1
Observation and rovibrational analysis of the intermolecular HCl libration band ν16 of HCN–HCl, DCN–HCl and H13CN–HCl

2
The high-resolution far-infrared absorption spectrum of the intermolecular HCl libration band ν16 (νB) of the gaseous molecular complex H12CN–HCl and the two isotopically substituted species H13CN–HCl and D12CN–HCl is recorded by means of static gas-phase Fourier transform far-infrared spectroscopy at 205 K using an electron storage ring source.

3
The rotational structure of the ν16 band has the typical appearance of a perpendicular type band of a linear polyatomic molecule.

4
The structure is analyzed using a standard semi-rigid linear molecule model including l-type doubling to yield the band origin ν0, together with values for the upper state rotational constant B′, the upper state quartic centrifugal distortion constant DJ and the value for the l-type doubling constant q6.

5
The values for the ground-state spectroscopic constants B″ and DJ for D12CN–H35Cl and H13CN–H35Cl are determined for the first time by ground state combination difference analyses.

6
A number of ν16 + ν17 − ν17 and ν16 + 2ν72 − 2ν72 hot bands are observed in the spectra and the sum of the anharmonicity constants X6,7 + g6,7 is estimated.

7
The observed decrease of the rotational constant B together with the simultaneous increase of the quartic centrifugal distortion constant DJ upon excitation of the HCl libration mode indicate that the hydrogen bond in the molecular complex is significantly destabilized upon intermolecular vibrational excitation.

8
The calculated harmonic force constants for the intermolecular hydrogen bond stretching vibration νσ for the ground state and the excited HCl libration state indicate that the excitation of the HCl libration mode destabilizes the intermolecular interaction between HCN and HCl by almost 20%.

9
The hydrogen bond is elongated by 0.030 Å upon excitation of the ν16 mode.

Introduction

10
The high-resolution infrared absorption spectrum of a bimolecular complex provides detailed information about the forces acting between the two molecular species defined by intermolecular potential energy surface (IPES), as well as how the intramolecular force fields within the subunits are modified by the incorporation into the bimolecular entity.

11
The accurate characterization of the IPES for a molecular complex requires the observation and analysis of a wide variety of gas phase spectra of fundamental bands, overtone bands, hot bands, combination bands, not only for the parent species but also for a sufficient number of isotopomers.

12
The low-wavenumber intermolecular vibrational degrees of freedom depend directly on the properties of the IPES and are therefore of special importance.

13
Infrared absorption spectra of hydrogen-bonded molecular complexes have been a subject of a long range of studies in the literature.

14
However, only a few numbers of these studies have dealt with the important intermolecular part of the vibrational spectrum for such molecular complexes.

15
The reason appears to be that there exist few suitable spectroscopic radiation sources in the far-infrared spectral region where the floppy intermolecular vibrations of molecular complexes are normally observed.

16
Measurements of rotationally resolved absorption spectra of intermolecular vibrations of gaseous molecular complexes require a spectral resolution close to the Doppler limit in the far-infrared spectral region.

17
However, the steep fall-off of energy with decreasing frequency for conventional black body radiators means that far-infrared spectroscopy generally is energy-limited at normal source temperatures in the sense that the resolution which can be attained is ultimately limited by the intensity of the source.

18
Far-infrared synchrotron radiation from an electron storage ring provides a far more brilliant source than conventional black body radiators, if the noise problems connected with electron beam oscillations can be solved.

19
A high-brightness source of far-infrared synchrotron radiation is therefore suitable for measurements in which high spectral resolution and high sensitivity is needed simultaneously.

20
This was demonstrated in our earlier publication on the intermolecular HCl libration band ν41of the linear hydrogen-bonded heterodimer OC–H35Cl.1,2

21
In the present work we investigate the high-resolution far-infrared absorption spectrum of the gaseous hydrogen-bonded heterodimer of HCN and HCl.

