TPD Peak Research Articles

Temperature-programmed desorption of H2 as a tool to determine metal surface areas of Cu catalysts

Temperature-programmed desorption of H2 (H2 TPD) is shown to be a useful new tool for the determination of Cu metal surface areas. The technique has been employed in on-line characterization of binary Cu/Al2O3 and ternary Cu/ZnO/Al2O3 catalysts in a combined TPD-microreactor setup. The catalysts were studied both after reduction and after water gas shift activity tests. A main H2 TPD peak is observed around 300 K which can be assigned to desorption from Cu metal surface sites. The H2 TPD method is compared with the N2O frontal chromatography method which has been extensively used in previous studies for the determination of Cu surface areas. It is found that the dissociative adsorption of N2O may induce significant changes in the catalyst structure leading to errors in the surface area determination. With the H2 TPD procedure such irreversible changes can be avoided. A further advantage of the H2 TPD method is the possibility of providing insight into the nature of the exposed Cu metal surfaces.

Reducibility and CO hydrogenation over Pt and Pt-Co bimetallic catalysts encaged in NaY-zeolite

Temperature programmed techniques (TPR, TPD) and X-ray diffraction (XRD) have been used to study ion migration and location as well as reducibility of platinum and cobalt ions encapsulated in Pt/NaY, Co/NaY and Pt-Co/NaY zeolites prepared by ion exchange. The temperature required to reduce Co2+ in NaY was significantly lowered by the presence of Pt and dependent upon the relative locations of Pt and Co ions in zeolite cages. The exact location was controlled by the calcination condition and the metal contents. For bimetallic catalyst with low Pt content (0.5 wt% Pt and 0.9 wt% Co), the TPR results indicated that reduction of Co2+ ions in the vicinity of Pt shifted toward lower temperature, while that of Co2+ staying alone was not affected. With high Pt loading (4.5 wt% Pt, 0.7 and 2.6 wt% Co), however, most of the Co2+ ions were reduced by means of Pt at temperature below 723 K after calcination at 573 K. The temperature for Pt reduction in bimetallic catalysts was somewhat higher than Pt/NaY and increased with Co atomic fraction, indicating that mixed oxide, PtCoxOy, might be formed during calcination. After reduction in hydrogen at 723 K, highly dispersed metal particles were formed. These fine particles were most probably confined inside zeolite cages as indicated by the absence of XRD peak for all samples after calcination and reduction. Surface composition of the bimetallic particles may be different for catalysts with similar Pt content but different Co loading. Accordingly, H/Pt ratios of 1.0 and 0.72 for catalysts with low and high Co content, respectively, were shown by hydrogen chemisorption. It was further supported by the increase in TPD peak intensity with Co loading in the high temperature range, which was related to the reoxidation of Co in bimetallic particles by surface hydroxyl groups. Preliminary results on CO hydrogenation demonstrated that activity and methanol selectivity were higher on Pt-Co bimetallic catalysts than either over monometallic Pt or Co catalyst, which was consistent with the Pt enhanced Co reduction and formation of Pt-Co bimetallic particles.

Oxidative leaching of Cu atoms from PdCu particles in zeolite Y

NaY-supported PdCu samples of various Cu/Pd ratios were prepared by ion exchange with Pd(NH 3) 4 2+ and Cu(NH 3) 4 2+ precursors. Reduction of Cu is enhanced by Pd; however, after calcination at 500°C the bare ions have migrated into sodalite cages, and reduction is only enhanced for those Cu 2+ ions which happen to share their cage with a Pd 2+ ion. Reduction with HZ results in bimetallic PdCu particles and protons of high Brønsted acidity. At 280°C, and in an inert atmosphere, these protons selectively reoxidize surface Cu atoms to Cu+ ions, which are still attached to the surface of the bimetal particles; they can be reduced at 210°C. Complete oxidation of the Cu component takes place at 500°C; the Cu 2+ ions leached from PdCu particles migrate to small zeolite cages, as indicated by EXAFS and a peak at 290°C in the profile of subsequent TPR. After complete leaching of Cu, the monometallic Pd particles are discerned from the original particles in PdCu/NaY and PdCu 2/NaY by their propensity to form hydrides that are detectable by TPD. No hydride is formed with bimetal particles with Cu/Pd > 0.5; particles in Pd 2Cu/NaY do form a hydride which is, however, distinguished from Pd hydride by its TPD peak position.

