Using Surface Science Techniques to Investigate the Interaction of Acetonitrile with Dust Grain Analogue Surfaces

Although pure acetonitrile is not found in astrophysical environments, studying pure ices is useful in order to determine the detailed physical and chemical behaviour of the ice and to provide a comparison with the behaviour in the presence of water ice. Figure 2 shows TPD data recorded following the adsorption of increasing amounts of acetonitrile on HOPG at 29 K. At the lowest exposures investigated, 0.5–2 L_m (Figure 2(a)), a decreasing peak temperature with increasing exposure is observed. This can be assigned to initially repulsive interactions between acetonitrile molecules on the HOPG surface. This effect has been seen for ultra-thin acetonitrile films adsorbed on SiO₂ (84) and for other molecules such as benzene on HOPG (38, 101). Behaviour of this type could also be explained by a distribution of binding sites with different adsorption energies, however this is unlikely in this case since HOPG is a very uniform surface.

For acetonitrile exposures between 3 L_m and 7 L_m (Figure 2(a)) the TPD data show an approximately constant peak temperature with increasing exposure, which is characteristic of first-order desorption of molecular acetonitrile, and indicates that monolayer desorption occurs over this range of exposures. The low temperature desorption, at approximately 127 K, and the approximately constant peak temperature indicate that acetonitrile forms a physisorbed monolayer on the HOPG surface at the intermediate exposures shown in Figure 2(a). The spectra seen in Figure 2(a) are in contrast to those seen for thin acetonitrile films adsorbed on SiO₂ (8, 84), where the acetonitrile peak temperature decreases with increasing coverage for all monolayer exposures. This difference is likely a consequence of the more uniform HOPG surface compared to the heterogenous SiO₂ surface studied by Abdulgalil et al. (8, 84).

Fig. 2.

TPD data for increasing exposures of acetonitrile on HOPG from 7–100 L_m can be seen in Figure 2(b). The data clearly show a single desorption feature which increases in temperature with increasing exposure. Traces in Figure 2(b) also show shared leading edges. These observations are indicative of the desorption of acetonitrile multilayers following zero-order kinetics. This is in good agreement with previous studies of multilayer acetonitrile adsorbed on a range of surfaces (7, 31, 82–84, 95).

Kinetic analysis of the TPD data shown in Figure 2 was performed using methods described previously (33). TPD data can be described by the Polanyi-Wigner equation (Equation (i)) where r_des is the rate of desorption, θ is the coverage, t is the time, A is the pre-exponential factor, n is the order of desorption, E_des is the activation energy for desorption, R is the gas constant and T is the temperature of the substrate.

Analysis using the Polanyi-Wigner equation enables us to obtain n, E_des and A. The order of desorption, n, can be obtained from rearrangement of Equation (i) as shown previously (33). This gives the order of desorption from a plot of ln[I(T)] (where I(T) is the recorded mass spectrometer signal at temperature T) as a function of ln[θ_rel] (where θ_rel is the relative coverage, given by the integrated area under the TPD curves) at a fixed desorption temperature. Figures 3(a) and 3(b) show examples of such plots for acetonitrile at fixed desorption temperatures of 125 K (monolayer order) and 131 K (multilayer order) respectively. Plots such as this are produced for a range of temperatures using data on the leading edge of the TPD curves; so called leading edge analysis (33). Average values for the order of the monolayer and multilayer desorption can then be obtained from these plots.

Fig. 3.

The order of desorption for monolayer acetonitrile was found to be 0.89 ± 0.05. The order for multilayer acetonitrile desorption was determined to be 0.08 ± 0.07. These values are as expected for monolayer and multilayer desorption of physisorbed species (33). These desorption orders were obtained for exposures of acetonitrile from 3–100 L_m. The lowest exposures (0.5–2 L_m) were omitted from the analysis of the order of desorption as they show behaviour associated with repulsive interactions and hence do not give meaningful data for the desorption order.

Once the desorption order has been determined, a plot of ln[I(T)]–nln[θ_rel] versus 1/T can be used to determine E_des (33). The gradient of this plot is equal to −E_des/R. An example of this plot for a 100 L_m exposure of acetonitrile is shown in Figure 3(c). Desorption energies can be determined for each exposure of acetonitrile to show the variation in desorption energy as a function of exposure, as seen in Figure 4.