22
In contrast to the wealth of spectroscopic information available for the closely related prototype dimer HCN–HF (see ref. 3 and references therein) and to a lesser extent for (HCN)2 (see ref. 4, and references therein) the spectroscopic information about the molecular complex HCN–HCl reported in the literature is rather limited.

23
The first observation of the isolated molecular complex between HCN and HCl in the gas phase was reported by Legon et al5. who used Fourier transform microwave (MW) spectroscopy conducted in a Fabry–Perot cavity with a pulsed nozzle molecular source.

24
In this study, accurate ground-state spectroscopic constants were determined for the parent isotopic molecular complex and three additional isotopomers containing D, 15N and 37Cl.

25
Interpretation of the spectroscopic constants led to the conclusion that the most stable configuration of the molecular complex of HCN and HCl is collinear or near collinear at equilibrium, establishing the presence of a hydrogen bond to the N atom of the HCN molecule.

26
The vibrationally averaged distance between the centers of mass of the two monomers was determined to be 3.9380 Å.

27
Bender et al6. observed and assigned the intramolecular C–H stretching vibration band ν1 of HCN–H35Cl using a static gas long-path absorption cell at 199 K interfaced with a tunable single-frequency color-center laser spectrometer.

28
Bender et al. also managed to make tentative assignments of the three hot bands ν1 + ν17 − ν17, ν1 + 2ν27 − 2ν27 and ν1 + 3ν37 − 3ν37 where ν17 is the doubly degenerate low-wavenumber intermolecular bending vibration (libration of HCN).

29
The assignment of the ν1 band was later slightly modified and extended by Block and Miller.7

30
The observed small red shift of 2.5 cm−1 for the band origin of ν1 of HCN–H35Cl relative to the C–H stretching vibration band of monomer HCN indicates that the hydrogen bond is rather unaffected by excitation of the ν1 mode.

31
In a recent study8 the present authors reported the observation and rovibrational analysis of the intramolecular H–Cl stretching vibration band ν2 of HCN–H35Cl.

32
A red shift of 107 cm−1 of the ν2 band of HCN–H35Cl relative to the H–Cl stretching vibration band for monomeric H35Cl together with a positive value of ΔB upon vibrational excitation of ν2 showed that the hydrogen bond strengthens upon intramolecular vibrational excitation of the ν2 mode.

33
The intermolecular hydrogen bond vibrations of the HCN–HCl complex have not been observed directly.

34
Bender et al. estimated the band origin of the intermolecular HCN libration band ν17 to be 41(3) cm−1 from the value of the l-type doubling constant.6

35
The same authors estimate the harmonic frequency for the intermolecular hydrogen bond stretching vibration ν4 (νσ) to be 100(5) cm−1 from the ground-state spectroscopic constants using the modified pseudo diatomic (MPD) approximation developed by Millen.9

36
The harmonic frequency for the intermolecular HCl libration band ν16 (νB) is computed to be 311 cm−1 in the ab initio study by Araújo et al.10

37
The present study reports the direct observation and rovibrational analysis of the intermolecular HCl libration band of HCN–HCl, DCN–HCl and H13CN–HCl by means of static gas-phase Fourier transform far-infrared spectroscopy and a far-infrared synchrotron radiation source.

Experimental

38
The present experiments are carried out at the infrared beam-line at MAX-Lab at Lund University.

39
A temperature controlled 200-L static absorption cell made of stainless steel is interfaced with a Bruker IFS 120 HR Fourier transform spectrometer (FTS).

40
The absorption cell has a White type multipass mirror system.

41
The base length of the absorption cell is 2.85 m and the optics provide a total optical path length of ca. 91.2 m.

42
The absorption cell is equipped with two sets of 3.2 mm CsI windows.

43
The cell temperature is measured in the middle and at both ends of the inner cell with standard Pt100 resistance thermometers.

