Thermodesorption

This techniqne, which is based on the exothermic natnre of adsorption, allows us, by using a programmed temperature increase, to analyze the chemical species initially bonnd to the snrface of a solid and which are desorbed under the inflnence of temperatnre. 

Detecting and/or analyzing desorbed molecnles is possible with the use of different detectors, such as thermistors, or analyzers, snch as chromatography or mass spectrometry. 

This analysis could also be conducted by maintaining the sample under a dynamic vacuum or inert gas vector flow. 

The choice of a detector depends most often on the nature and the quality of the expected information  

The experimental results presented in this chapter were all obtained through mass spectrometry. 

---

Occurrence of the temperature inhomogeneity along a film during its heating was reported recently (45f). Different results were obtained when temperature measurement and control were effected only in one point, and in three points of the desorption cell, which illustrates that caution should be observed in this respect when studying thermodesorption from the films. This caution is obviously yet more essential when working with powdered materials. 

For estimates of both Ed and fcd in the Arrhenius equation, in principle two different points on a desorption peak or two runs with different heating factors o2 are required. One obvious point is the maximum of the peak, and very often only this is used while the value of kd is supposed to be of the order of magnitude 1012 to 1013 sec-1. As seen from Eq. (28), the location of Tm depends but weakly on fcd as compared to its dependence on Ed, so that an uncertainty in the value of kd of one order of magnitude does not affect the estimated value of Ed appreciably. This has been clearly illustrated by analogue simulation of the thermodesorption processes (104). On the other hand, the said fact causes the estimates of kd to be very uncertain. A recently published computational analysis of the peak location behavior shows the accuracy of the obtained values of Ed (105). 

This is corroborated by a thermodesorption study of NH3 adsorbed at 373 K, which shows that the quantity of adsorbed NH3 is only slightly higher on PdyAl203 (4 7 1ft" mol NH3/g solid) th on Pd/Zr02 (3.3 10" mol NH3/g solid) The different behavior between PCI/AI2O3 and Pd/Zr02 cannot therefore be explained only by difference in the acidic properties... 

DAIMLER BENZ, Arbeitsvorschrift zur Bestimmung von gasformigen und kondensierbaren Emissionen aus Fahrzeuginnenausstattungsmaterialien mit Thermodesorption, Prilfanweisung PB VWT 709 (1997). 

This adsorption reaction has been extensively studied on most noble metals, especially on Rh by NO thermodesorption [64-66], On Rh/Zr02 [65], it was shown that N2 left the surface from two separate features a sharp jB1 peak at 170°C due to N2 desorption... 

The acidic character of 5A zeolite as a function of the calcium content has been explored by different techniques propylene adsorption experiments, ammonia thermodesorption followed by microgravimetry and FTIR spectroscopy. Propylene is chemisorbed and slowly transformed in carbonaceous compounds (coke) which remain trapped inside the zeolite pores. The coke quantities increase with the Ca2+ content. Olefin transformation results from an oligomerization catalytic process involving acidic adsorption sites. Ammonia thermodesorption studies as well as FTIR experiments have revealed the presence of acidic sites able to protonate NH3 molecules. This site number is also correlated to the Ca2+ ion content. As it has been observed for FAU zeolite exchanged with di- or trivalent metal cations, these sites are probably CaOH+ species whose vas(OH) mode have a spectral signature around 3567 cm"1. 

Figure 1 Propylene cokage kinetics (350°C, 1 Figure 2 Ammonia thermodesorption profiles bar) for 5A samples. for 5A samples...

These two last remarks have also been confirmed by the ammonia thermodesorption profiles shown on Fig.2. After physisorbed ammonia desorption at 50°C under primary vacuum, it remains adsorbed 5.9 and 7.6 ammonia molecules by unit cell for 5A 67 and 5A 86 samples, respectively. At 250°C, 80 and 76 % of ammonia adsorbed has been evacuated from the 5A 67 and 5A 86 samples, respectively. These results also indicate the presence of strong acid sites. The number of these sites seems to be correlated to the Ca2+ ion content. 