Fig. 4.

The inset to Figure 4 clearly shows a decreasing desorption energy with increasing exposure for the very lowest exposures of acetonitrile on HOPG. Following this, the desorption energy increases to the multilayer desorption energy value. A decrease in desorption energy with increasing exposure was also seen by Abdulgalil et al. for acetonitrile adsorption on SiO₂ (84). The average value of the desorption energy of multilayer acetonitrile is determined to be 43.8 ± 1.7 kJ mol^−1. For exposures less than 5 L_m, assigned to monolayer desorption, the desorption energy ranges from 28.8–39.2 kJ mol⁻¹. The monolayer energy determined here is somewhat lower than that determined by Abdulgalil et al. (84) (35–50 kJmol⁻¹) and by Bertin et al. (95) (44.4 ± 2.8 kJ mol⁻¹) for monolayer desorption from SiO₂ and α-quartz respectively. This is most likely due to the different binding energy of acetonitrile with the carbonaceous HOPG surface compared to that on a silicate surface. The multilayer desorption energy determined here is in reasonable agreement with that determined by Abdulgalil et al. (84) (38.2 ± 1 kJ mol⁻¹).

Once the desorption order and activation energy for desorption have been determined, it is possible to obtain values for the pre-exponential factor for desorption, A (33). The average pre-exponential factor for the desorption of monolayer acetonitrile from HOPG was determined to be 3.4 × 10^{15 ± 0.5} s⁻¹. This value is as expected for the desorption of a monolayer species from a surface. The value obtained for multilayer desorption of acetonitrile from HOPG is 1.4 × 10^{32 ± 0.3} molecules m⁻² s⁻¹. This value is in good agreement with that previously obtained for the desorption of acetonitrile from SiO₂ (84).

RAIR spectra for acetonitrile adsorbed on HOPG at 29 K are shown in Figure 5. The top spectrum shows that recorded upon adsorption at base temperature. The spectrum shown is for the adsorption of 80 L_m of acetonitrile on HOPG.

Fig. 5.

Adsorption from 0.5–80 L_m is not shown here, but shows that bands increase in intensity with increasing amount of acetonitrile on the surface. No frequency shifts are observed for increasing amounts of acetonitrile. This indicates that physisorption is taking place, as already shown by the TPD spectra in Figure 2. The assignments of the bands shown in Figure 5 are given in Table I and are made by comparison with the literature. Bands in the high wavenumber region are assigned to the symmetric and antisymmetric methyl group stretching modes at 3001 cm⁻¹ and 2941 cm⁻¹ respectively. The C≡N stretching mode is observed at 2253 cm⁻¹ and is the most intense band observed in the spectrum. In the lower wavenumber region, a band at 1373 cm⁻¹ is assigned to the symmetric methyl group deformation mode, while bands at 1412 cm⁻¹ and 1448 cm⁻¹ are also assigned to methyl group deformation modes. The band at 1045 cm⁻¹ is assigned to the methyl group rocking mode (not shown) and a further mode at 914 cm⁻¹ (also not shown) is assigned to the C–C stretching mode.

Figure 5 also shows the results of annealing the adlayer of acetonitrile adsorbed on HOPG at 29 K. The acetonitrile was annealed in 10 K increments up to its desorption temperature. No changes were observed in the spectra until 100 K and hence these spectra have been omitted from Figure 5. Following annealing to 100 K, there is an increase in intensity of the C≡N stretch and a small decrease in the wavenumber of this band to 2251 cm⁻¹. This change is accompanied by intensity changes and frequency shifts for other bands in the spectrum. For example, the methyl group stretching mode at 2941 cm⁻¹ moves downwards to 2939 cm⁻¹ following annealing to 100 K; the band at 1448 cm⁻¹ sharpens and increases in wavenumber to 1454 cm⁻¹; and the band at 1045 cm⁻¹ decreases in wavenumber to 1038 cm⁻¹. Further annealing to 110 K shows an increase in intensity of all of the bands in the spectrum, along with a splitting of the bands at 1410 cm⁻¹ and 1373 cm⁻¹. No further changes in the spectrum are observed and all modes have disappeared from the spectrum by 130 K. The lack of bands in the spectrum following annealing to 130 K is evidence for desorption of acetonitrile by this temperature. This temperature is slightly lower than that observed in the TPD spectra, seen in Figure 2, due to the different nature of heating in both cases.