44
A computer emulated PID temperature controller regulates the current supplied to the three different resistive heaters welded to the outside of the inner cell and maintains the cell temperature of 205 ± 0.25 K during the experiment.

45
This temperature is close to the condensation point of HCN at the pressure used for the experiment.

46
HCN and DCN are prepared by dropwise addition of diluted H2SO and D2SO4 (99.8 atom% diluted in 99.3 atom% D2O) onto KCN in vacuo and condensation of the gaseous product.

47
The hydrogen cyanide samples are dried by vacuum distillation through a column containing the P2O5 drying agent.

48
Impurities of CO2 and (CN)2 are then removed from the samples by fractional distillation.

49
The partial pressures of HCN and natural isotopic HCl in the absorption cell are 2 Torr and 5 Torr, respectively.

50
The radiation source is synchrotron radiation from the electron storage ring MAX-I at MAX-lab.11,12

51
MAX-I is a 550 MeV electron storage ring with a 250 mA maximum ring current and a mean lifetime of 4 h.

52
The transfer optics are described elsewhere.11

53
The CsI exit window from the electron storage ring is mounted at Brewster’s angle to the horizontally polarized radiation.

54
The plane of polarization is then converted to vertical, since far-infrared beam splitters are more effective with this polarization.

55
The electron storage ring is a high-brightness source of broadband infrared radiation, covering the full far-infrared spectral region.

56
Synchrotron radiation is very close to a point source and is very suitable for high-resolution infrared spectroscopy.11–20

57
The radiation output from the electron storage ring relative to the radiation output from a conventional globar source is characterized in refs. 11 and 12.

58
The interferometer is equipped with a 6 μm multilayer beam splitter which works well over the entire spectral range 50−600 cm−1.

59
The detector is a liquid He cooled Si-bolometer operating at 1.7 K (Infrared Laboratories, Inc)..

60
The detector element is small, and allows us to use high scanning speeds.

61
The high scanning speeds make the bolometer less sensitive to motions of the electron beam.

62
A cold band pass optical filter (0–370 cm−1) is mounted in the bolometer in order to reduce the photonic noise level and detector nonlinearity effects in the final spectra.

63
The combination of optical filters at the bolometer, the 6 μm multilayer beam splitter and the CsI cell and exit windows gives a band pass in the region 190–370 cm−1.

64
The FTS instrument resolution (RES) is defined as RES = 0.9/(MOPD), where MOPD is the maximum optical path difference in the interferometer.

65
Sample interferograms are recorded with a resolution of 0.005 cm−1 (MOPD = 180 cm).

66
The recorded sample interferograms are transformed using Mertz phase correction and boxcar apodization.

67
A total scan time for each spectrum of ca. 10 h is achieved.

68
Throughout the measurements the FTS is evacuated by a vacuum pump reducing absorption by atmospheric gasses in the optical path inside the FTS.

69
Background interferograms of the evacuated absorption cell are recorded with a resolution of 0.08 cm−1.

70
These are transformed and interpolated onto a wavenumber grid matching that of the sample spectra using a zero-filling factor of 16.

71
This background resolution is appropriate in order to cancel out the most dominant interference fringes.

72
The resulting signal-to-noise ratio in the final absorbance spectra is about 20 : 1 for the most intense observed transitions.

73
The absolute wavenumber scale of the absorption spectrum is calibrated against published rotational lines of H2O21 observed in the spectrum.

74
The H2O lines appear in our spectra due to residual water vapor in the evacuated interferometer tank.

75
The accuracy of these water line positions is estimated to be 0.0002 cm−1.

76
Line positions from the absorbance spectra are generated using the Microcal Origin 7.0 software package (Microcal software, Inc)..

