Intermolecular coulombic decay (ICD) is an efficient and fast charge migration process following the ionization of the inner valence electrons.⁸ The ICD process can be described with respect to a hetero-dimer A⋯B interacting via intermolecular forces (such as hydrogen bonding) as the following. If the adiabatic ionization energy (AIE) of A is greater than that of B, and if the ionization of A is achieved, the electron ejection occurs from B instead of the initially excited (ionized) molecule A. The excess energy, which is the difference between the AIEs of A and B, can lead to doubly ionizing B or fragmentation of B and other reactive pathways. The ICD process, as described above, is much more efficient than the excitation energy transfer process as the acceptor state lies in the continuum. It has been recently reported that the ICD processes can be relevant in a number of chemical and biological systems interacting with a variety of radiation sources.⁹ In this work, DFPHA with the AIE of about 9.11 eV can be considered as fragment A, while the four amines, having AIEs in the order MA (9.04 eV) > DMA (8.32 eV) > TMA (7.82 eV) > TEA (7.5 eV), can be considered as fragment B.^10–12 The DFPHA⋯amine complexes offer an opportunity to investigate ICD processes systematically in non-covalently bonded systems as the energy gap between the two states involved in the ICD process can be tuned. This article also deals with the consequences of the ICD process following resonant two photon ionization of the DFPHA⋯amine complexes.

2. Methods

The details of the experimental setup have been described elsewhere.¹³ Spectra were acquired using a time-of-flight (TOF) mass spectrometer coupled with the supersonic jet expansion technique. Molecular complexes were created by passing helium buffer gas at 2 atm over 2,6-difluorophenylacetylene (Aldrich) and MA, DMA, TMA (50% aqueous solutions; Aldrich) and TEA (Aldrich) kept at room temperature and expanding through a 0.5 mm diameter pulsed nozzle (Series 9, Iota One; General Valve Corporation) operating at 10 Hz into a vacuum chamber at 10⁻⁶ Torr. The expanded molecular beam was crossed with the frequency doubled output of a dye laser (Narrow Scan GR; Radiant Dyes) operating with the Rhodamine-19 dye, pumped with second harmonic of a Nd:YAG laser (Surelite I-10; Continuum). LIF excitation spectra were recorded by monitoring the total fluorescence using a photo-multiplier tube (PMT – 9780SB + 1252-5F; Electron Tubes Limited) and a filter (WG-320) combination, while scanning the UV laser frequency. On the other hand, one-color resonant two-photon ionization (1C-R2PI) spectra were recorded by monitoring the appropriate mass ion signal using a two stage Wiley-McLaren TOF spectrometer and a channel electron multiplier (CEM-KBL-25RS; Sjuts Optotechnik) and a preamplifier (SR445A; Stanford Research Systems). The signal from the PMT/CEM was digitized by a digital storage oscilloscope (TDS-1012; Tektronix), which is interfaced to a personal computer using a data acquisition program written in LabView. The integrated signal intensity is plotted against the wavelength to get the LIF/1C-R2PI spectrum. The infrared spectra in the acetylenic C–H stretching region were obtained using the fluorescence-dip infrared (FDIR)/ion-dip infrared (IDIR) spectroscopic technique.¹⁴ The identification and assignment of a particular band in the 1C-R2PI spectrum among several other bands were carried out using IR-UV hole-burning spectroscopy. In our experiments the source of tunable IR light is an idler component of LiNbO₃ OPO (Custom IR OPO; Euroscan Instruments) pumped with an injection-seeded Nd:YAG laser (Brilliant-B; Quantel). The typical bandwidth of both UV and IR lasers is about 1 cm⁻¹ and the absolute frequency calibration is within ±2 cm⁻¹.

A detailed conformational search was followed by geometry optimization by the MP2 method using cc-pVDZ and cc-pVTZ basis sets (unless specified). In order to check the accuracy of the structures of the DFPHA–TMA complexes, they were also optimized by ωB97X-D,¹⁵ B2PLYP-D¹⁶ and B2PLYP-D3¹⁷ methods using the cc-pVTZ basis set. The vibrational frequency calculations were carried out using the MP2/cc-pVDZ optimized structures at the same level of theory. A scaling factor of 0.954 was chosen to match the experimental vibrational frequencies of the monomers and the same scaling factor was used for the complexes.

The stabilization energy was determined as the difference between the dimer energy and the sum of monomer energies. The CCSD(T)/CBS stabilization energies were determined as the sum of MP2/CBS energies and the CCSD(T) correction term according to eqn (1).