The observed changes in the RAIR spectra following annealing to 110 K can be assigned to the crystallisation of acetonitrile that occurs upon annealing. Acetonitrile crystallisation has been studied in detail by Hudson (92) and the spectra shown in Figure 5 are in good agreement with the high temperature crystalline phase described by Hudson. The crystallisation of solid phase organic species has been observed previously for a range of molecules (42, 92, 100, 102). Recording infrared spectra for ices adsorbed on model grain surfaces helps to allow the identification and assignment of spectra recorded in astrophysical environments. For example, the presence of acetonitrile has been observed in Titan’s atmosphere and hence accurate infrared spectra can help with assignment of observational data (7, 67, 92, 93).

As described in the introduction, astrophysical ices contain large amounts of water ice, with the exact amount and ice phase depending on the region of space. Given that astrophysical ices are primarily composed of water ice, it is necessary to investigate the interaction of astrophysically relevant molecules with various configurations of water ice as shown in Figure 1(b).

Figure 6 shows TPD data resulting from the adsorption of thin films of acetonitrile in the presence of water ice in different configurations. It is clear from Figure 6 that the presence of ASW has a significant effect on the desorption behaviour of acetonitrile compared to pure acetonitrile ices (Figure 2). Even at the lowest exposures, three features are observed in the TPD data.

Fig. 6.

The highest temperature species at approximately 158 K can be assigned to the co-desorption of acetonitrile with the bulk water ice. This assignment is confirmed by the fact that this peak has the same desorption temperature as the main water desorption. This observation of co-desorption has also been seen for other molecules (100, 102, 103) and occurs as the molecules become trapped in the bulk of the water ice when the ASW-CI phase transition occurs (31, 103).

The sharp acetonitrile desorption feature observed at approximately 146 K is assigned to volcano desorption of acetonitrile that becomes trapped in the pores of the water ice as the ASW to CI phase transition occurs at ~145 K. The observation of the volcano and co-desorption peaks provides evidence that the acetonitrile likely diffuses into the water ice surface, prior to the ASW to CI phase transition occurring. Volcano and co-desorption have previously been seen for a number of molecules, both smaller volatiles (31, 33, 53, 104) and larger organic species (32, 50, 99, 100), adsorbed on and in ASW. The lowest temperature acetonitrile desorption peak seen in Figure 6(c) can be assigned to the desorption of acetonitrile directly from the amorphous water ice surface.

Significant desorption is only seen directly from the ASW surface once the pores of the ASW are saturated and hence this lowest temperature peak only becomes significant for the higher exposures of acetonitrile on ASW. Further increasing the amount of acetonitrile on the ASW leads to the formation of multilayer acetonitrile with desorption behaviour following that of the pure acetonitrile shown in Figure 2.

In contrast to the behaviour of acetonitrile on ASW (Figure 6(c)), adsorption on CI shows simpler desorption behaviour (Figure 6(b)). There are no desorption features that can be assigned to volcano or co-desorption, as expected since the water ice in Figure 6(b) is grown in the crystalline phase. Instead, desorption directly from the CI surface is the only desorption peak observed in the spectrum. Figure 7 shows a comparison between the desorption of acetonitrile from the bare HOPG surface (Figure 7(c) and7(d)) and from the CI surface (Figure 7(a) and 7(b)). There is clearly a difference in the desorption behaviour of acetonitrile bonded to CI compared to acetonitrile bonded to HOPG. In particular, for the lowest exposures, equivalent amounts of acetonitrile desorb at a higher temperature from CI than from HOPG. For example 2 L_m of acetonitrile desorbs from HOPG at a temperature of ~126 K compared to a temperature of ~139 K from CI. In addition, the peak profiles for acetonitrile desorbing from CI are considerably broader compared to those seen for desorption directly from HOPG.

Fig. 7.