77
The precision of the line positions is estimated to be ten times better than the spectral resolution.

Results and analysis

HCN–H35Cl and HCN–H37Cl

78
The high-resolution far-infrared absorption spectrum of the mixture of natural isotopic HCl and HCN shows extensive rotational structures which are assigned to the intermolecular HCl libration band ν16 of the HCN–H35Cl and HCN–H37Cl heterodimers and hot band progressions for the parent isotopomer.

79
The appearance of the main band is typical for a linear molecule perpendicular type band (Σ–Π) with a rather large negative value for ΔB, as may be expected for the HCl libration motion of the HCN–HCl complex.

80
This observation together with the fact that the HCN–HCl system has high symmetry in the equilibrium configuration suggest that a standard semi-rigid linear molecule model is suitable for the present spectral analysis.

81
The central Q-branch of the band is located at 331.4 cm−1 and degrades widely at lower wavenumbers.

82
The Q-branch is accompanied by weaker P- and R-branches; the R-branch forms a band head near 335.4 cm−1 while the P-branch shows increasing J-spacing at lower wavenumbers.

83
The Q-branch and the P-branch are assigned to J = 65 and 53, respectively.

84
For the R-branch the band head appears for J = 55, 56.

85
This branch is therefore assigned up to J = 46.

86
Fig. 1 shows part of the J-assignment of the Q-branch and Fig. 2 shows part of the J-assignment and the observed band head for the R-branch.

87
The total number of rovibrational transitions assigned to the ν16 band of HCN–H35Cl is listed in Table S1 which is available as ESI.

88
A number of additional weak absorption lines is observed in the Q-branch region for HCN–H35Cl.

89
Fifteen of these weak lines are picked out and assigned to Q-branch transitions with J ∈ [27;50] for HCN–H37Cl by use of computer-assisted Loomis–Wood diagrams.22

90
It is not possible to identify any P- and R-branch lines for the 37Cl isotopomer and the Q-branch J-assignment cannot be verified by ground state combination differences (GSCDs).

91
The assigned Q-transitions in the ν16 band of HCN–H37Cl are listed in Table S2 as ESI.

92
In order to determine the values of the upper state spectroscopic constants B′ and DJ (v6 = l) for HCN–H35Cl the standard semi-rigid linear molecule model based on the following rovibrational energy expressions for the vibrational ground state, E″, and excited vibrational state, E′, is used:E″ = BJ(J + 1) − DJJ2(J + 1)2E′ = ν0 + BJ(J + 1) − DJJ2(J + 1)2 ± (1/2)q6J(J + 1)In the fit the ground-state constants B″and DJ are constrained to the MW values for HCN–H35Cl reported in .ref. 5

93
It appears that the effective ΔB value obtained from the Q-branch fit is larger than the ΔB value from the P, R-branch fit.

94
This indicates the presence of l-type doubling23 which is accounted for in eqn. (2).

95
The plus sign is used for the analysis of the Q-branch; the minus sign is used for the analysis of the P,R-branch system.

96
Including the l-type doubling constant q in the model the P, Q, and R-branch transitions are analyzed simultaneously.

97
A least-squares fit of a total number of 95 rovibrational transitions determines the values of ν0, ΔB, ΔDJ and q6 listed in Table 1.

98
The J-assignment of the Q-branch and the determination of the spectroscopic constants ν0, ΔB and ΔDJ for the HCN–H37Cl isotopomer are based on a fit of the 15 observed transitions to the conventional diatomic molecule expression given in eqn. (3):Q(J) = ν0 + ΔBJ(J + 1) − ΔDJJ2(J + 1)2Since the polynomial fit is based solely on observed Q-branch transitions the determined value of ΔB is affected by l-type doubling.

99
The actual value is therefore an effective rotational constant ΔB ± 0.5q6.

100
The J-assignment leading to the fit with minimum standard deviation is chosen.

101
The resulting spectroscopic constants are listed in Table 1.

102
In the analysis of the ν1 band for HCN–HCl by Bender et al6. the hot band progressions ν1 + ν17 − ν17, ν1 + 2ν27 − 2ν27 and ν1 + 3ν37 − 3ν37 from the low-wavenumber hydrogen-bond bending vibration ν17 were observed.6

103
A similar hot band pattern consisting of rovibrational structures centered at ca. 325, 320 and 315 cm−1 is observed in the present spectrum.