The method takes advantage of the fact that both the CCSD(T) and MP2 methods exhibit approximately the same basis set dependence.¹⁸ Two sets of CCSD(T)/CBS extrapolation were carried out: (a) with MP2/aug-cc-pVDZ (aVDZ), MP2/aug-cc-pVTZ (aVTZ) and CCSD(T)/aVDZ energies for the geometries obtained at the MP2/cc-pVDZ level, and (b) with MP2/cc-pVDZ, MP2/cc-pVTZ and CCSD(T)/cc-pVDZ energies for the geometries obtained at MP2/cc-pVTZ level of theory. The extrapolations were performed using the method proposed by Helgaker and co-workers,¹⁹ wherein the electron-correlation error is proportional to N⁻³.²⁰ In all the cases the BSSE corrected energies were used for the extrapolation.²¹ The interaction between DFPHA and TMA was analysed using atoms-in-molecules (AIM)²² and the simplest approach of Symmetry Adapted Perturbation Theory viz., SAPT0.²³

The equation of motion ionization potential coupled cluster with single and double excitation (EOM-IP-CCSD) method is very accurate and reliable for describing the closely spaced electronic states and ionization levels of the molecules,²⁴ and has already been successfully applied to calculate the ionization energies of various neutral and ionized species.²⁵ This was, therefore, used as our method of choice for the calculation of ionization energies. For the adiabatic ionization energy (AIE) calculations, geometries of all monomers and the binary complexes in their ground state were optimized at the MP2/cc-pVDZ level of theory. For the cationic state the MP2 calculations are unreliable due to spin contamination, and therefore the geometries were optimized using a long range corrected ωB97X functional with empirical dispersion correction (ωB97X-D) using the cc-pVDZ basis set.¹⁵ Furthermore, the frequency calculations for both the ground and cationic states at ωB97X-D/cc-pVDZ level of theory were used to account for the zero-point energy (ZPE) corrections. The optimized geometries of the cations were then subjected to EOM-IP-CCSD calculations to estimate the AIEs. The lowest AIEs were calculated as the difference between the EOM-IP-CCSD energy of the first ionized state at the cation geometries and the CCSD energy at the neutral ground state geometries. For the CCSD and EOM-IP-CCSD calculations cc-pVTZ and 6-311++G(d,p) basis sets were used. EOM-IP-CCSD/cc-pVTZ calculations use the resolution of identity (RI) approximation with the auxiliary RIMP2-cc-pVTZ basis set at the CCSD level.²⁶ The frozen natural orbital (FNO) approach was used in the calculation for the complexes (dimers) to largely reduce the computational cost using a natural orbital population threshold of 99.5%.²⁷ Geometry optimization and frequency calculations were carried out using the GAUSSIAN 09 suite of programs,²⁸ and all the calculations required for the IP calculations were carried out using Q-Chem-4.0.1 ab initio package.²⁹ All the structures and the vibrations were visualized using GaussView-5 and ChemCraft.^30,31

3. Results and discussion

A. Spectra and intermolecular structures

The 1C-R2PI spectra of various amine complexes of DFPHA are shown in Fig. 1 and Table 1 lists the observed S₁ ← S₀ band-origin excitation energies along with the shifts relative to the DFPHA monomer. Also shown in Fig. 1 (trace A) is the LIF spectrum of DFPHA and the strong band at 35 725 cm⁻¹ is assigned to the S₁ ← S₀ band-origin. The 1C-R2PI spectrum of DFPHA could not be recorded as its ionization potential is greater than double of the S₁ ← S₀ band-origin excitation energy of 71 450 cm⁻¹ (8.859 eV). Fig. 1B depicts the 1C-R2PI spectra of the DFPHA–MA complex recorded by monitoring the parent mass signal at 169 Da. This spectrum shows a single band which corresponds to the band-origin of S₁ ← S₀ excitation of the DFPHA–MA complex at 35 776 cm⁻¹ (A₁). In the case of the DFPHA–DMA complex, 1C-R2PI spectra could be recorded either by monitoring the parent (183 Da) or fragment (45 Da) mass signals. The two spectra (traces C and D of Fig. 1) are identical with band-origin of S₁ ← S₀ excitation at 35 769 cm⁻¹ (A₂). For the TMA and TEA complexes of DFPHA the parent mass signal could not be detected following the 1C-R2PI process and only fragment mass signals corresponding to TMA (59 Da) and TEA (101 Da) were observed, respectively. The 1C-R2PI spectra of DFPHA complexes with TMA and TEA were recorded by monitoring the fragment mass signals which are also shown in Fig. 1 (traces E and F, respectively). The 1C-R2PI spectra of the DFPHA–TMA and DFPHA–TEA complexes show several bands, which have been assigned to two isomers in each case, A₃ and A₄ for the TMA complex and A₅ and A₆ for the TEA complex, using IDIR and hole-burning spectroscopic methods, vide infra.

Fig. 1 (A) LIF spectrum of DFPHA. 1C-R2PI spectra of (B) DFPHA–MA@169 amu (C) DFPHA–DMA@183 amu (D) DFPHA–DMA@45 amu (E) DFPHA–TMA@59 amu (F) DFPHA–TEA@101 amu. In (A) the asterisk indicates the band-origin of the DFPHA monomer. Bands marked A₁–A₆ are due to various amine complexes.

