J. Howard Purnell

Selective chemical conversions using sheet silicate intercalates: Low-temperature addition of water to 1-alkenes

By refluxing hex-1-ene, hept-1-ene, or oct-1-ene in hexane solution with one of a range of cation-exchanged montmorillonites, the alkenes are converted to the corresponding bis-sec-alkyl ethers. The structures of the ethers may be demonstrated by a combination of gas-liquid chromatography and spectroscopy, but they cannot practicably be prepared by any other method. A number of cations facilitate conversion, the most efficient being Cu2+, Fe2+ and Fe3+, Cr3+, and Al3+: Intercalation is a prerequisite for the reaction which involves transfer of oxygen from hydration shell water held in the interlamellar region to the alkene. Collapsed (strongly dehydrated) montmorillonites do not facilitate the conversion which, for the expanded silicates, is characterized by an optimal water content for a given cation. Thus for Cu2+-exchanged bentonite the optimal metal-water ratio is 1:12 and under these circumstances 100% conversion of usable interlamellar water to ether is achieved. Use of other sheet structures, e.g., Cu(UO2)2(PO4)2 · 6H2O, failed to convert the alkenes, while with Cu2+-exchanged synthetic hectorites and fluorohectorites the reaction was not clean, the ether being only one of a number of products. It has not yet proved possible to make the reaction self-sustaining by constant addition of water.

Organic reactions catalysed by sheet silicates: ether formation by the intermolecular dehydration of alcohols and by addition of alcohols to alkenes

Abstract Ion-exchanged montmorillonites can be used as heterogeneous catalysts for the dehydration of liquid alcohols in pressure vessels at 200 °C. In contrast to the situation with homogeneous, aqueous acid catalysts, primary alcohols undergo preferential intermolecular dehydration to give good yields of di(alk-1-yl) ethers with these sheet silicate catalysts, whereas secondary and tertiary alcohols undergo the expected facile intramolecular dehydration to the corresponding stable alkenes. With the acid clay catalysts, primary diols give either oligomeric or cyclic ether products by intermolecular dehydration, whereas benzyl alcohol undergoes intermolecular dehydration to produce a polymeric material, poly(phenylene methylene), in the presence of ion-exchanged montmorillonites. Primary alcohols can be induced to add onto highly substituted alkenes to produce mixed primary/tertiary ethers in excellent yields. Less highly substituted alkenes give much lower yields of mixed ethers. In general, these catalysts are more selective than concentrated mineral acids and have the distinct advantage of being much easier to separate from the products.

Organic reactions catalysed by sheet silicates: ester production by the direct addition of carboxylic acids to alkenes

Abstract The addition of aliphatic acids to a wide Variety of alkenes to form esters is heterogeneously catalysed by ion-exchanged sheet silicates. Cation exchange of the silicate is a prerequisite since natural montmorillonite is a poor catalyst for these reactions. Exchange with almost any multivalent ion induces some degree of catalytic activity but Al 3+ , Cr 3+ and H + ion-exchanged clays proved the most efficient catalysts in the present context. The reactions were carried out in pressure vessels, the reaction temperature range depending on the nature of the alkene employed. Esterification is in competition with alkene oligomerisation, and the synthesis of t-alkyi esters failed at temperatures much above room temperature due to the rapidity of the competing process. All other ester types, however, could be synthesised at temperatures up to 200 °C, and only above this temperature does the competitive process become significant. Initial reaction rates of esterification increase rapidly with temperature although, as would be expected, final equilibrium yields diminish. The number of ester products which may be obtained in a given reaction is determined by the number of stable carbocation intermediates which each alkene can produce. In those reactions where single carbocation intermediates are to be expected, excellent yields of a single ester are often obtained. For example ethyl ethanoate was prepared in 92% yield (based on ethanoic acid) with 96% selectivity, from ethanoic acid and excess ethene using Al 3+ ion-exchanged montmorillonite as the catalyst in a 40 h reaction at 200°C.

