Abstract
Aldehydes are frequently used either as flavours and fragrances or as intermediates in numerous chemical syntheses, such as the production of dies, pharmaceutics, and agrochemicals. The existing aldehyde synthesis methods all have specific drawbacks. Moreover, they all have in common the production of stoichiometric amounts of waste. Therefore, large interest exists in an economically and environmentally acceptable alternative for the production of aldehydes. As they are either easily available or easy to synthesise, carboxylic acids would be suitable as raw material for this alternative synthesis. A welcome synthesis route would, therefore, be the direct catalytic reduction of carboxylic acids to aldehydes. However, until now direct reduction has only been successful with aromatic acids. The aim of the investigation described in this thesis is to extend this attractive catalytic aldehyde production route to all kinds of carboxylic acid.
The literature presents only a very limited insight into the behaviour of carboxylic acids on catalysts. On metallic surfaces, acids are mostly decom¬posed completely. On oxidic surfaces, aromatic acids are known to form aldehydes, but aliphatic acids usually yield ketones. So far, no proposals have been advanced concerning the mechanism of aldehyde formation from aromatic acids. However, numerous mechanisms have been suggested describing the ketone formation, albeit no agreement has been reached yet on this subject. By this thesis, an attempt is made to obtain aldehydes from aliphatic acids and to reveal the mechanism involved as well as the catalysts and reaction conditions needed. Simultaneously, a part of the study has been devoted to solving the mechanism of the ketonisation reaction, as a better understanding of this undesired side-reaction should help in suppressing it.
As model compounds, acetic -, propionic -, isobutyric -, and pivalic acid have been used, which have a number of α-hydrogen atoms ranging from three to zero, respectively. Their reactions on catalysts have been tested under flow conditions, at atmospheric pressure, and in the temperature range from 50 to 450°C. Analysis of the effluent gas has been done by gas chromatography and mass spectrometry.
First, acetic acid was tested over a broad range of oxidic catalysts. Contrary to what the existing literature claims, it appears that, when using oxides with a moderate metal-oxygen bond strength, acetic acid can be hydrogenated to aldehyde. A plot of the aldehyde selectivity as a function of the metal-oxygen bond strength of the catalyst yields a volcano-shaped curve, suggesting the involvement of a Mars and Van Krevelen mechanism. The most common side-product was acetone, the formation of which occurred clearly via two different reaction path-ways. Oxides with a low lattice energy formed bulk acetates, decomposition of which yielded acetone. On oxides with a high lattice energy, a surface reaction to acetone occurred.
Two catalysts exhibited a very high selectivity: copper - and iron oxide. During the reaction, both these catalysts probably consisted of a mixture of metal and oxide.
To investigate further the precise roles of the metal and the oxide, platinum was added to various oxides. Although platinum itself was not active in aldehyde formation, addition of it to oxides remarkably increased the aldehyde selectivity and decreased the reaction temperature. However, the above-mentioned volcano-shaped correlation was not affected by platinum addition and the hydrogenation reaction still proceeded on the oxide and not on the metal. All reasonable explanations for these phenomena can be rejected, except the possibility that platinum is responsible for the activation of hydrogen, which subsequently spills over to the oxide, where the actual hydrogenation reaction takes place.
Since iron oxide was the most selective catalyst, it was subjected to a more detailed study. As already mentioned, it appeared that a selective iron catalyst consisted of both a metallic and an oxidic phase. The metallic iron probably fulfils the same role as platinum does in the above-described catalysts. The importance of the reductant was studied further. Carbon monoxide could not substitute hydrogen, and the hydrogen/acid ratio had to be high enough to keep the catalyst in its, partly reduced, selective state. In the literature, no aldehyde formation from aliphatic acids was ever reported on iron oxide, as always a too low hydrogen/acid ratio was used.
By using acids with a different number of α-hydrogen atoms, it has been revealed that the presence of α-hydrogen is needed for the surface reaction to ketones. An acid without α-hydrogen atoms cannot ketonise and is, therefore, more easily hydrogenated to its corresponding aldehyde. This explains why all patents dealing with carboxylic acid hydrogenation report only on aromatic acids, which have no α-hydrogen atoms. Shortening the contact time during the ketonisation reaction favoured ketene formation, indicating that ketene or a ketene-like species functions as an intermediate in the ketonisation. As the formation of this intermediate demands the abstraction of an α-hydrogen atom, it is clear why α-hydrogen is needed.
Carbon-13 and deuterium-labelling experiments elucidated the mechanism of the ketonisation reaction in more detail. The intermediate involved is, most likely, oriented parallel to the surface and interacts with the surface on two sides: on one side via the carboxyl group and on the other side via the α-carbon. For the latter interaction α-hydrogen must be split off first. The carbon-carbon bond of this intermediate can easily be broken, yielding CO2 and a reactive alkylene species, which reacts with a neighbouring carboxylate to a ketone.
The involvement of a Mars and Van Krevelen mechanism in the selective hydrogenation to aldehydes was confirmed convincingly by prereduction experiments: the creation of extra oxygen-vacancies enhanced the selectivity. Oxygen-labelling experiments gave less unambiguous results, but they seem to sustain the same mechanism.