Table 1 Observed S₁ ← S₀ electronic transitions and the acetylenic C–H stretching frequency (cm⁻¹) and their shifts for various DFPHA complexes

	E(S1 ← S0)	ΔE(S1 ← S0)	ν(C–H)	Δν(C–H)
DFPHA	35 725	—	3335	—
DFPHA–MA (A1)	35 776	51	3183	−152
DFPHA–DMA (A2)	35 769	44	3150	−185
DFPHA–TMA1 (A3)	35 765	40	3116	−219
DFPHA–TMA2 (A4)	35 652	−73	3331	−4
DFPHA–TEA1 (A5)	35 745	20	3136	−199
DFPHA–TEA2 (A6)	35 627	−98	3333	−2

---

In an attempt to identify the mode of interaction between the various amines and DFPHA infrared spectra in the acetylenic C–H stretching region were recorded and the results are presented in Fig. 2. Table 1 also lists the observed bands of the acetylenic C–H stretching frequency of the DFPHA moiety in various amine complexes and they shift relative to the DFPHA monomer. The FDIR spectrum of DFPHA (Fig. 2A) shows two strong bands at 3332 and 3340 cm⁻¹, accompanied by two weaker bands at 3347 and 3351 cm⁻¹. The appearance of multiple peaks in the acetylenic C–H stretching region of PHA and its substituted analogues has been attributed to Fermi resonance and other higher order anharmonic coupling terms.^1,32 With the assumption that the two intense bands are entirely due to Fermi resonance coupling, a simple two-state de-perturbation model places the unperturbed acetylenic C–H stretching vibration of DFPHA at 3335 cm⁻¹ with a coupling constant of 4 cm⁻¹.¹³ The IDIR spectra of DFPHA–MA (A₁; Fig. 2B) and DFPHA–DMA (A₂; Fig. 2C) complexes, recorded by monitoring the parent mass signals, show single bands at 3183 and 3150 cm⁻¹, respectively. The IDIR spectrum of the DFPHA–DMA complex was also recorded by monitoring the DMA fragment (45 Da) and was identical to the one shown in Fig. 2C (not shown). The interaction of MA and DMA with DFPHA leads to the lowering of the acetylenic C–H stretching vibration by 152 and 185 cm⁻¹, respectively. Therefore the A₁ and A₂ complexes can be unambiguously assigned to C_Ac–H⋯N hydrogen-bonded complexes.

Fig. 2 (A) FDIR spectrum of DFPHA. IDIR spectra of (B) DFPHA–MA, (C) DFPHA–DMA (D) DFPHA–TMA1, (E) DFPHA–TMA2, (F) DFPHA–TEA1 and (G) DFPHA–TEA2 complexes. Traces (B)–(G) correspond to IDIR spectra of complexes A₁–A₆ marked in Fig. 1.

The 1C-R2PI spectra of TMA and TEA complexes of DFPHA show multiple bands, which are assigned to two isomers in each case (see Fig. 1). The IDIR spectrum of the A₃ band, depicted in Fig. 2D, has a broad band centred at 3116 cm⁻¹ accompanied by a weak band at 3181 cm⁻¹. Similarly, the IDIR spectrum of the A₅ band (Fig. 2F) shows a strong band at 3136 cm⁻¹, along with a weak band at 3191 cm⁻¹. In these two spectra the bands at 3116 cm⁻¹ (Fig. 2D) and 3136 cm⁻¹ (Fig. 2F) can be assigned to the acetylenic C–H stretching vibration of the DFPHA–TMA and DFPHA–TEA complexes, which will correspond to low frequency shifts of 219 and 199 cm⁻¹, respectively. Once again, the A₃ and A₅ complexes can be assigned to C_Ac–H⋯N hydrogen-bonded structures straightforwardly. Furthermore, in these two spectra weak bands at 3181 cm⁻¹ (Fig. 2D) and 3191 cm⁻¹ (Fig. 2F) can probably be assigned to a combination band over the hydrogen-bonded C–H stretching vibration in both the complexes.

The IDIR spectra presented in Fig. 2 clearly indicate the formation of C_Ac–H⋯N hydrogen-bonded complexes of DFPHA with MA (A₁), DMA (A₂), TMA (A₃) and TEA (A₅), (calculated optimized structures shown in Fig. 3). One of the interesting observations that can be made from the IDIR spectra is the lowering of the acetylenic C–H stretching frequency for the complexes of MA, DMA, TMA and TEA, which is 152, 185, 219 and 199 cm⁻¹, respectively. Clearly, the lowering of the acetylenic C–H stretching frequency for the TMA complex (219 cm⁻¹) is greater than that of TEA complex (199 cm⁻¹). This observation is very surprising given the fact that TEA is a stronger base than TMA and is expected to form stronger hydrogen bonds, which in turn is expected to show a larger shift in the acetylenic C–H stretching frequency. In the case of hydrogen-bonded complexes of phenol the lowering of the O–H stretching frequency for the TMA and TEA complexes is 590 and 670 cm⁻¹, respectively.³³ MP2/aug-cc-pVDZ calculated structures reveal that the H⋯N hydrogen-bonding distances are in the order MA > DMA > TMA < TEA, which is in accordance with the observed lowering of the acetylenic C–H stretching frequency (see Fig. 3).

Fig. 3 Experimentally observed C_Ac–H⋯N hydrogen-bonded complexes of DFPHA with MA, DMA, TMA and TEA. The H⋯N hydrogen-bonded distances (Å) and acetylenic C–H stretching frequencies (cm⁻¹) are shown for structures optimized at MP2/cc-pVDZ level of theory.