The decomposition of H 2S on Ni(110)

Adsorbed H 2S decomposes on Ni(110) to form primarily surface S and H for coverages of less than 0.5 ML. The hydrogen evolves in two separate TPD peaks, characteristic of hydrogen recombination and desorption from the clean surface and from regions perturbed by chemisorbed sulfur. XPS and HREELS indicate the presence of SH and possibly H 2S groups on the surface at 110 K. The XPS data indicates that for coverages less than about 0.5 ML, the concentration of molecular H 2S is small, but it is difficult to asess the coverage of SH groups. However, all of the molecular species decompose prior to hydrogen desorption (for high coverage, 180 K). Physisorbed H 2S is observed on the surface for coverages greater than about 0.5 ML. The sulfur Auger lineshape was observed to be a function of both coverage and temperature. The changes in the lineshape were attributed to perturbations in local bonding interactions between the S and the Ni surface, perhaps involving some change in either bonding sites or distances but not involving SH bond scission. The decomposition reaction was modeled using a bond order conservation method which successfully reproduced the experimental results.

Strong adsorbate-adsorbate interactions in the H 2O/Pd(110) system

The chemisorption of water on clean Pd(11O) and its coadsorption with CO 2 or O was studied using HREELS and TPD. On the clean surface HREELS spectra reveal two distinct arrangements below the onset of multilayer growth while there is only one TPD peak in this coverage range. We propose that at low coverages tetramers are formed whereas at intermediate coverages these tetramers coalesce into islands with the distorted hexagonal bilayer structure. The bilayer model for the saturated chemisorbed layer is carefully scrutinized and found to be in agreement with all the experimental results for Pd(110). Water coadsorption with CO 2 is dominated by a charge transfer from H 2O to CO 2 which is mediated by the metal surface. This leads to the formation of a bent CO 2 δ− species. The interaction between water and oxygen is a strong function of oxygen coverage. At low coverage a direct interaction leads either to the dissociation of water or to the formation of an O-H 2O complex. When the surface is covered with oxygen, the strength of the water surface bond is significantly reduced compared to the clean surface.

Selectivity of hydrogen chemisorption on clean and lead modified palladium particles; a TPD and photoemission study

This work describes hydrogen chemisorption on clean and lead modified palladium particles obtained from decomposition of PdO. TPD is used as a chemical probe to test the surface properties of several states of metallic palladium relevant in practical selective hydrogenation catalysts. These states differ in oxygen content and the presence of a lead modifier. XPS and UPS data serve as a basis for identifying the surface properties. TPD spectra show a very broad low temperature peak-likely bulk hydride decomposition-and a sharp TPD peak between 330 and 380 K. This latter can be devided into three rather poorly separated subpeaks; addition of Pb does not shift peak maxima but decreases the central subpeak and eliminates the high temperature peak completely. This points to the interaction of Pb with specific surface sites rather than to bulk alloy formation. The enhancement of selectivity in hydrogenation obtained from lead modification is considered as a geometric site blocking effect rather than to arise from a bulk modification of the valence electronic structure of palladium metal.