The above observation on the acidity of the Si-(OH)-Fe groups in the zeolites of different Si/Fe was confirmed by the IR experiments for the ammonia thermodesorption. 

Hydrogen, which covers the internal surface of PS, can also be used to estimate its structural dimensions. IR measurements indicated a stoichiometry of roughly SiH for electrochemically prepared micro PS [Be2]. If dihydride groups are assumed to cover the internal surface, every second atom must be a surface atom. This is the case for a cube of about 1000 atoms that has a diameter of approximately 2 nm. A stoichiometry of SiH04 obtained by thermodesorption measurements points to a crystallite diameter in the order of 4nm [Pe2]. The chemical composition for a hydride coverage surface and for a 0.5 nm thick native oxide layer are given in Table 6.1. 

Fig. 4 (a) Stepwise dehydrocyclization of -hexane (21, 62). (b) Temperature programmed desorption of benzene originating from various adsorbates over Pt-AljOs. Temperature of adsorption 25°C. Rate of heating 23°C per minute. Detector monopolar mass spectrometer, the ordinate corresponds to the I intensity of mass number 78, in arbitrary units. For clarity, the thermodesorption curves for other compounds (starting hydrocarbon) hexene from hexa-dienes and hydrogen have not been shown (62c). 

Whereas a single TPR peak of benzene (T a = 230°C) was reported for n-hexane (62b), a peculiar triple benzene peak system (with values of 210°, 230°, and 255°C, respectively) was observed with 1,3-, 1,4- and 2,4-hexadienes as adsorbates (62c) (Fig. 4b). The middle peak (the only one for n-hexane) was absent during thermodesorption of benzene, therefore this can be tentatively assigned to the ring closure step (as discussed in Section II,B,2). 

The various TPR peaks may correspond to different active sites. One hypothesis assumed cyclization over metallic and complex (Section II,B,4) platinum sites (62e) the participation of various crystallographic sites (Section V,A) cannot be excluded either. Alternatively, the peaks may represent three different rate determining steps of stepwise aromatization such as cyclization, dehydrogenation, and trans-cis isomerization. If the corresponding peak also appears in the thermodesorption spectrum of benzene, it may be assumed that the slow step is the addition of hydrogen to one or more type of deeply dissociated surface species which may equally be formed from adsorbed benzene itself (62f) or during aromatization of various -Cg hydrocarbons. Figure 11 in Section V,A shows the character of such a species of hydrocarbon. 

In a further approximation, not only the amount but also the position of surface hydrogen should be considered. Thermodesorption studies showed... 

A good-quality CeMCM-41 material with Si/Ce=50 was synthesized by hydrothermal method. For the purpose of comparison a pure siliceous MCM-41 was prepared using the same composition without cerium. Thermogravimetric curves for the synthesized uncalcined samples exhibit shape characteristic for the MCM-41-type materials. The specific surface area of CeMCM-41 evaluated from nitrogen adsorption was equal to 850 m2/g, whereas the pore width and mesopore volume of this material were equal to 3.8 nm and 0.8 cm3/g, respectively. In contrast to the pure silica MCM-41, the CeMCM-41 material exhibits medium and strong acid sites as revealed by thermogravimetric studies of n-butylamine thermodesorption. 

The evaluation of the acid properties of calcined materials was based on the assumption that n-butylamine molecules interact with all acid sites, and the total acidity of the sample studied can be determined from the maximum amount adsorbed. Shown in Fig. 2 are thermogravimetric curves for n-butylamine thermodesorption, which were used to evaluate the amount of medium and strong acid sites for both samples. Thermodesorption of n-butylamine from CeMCM-41 exhibits three distinct ranges (i) desorption of physically adsorbed amine bellow 230 °C (ii) desorption of n-butylamine from medium acid sites at 230 - 410 °C (0.25 mmol/g), and (iii) its desorption from strong acid sites at 410 - 590 °C (0.21 mmol/g). However, only one weight loss was observed for pure silica MCM-41 due to thermodesorption of physically adsorbed amine, indicating negligible acidity of this material. 

