Chemisorption of methanol

Lopes MIS, Beden B, Hahn F, Leger J-M, Lamy C. 1991. On the nature of the adsorbates resulting from the chemisorption of methanol at a platinum electrode in acid medium An EMIRS study. J Electroanal Chem 313 323-339. 

The electrodes in the direct methanol fuel cell (DMFC) (i.e. the anode for oxidising the fuel and the cathode for the reduction of oxygen) are based on finely divided Pt dispersed onto a porous carbon support, and the electro-oxidation of methanol at a polycrystalline Pt electrode as a model for the DMFC has been the subject of numerous electrochemical studies dating back to the early years ot the 20th century. In this particular section, the discussion is restricted to the identity of the species that result from the chemisorption of methanol at Pt in acid electrolyte. This is principally because (i) the identity of the catalytic poison formed during the chemisorption of methanol has been a source of controversy for many years, and (ii) the advent of in situ IR culminated in this controversy being resolved. 

The authors first experiments were intended to give some quantitative appreciation of the coverage of the CO adsorbate arising from the chemisorption of methanol and were based on a comparison of the CO and CH, OH... 

Figure 3.35 shows the potential dependence of the integrated band intensity of the linear CO observed in the experiment described above and the corresponding variation in the methanol oxidation current. The latter was monitored as a function of potential after the chemisorption of methanol under identical conditions to those employed in the IRRAS experiments. As can be seen from the figure the oxidation of the C=Oads layer starts at c. 0.5 V and the platinum surface is free from the CO by c. 0.65 V. The methanol oxidation current shows a corresponding variation with potential, increasingly sharply as soon as the CO is removed strong evidence in support of the hypothesis that the adsorbed CO layer established at 0.4 V acts as a catalytic poison for the electro-oxidation of methanol. 

Figure 6. Comparison of various in situ techniques for the chemisorption of methanol on platinum.

Chemisorption of methanol or hexanol and formation of surface esters on heat-treated Ti02 were described by Isirikyan, Kiselev, and Ushakova (306). 

The spectra obtained from the chemisorption of methanol onto catalyst above 100°C indicated the progressive oxidation of methoxy species to formate via dioxymethylene/HCHO and finally to CO, CO2, and H2. Phenol adsorbed on the surface Lewis acid-base pair site and dissociated to phenolate anion and proton. The formation of phenolate anion and proton were discerned from the strong intense C-0 stretching vibration and the disappearence of phenolic 0-H stretching vibration, respectively. Importantly, there were series of definite low intensity bands between 2050 and 1780 cm" that were identified as the out-of-plane aromatic C-H bending vibrations [79, 84-85]. These bending vibrations are possible only if the phenyl ring of phenol is perpendicular to the catalyst surface. 

This may be explained by the bifunctional theory of electrocatalysis developed by Watanabe and Motoo [14], according to which Pt activates the dissociative chemisorption of methanol to CO, whereas Ru activates and dissociates water molecules, leading to adsorbed hydroxyl species, OH. A surface oxidation reaction between adsorbed CO and adsorbed OH becomes the rate-determining step. The reaction mechanism can be written as follows [15] ... 

Even more striking evidence is provided by the following experiment. If the chemisorption of methanol involves the abstraction of a hydrogen atom, so that the chemisorbed species is CH3O, its amount on the tip at pressures of 10 mm. must far exceed the number of methanol molecules in the gas phase near the tip where ionization can occur, since the density... 

The absence of any direct, i.e. molecular, means of identifying the adsorbed species in situ rendered the controversy unresolvable and it remained undecided over the ensuing fifteen years. However, in 1981 Beden et til. published EMIRS spectra that were destined to have a major impact on this dispute, as discussed in section 2.1.6. This early paper concluded that C=Oads is the dominant strongly adsorbed species (poison) and it is present at high coverage. Some Pt CO is also present but there is no evidence of COHads under the experimental conditions employed (non steady-state, potential perturbation at 8.5 Hz and with the dissociative chemisorption of methanol slow). The principal assignments of the paper were very quickly verified by Russell et al. (1982) using the IRRAS technique. 

Surface hydroxyl groups on an oxide can usually be replaced by other organic functional groups, thereby altering the polarity or hydrophobicity of the surface. One simple process involves the dissociative chemisorption of methanol on silica. 

Currents are significantly higher on Pt(lll) if the sulphuric acid is replaced by perchloric acid in the presence of methanol, the hydrogen region between 0.0 and 0.4 V is now significantly suppressed (by about 40%) and the butterfly at 0.8 V is absent, suggesting that in perchloric acid there much more rapid and complete chemisorption of methanol. 

Although the data of Herrero et al. [34] were interpreted in terms of a parallel reaction scheme model, such a model is certainly not established by their treatment, and Vielstich and Xia [36] have criticised such a model on the basis of their Differential Electrochemical Mass Spectroscopy (DEMS) data [37]. At least below a potential of 420 mV, the very sensitive DEMS technique detects no C02 evolved from a polycrystalline particulate Pt electrode surface on chemisorption of methanol indeed, the only product detected other than adsorbed CO, in very small yield (one or two orders of magnitude smaller), is methyl formate from the intermediate oxidation product HCOOH. This is graphically illustrated in Fig. 18.2 in which the clean electrode is maintained at 50 mV, a 0.2M methanol/O.lM HCIO4 electrolyte introduced, and the electrode swept at 10 mV s I anod-... 