The IDIR spectra of the other DFPHA–TMA (A₄ band; Fig. 2E) and DFPHA–TEA (A₆ band; Fig. 2G) complexes are very similar with a slightly broad band at around 3332 cm⁻¹ along with a shoulder at 3340 cm⁻¹. These two spectra are similar to that of the DFPHA monomer (Fig. 2A), albeit broadened by a few cm⁻¹, and indicate that the mode of interaction of TMA and TEA with DFPHA might be very similar. Furthermore the 1C-R2PI spectra of DFPHA complexes of TMA and TEA are very similar in appearance, which also suggests that the mode of interaction of TMA and TEA with DFPHA is very similar. The assignment of the different bands appearing in the 1C-R2PI spectra of the DFPHA complexes with TMA (Fig. 1E) and TEA (Fig. 1F) to various structural isomers was carried out by IR-UV hole-burning spectroscopy and the results are presented in Fig. 4. Fig. 4A shows the 1C-R2PI spectrum of the DFPHA–TMA complex, which was recorded by monitoring the fragment mass signal at 59 Da. The hole-burnt spectrum was recorded by pumping the acetylenic C–H stretching vibration of the A₄ complex at 3331 cm⁻¹ and scanning the UV laser energy and is shown in Fig. 4B. All the bands corresponding to complex A₄ show diminished intensities in the hole-burnt spectrum in comparison with the 1C-R2PI spectrum (indicated by arrows). The intensity of the band corresponding to complex A₃ remains unaltered. Fig. 4C is the difference spectrum generated by subtracting the 1C-R2PI spectrum from the hole-burnt spectrum. The bands corresponding to complex A₄ appear as dips in the difference spectrum. Traces D, E, and F of Fig. 4 depict the 1C-R2PI spectrum of the DFPHA–TEA complex, the IR-UV hole-burnt spectrum recorded by pumping the acetylenic C–H stretching vibration of complex A₆ at 3333⁻¹, and the difference spectrum, respectively. The presence of two distinct isomers for the DFPHA–TMA complex (A₃ and A₄) and the DFPHA–TEA complex (A₅ and A₆) can be inferred from the hole-burnt spectra.

Fig. 4 (A) 1C-R2PI excitation spectrum of the DFPHA–TMA complex. (B) IR-UV hole-burnt spectrum recorded by pumping acetylenic C–H stretching vibration of the A₄ complex at 3331 cm⁻¹. (C) Difference spectrum B–A. (D) 1C-R2PI excitation spectrum of the DFPHA–TEA complex. (E) IR-UV hole-burnt spectrum recorded by pumping acetylenic C–H stretching vibration of the A₆ complex at 3333 cm⁻¹. (F) Difference spectrum B–A.

Since the assignment of the DFPHA complexes of MA (A₁) and DMA (A₂) to C_Ac–H⋯N hydrogen-bonded structures is straightforward, calculations of these two complexes are not presented here. Furthermore, the A₃ and A₅ complexes of TMA and TEA, respectively, have also been assigned to the C_Ac–H⋯N hydrogen-bonded structures. The structures of the DFPHA–TMA complex optimized at the MP2/cc-pVTZ level are shown in Fig. 5. The intermolecular structures optimized at the MP2/cc-pVDZ and MP2/cc-pVTZ levels are almost identical with the exception of S8 structure, in which the orientation of the methyl group is slightly different. Furthermore, the coordinates of various structures of the DFPHA–TMA complex optimized at various levels of theory (MP2/cc-pVDZ, MP2/cc-pVTZ, ωB97X-D/cc-pVTZ, B2PLYP-D/cc-pVTZ and B2PLYP-D3/cc-pVTZ) are listed in Table S1 (ESI†). The effect of various theoretical methods on the intermolecular structures was found to be minimal. The atoms-in-molecules (AIM) analysis was used to characterize the interactions present in various structures of the DFPHA–TMA complex. Molecular graphs derived from the AIM analysis are shown in Fig. S1 (ESI†).

Fig. 5 Structures of DFPHA–TMA and DFPHA–TEA complexes optimized at MP2/cc-pVTZ level of theory. The stabilization energies (kJ mol⁻¹) calculated at CCSD(T)/CBS//MP2/cc-pVDZ for the TMA complex and MP2/aug-cc-pVDZ//MP2/cc-pVDZ for the TEA complex are shown in parentheses.

For the DFPHA–TMA complex, the first structure S1 is a C_Ac–H⋯N hydrogen-bonded structure. In S2 the aromatic C–H group in the para position to the acetylenic group interacts with TMA leading to the formation of a C_Ar–H⋯N hydrogen bond. Similarly, S3 is characterized by the C_Ar–H⋯N hydrogen bond with the aromatic C–H group in the meta position interacting with the nitrogen atom of TMA, along with two C_Me–H⋯F hydrogen bonds formed by two hydrogen atoms of two methyl groups interacting with the fluorine atom in the ortho position. Structures S4 and S5 can be considered as Lp⋯π complexes, wherein the primary interaction is between the nitrogen atom of the TMA and the electron deficient phenyl ring of DFPHA. Furthermore, the S4 structure also incorporates C_Me–H⋯π interaction. The structures S6 and S7 contain primarily C_Me–H⋯π contacts with the extended π electron density of DFPHA along with C_Me–H⋯F interaction in the S7 structure. Finally, it is very interesting to note that the S8 structure incorporates interaction between the carbon atom and the π electron density of DFPHA. Table 2 lists ZPE and BSSE corrected MP2/CBS and CCSD(T)/CBS stabilization energies extrapolated using aug-cc-pVNZ and cc-pVNZ (N = D,T) basis sets.