104
These features seem to consist of Q-branch lines only.

105
It is not possible to assign corresponding P- and R-branch lines which makes it difficult to establish the definitive J-assignments of the Q-branch lines.

106
From a tentative J-assignment of the 325 cm−1 feature, which is assigned to the ν16 + ν17 − ν17 hot band, the band origin is estimated to be 325.690(5) cm−1.

107
The following sum of anharmonicity constants is estimated based on the band origin of the main band located at 331.400 cm−1:X6,7 + g6,7 = −5.710(5) cm−1The spectral feature observed at ca. 320 cm−1 is assigned to the hot band ν16 + 2ν27 − 2ν27 and the band origin is estimated to be 320.155(10) cm−1 from a tentative J-assignment of the Q-branch.

108
The following sum of anharmonicity constants is estimated assuming as in ref. 6 that this hot band involves the l = 2 component:X6,7 + g6,7 = −5.622(5) cm−1which is in reasonable agreement with the result for ν16 + ν17 − ν17.

109
We predict the band origin of the ν16 + 3ν37 − 3ν37 (l = 3) hot band at 314.4 cm−1 from the average value of X6,7 + g6,7 of −5.67 cm−1.

110
It is difficult to locate the beginning of the Q-branch due to the weak intensity of this hot band.

111
However, an estimate of the band origin of 314.60(15) cm−1 seems reasonable, and this is taken as a further confirmation that our assignment of the hot band progressions is consistent.

DCN–H35Cl and DCN–H37Cl

112
The far-infrared absorption spectrum of the mixture of DCN and natural isotopic HCl shows rotational resolved structures which are almost identical to the structures observed for the HCN/HCl mixture and are assigned to the ν16 band of DCN–H35Cl and DCN–H37Cl and weak hot band progressions of DCN–H35Cl.

113
The central Q-branch of the band for DCN–H35Cl is located at 331.7 cm−1 and the R-branch forms a band head near 335.62 cm−1.

114
Fig. 3 shows part of the J-assignment of the Q-branch.

115
The Q-branch and the P-branch are assigned to J = 71 and 61, respectively.

116
For the R-branch the band head appears for J = 58, 59.

117
This branch is assigned up to J = 51.

118
The total number of rovibrational transitions assigned to the band for DCN–H35Cl is listed in Table S3, provided as ESI.

119
An additional number of weak absorption lines is observed in the Q-branch region for DCN–H35Cl.

120
A series of 31 weak lines is picked out and assigned to the Q-branch with J ∈ [12;70] for the DCN–H37Cl isotopomer.

121
The assigned transitions for DCN–H37Cl are listed in Table S4, provided as ESI.

122
The present P- and R-branch assignments allow us to perform a ground-state combination difference (GSCD) analysis for DCN–H35Cl.

123
A total number of 26 GSCDs is formed with J values in the range of 11 to 49 and the resulting values for the ground-state constants B″ and DJ of DCN–H35Cl are listed in Table 1.

124
A total number of 132 assigned P-, Q- and R-branch transitions are subsequently analyzed simultaneously using the standard semi-rigid linear molecule model based on eqns. (1) and (2) to obtain values of the band origin ν0 and the upper state rotational constant B′, the upper state quartic centrifugal distortion constant DJ and the l-type doubling constant q6.

125
In this fit, the ground-state spectroscopic constants are constrained to the values obtained from the GSCD analysis.

126
The resulting values of ν0, ΔB and ΔDJ and q6 are listed in Table 1.

127
The assignment and rovibrational analysis of the 31 observed Q-branch lines in the band for DCN–H37Cl are based on the conventional diatomic molecule expression given in eqn. (3).