The broader peak profile for acetonitrile desorbing from CI suggests that there is a range of desorption energies and a distribution of binding sites for the acetonitrile on the CI surface. These observations are in agreement with previous investigations of acetonitrile on CI by Bertin et al. (95) which have also been assigned to the observation of a range of desorption energies for acetonitrile on the CI surface. As seen in Figure 7(a), as the amount of acetonitrile on the CI surface increases above 3 L_m there is evidence for an additional peak growing into the TPD at lower temperature. Figure 7(b) shows that this feature can be assigned to the growth of multilayers of acetonitrile on the CI surface as also seen in Figure 7(d) for acetonitrile adsorbed directly on HOPG.

Leading edge analysis can also be applied to the TPD curves shown in Figures 6(b) and 7(a) to provide an estimate of the desorption energy of acetonitrile from the CI surface. Analysis of the TPD curves for 1 L_m and 2 L_m of acetonitrile adsorbed on CI, with an assumed desorption order of one, give desorption energies of 36.8 ± 0.8 kJ mol⁻¹ and 36.6 ± 0.7 kJ mol⁻¹ respectively. Note that, as already discussed, the shape of the TPD curves for monolayer acetonitrile desorbing from CI indicates that there are a range of desorption energies dictating the desorption. These desorption energy values hence give an estimate of the increase in the desorption energy from the CI surface, compared to the HOPG surface, and do not give the full range of desorption energies across all binding sites. Analysis of multilayer TPD data for acetonitrile adsorbed on CI give the same desorption energies as already determined for pure acetonitrile, as expected. For exposures ≥3 L_m, Figure 7(b) shows that the additional low temperature feature assigned to the growth of multilayers is already growing into the TPD curves. Hence it is not appropriate to use a desorption order of one to obtain desorption energies for these spectra. An attempt was also made to determine desorption energies for acetonitrile monolayers desorbing from the ASW surface. However, the complexity of the data are such that it was not possible to determine reliable energies of desorption for this system. Nonetheless, a comparison can be made with the data for monolayer acetonitrile on CI (Figure 7(a) and 6(c)), which show that the monolayer on both water ice surfaces shows similar features. Hence, it is expected that the monolayer acetonitrile on ASW desorbs with a wide range of energies, and from a broad range of binding sites. As for desorption from CI, the desorption energy from the ASW surface is higher than from HOPG, and is likely to be around ~37 kJ mol⁻¹, as determined from leading edge analysis of the data resulting from desorption of acetonitrile from CI.

The desorption energy values for 1 L_m and 2 L_m of acetonitrile on CI are significantly higher when compared to equivalent exposures adsorbed directly on HOPG (shown in Figure 4) and correspond to a stronger binding energy of the acetonitrile on the CI surface. This can be assigned to an interaction between the C≡N of the acetonitrile and the water ice, as confirmed by RAIR spectra (shown later). This observation of a higher binding energy on CI is in agreement with previous observations (95).

Figure 6(a) also shows the desorption of acetonitrile from mixed ices, formed by co-deposition of acetonitrile and ASW. It is clear from Figure 6(a) that the behaviour of the mixtures is very similar to that of acetonitrile adsorbed on ASW (Figure 6(c)). This behaviour of acetonitrile in the presence of water ice is consistent with that observed for other organic molecules that have weaker interactions (when compared with strongly hydrogen-bonding species such as methanol) with water ice such as methyl formate (103) and ethyl formate (99). This behaviour occurs as the molecule forms a weak interaction with the polar water ice surface, as shown by the different desorption energy when compared to that on HOPG.

Figure 6 also shows that the acetonitrile has an effect on the temperature of the ASW-CI phase transition. Figure 6(c) shows the volcano desorption peak decreasing in temperature with increasing acetonitrile exposure on the ASW surface. The same effect is also seen for mixed ices (Figure 6(a)) where the ice containing a higher percentage of acetonitrile shows a lower temperature volcano desorption. This effect has been reported previously for species that strongly hydrogen bond to water such as methanol (105). It has also been seen for more weakly bonded species that do still interact with water, such as ethyl formate adsorbed in the presence of water ice (99), and provides evidence for an interaction between acetonitrile and water.