Table 2 ZPE and BSSE corrected CBS extrapolated stabilization energies (kJ mol⁻¹) and scaled vibrational frequencies and their shifts (cm⁻¹) calculated at MP2/cc-pVDZ level of theory

Structure	MP2/CBS^a	MP2/CBS^b	CCSD(T)/CBS^a^,^c	CCSD(T)/CBS^b	ν(C–H)	Δν(C–H)
a aug-cc-pVNZ (N = D, T) basis with MP2/cc-pVDZ optimized geometries. b cc-pVNZ (N = D, T) basis with MP2/cc-pVDZ optimized geometries. c Best estimate.
S1	−19.1	−16.9	−17.2	−14.8	3096	−239
S2	−13.7	−13.9	−12.9	−12.3	3335	+1
S3	−16.8	−15.2	−15.3	−13.5	3332	−2
S4	−22.1	−20.5	−15.7	−15.1	3332	−2
S5	−22.9	−21.7	−15.9	−16.3	3333	−1
S6	−15.1	−14.0	−11.4	−10.4	3332	−2
S7	−15.2	−14.3	−11.7	−10.9	3332	−2
S8	−10.2	−11.8	−7.9	−9.2	3334	−1

---

The IR spectrum of the DFPHA monomer (Fig. 2A) shows the presence of Fermi resonance coupling. In the case of PHA and its substituted analogues, the Fermi resonance coupling involves the acetylenic C–H stretching vibration and a combination of one quantum of C≡C stretch and two quanta of the acetylenic C–H out-of-plane bend and is localized on the acetylenic moiety, which implies that the Fermi resonance coupling is localized on the acetylenic moiety. As stated earlier the DFPHA–TMA A₃ complex was assigned to the C_Ac–H⋯N hydrogen-bonded structure, rather straightforwardly. The IR spectrum of the DFPHA–TMA A₄ complex (Fig. 2E) also shows Fermi resonance coupling, albeit broadening by a few cm⁻¹, which indicates that the interaction of TMA does not perturb the acetylenic moiety in DFPHA thereby retaining the Fermi resonance coupling to a large extent. From Fig. 5 it can be inferred that any of the seven S2–S8 structures fit this description. However, unlike the assignment of S1 structure to the A₃ complex, the experimental observations in the present case cannot unambiguously distinguish between any of the structures. The CCSD(T)/CBS extrapolation with aug-cc-pVNZ (N = D,T) is widely used,³⁴ therefore it is the best estimate for the stabilization energies of various structures of the DFPHA–TMA complex. This is further confirmed by the fact that the cc-pVTZ//cc-pVDZ and cc-pVTZ//cc-pVTZ energies are comparable (see Table S1, ESI†). Therefore based on the calculated stabilization energy considerations the observed DFPHA–TMA A₄ complex is assigned to S5 structure with Lp⋯π interaction. The energy decomposition analysis using the SAPT0 method was also carried out for various structures of the DFPHA–TMA complex and the results are listed in Table 3. The stabilization of the C_Ac–H⋯N complex (S1 structure assigned to the A₃ complex) is dominated by the electrostatic component, while the Lp⋯π complex (S5 structure assigned to the A₄ complex) is dominated by the dispersion component. Furthermore, it can also be noted from Table 3 that the electrostatics and dispersion components are overwhelming for the S1 and S5 structures among all the structures considered.

Table 3 SAPT0 interaction energy components (kJ mol⁻¹) for various structures of the DFPHA–TMA complex

Structure	E elec	E ind	E disp	E exch	E ch.tr.	E SAPT0
S1	−32.8	−8.2	−13.3	29.8	−1.5	−24.5
S2	−18.2	−5.0	−16.4	21.1	−0.8	−18.5
S3	−23.1	−6.4	−17.0	24.9	−1.1	−21.6
S4	−18.5	−6.2	−37.6	39.9	−0.6	−22.4
S5	−14.9	−4.8	−39.6	36.0	−0.5	−23.3
S6	−11.2	−4.6	−28.3	29.7	−0.6	−14.4
S7	−10.7	−3.6	−27.5	27.6	−0.5	−14.1
S8	−7.8	−3.6	−20.6	21.1	−0.2	−10.9

---

The TEA monomer can manifest in 6 different conformations, and the structures and relative energies of these conformations are shown in Fig. S2 (ESI†). Since the experiments were carried out under jet-cooled conditions only the lowest energy conformation was used in all subsequent calculations. The structures of the DFPHA–TEA complex were optimized at MP2/cc-pVDZ level of theory and are also shown in Fig. 5. Due to limitation in computational resources the stabilization energies of the DFPHA–TEA complex were calculated at the MP2/aug-cc-pVDZ level and were corrected for ZPE and BSSE. The stabilization energies of various structures of the DFPHA–TEA complex are shown in Fig. 5 along with the structures. In the case of the DFPHA–TEA complex, once again, the A₅ complex is assigned to C_Ac–H⋯N hydrogen-bonded structure (S9) based on the appearance of the IR spectrum (Fig. 2F). Among the calculated structures of the DFPHA–TEA complex S13 is the global minimum, which incorporates Lp⋯π and C_Me–H⋯F/C_Me–H⋯π interactions. In the S13 structure the acetylenic moiety is minimally perturbed, and therefore, it is a possible structure for the A₆ complex. It can be seen from Fig. 1 and 2 that the appearance of the 1C-R2PI and the IDIR spectra of the DFPHA–TEA are almost identical to that of the DFPHA–TMA A₄ complex. Therefore the structure of the DFPHA–TEA A₆ complex is assigned to the S13 structure, which is similar to the DFPHA–TMA A₄ complex, wherein the nitrogen atom of TEA interacts with the electron deficient phenyl ring.