128
The J-assignment leading to the fit with minimum standard deviation is chosen and the resulting spectroscopic constants are listed in Table 1.

129
A hot band pattern with rotational resolved features at ca. 325 and 320 cm−1 is observed in the absorption spectrum.

130
These features are assigned to the ν16 + ν17 − ν17 and ν16 + 2ν27 − 2ν27 hot bands for DCN–H35Cl and are assumed to consist of Q-branch lines.

131
Tentative J-assignments of these Q-branches yield the band origins of 326.50(10) cm−1 and 321.60(20) cm−1, respectively.

132
The observed shifts of 5.20(10) and 10.10(20) cm−1 relative to the main band give the following average value for the sum of anharmonicity constants:X6,7 + g6,7 = − 5.12(14)

H13CN–H35Cl and H13CN–H37Cl

133
The central Q-branch of the ν16 band of H13CN–H35Cl is located at 331.49 cm−1 and the R-branch forms a band head near 335.45 cm−1.

134
The Q-branch and the P-branch are assigned to J = 72 and 56, respectively.

135
For the R-branch the band head appears for J = 56, 57.

136
This branch is assigned up to J = 39.

137
The rovibrational transitions assigned to the band for H13CN–H 35Cl are listed in Table S5, provided as ESI.

138
A series of 27 weak lines is picked out and assigned to the Q-branch with J ∈ [19;70] for the ν16 band of H13CN–H37Cl.

139
The assigned transitions for H13CN–H37Cl are listed in Table S6 as ESI.

140
A GSCD analysis for H13CN–H35Cl based on 16 GSCDs with J in the range 6 to 39 determines the values for the ground-state constants B″ and DJ listed in Table 1.

141
An upper state analysis of 119 assigned P-, Q- and R-branch transitions is subsequently carried out with the ground-state spectroscopic constants constrained to the values obtained from the GSCD analysis.

142
The values of the spectroscopic constants for H13CN–H35Cl are listed in Table 1.

143
The assignment and rovibrational analysis of the 27 observed Q-branch lines in the ν16 band of H13CN–H37Cl is based on the conventional diatomic molecule expression given in eqn. (3).

144
The J-assignment leading to the fit with minimum standard deviation is chosen and the resulting spectroscopic constants are listed in Table 1.

145
A single Q-branch structure is observed in the spectrum around 326 cm−1.

146
The Q-branch is assigned to ν16 + ν17 − ν17 hot band for H13CN–H35Cl and the sum of anharmonicity constants X6,7 + g6,7 is estimated to be −5.42(14) cm−1.

Discussion

147
The appearance of the rotational structure of the HCl libration band ν16 and, in particular, the presence of l-type doubling observed in the far-infrared absorption spectra confirms the collinearity of the atoms in the molecular complex HCN–HCl.

148
The observed decrease of the rotational constant B together with the simultaneous increase of the quartic centrifugal distortion constant DJ upon excitation of the HCl libration mode indicate that the hydrogen bond in the molecular complex is significantly destabilized upon intermolecular vibrational excitation.

149
In terms of electrostatics the observed destabilization may be understood from the angular dependence of the dominant dipole–dipole interaction energy between HCN and HCl.

150
The dipole–dipole interaction between HCN and HCl can be described by a simple potential energy function:where μHCN and μHCl are the electric dipole moments of HCN and HCl, respectively, θHCl and θHCN are the libration angles (the angles between the two monomeric species and the line between the centers of mass), (ϕ1ϕ2) is the dihedral angle and r is the center of mass separation.

151
In the harmonic oscillator approximation based on the HCl libration band origin and the rotational constant for HCl the classical HCl bending amplitude θHCl increases from 14.4° in the ground state to 20.3° upon excitation of the HCl libration mode for H12CN–H35Cl.

