The Heriot-Watt Astrochemistry Research Group

HWU Logo





Thermal desorption of H2O/CO ices

Layered and mixed ices containing water and carbon monoxide were deposited on the surface and then heated. When the water was deposited on top of a layer of CO three CO desorption peaks are seen in the CO TPD trace. The first desorption peak occurs at ~20-60K, corresponding to desorption from the surface of the ice. This implies that some of the CO has diffused through the highly porous water layer. Two desorption features can be seen in this region, showing that there are two different binding sites present.


Figure 1. TPD of 13CO (0.07 µg cm-2) deposited at 8 K (a) on a gold substrate; (b) on Ilda (30 µg cm-2) deposited at 120 K; (c) on Ilda (30 µg cm-2) deposited at 70 K; (d) on Ihda (30 µg cm-2) deposited at 8 K; and (e) deposited simultaneously with H2O (30 µg cm-2) at 8K. Heating rate = 0.08 K s-1. Traces are offset for clarity.


The two peaks at 140 K and 160 K occur at too high a temperature to be carbon monoxide desorbing from water ice, and the peaks are co-incident with a phase change in, and desorption of the water ice. New pathways to the surface are created during the phase change allowing desorption of some of the previously trapped CO in a desorption process called a "molecular volcano". A qualitatively similar TPD trace is seen when the water and carbon monoxide are dosed into the chamber at the same time, creating a mixed ice.

When the CO is deposited on top of a layer of water grown at 8 K, peaks due to both trapped CO and CO desorbing from the surface are seen. Since there is no physical barrier to CO desorption at the time of its deposition, there must be a structural change in the ice film which occurs before the CO finishes desorbing at ~60 K. This is confirmed by depositing a layer of CO on water ices grown at 30, 45 and 70 K. The fraction of trapped CO is reduced as the water deposition temperature is increased. Ice deposited at 8 K has a more porous structure than ice deposited at 120 K. The more porous ice is sometimes called Ihda and the less porous ice Ilda.

RAIRS experiments were also carried out on the H2O/CO ices, the results of which are shown in Fig. 2. Two sharp features are seen in the CO stretching region at 2139 cm-1 and 2142 cm-1 when CO is adsorbed onto water ice grown at 8 and 80 K. The peaks are attributed to the orthogonal and transverse vibrations of solid CO respectively, indicating that multilayers of solid CO are present. When the water and carbon monoxide are co-dosed, forming a mixed ice with ~5% CO, a broad peak is seen at 2139 cm-1 with a smaller feature at 2153 cm-1. The absence of the splitting in the co-deposited spectra indicates that the 2139 cm-1 and 2153 cm-1 features are due to carbon monoxide in a water matrix.

The ices were also warmed, held at the higher temperature for 5 minutes, and then cooled back down to 8 K to further investigate the effects of heating on the ices. When CO on Ihda was heated to 15 K the peaks due to multilayer CO decreased and had disappeared completely by 20 K. At the same time the growth of broad peaks at 2153 cm-1 and 2139 cm-1 can be seen, leading to the same two peak profile which is seen for co-deposited ices.

However, when the CO is adsorbed onto Ilda only a small reduction in the split peak is seen at 20 K and the 2153 cm-1 has only grown slightly. Evidence of multilayer CO does not disappear until the ices are annealed to 50 K, which is consistent with the TPD which shows desorption from Ihda is complete by 50 K.

For mixed ices the RAIRS shows evidence of CO being present up to ~130-135 K, which is in line with the TPD traces showing CO desorption at ~140 and 160 K.



Figure 2. RAIR spectra of the 12CO stretch region of (a) 0.35 µg cm-2 CO adsorbed over 57 µg cm-2 H2O at 8K, (b) 0.35 µg cm-2 CO adsorbed at 8 K over 57 µg cm-2 H2O deposited at80 K, (c) 57 µg cm-2 of a ~ 5:100 mixture of CO:H2O gas mixture adsorbed at 8 K. The sample have been annealed for 5 minutes at the temperatures indicated. Features marked * in the 45 Kspectrum of Figure 2a are due to gas phase CO transitions, resulting from contamination of the dryair in the purge of the IR optics. The increased baseline noise in the 45 K, 80 K and 130 K spectrain Figure 2a result from a slight mismatch between the sample spectrum and background spectrumused in data processing in a lengthy experiment.


Figure 3. Cartoon showing the trapping of CO within the H2O ice matrix and subsequent desorptions - (i) molecular volcano at the H2O amorphous crystalline phase transition and (ii) desorption of remaining CO at the same temperature as the H2O.

This research is included in the following publications:

Carbon monoxide entrapment in interstellar ice analogs
M. P. Collings, J. W. Dever, H. J. Fraser, M. R. S. McCoustra and D. A. Williams
Astrophysical Journal, 583, 1058, 2003

Laboratory Studies of the Interaction of Carbon Monoxide with Water Ice
M. P. Collings, J. W. Dever, H. J. Fraser and M. R. S. McCoustra
Astrophysics and Spaces Science, 285, 633, 2003