All four amines (MA, DMA, TMA and TEA) form C_Ac–H⋯N hydrogen-bonded complexes with DFPHA. Furthermore, TMA and TEA also form complexes by Lp⋯π interaction. The formation of Lp⋯π complexes can be attributed to the reduced π electron density due to fluorine substitution on the phenyl ring and enhanced Lewis basicity of TMA and TEA relative to MA and DMA. The observation of both C_Ac–H⋯N and Lp⋯π complexes and ab initio calculations suggest that the stabilization energies are comparable and can coexist.

B. Intermolecular coulombic decay

The AIEs were calculated for DFPHA and the four amines MA, DMA, TMA and TEA using the EOM-IP-CCSD method and cc-pVTZ and 6-311++G(d,p) basis sets and the values are listed in Table 4. The basis set dependence is weak predicting the AIEs to be within 0.1 eV of the computed values. In order to set the confidence limit the AIEs of PHA and three singly fluorine substituted fluorophenylacetylenes viz. 2-fluorophenylacetylene (2FPHA), 3-fluorophenylacetylene (3FPHA) and 4-fluorophenylacetylene (4FPHA) were also calculated as the experimental AIEs of these four molecules are known with very high accuracy (also listed in Table 4).³⁵ Both basis sets underestimate the AIEs by about 0.1–0.2 eV. In the case of MA the experimental AIE measured by the mass analysed threshold ionization method is 9.04 eV,¹⁰ while the two set of calculations underestimate the AIE by about 0.2–0.3 eV.

Table 4 Calculated AIEs (eV) for various amines, phenylacetylenes and DFPHA–amine complexes. Experimental AIEs are also listed wherever available

	cc-pVTZ	6-311++G(d,p)	Expt
a Ref. 35. b Ref. 10. c Ref. 11. d Ref. 12.
DFPHA	9.11	9.14	>8.86
PHA	8.68	8.63	8.82^a
2FPHA	8.77	8.76	8.92^a
3FPHA	8.87	8.85	9.00^a
4FPHA	8.60	8.60	8.80^a
MA	8.81	8.70	9.04^b
DMA	8.10	7.99	8.32^c
TMA	7.64	7.53	7.82^c
TEA	7.43	7.33	7.50^d
DFPHA–MA (D1)	8.93	8.95	—
DFPHA–MA (D2)	8.65	8.68	—
DFPHA–MA (D3)	8.19	8.18	—
DFPHA–MA (D4)	8.15	8.10	—
DFPHA–DMA (CAc–H⋯N)	8.39	8.36	—
DFPHA–TMA (CAc–H⋯N)	8.25	8.22	—
DFPHA–TEA (CAc–H⋯N)	—	8.11	—

---

Fig. 6 shows the energy level schematic for the AIEs of DFPHA and all the four amines. The calculated AIE for DFPHA is 9.11 eV, which is higher than the energy available in the 1C-R2PI process (2 × 35 725 cm⁻¹ = 8.86 eV), therefore the 1C-R2PI spectrum of the DFPHA monomer could not be recorded. It is well known that the AIEs of the hydrogen-bonded complexes of substituted benzenes are usually lower than the corresponding monomers (substituted benzene). For example, the AIE of the phenol–water complex (7.94 eV) is lower than phenol (8.51 eV) which has been attributed to the difference in the hydrogen bond energies in the neutral and cationic ground states.³⁶ Extending this argument to the present case, it can be inferred that the AIEs of hydrogen-bonded complexes of DFPHA with MA, DMA, TMA and TEA would be lower than the DFPHA monomer and therefore could be ionized using the 1C-R2PI process. Fig. 7 shows four distinct structures optimized for the DFPHA–MA complex in the cationic state and Table 4 lists their AIEs. The AIE depends strongly on the intermolecular structure with C_Ac–H⋯N hydrogen-bonded structure (D1) having the highest AIE while the Lp⋯π structure (D4) having the lowest AIE. Table 4 also lists the AIEs of C_Ac–H⋯N hydrogen-bonded structures of DFPHA complexes with DMA, TMA and TEA. In the case of the DFPHA–MA complex the 1C-R2PI process can access any of the structures with AIE less than 8.86 eV leading to the formation of the DFPHA–MA complex cation. It must be pointed out here that the 1C-R2PI ensures that initially excitation is localized on the DFPHA moiety in all the amine complexes. As noted earlier no fragmentation was observed for the DFPHA–MA complex following the 1C-R2PI process since the appearance energy of the MA cation would be at least 9.04 eV (AIE), which is higher than the experimentally available energy of 8.86 eV. However in the case of DFPHA complexes with DMA, TMA and TEA complexes the experimentally available energy of 8.86 eV is higher than the AIE of respective amines.