TPD and IR spectroscopic studies of CO, CO 2 and H 2 adsorption on LaCrO 3

The surface interaction of H 2, CO and CO 2 with oxidized and reduced LaCrO 3 were studied by means of temperature programmed desorption and infrared spectroscopies. CO single adsorption at 298 K gave rise to CO TPD peaks (375–765 K) assigned to linear and bridged species and to CO 2 peaks (425–810 K) assigned to bidentate carbonates (infrared bands at 1655, 1625, 1290 and 1055 cm −1). CO 2 single adsorption yielded CO 2 TPD peaks (365–730 K) assigned to bidentate carbonates (infrared bands at 1615, 1310 and 1045 cm −1). CO and CO 2 desorption spectra after CO adsorption and after H 2CO successive adsorption on LaCrO 3 reduced at 923 K showed that hydrogen preadsorption does not affect in any significant way the subsequent CO adsorption. However, the TPD spectrum of H 2 obtained after H 2 adsorption is substantially modified by CO preadsorption (COH 2 sequence) or by CO subsequent adsorption (H 2CO sequence). Significant differences were also found in CO and CO 2 TPD spectra after CO single adsorption and after COH 2 successive adsorption. These differences are interpreted to be due to a strong interaction between CO and hydrogen in the adsorbed state. These results are consistent with the formation of an oxygenated surface species after the COH 2 adsorption sequence.

Dealumination of ZSM-5 — Characterization by TPD, DTA/DTG, XPS and Adsorption Measurements

A series of ZSM 5 samples obtained by dealumination is studied by TPD, DTA/DTG, XPS and adsorption measurements. The amount of adsorbed water at p/p0 ≦ 0.1 and the amount of chemisorbed NH3 corresponding to the TPD peak at 693–723 K correlate with the number of framework aluminium atoms per unit cell. At strong dealumination an enrichment of non-framework aluminium in the surface layers of the crystals was found. The HZSM-57 zeolites are transformed into silicalite-like samples during the course of the DTA experiment in the temperature range 973–1273 K.

Preparation of mono- or zerovalent nickel by single or successive one-electron-transfer steps in the photoreduction of silica-supported nickel catalysts

Ni/SiO/sub 2/ supported catalysts prepared by competitive cation exchange can be reduced by UV irradiation in hydrogen at 77 K to lead by a one-electron process to Ni/sup +/ species. These Ni/sup +/ ions which exhibit an EPR signal at g/sub 1/ = 2.68, g/sub 2/ = 2.32, and g/sub 3/ = 2.007 and a band at 838 nm in the UV-vis and near-IR reflectance spectra bind molecular hydrogen to form (similarly ordered Ni(H/sub 2/))/sup +/ pseudotetrahedral surface complexes. The photoreduction process in carbon monoxide occurs appreciably only at 25/sup 0/C and depends on the CO pressure. It is a one-electron-reduction process, and (Ni(CO)/sub n/)/sup +/ species (n = 2-4) are observed by EPR and IR for pressures above 10 Torr. For lower pressures, the formation of metal is observed by IR and ferromagnetic resonance via the reduction by two consecutive one-electron-transfer steps. The first step proceeds from the photoproduction of (Ni/sup +/-O/sup -/)* excited states. The second step is due to the release of electrons from a reservoir which has been tentatively identified to pseudo-carbonates characterized by IR bands at 1750-1850 cm/sup -1/ and by TPD peaks of CO at 803 K and of CO/sub 2/ at 513 andmore » 803 K. The formation of the latter species is the result of the quenching of the excited state by CO molecules which react with the O/sup -/ activated surface oxygens to lead to CO/sub 2//sup -/ which with O/sup 2 -/ ions generates CO/sub 3//sup 3 -/ radical ions.« less

Measurements of weakly bound hydrogen on Pd(111) between 5 and 80 K

The adsorption of hydrogen and deuterium onto a low-temperature surface of Pd(111) has been studied experimentally using electron-stimulated desorption (ESD) and temperature programmed desorption spectrometry (TPD). At a surface temperature TS=7.5±1 K, exposure to D2 leads to multilayer formation which is reflected in the desorption of deuterium ion clusters D+n by ESD (Ee−=100 eV). Cluster sizes up to n=21 have been observed which, except for n=2, consist exclusively of odd numbers of atoms. The temperature dependence of the ESD signal following exposure at TS=7.5±1 K suggests that by 9.5±1 K the multilayer desorbs leaving a diatomic molecular species which, in the high-coverage limit, is stable up to ∼20 K. Exposure to hydrogen at these low temperatures gives rise to two TPD peaks: the β (chemisorption) peak and a new TPD peak denoted γ (near 60 K at low coverage) which is assigned to desorption of a surface molecular hydrogen species. Mixed isotope experiments show that there is no mixing in the γ state consistent with this assignment. With increasing exposure, the γ peak grows in intensity and shifts rapidly towards lower temperature. TPD peak shape analysis gives a binding energy for the molecular state at low coverage of ∼0.020±0.010 eV and a corresponding preexponential of approximately 102 s−1.