152
The electric dipole moment of HCl is tilted away from the intermolecular symmetry axis in the molecular complex as the HCl subunit carries out the librational motion and the dipole–dipole interaction becomes more unfavorable owing to the enhanced non-linearity of the configuration.

153
In an attempt to quantify this destabilizing effect we use the relations between the determined centrifugal distortion parameters and the parameters of the analytical Morse potential function derived by Larsen et al.1

154
In this work the values of the Dunham parameters B, DJ and HJ were related to the Morse potential parameters r, a and De and used to characterize the intermolecular potential energy surface for the stretching of the hydrogen bond in the linear OC–H35Cl complex.

155
The center of mass separation r between HCN and HCl is related to the rotational constant according to:where B is the rotational constant of the molecular complex, IHCN and IHCl are the moments of inertia for the HCN and HCl subunits and μ is the reduced mass of the molecular complex.

156
The values for the moments of inertia of the subunits are taken from .refs. 24–26

157
The harmonic frequency ω4 for the intermolecular stretching vibration of the hydrogen bond is related to the rotational constant and quartic centrifugal distortion constant DJ by eqn. (6) assuming that the dominant contribution to the centrifugal distortion of the center of mass distance comes from the hydrogen bond stretching vibration:

158
In Table 2 the estimated center of mass distance and harmonic frequency for the hydrogen bond stretching vibration are listed for the ground state and the excited HCl libration state for H12CN–H35Cl, D12CN–H35Cl and H13CN–H35Cl.

159
The corresponding harmonic force constants k4 which are correlated directly to the dissociation energies of the molecular complexes are also listed in Table 2.

160
It appears that the hydrogen bond elongates by 0.030 Å upon excitation of the HCl libration mode for all three complexes.

161
The harmonic frequency for the hydrogen bond stretching vibration for H12CN–H35Cl decreases from ca. 100(5) to 90(5) cm−1 upon excitation of ν16 and the corresponding harmonic force constant thereby decreases by 19%.

162
These values are the same as those obtained using Millen’s modified pseudodiatomic approximation.9

163
The estimated force constants suggest that the excitation of the HCl libration mode destabilizes the intermolecular interaction between HCN and HCl by almost 20%.

164
This very pronounced destabilizing effect due to excitation of the high-wavenumber intermolecular bending mode νB of the molecular complex has also been observed for other hydrogen bonded systems like HCN–HF,27 OC–HCl1 and (H2O)3.28

165
In the strong HCN–HF dimer the hydrogen bond weakens by ca. 20% upon excitation of the HF libration mode and for the weaker OC–HCl dimer the hydrogen bond is destabilized by almost 30% upon excitation of the HCl libration mode.

166
Furthermore Keutsch et al28. showed that the lifetime of the doubly hydrogen bonded water trimer decreases by three orders of magnitude upon single excitation of the intermolecular out-of-plane libration mode.

167
Unfortunately, the Morse model described in ref. 1 cannot give values for the dissociation energy De of the molecular complex without values for the sextic centrifugal distortion constant HJ.

168
In an attempt to quantify the relative dissociation energy from the determined values for the ground-state spectroscopic constants Legon et al5. used the Lennard-Jones 6/12 potential function.

169
This potential energy function may not provide an entirely satisfactory description of the radial behavour of the intermolecular hydrogen bonding interaction but has the clear advantage that it allows the dissociation energy of the molecular complex to be estimated from the values of B and DJ alone.

170
Legon et al. estimate the dissociation energy for HCN–HCl to be 1199 cm−1 in the simple pseudodiatomic approximation.

171
Standard ab initio molecular orbital theory29 calculations are carried out using the program package GAUSSIAN 9830 in order to validate the empirical value for the dissociation energy.