Fig. 6 Schematic energy level diagram depicting AIEs of DFPHA, MA, DMA, TMA and TEA. The horizontal line at 8.86 eV represents experimentally available energy.

Fig. 7 Structures of DFPHA–MA complexes in the cationic state optimized at ωB97X-D/cc-pVDZ level of theory. The calculated AIEs (in eV) are shown in parentheses.

In the case of the DFPHA–DMA complex both the parent ion mass signal (183 Da) and the fragment ion mass signal were (45 Da) observed. Since the AIE of DMA is 8.33 eV, the appearance energy of the DMA cation from the DFPHA–DMA complex would be at least 8.33 eV, which is about 0.54 eV lower than available energy. On the other hand for the DFPHA complexes of TMA and TEA only the fragment ions of TMA (59 Da) and TEA (101 Da) were observed. The excess energy available over the appearance energy of TMA and TEA cations is 1.04 and 1.36 eV, respectively (see Fig. 6). For the DMA, TMA and TEA complexes the available excess energy increases and concurrently the density of states is also expected to increase both on account of energy and also the size of the molecule, which leads to extensive fragmentation in the case of TMA complexes following ICD. Therefore it can be inferred that the efficacy of the ICD process depends on the density of acceptor states.

4. Conclusions

In summary, we have investigated amine complexes of DFPHA using 1C-R2PI and IR-UV double resonance spectroscopic techniques. All the four amines form C_Ac–H⋯N hydrogen-bonded complexes, while TMA and TEA also form complexes which incorporate Lp⋯π interaction. One of the most interesting observations is that the shift in the C_Ac–H stretching frequency of the DFPHA moiety in the TEA complex is lower than the TMA complex. In the case of the DFPHA–TMA complex the Lp⋯π structure is about 1.5 kJ mol⁻¹ higher in energy than the C_Ac–H⋯N hydrogen-bonded structure, while the trend is reversed in the case of the DFPHA–TEA complex. These results point out that the Lp⋯π interactions can be competitive with normal hydrogen bonds. The SAPT0 energy decomposition analysis indicates that the C_Ac–H⋯N hydrogen bonding is electrostatics driven, while the Lp⋯π interaction is favored by dispersion interaction. Finally, the efficiency of fragmentation following an intermolecular coulombic decay process depends on the density of the acceptor states.

Author contributions

The problem was formulated by GNP. SIM and AD carried out experiments and some initial calculations in consultation with GNP. Most of the calculations were carried out by SS in consultation with DG. The results were interpreted jointly by GNP and DG along with SIM, AD and SS.

Acknowledgements

SIM thanks UGC and SS thanks IITB for their research fellowship. This material is based upon the work supported by Department of Science and Technology (Grant No. SR/S1/PC/23/2008) and Board of Research in Nuclear Science (Grant No. 2012/34/14) to GNP. High performance computing facility of IIT Bombay is gratefully acknowledged. DG would like to thank CSIR-National Chemical Laboratory for Startup grant and CSIR for funding from XIIth five year plan on Multiscale modelling (CSC0129).