The in the literature advanced mechanisms concerning the ketonisation reaction are not compatible with the results presented in this thesis. However, the mechanism proposed in chapter 9 and represented by reaction scheme 2 (p. 114) explains not only the results presented in this thesis, but also those presented in the literature. It is, therefore, the most plausible mechanism for the ketonisation reaction presented so far.
The literature presents only a very limited insight into the behaviour of carboxylic acids on catalysts. On metallic surfaces, acids are mostly decom¬posed completely. On oxidic surfaces, aromatic acids are known to form aldehydes, but aliphatic acids usually yield ketones. So far, no proposals have been advanced concerning the mechanism of aldehyde formation from aromatic acids. However, numerous mechanisms have been suggested describing the ketone formation, albeit no agreement has been reached yet on this subject. By this thesis, an attempt is made to obtain aldehydes from aliphatic acids and to reveal the mechanism involved as well as the catalysts and reaction conditions needed. Simultaneously, a part of the study has been devoted to solving the mechanism of the ketonisation reaction, as a better understanding of this undesired side-reaction should help in suppressing it.
As model compounds, acetic -, propionic -, isobutyric -, and pivalic acid have been used, which have a number of α-hydrogen atoms ranging from three to zero, respectively. Their reactions on catalysts have been tested under flow conditions, at atmospheric pressure, and in the temperature range from 50 to 450°C. Analysis of the effluent gas has been done by gas chromatography and mass spectrometry.
First, acetic acid was tested over a broad range of oxidic catalysts. Contrary to what the existing literature claims, it appears that, when using oxides with a moderate metal-oxygen bond strength, acetic acid can be hydrogenated to aldehyde. A plot of the aldehyde selectivity as a function of the metal-oxygen bond strength of the catalyst yields a volcano-shaped curve, suggesting the involvement of a Mars and Van Krevelen mechanism. The most common side-product was acetone, the formation of which occurred clearly via two different reaction path-ways. Oxides with a low lattice energy formed bulk acetates, decomposition of which yielded acetone. On oxides with a high lattice energy, a surface reaction to acetone occurred.
Two catalysts exhibited a very high selectivity: copper - and iron oxide. During the reaction, both these catalysts probably consisted of a mixture of metal and oxide.
To investigate further the precise roles of the metal and the oxide, platinum was added to various oxides. Although platinum itself was not active in aldehyde formation, addition of it to oxides remarkably increased the aldehyde selectivity and decreased the reaction temperature. However, the above-mentioned volcano-shaped correlation was not affected by platinum addition and the hydrogenation reaction still proceeded on the oxide and not on the metal. All reasonable explanations for these phenomena can be rejected, except the possibility that platinum is responsible for the activation of hydrogen, which subsequently spills over to the oxide, where the actual hydrogenation reaction takes place.
Since iron oxide was the most selective catalyst, it was subjected to a more detailed study. As already mentioned, it appeared that a selective iron catalyst consisted of both a metallic and an oxidic phase. The metallic iron probably fulfils the same role as platinum does in the above-described catalysts. The importance of the reductant was studied further. Carbon monoxide could not substitute hydrogen, and the hydrogen/acid ratio had to be high enough to keep the catalyst in its, partly reduced, selective state. In the literature, no aldehyde formation from aliphatic acids was ever reported on iron oxide, as always a too low hydrogen/acid ratio was used.
By using acids with a different number of α-hydrogen atoms, it has been revealed that the presence of α-hydrogen is needed for the surface reaction to ketones. An acid without α-hydrogen atoms cannot ketonise and is, therefore, more easily hydrogenated to its corresponding aldehyde. This explains why all patents dealing with carboxylic acid hydrogenation report only on aromatic acids, which have no α-hydrogen atoms. Shortening the contact time during the ketonisation reaction favoured ketene formation, indicating that ketene or a ketene-like species functions as an intermediate in the ketonisation. As the formation of this intermediate demands the abstraction of an α-hydrogen atom, it is clear why α-hydrogen is needed.
Carbon-13 and deuterium-labelling experiments elucidated the mechanism of the ketonisation reaction in more detail. The intermediate involved is, most likely, oriented parallel to the surface and interacts with the surface on two sides: on one side via the carboxyl group and on the other side via the α-carbon. For the latter interaction α-hydrogen must be split off first. The carbon-carbon bond of this intermediate can easily be broken, yielding CO2 and a reactive alkylene species, which reacts with a neighbouring carboxylate to a ketone.
The involvement of a Mars and Van Krevelen mechanism in the selective hydrogenation to aldehydes was confirmed convincingly by prereduction experiments: the creation of extra oxygen-vacancies enhanced the selectivity. Oxygen-labelling experiments gave less unambiguous results, but they seem to sustain the same mechanism.
The in the literature advanced mechanisms concerning the ketonisation reaction are not compatible with the results presented in this thesis. However, the mechanism proposed in chapter 9 and represented by reaction scheme 2 (p. 114) explains not only the results presented in this thesis, but also those presented in the literature. It is, therefore, the most plausible mechanism for the ketonisation reaction presented so far.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Awarding Institution |
|
Supervisors/Advisors |
|
Thesis sponsors | |
Award date | 12 Oct 1995 |
Place of Publication | Leiden |
Print ISBNs | 978-90-386-3504-0 |
Publication status | Published - 12 Oct 1995 |
Externally published | Yes |