Synthesis, characterization, and catalytic activity of beidellite–montmorillonite layered sillcates and their pillared analogues

A Synthetic beidellite-smectite, characterized by a range of techniques, including high-resolution 27Al ans 29Si solid-state n.m.r. spectroscopy, shows interesting catalytic activity (in secondary amine formation from cyclohexylamine, in ester production from hex-1-ene and acetic acid, and in ether synthesis from pentanol): the selectivities differ significantly from those of montmorillonite-smectites.

Reaction of hydrogen atoms with ethane

The reaction between hydrogen atoms and ethane has been studied in a flow-discharge system over the wide temperature range 503–753 K at pressures between 8 and 16 Torr. The major products are methane and ethane, reformed by methyl recombination. Minor products are propane and ethylene with traces of n-butane.The detailed mechanism has been established and computer calculations have been used to derive the set of best-fit rate parameters which reproduce all the experimental results. The results of this work yield the result k1/cm3 mol–1 s–1= 1014.27 ± 0.13 exp (–40.9 ± 1.6 kJ mol–1/RT); H + C2H6→ H2+ C2H5. (1)A survey of all published data some of which have been revised by us to take account of wrongly assumed stoichiometry in the original work, shows that, over a range of 1000 K, the data can be represented by the Arrhenius expression, k1/cm3 mol–1 s–1= 1014.12 ± 0.09 exp (–39.2 ± 0.9 kJ mol–1RT). There is thus no reason to suppose curvature in this plot as has been suggested.Values of the rate constants for the reactions H + C2H5→ 2CH3(2), CH3+ H → CH4(5), and, H + C2H5→ H2+ C2H4(11), are found to be k2/cm3 mol–1 s–1= 1013.57, k5/cm3 mol–1 s–1= 1012.04 at 8 Torr, 1012.20 at 12 Torr and 1012.34 at 16 Torr and k11/cm3 mol–1 s–1= 1012.23.We have reassessed our earlier data on the reaction of hydrogen atoms with ethylene in the light of the recent “low” values for the rate constant of ethyl recombination. From this, we find values for k2 and k5 at 290 K which are, respectively, lower and higher than the corresponding values in the range 503–753 K. It is shown that the slight temperature dependence observed is consistent with the order of reactions (2) and (5).

Organic reactions catalysed by sheet silicates: intermolecular elimination of ammonia from primary amines

Abstract (1) Certain ion-exchanged montmorillonites catalyse the intermolecular elimination of ammonia from primary amines to produce dialkylamines. The yields are significant with cycloalkylamines and benzylamine but poor in the cases of alkan-1-amines and alkan-2-amines. (2) Cyclic amines, such as pyrrolidine, also evolved ammonia to give coupled, ring-opened products such as 1,4-di-(1-pyrrolidyl)butane. (3) However, those ion-exchanged sheet silicate catalysts which catalyse the addition of either water, or alcohols, or thiols or carboxylic acids to alkenes failed to effect the addition of amines to the corresponding alkenes. (4) The products from these amine reactions are consistent with the involvement of interlamellar protons, but the reactions have no counterparts in the proton-catalysed solution chemistry of amines. Similar products have been observed previously, however, over the surfaces of finely divided nickel, palladium and ruthenium.

Gas–liquid partition coefficients of n-alkane solutes at infinite dilution in binary n-alkane solvents

Theoretical expressions due to Janini and Martire and to Orwoll and Flory are extended to allow calculation of the value of the infinite dilution activity coefficient of an n-alkane dissolved in a mixture of two other n-alkanes of any composition. Activity and partition coefficients are calculated and compared with experimental (g.l.c.) values for C5-C8 n-alkanes dissolved in C18H38+ C36H74 mixtures of mole fraction of 0 to 1. The theoretical treatments provide essentially identical numerical values and predict a very small negative curvature of plots of relative partition coefficients against volume fraction of C18H38. The predicted deviation from linearity is, at most, 1 %, a value within which the experimental data are linear. To this extent the theory provides a quantitative description of the experimental findings although curvature cannot be unequivocally identified. However, since it has been shown elsewhere that the mixtures of a remarkable range of solute and mixed solvent types also provide data showing apparent linearity of such plots, the results obtained here have interesting implications in that the underlying theories are specific to random mixtures of chain molecules. It is concluded that comprehensive experimental studies, with non-alkane systems, designed to provide more accurate data than have been common are now essential in order more precisely to define the problem.