CONVERSION OF Rh(III) SPECIES ON Rh–Y ZEOLITE DURING EXPOSURE TO CARBON MONOXIDE REVEALED BY TPD TECHNIQUE

Abstract On the exposure of fresh Rh–Y zeolite to CO, the formation of the rhodium species responsible for the TPD peak of CO at 473–500 K, which was followed by the conversion into the second type rhodium species responsible for the peak at 563–583 K, was observed.

The adsorption of NO on Ru(001) and its co-adsorption with oxygen studied by vibrational spectroscopy

The adsorption of NO on Ru(001), and its co-adsorption with oxygen, has been studied by LEED, TPD, EELS and IRAS (with particular emphasis on the vibrational spectroscopies) over a wide range of temperatures. It has been well established (ref.1,2) that the adsorption of NO on Ru(001) at room temperature is initially dissociative with molecular adsorption taking place only after a dissociative layer is formed. It therefore seemed appropriate to study the effect of oxygen co-adsorption on the adsorption of NO under well defined conditions. The dissociation layer is characterized by a (2×2) LEED structure and is found to influence the subsequent molecular adsorption of NO in exactly the same way as a saturated, pre-chemisorbed layer of oxygen. One effect of oxygen co-adsorption is to suppress the v 1 intensity in the desorption spectra (ref.1) (the v 2 peak remains essentially unchanged), while simultaneously producing a new TPD peak of NO ( v 1(O)) with lower binding energy. In the corresponding EELS spectra these two species ( v 1 and v 1 (O)) are clearly distinguished. Our inability to observe anything other than the absorption band at ∼1800cm −1 in the IRAS experiments led us to repeat the EELS experiments (ref.2) together with TPD to elucidate more clearly the nature of the v 1 molecular species (ref.1). The results indicate that the adsorption of NO at low temperature (∼ 95K) initially produces a species of NO with an N-O stretch frequency of 1400cm −1. This is the only species observed up to an exposure of 0.5 Langmuirs, and on warming to room temperature it dissociates completely to produce the (2×2) pattern described above. Pre-adsorption of oxygen prevents the formation of this species of NO. Only subsequent to the saturation of this species during adsorption at low temperature do two bands at ∼1490cm −1 and ∼1810cm −1, associated with the v 1 and v 2 molecular species (ref.1), appear. We suggest that the low temperature, low coverage species is lying down, and the molecular species v 1, and v 2 are both adsorbed on “on top” sites but corresponding to the bent and linear forms of the NO molecule, respectively.

Étude de l'interaction platine-hydrogène. I. Désorption thermique éclair de H 2 d'un filament de Pt après adsorption sous forte pression

An ultrahigh vacuum flash thermal desorption apparatus has been designed in order to allow drawing of H 2—TPD curves from a Pt polycrystalline wire, after chemisorption at low (10 t̄7 Torr) and high pressures under a similar pressure used during investigations of supported Pt catalysts (100 Torr). In the absence of any contamination of the Pt wire by CO or H 2, a single H 2—TPD peak is observed, at about 400 K, that is approximately at the same temperature as over Pt supported catalysts, after the H 2 chemisorption at 300 K. Unlike what happens over the Pt/support catalysts, the cooling of the Pt wire in H 2 from 800 to 300 K does not give rise to any H 2 thermal desorption at 700–800 K during the subsequent heating in vacuum. In conditions where the Pt wire is contaminated with CO or H 2O, these gases are thermally desorbed, at about 520 K and 600–630 K, respectively; simultaneously some hydrogen is detected by the quadrupole mass analyzer.