172
Geometry optimization procedures are performed for HCN, HCl and HCN–HCl using CCSD(T)31 methodology in combination with correlation consistent polarized valence double-, triple- and quadruple zeta basis sets including polarization and diffuse functions.32

173
Single-point energy calculations on selected geometries are performed in order to calculate BSSE corrections being assessed using the counterpoise (CP) correction, as suggested by Boys and Bernadi.33

174
The calculated dissociation energies for the HCN–HCl dimer corrected for the BSSE are listed in Table 3.

175
In order to estimate the basis set limit we use the results of Dunning.34

176
In this work, the BSSE corrected values of De for (H2O)2 and (HF)2 were calculated using the CCSD(T) level of theory in combination with the aug-cc-pVnZ basis sets, n = 1–4.

177
The conclusion of this work was that the basis set convergence errors for (H2O)2 and (HF)2 for a specific basis set are very similar.

178
For (H2O)2, the CCSD(T)/aug-cc-pVTZ level of theory yields 94% of the value for the basis set limit.

179
Assuming this to be true also for the HCN–HCl dimer we predict a dissociation energy De of −1530 cm−1 in the complete basis set limit.

180
The computed value for the center of mass distance at equilibrium is 3.910 Å at the CCSD(T)/aug-cc-pVTZ level of theory which is close to the empirical value.

181
It is interesting to note that the band origins for the HCl librations of H13CN–H35Cl and DCN–H35Cl are slightly blueshifted relative to the band origin for the HCl libration of HCN–H35Cl.

182
In the harmonic oscillator approximation we expect that the HCl libration frequencies are lower for the heavier isotopomers.

183
We consider the following potential energy function for the two libration modes given in eqn. (9):where θHCl and θHCN are the libration angles and φ1φ2 is the dihedral angle.

184
We estimate the importance of the third term in the potential energy function involving the mixed harmonic force constant k67 assuming that the HCN libration and the coupling between the libration modes are dominated by the dipole–dipole interaction (k67 = 1/2k7).

185
The values of the force constant k7 are estimated from Bender et al.6

186
In this way we estimate how much the HCl libration frequencies are blueshifted due to the coupling with the HCN libration.

187
The estimated differences in these shifts upon isotopic substitution are much smaller than the observed blueshifts and it seems that the fourth term in the potential energy function involving the anharmonic force constant k6677 is more important for the couplings.

188
If we assume that this anharmonic term also determines the shifts of the observed ν16 + ν17 − ν17 hot bands, we can estimate the HCl libration shifts expected from the k6677 term alone.

189
We obtain 0.539 and 0.073 cm−1 for the frequency shifts upon isotopic substitution of D and 13C, respectively, compared to the observed values of 0.302 86(33) and 0.088 46(33) cm−1.

Conclusions

190
In the present study, we report the direct observation and rovibrational analysis of the intermolecular HCl libration bands ν16 for gaseous H12CN–H35Cl, H13CN–HCl and D12CN–H35Cl by means of static gas-phase Fourier transform far-infrared spectroscopy at 205 K using an infrared synchrotron radiation source.

191
A limited number of Q-branch lines are observed and assigned for the ν16 bands of the corresponding 37Cl species in natural abundance.

192
A number of ν16 + ν17 − ν17 and ν16 + 2ν27 − 2ν27 hot bands are observed in the spectra and sums of the anharmonicity constants X6,7 + g6,7 are estimated.

193
The relative HCl libration band origins seem to be determined from the anharmonic coupling to the HCN libration modes.

194
The observed decrease of the rotational constants B together with the simultaneous increase of the quartic centrifugal distortion constants DJ upon excitation of the HCl libration modes indicate that the hydrogen bond in the molecular complexes is significantly destabilized upon excitation of the intermolecular HCl libration modes.

195
The estimated force constants for the intermolecular hydrogen bond stretching vibration suggest that the excitation of the HCl libration mode destabilizes the intermolecular interaction between HCN and HCl by almost 20%.

196
The hydrogen bond elongates by 0.030 Å upon excitation of the ν16 mode for all three complexes.