References

1. (a) P. C. Singh and G. N. Patwari, J. Phys. Chem. A, 2008, 112, 4426–4431; (b) P. C. Singh and G. N. Patwari, J. Phys. Chem. A, 2008, 112, 5121–5125; (c) R. Sedlak, P. Hobza and G. N. Patwari, J. Phys. Chem. A, 2009, 113, 6620–6625.
2. S. Maity, M. Guin, P. C. Singh and G. N. Patwari, ChemPhysChem, 2011, 12, 26–46.
3. A. Dey, S. I. Mondal and G. N. Patwari, ChemPhysChem, 2013, 14, 746–753.
4. S. Maity, A. Dey, G. N. Patwari, S. Karthikeyan and K. S. Kim, J. Phys. Chem. A, 2010, 114, 11347–11352.
5. S. Maity, G. N. Patwari, S. Karthikeyan and K. S. Kim, Phys. Chem. Chem. Phys., 2010, 12, 6150–6156.
6. (a) Q. Gou, G. Feng, L. Evangelisti and W. Caminati, Angew. Chem., Int. Ed., 2008, 47, 3430–3434; (b) M. Egli and S. Sarkhel, Acc. Chem. Res., 2007, 40, 197–205; (c) G. J. Bartlett, R. W. Newberry, B. VanVeller, R. T. Raines and D. N. Woolfson, J. Am. Chem. Soc., 2013, 135, 18682–18688.
7. (a) J. P. Gallivan and D. A. Dougherty, Org. Lett., 1999, 1, 103–105; (b) Y. Danten, T. Tassaing and M. Besnard, J. Phys. Chem. A, 1999, 103, 3530–3534; (c) N. Mohan, C. H. Suresh, A. Kumar and S. R. Gadre, Phys. Chem. Chem. Phys., 2013, 15, 18401–18409; (d) S. K. Singh, S. Kumar and A. Das, Phys. Chem. Chem. Phys., 2014, 16, 8819–8827.
8. (a) L. S. Cederbaum, J. Zobeley and F. Tarantelli, Phys. Rev. Lett., 1997, 79, 4778–4781; (b) J. Zobeley, L. S. Cederbaum and F. Tarantelli, J. Chem. Phys., 1998, 108, 9737–9751.
9. (a) B. Boudaiffa, P. Cloutier, D. Hunting, M. A. Huels and L. Sanche, Science, 2000, 287, 1658–1660; (b) P. H. P. Harbach, M. Schneider, S. Faraji and A. Dreuw, J. Phys. Chem. Lett., 2013, 4, 943–949.
10. S. J. Baek, K.-W. Choi, Y. S. Choi and S. K. Kim, J. Chem. Phys., 2003, 118, 11040–11047.
11. A. Bodi, B. Sztaray and T. Baer, Phys. Chem. Chem. Phys., 2006, 8, 613–623.
12. J. E. Mathis and R. N. Compton, J. Chem. Phys., 1996, 104, 8341–8347.
13. P. C. Singh and G. N. Patwari, Curr. Sci., 2008, 95, 469.
14. (a) R. H. Page, Y. R. Shen and Y. T. Lee, J. Chem. Phys., 1988, 88, 5362–5376; (b) S. Tanabe, T. Ebata, M. Fujii and N. Mikami, Chem. Phys. Lett., 1993, 215, 347–352; (c) A. Fujii, G. N. Patwari, T. Ebata and N. Mikami, Int. J. Mass Spectrom., 2002, 220, 289–312.
15. J. D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 2008, 10, 6615–6620.
16. T. Schwabe and S. Grimme, Phys. Chem. Chem. Phys., 2007, 9, 3397–3406.
17. L. Goerigk and S. Grimme, J. Chem. Theory Comput., 2011, 7, 291–309.
18. P. Jurecka and P. Hobza, Chem. Phys. Lett., 2002, 365, 89–94.
19. A. Halkier, T. Helgaker, P. Jorgensen, W. Klopper, H. Koch, J. Olsen and A. K. Wilson, Chem. Phys. Lett., 1998, 286, 243–252.
20. S. K. Min, E. C. Lee, H. M. Lee, D. Y. Kim, S. Kim and K. S. Kim, J. Comput. Chem., 2008, 29, 1208–1221.
21. S. Scheiner, Comput. Theor. Chem., 2012, 998, 9–13.
22. R. F. Bader, Atoms in Molecules: A Quantum Theory, Oxford University Press, Oxford, UK, 1990.
23. B. Jeziorski, R. Moszynski and K. Szalewicz, Chem. Rev., 1994, 94, 1887–1930.
24. (a) J. F. Stanton and J. Gauss, J. Chem. Phys., 1994, 101, 8938–8944; (b) P. A. Pieniazek, S. A. Arnstein, S. E. Bradforth, A. I. Krylov and C. D. Sherrill, J. Chem. Phys., 2007, 127, 164110; (c) A. I. Krylov, Annu. Rev. Phys. Chem., 2008, 59, 433–462.
25. (a) T. Das and D. Ghosh, J. Phys. Chem. A, 2014, 118, 5323–5332; (b) D. Ghosh, A. Roy, R. Seidel, B. Winter, S. Bradforth and A. I. Krylov, J. Phys. Chem. B, 2012, 116, 7269–7280.
26. F. Weigend and M. Haser, Theor. Chem. Acc., 1997, 97, 331–340.
27. A. Landau, K. Khistyaev, S. Dolgikh and A. I. Krylov, J. Chem. Phys., 2010, 132, 014109.
28. M. J. Frisch, et al., Gaussian-09, Revision A.02, Gaussian, Inc., Wallingford, CT, 2009.
29. (a) Y. Shao, et al. , Phys. Chem. Chem. Phys., 2006, 8, 3172–3191; (b) A. I. Krylov and P. M. W. Gill, WIREs Comput. Mol. Sci., 2013, 3, 317–326; (c) Y. Shao, et al. , Mol. Phys., 2014, 9, 1–32.
30. T. Dennington, T. Keith and J. Millam, GaussView, Version 5, Semichem Inc., Shawnee Mission, KS, 2009.
31. http://www.chemcraftprog.com .
32. (a) G. W. King and S. P. So, J. Mol. Spectrosc., 1971, 37, 543–570; (b) J. A. Stearns and T. S. Zwier, J. Phys. Chem. A, 2003, 107, 10717.
33. (a) A. Iwasaki, A. Fujii, T. Ebata and N. Mikami, J. Phys. Chem., 1996, 100, 16053–16057; (b) A. Fujii, T. Ebata and N. Mikami, J. Phys. Chem. A, 2002, 106, 8554–8560.
34. (a) S. J. Grabowski, W. A. Sokalski and J. Leszczynski, J. Phys. Chem. A, 2005, 109, 4331–4341; (b) K. Berka, R. Laskowski, K. E. Riley, P. Hobza and J. Vondrasek, J. Chem. Theory Comput., 2009, 5, 982–992; (c) A. Bhattacherjee, Y. Matsuda, A. Fujii and S. Wategaonkar, ChemPhysChem, 2013, 14, 905–914.
35. V. S. Shivatare, A. Kundu, G. N. Patwari and W.-B. Tzeng, J. Phys. Chem. A, 2014, 118, 8277–8286.
36. O. Dopfer, G. Reiser, K. Muller-Dethlefs, E. W. Schlag and S. D. Colson, J. Chem. Phys., 1994, 101, 974–989.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4cp03445g

---