The CeMCM-41 material studied had much higher quality than the corresponding MCM-41 sample synthesized under the same conditions. While both materials exhibited analogous adsorption properties with respect to nitrogen, their interaction with n-butylamine was different. Thermogravimetric analysis of w-butylamine thermodesorption showed that CeMCM-41 possessed medium and strong acid sites in contrast to the pure silica MCM-41, the acidity of which was negligible. Thus, incorporation of cerium to MCM-41 seems to improve its hydrothermal stability and enhance the adsorption and catalytic properties. 

The separation of micropore and mesopores plus the outer surface area was done by the combination of elution and flash thermodesorption [9]. With the latter it is possible to determine the contribution of the micropores to sorption separately as micropore desorption requires a higher activation energy because of the above mentioned effects. After injection of an organic vapour or a selection of a particular concentration of the gas mixture, adsorption took place on the sample in the column. The following carrier gas eluted the adsorbate and the peak was recorded. 

For thermodesorption, the sample was heated immediately to 473 K after the elution peak reached the baseline again. The peak which occurred had the same shape as the elution peak and was analysed again with the above mentioned methods. 

The very weak thermodesorption peak near 600 sec. supports this assumption. 

The 13X shows a strong themtodesorption peak whilst the 3A shows none. The explanation is found in the size of the cyclohexane which has a critical diameter of 6 A. Therefore it has access to the pores of the 13X with 10 A diameter whilst there is no access in the pores of the 3A that has a diameter of 3 A. In this case the thermodesorption isotherm should be similar to the static gravimetric obtained one as shown in Figure 4. 

Figure 4. Comparison of static-grav. measured isotherm and 1GC thermodesorption isotherm for zeolite 13X...

It is especially easy to recognize that thermodesorption shows a completely different picture for both samples. This suggests a difference in micropore structure. This is confirmed by Figure 6 that shows the micropore size distribution calculated from the second peak isotherm. 

This is easy to explain because values derived from the elution partial isotherm only pay regard to the amount adsorbed in the mesopores and the outer surface area. By contrast, the static method is not able to distinguish between these contributions and the micropore part of adsorption. Therefore the obtained values are higher but have no physical meaning whereas the elution values give a more realistic picture. The results for the standard carbon are very similar to the static values. This means that there are almost no micropores and the sorption processes take place in the mesopores and the outer surface area. This is confirmed by the huge difference in the thermodesorption peak of both materials. 

Dynamic sorption methods are giving fast and accurate results in the characterisation of a variety of different materials. Moreover the combined elution/thermodesorption IGC method permits a separation between micropore on one hand and mesopore and outer surface area on the other hand. The obtained results give a much more realistic picture of adsorption properties of highly micropores materials. 

In addition, some samples also gave a signal at 7.1 ppm from the residual NH4 cations the amount of the latter was determined by thermodesorption and subtracted from the intensity of line (3). Thus the sum of intensities of (2) and (3) gave the true total content of acidic hydroxyl groups. They had T2 of 60-75 ftscc, while sample 500 SB contained an additional free induction decay (FID) component due to extra-framework hydroxyls. 

The isotherms on silane-treated surfaces are different from one another, considering the precision of the method, but these differences are minor, and are not worthy of detailed analysis. The main inference from analysis of physical adsorption is that silane deposition mitigates the sorptive properties of glass substrates, which is in accord with results of previous work [2-5]. This trend also correlates with results from water thermodesorption, which demonstrate that water-sized glass has the highest sorptive capacity for water, while silane-treated samples exhibit diminished sorptive properties. 

---