Carbon monoxide thermal desorption

Figure 2.2. Thermal desorption spectra of carbon monoxide, measured mass spectrometically at mass 28 (atomic units, a.u.), on a platinum (100) surface upon which potassium has been pre-adsorbed to a surface coverage of 0K.7 Reprinted with permission from Elsevier Science.

We have undertaken a series of experiments Involving thin film models of such powdered transition metal catalysts (13,14). In this paper we present a brief review of the results we have obtained to date Involving platinum and rhodium deposited on thin films of tltanla, the latter prepared by oxidation of a tltanliua single crystal. These systems are prepared and characterized under well-controlled conditions. We have used thermal desorption spectroscopy (TDS), Auger electron spectroscopy (AES) and static secondary Ion mass spectrometry (SSIMS). Our results Illustrate the power of SSIMS In understanding the processes that take place during thermal treatment of these thin films. Thermal desorption spectroscopy Is used to characterize the adsorption and desorption of small molecules, In particular, carbon monoxide. AES confirms the SSIMS results and was used to verify the surface cleanliness of the films as they were prepared. 

On a third sample, thermal desorption of carbon monoxide was carried out before and after annealing to 760 K and after sputtering into the region where the Ti" " was a minimum (l.e., about 180 seconds in Fig. 3). The TDS results are summarized in Table I. The last column gives the Integrated peak desorption area for carbon monoxide. 

Thermal desorption spectra of carbon monoxide on polycrystalline and on single crystal platinum are well known from experiments in the gas phase [48,49], The system is therefore appropriate to test the experimental setup. 

Adsorbed carbon monoxide on platinum formed at 455 mV in H2S04 presents a thermal desorption spectrum as shown in Fig. 2.4b. As in the case of CO adsorption from the gas phase, the desorption curve for m/e = 28 exhibits two peaks, one near 450 K for the weakly adsorbed CO and the other at 530 K for the strongly adsorbed CO species. The H2 signal remains at the ground level. A slight increase in C02 concentration compared to the blank is observed, which could be due to a surface reaction with ions of the electrolyte. Small amounts of S02 (m/e = 64) are also observed. 

Figure 2.9 Thermal desorption of carbon monoxide from two rhodium surfaces in ultrahigh vacuum, as measured with the experimental set-up of Fig. 2,10. Each curve corresponds to a different surface coverage of CO. At low coverages CO desorbs in a single peak indicating that all CO molecules bind in a similar configuration to the surface. At higher coverages, an additional desorption peak appears, indicative of a different adsorption geometry (courtesy of M.J.P. Hopstaken and W.E. van Gennip [141).

After the adsorption of oxygen, the sample is replaced under vacuum at 30°C. No desorption occurs, and no thermal effect is registered. Carbon monoxide may be adsorbed again, and its adsorption is followed by a decrease of the electrical conductivity of the sample from 1.6 X 10"7 to 1.6 X 10 10 (ohm cm.)-1. The differential heats measured during the... 

Thermal desorption studies have indicated that, as with massive gold, adsorbed carbon monoxide desorbs from small gold particles below room temperature. Thus, with model AU/AI2O3 there were two desorption... 

Figure 28. Thermal desorption (heating in flowing nitrogen, heating rate 3Ks 1) of CO2 and carbon monoxide from heavily oxidized fullerene black (treated in 10% ozone, oxygen at 333 K. in water). An IMR-MS detector was used for unperturbed gas analysis and simultaneous detection of other desorption products (see text and Fig. 29). The value for the bum-off temperature is defined as the temperature where a weight loss of 3% had occurred.

Carbon dioxide and hydrogen also interact with the formation of surface formate. This was documented for ZnO by the IR investigation of Ueno et al. (117) and, less directly, by coadsorption-thermal decomposition study (84). Surface complex was formed from C02 with H2 at temperatures above 180°C, which decomposed at 300°C with the evolution of carbon monoxide and hydrogen at the ratio CO Hs 1 1. When carbon dioxide and hydrogen were adsorbed separately, the C02 and H2 desorption temperatures were different, indicating conclusively that a surface complex was formed from C02 and H2. A complex with the same decomposition temperature was obtained upon adsorption of formaldehyde and methanol. Based upon the observed stoichiometry of decomposition products and upon earlier reported IR spectra of C02 + H2 coadsorbates, this complex was identified as surface formate. Table XVI compares the thermal decomposition peak temperatures and activation energies, product composition, and surface... 

Toluene surface coordination chemistry was quite different from that of benzene. Toluene chemisorption on all the clean surfaces was thermally irreversible. In addition, toluene was not displaced from these surfaces by trimethylphosphine nor by any other potentially strong field ligand examined to date, e.g., carbon monoxide or methyl isocyanide. In the thermal decomposition of toluene on these surfaces (attempted thermal desorption experiments), there were two thermal desorption maxima for H2 (or D2 from perdeuterotoluene) with the exception of the Ni(110) surface. This is illustrated in Figure 6 for Ni(lll)-C7Dg. 

Significant adsorption of carbon monoxide first appears after activation at 300°. After activation at 373°, for example, the first 0.1 ml appeared to be completely adsorbed. One-half of another 0.1 ml injected 15 minutes later was adsorbed. Several subsequent injections at 15-minute intervals were similarly about one-half adsorbed. However, after an 80-minute interval, all of an injected 0.1 ml was adsorbed. It is clear that about 25/i mole/gm Ct203 of carbon monoxide is rapidly chemisorbed but that much of the material adsorbed is held rather weakly and desorbs between injections, but so slowly that the rate of desorption is not readily measurable by thermal conductivity measurement. 