We may summarise this sub-section by re-iterating the fact that the CO adlayers formed on oxidation of methanol clearly closely resemble those formed by adsorption of CO from aqueous solution, but they also show differences. Not only are coverages usually lower, but there is at least some indications that in addition to the terminal and bridge-bonded COads species expected, there are probably other carbon-containing species on the surface that affect the way in which methanol is oxidised. In addition, whilst there is some evidence for the formation of islands of adsorbed CO on chemisorption of methanol, these islands appear to be smaller than those that form on adsorption of CO itself, and particularly in the case of sulphuric acid, appear to be strongly associated with adsorbed anions. 

The most natural explanation of this is that doubly bonded >CHOH on the surface of the platinum particles is oxidised to formic acid given that subsequent to this process, CO bonded at step sites is observed to form, this suggests that the >CHOH species are actually formed at the steps at low potentials indeed, it would appear likely that the step sites may well facilitate the oxidative chemisorption of methanol, and the doubly bonded >CHOH species are then intermediates in the overall chemisorption to adsorbed CO. The removal of these species from the step sites by oxidation leads to migration of CO from the terraces to the steps, and the reduction in CO coverage on the terraces then leads to conversion of some of the linearly bonded CO to multiply bonded forms. 

We can conclude this second section by asserting the following chemisorption of methanol gives rise to linear and bridge bonding CO, the relative proportions depending on the crystallographic face exposed. The... 

We have seen that the process of oxidation of methanol involves the formation of chemisorbed fragments, predominantly COads and (probably) =C-OH. At lower potentials (E < 0.5 V), chemisorption of methanol on a clean platinum surface is faster than subsequent oxidation of the chemisorbed fragments to C02, but all investigators have reported that a steady-state can be established, in which a small residual current flows. It is less clear what the rate-limiting step is for this residual current, and intensive studies were first carried out by Bagotzky and Vassiliev [5] to attempt to distinguish the mechanism. For 1 M methanol/0.05 M H2S04 on smooth... 

Figure 11. Infrared reflectance spectra (EMIRS) of the adsorbed species resulting from the chemisorption of methanol on a platinum polycrystalline electrode in acid medium (10 M CH3OH in 0.5 M HCIO4, room temperature).

Gas-phase synthesis of 2MN can be carried out efficiently over H-ZSM-5 and H-ZSM-11 type zeolites. The results are consistent with the Rideal type mechanism for alkylation of naphthalene with methanol. The first step in the alkylation reaction of naphthalene is the chemisorption of methanol on the Bronsted acid sites. Methoxy groups are formed on the surface and according to TPD analysis, naphthalene reacts with them impacting, directly from the gas phase. The reaction seems to occur on the external surface of the crystallites of the medium pore zeolites. Using large pores zeolites, the reaction also takes place also in the channel space, and the selectivity of B derivatives is suppressed. 

Until now, for methanol oxidation the best bimetallic catalyst was found to be Pt-Ru. Several papers deal with the electro-oxidation of methanol at Pt-Ru bimetallic system dispersed in polyaniline [33,46]. From results with bulk alloys, the optimum Pt/Ru ratio of around 6 1 to 4 1 was found [49] and confirmed [50]. The electroactivity of Pt-Ru-modified polyaniline is much better than that displayed by pure Pt particles dispersed into the PAni film. The optimum composition of the Pt-Ru bimetallic system was confirmed from these results [33]. The decrease of the poisoning phenomenon is the consequence of a low coverage in adsorbed CO resulting from the chemisorption of methanol. This was checked by considering the oxidation of CO at the same Pt-Ru/PAni-modified electrode [34], which occurs at low overvoltages (150 mV) in the presence of Ru. 

The increase in the island dispersion, as it increases the ruthenium coverage, removes platinum sites that were present on the bimetallic surface before the experiment reported above was executed. In other words, the number of ensembles of Pt sites available,65,66 e.g., for chemisorption on the Pt sites of the Pt( 111 )/Ru surface was reduced. A typical case where this development is important is the process of dissociative chemisorption of methanol, as it requires as many as three adjacent sites for methanol dissociation (dehydrogenation) to chemisorbed CO65-67 As methanol oxidation to CO2 predominantly occurs via the CO formation process,67,68 the overall rate of methanol oxidation may be affected by an increase in ruthenium coverage at the expense of the number of collective Pt sites required for methanol decomposition to CO.65,66... 

Carbenic mechanisms. Venuto and Landis [10] were the first to address the question of mechanism of hydrocarbon formation from methanol over zeolites, in this case zeolite X [11]. These workers proposed a scheme involving a-elimina-tion of water and polymerization of the resultant methylcarbenes to olefins. Swabb and Gates [12], elaborating on Venuto-Landis, proposed that concerted action of acid and basic sites in the zeolite (mordenite) facilitates a-elimina-tion of water from methanol. According to Salvador and Kladnig [13], carbenes are generated through decomposition of surface methoxyls (a-el imination of silanol) formed initially upon chemisorption of methanol on the zeolite (zeolite Y). Hydrocarbons are assumed to form, in the latter two schemes, also by carbene polymerization. 

---