Thermogravimetric study of the intercalation of acetic acid and of water by Al3+-exchanged montmorillonite

Abstract Thermogravimetric study of Al3+-exchanged montmorillonite after exposure to acetic (ethanoic) acid vapour or liquid reveals two desorption peaks in the range up to 300 °C. The first, centred around 100 °C, is indistinguishable on this account, as our numerous studies have shown, from the similar desorption of water and many other liquids from the original material at about this temperature. The second, centred around 230 °C, is characteristic of the acid-clay system. We establish that the first desorption is also that of acetic acid that is physically sorbed, whereas the higher temperature desorption is associated with chemisorbed acid. Uptake rate and equilibration studies with acid and with water vapours yield enthalpies of intercalation for the physically adsorbed species consistent with liquefaction accompanied, in the case of water, by hydrogen bonding to structural hydroxyls and, possibly, in the case of the acid, with cyclic dimerization. Such intercalation is very slow, 10 h being required to achieve equilibrium at 30 °C, for example. Uptake of the chemisorbed acid is, by contrast, very rapid and reaches a constant value of ca. 90 mg g−1 dry clay at all values of vapour pressures, uptake times and temperatures (30–90°C) employed. The acid in this regime, although obviously strongly bound, can be competitively displaced by water vapour when this is in great excess. Partition coefficients are calculated and these establish that the intercalated weight ratio (at equal partial pressure) is about 1.3:1.0 in favour of water vapour over the temperature range. This establishes that, in acid-clay systems, where only even modest effort is made to exclude residual water, the acid effectively expels all water. Thus, TGA of the acetic acid-clay system provides a novel opportunity of separately determining sorption isotherms for the physically and chemisorbed acid. The preliminary results reported here establish that the chemisorbed acid dominates the primary uptake, to form the equivalent of a single layer. Subsequently, physical sorption proceeds to produce so-called second and third layers intercalated in the montmorillonite.

Stability constants of hydrazoic acid–tributyl phosphate complexes in hexadecane solution

The gas-chromatographic technique of elution by characteristic point (ECP) has been used to study the complexing of hydrazoic acid with tributylphosphate (TBP) in hexadecane solution. The study covered the temperature range, 298–338 K, with hydrazoic acid concentrations in solution up to 0.55 mol dm–3. The dilution–depletion theory previously advanced by the authors is shown, again, to accommodate the results very well and stability constants for the HN3–TBP complex in the range 40–100 dm3 mol–1 have been determined; the associated enthalpy and entropy are found to be –19 200 J mol–1 and –26.3 J mol–1 K–1, respectively. The results correlate well with those previously reported for HCl–TBP complexation and substantiate the description of HN3 as a halogenoid hydracid. The study further confirms the considerable utility of the relatively simple ECP technique in the study of strongly complexing systems at finite concentrations in solution.

Prediction of infinite dilution activity coefficients in binary n-alkane mixtures

The Janini–Martire modification of Prigogines theory of chain molecule mixtures is further tested via a comparison of predicted infinite dilution activity coefficients (γ∞1) and currently available experimental values. The systems included cover a wider range of molecular size ratios and of temperature than heretofore considered. The extent of agreement is within experimental certainty but as previously noted, Orwoll–Flory theory predicts values which are consistently slightly lower. Both theories, however, predict the same relative values of γ∞1 for any solute with a series of n-alkane solvents. The success of these theories for binary systems is regarded as adequate to provide a good basis for their extension to ternary n-alkane systems for which a surprisingly simple empirical relationship has recently been identified.