Evacuation of the oxygen atmosphere down to 10- torr does not produce a desorption (24) and no thermal effect is registered (68). COa fads) is therefore a stable species at room temperature, in the presence of oxygen or under vacuum. Its spontaneous decomposition to yield carbon dioxide is not possible at room temperature but is observed at 200° (24). A subsequent adsorption of carbon monoxide is possible on the samples of NiO(200°) and NiO(250°) (Table V) and, in both cases, carbon dioxide is then found in the cold trap. This adsorption of carbon monoxide decreases the electrical conductivity of the samples which, however, remains higher than the conductivity of the initial nickel oxide [7 X 10-i< ohm-i cm-i for NiO(200°) 1.6 x lO- o ohm-i cm-i for NiO(250°)] (25, 41). It was concluded from these experiments that a fraction of C03 (ads) ions is decomposed at room temperature by carbon monoxide and that the interaction product is carbon dioxide, which is, at least partially, desorbed to the gas phase (0.5 cm /gm) (25). 

A to G (Fig. 33) were obtained by integration of curves A to G (Fig. 34). Evolution of the profile of the calorimetric curves indicates that the reactivity of the oxide toward carbon monoxide increases progressively with the extent of reduction. From curve A (Fig. 34), it appears that the reaction of dose A is a relatively slow exothermic process. Curves B to F (Fig. 34) are more complex. Analysis of these curves shows that three thermal phenomena occur during the reaction of doses B to F (i) a fast exothermic process whose intensity increases with the extent of reduction, (ii) a slower exothermic process similar to that observed for dose A, whose intensity decreases from curve B to F and, (iii) a slow endothermic process which is evidently the desorption of carbon dioxide. Both exothermic processes are related to the adsorption of carbon monoxide and to the surface reduction of the solid. 

The region of the cyclic voltammogram, corresponding to anodic removal of Hathermal desorption spectra of platinum catalysts. However, unlikely the thermal desorption spectra, the cyclic-voltammetric profiles for H chemisorbed on Pt are usually free of kinetic effects. In addition, the electrochemical techniques offer the possibility of cleaning eventual impurities from the platinum surface through a combined anodic oxidation-cathodic reduction pretreatment. Comparative gas-phase and electrochemical measurements, performed for dispersed platinum catalysts, have previously demonstrated similar hydrogen and carbon monoxide chemisorption stoichiometries at both the liquid and gas-phase interfaces (14). 

In order to determine how much of the platinum surface is exposed and remains uncovered, the adsorption and subsequent thermal desorption of carbon monoxide was utilized. This molecule, although readily adsorbed on the metal surface at 300 K at low pressures, does not adsorb on the carbonaceous deposit. The results indicate that up to 10-15% of the surface remains uncovered metal sites decreases slowly... 

Various mass spectroscopies are applicable ex situ in order to obtain molecular information. Secondary ion mass spectroscopy (SIMS) can be used [124, 125]. Thermal desorption mass spectroscopy is a viable alternative as a less intrusive and more surface sensitive tool [126, 127]. So far, the latter method has been applied exclusively to adsorbed hydrogen and carbon monoxide formed in an electrochemical reaction of organic CHO-compounds. 

Bectrodes containing the desired probe adlayer were transferred from vacuum directly to an attached electrochemical cell. Carbon monoxide was adsorbed either by solution or gas phase dosing, while formic acid was supplied only through solution. Vacuum measurements included thermal desorption. Auger spectroscopy, and low energy electron diffraction and were primarily intended to establish coverages and the state of the probe adlayer. In this paper we focus on the electrochemical measurements, as these highlight the influence of the probe adlayer on the reaction. 

Thermal desorption spectroscopy provides information about the heat of adsorption of atoms and molecules at solid surfaces [13]. When heating the solid at a well defined rate (in the range of 4-20°/sec), there is a specific temperature at which the desorption rate is a maximum and from which the heat of desorption can be obtained. When the thermal desorption spectrum is taken at different surface coverages, as shown for carbon monoxide in figure 15, one observes a shift in the peak desorption temperature indicating a variation in the heat of adsorption with coverage. 

Fig. 15. Thermal desorption spectra of carbon monoxide on Rh(lll) measured as a function of coverage following adsorption near 300 K. The crystal heating rate was linear at 15 K/sec. Note the desorption peak temperature shift as a function of coverage...

Fig. 30. Carbon monoxide dissociation on a Rh(l 11) surface as a function of potassium coverage as determined by thermal desorption isotope scrambling experiments with C 0 and C ...

When atoms or molecules adsorb on these high Miller index surfaces, they have available to them now several additional sites where their binding could be different. This is clearly indicated by thermal desorption spectroscopy studies. Figure 35 shows that carbon monoxide exhibits two desorption peaks at high coverages while at low coverages only the higher temperature desorption peak is present. The two desorption peaks,when compared with flat surfaces, can be attributed to carbon monoxide adsorbed at stepped... 

Fig. 35. Thermal desorption spectra of carbon monoxide from a Pt(533), stepped crystal face as a function of coverage. The two peaks are indicative of CO bonding at step and terrace sites. The higher temperature peak corresponds to CO bound at step sites...

Barth R, Pitehai R, Anderson RL, Verykios XE. Thermal desorption-iu ared study of carbon monoxide adstuptitm by alumina-supported platinum. J Catal. 1989 116 61. 

However, the quantum treatment of these processes increases the role of the energy transfer from the model heat bath to the reaction co-ordinate, resulting in thermal desorption of excited carbon monoxide adsorbates. The other frustrated translational or rotational modes may also play an intermediate role in energy transfer from excited carbon monoxide stretching mode into the adsorption bond mode (CO-surface). These interactions are important because direct coupling between the earbon monoxide internal stretching mode and the frustrated translational mode, responsible for desorption, is very weak. Taking into account these interactions, this leads to drastic acceleration of desorption [5, 49]. 

---