Oxidation and reduction reactions are widely used in chemical industry for the production of chemicals. As many of these processes are highly exothermic, a lot of heat is produced. Because this heat cannot be fully recovered the processes often have a low energy e_ciency. A second issue many of the current processes encounter is low selectivity. This leads to loss of raw materials, and expensive puri_cation steps are needed as downstream processing. The ine_ciency of these processes brings about opportunities in electrochemical production processes. A possible production process is cogeneration of chemicals and electricity, also called electrogenerative hydrogenation. Such a process uses a fuel cell-type reactor. A recent innovation on fuel cell design, co-laminar ow cells, may be bene_cial for cogeneration processes due to their interesting features, such as membraneless construction and pH exibility. A co-laminar ow cell has not been used as a reactor for cogeneration purposes before. The goal of this work is therefore to assess the possibility of cogeneration of chemicals and electricity in this type of microuidic fuel cell. The electrochemical hydrogenation of nitrobenzene is used as a case study. Due to its many possible end-products, as a _rst step in this research a suitable electrocatalyst and reaction conditions have to be found. After these have been found the investigation can move on to the reactor design.
In the _rst phase of the research, carbon supported metal nanoparticle electrocatalysts are studied in as electrocatalysts for the nitrobenzene reduction. As nanoparticles, Cu, Pt an Pt{Cu alloys are studied. Carbon nanotubes, both functionalised and non-functionalised, and two types of activated carbon are considered as possible nanoparticle supports. Di_erent reduction methods are used to obtain the electrocatalysts, and the structure and composition of said electrocatalyts are examined by means of TEM, XRD, XPS and EDX analysis. The electrocatalytic behaviour of the electrocatalysts for the electrochemical reduction of nitrobenzene is then studied by LSV measurements. Cu nanoparticle electrocatalysts proved to have the highest activity for the nitrobenzene reduction, both on activated carbon and on non-functionalised carbon nanotubes. Chronoamperometry and Kout_ecky-Levich analysis veri_ed that both catalysts are selective for the nitrobenzene reduction towards phenylhydroxylamine. Comparing the _rst and 1000th cycle of a CV measurement the catalysts also con_rmed to have a high stability. The activated carbon nanoparticle electrocatalyst featured smaller particles (_ 3nm). Combined with the fact that activated carbon is less expensive and less toxic than carbon nanotubes, this electrocatalyst is chosen as electrocatalyst for the new reactor design. The construction and extensive study of the microuidic reactor for cogeneration entails the second phase of the research. A simple construction is used _rst to prove the concept of cogeneration. Initially the electrochemical behaviour of this reactor, not taking into account the formed products, is studied via LSV measurements. These measurements are used to _nd a suitable concentration and range of ow rates to study the actual cogeneration process. This process is studied by applying an external load of 100 and 1000 to the reactor, during which the reactor produces current and products. The formed products are examined by UV-Vis. It is found that the reactor is capable of producing current and phenylhydroxylamine, but due to oxidant crossover aniline is also produced by the reactor, which is formed at the anode. The methanol oxidation, which is used as a counter reaction, introduced unstable reactor performance due to its self-poisoning behaviour. Therefore, while the proof-of-concept is given, the actual cogeneration performance of the reactor is still sub par. As a _nal step in this research an optimisation is started, by altering the design in three distinct ways. The _rst way the design is changed is the use of ow-through porous electrodes, which increase the mass transport towards the electrodes and the available surface area for reaction. Secondly, by using four parallel channels, the reactor is capable of processing increased ow rates. Finally, a current collector design is used which enables easier leak-free assembly. Due to problems of electrolyte mixing and Ohmic resistance these optimisations did not produce an increase in reactor performance, which means further research is required to increase the cogeneration e_ciency. Overall, the process of cogeneration in a microreactor is promising. This is especially true if the problems with the instability of the reactor performance can be solved, for instance by changing the anode reaction. Because of the low volume that can be processed by the reactor, the applications are most likely in the _eld of _ne chemicals.