Further evidence for an interaction between acetonitrile and water ice is shown in Figure 8, which shows the C≡N stretch of acetonitrile in different ice configurations at 29 K. From Figure 8 it is clear that in the presence of ASW, either in a layered or mixed ice configuration, the C≡N stretch has a different line shape when compared to that observed for pure acetonitrile on HOPG. For the mixed ice, the C≡N stretch shifts to 2266 cm⁻¹, compared to 2253 cm⁻¹ for the pure ice, showing evidence of a direct interaction between the acetonitrile and water ice. In contrast, the C≡N stretch for acetonitrile adsorbed on top of ASW shows features for both the pure acetonitrile (2253 cm⁻¹) and for the acetonitrile bonded to water ice (2266 cm⁻¹). Acetonitrile bonded to CI has the same vibrational frequency as that seen for pure acetonitrile. This is unsurprising since the spectra shown in Figure 8 are for the adsorption of 20 L_m of acetonitrile in different ice configurations. At this exposure, TPD spectra shown in Figure 7 also show that acetonitrile on CI and on HOPG exhibit similar behaviour. The sensitivity of our RAIRS experiment is less than that of the TPD experiment and hence it is not possible to record RAIR spectra for the very lowest exposures of acetonitrile on the surface.

Fig. 8.

Figure 9 shows the results of annealing an acetonitrile adlayer adsorbed on top of ASW (Figure 9(a)) and a mixed acetonitrile water layer (Figure 9(b)), focusing on the C≡N region of the spectrum. This is the only spectral region that shows appreciable differences when compared with spectra that result from annealing a pure acetonitrile ice (Figure 5). For acetonitrile adsorbed on top of ASW, annealing the ice leads to the appearance of a high wavenumber shoulder at 2266 cm⁻¹ (Figure 9(a)), initially becoming more prominent following annealing to 110 K. The main C≡N stretch at 2253 cm⁻¹ also sharpens and shifts down in wavenumber slightly to 2251 cm⁻¹. This effect was also seen for the annealing of pure acetonitrile ice (Figure 5) and can be assigned to the crystallisation of the acetonitrile layer adsorbed on top of the ASW. Subsequent annealing of this system to 130 K leads to a complete change in the RAIR spectrum. The sharp feature at 2251 cm⁻¹ disappears from the spectrum and the only band that remains is a very broad feature centred at 2266 cm⁻¹.

Fig. 9.

Comparing the spectrum at 130 K in Figure 9(a) with the spectrum at 130 K for the annealing of pure acetonitrile (Figure 5) shows that this peak can be assigned to acetonitrile trapped within water ice. Annealing of pure acetonitrile ice to 130 K (Figure 5) leads to complete desorption and no peaks are observed in the RAIR spectrum. However, TPD data for acetonitrile adsorbed on top of ASW show the presence of volcano and co-desorption peaks due to acetonitrile trapped within the ASW structure. Hence, the infrared band seen at 130 K following the annealing of the layered ice can be assigned to acetonitrile trapped within the water ice structure. This band occurs at the same wavenumber as that seen for acetonitrile in a mixed ice at base temperature (Figure 8), further confirming the assignment of this peak to acetonitrile trapped within, and interacting with, water ice. Further annealing to 140 K leads to complete desorption of the acetonitrile from the surface. Note that TPD data for the desorption of acetonitrile from the layered ice show slightly higher desorption temperatures than 140 K. This is due to the different heating methods used in the infrared and TPD experiments.

Figure 9(b) shows RAIR spectra resulting from the annealing of a mixed ice. For this ice, there is no evidence of crystallisation of acetonitrile. This is unsurprising since the acetonitrile is within the water ice matrix and hence cannot nucleate to form a crystalline structure. As seen in Figure 9(b), no changes are observed in the RAIR spectrum until the ice has been annealed to above 100 K. Following annealing to above 100 K the spectrum gradually changes to give a very broad feature with bands centred around 2266 cm⁻¹ and 2280 cm⁻¹. Whilst the 2266 cm⁻¹ band has already been assigned, it is not clear what the exact origin of the band at 2260 cm⁻¹ is. However, it is likely that this also occurs due to an interaction between the acetonitrile and the water ice, although the exact assignment of the band requires further investigation. As for the acetonitrile on ASW, there is clear evidence for the presence of acetonitrile trapped within the water ice as bands are observed up to 140 K in the spectrum, which is considerably higher than seen for pure acetonitrile.