In our paper “Overcoming chemical equilibrium limitations using a thermodynamically-reversible chemical reactor” we show how using a non-stoichiometric oxygen carrier material it is possible to overcome the overall equilibrium limitations of a conventional reactor. This requires the transfer of key chemical information across the half cycles of a periodically-fed reactor, achieving full conversion and producing pure, separated product streams.
One of the most challenging parts of this study was showing how the oxygen content of the material changed during reaction. The material stores the chemical information that is transferred between half cycles. We decided to perform this by following dynamically the lattice parameter of the oxide using XRD (the lattice parameter is related to oxygen content). But for this to be possible with a suitable time constant and sufficiently high precision we needed to perform this at a synchrotron. In addition we wanted, while performing these experiments, to run a reactor and produce hydrogen. This necessitated a trip to ID22 of the European Synchrotron Radiation Facility, ESRF.
So we designed and built a custom furnace and flow system (weighing 200 kg!) capable of heating the reactor (which contained a few grams of oxygen carrier). The furnace also need to allow the x-rays to reach the carrier and pass through to the detectors. The furnace took up to 8 hours to heat and cool so it was designed to operate with two parallel reactors so, should problems occur with the first reactor, a second reactor was at temperature and ready to undergo reaction and measurement.
Custom furnace with slits to allow both reactor beds to be studied
Then the setup, along with the flow system, gas analysis equipment, samples and spare parts were packed away and loaded into a van and we drove (some of the team with less stamina flew!) the 1500 km from Newcastle upon Tyne in the north of England down to Grenoble in the south of France.
The flow system in the van ready for transportation
We arrived on a machine day which meant we had 24 hours to rebuild the furnace and flow system on top of (and hanging from!) the positioner table, carry out leak tests and heat up the reactor before our 72 hours of beam time began.
Dr Ray, Dr Mak and Dr Papaioannou reconstructing the experiment set up inside ID22
Problems plagued the initial 24 hours of beam time with movement of the positioner table leading to cracks forming in the seals of both heated reactors. Once we had reinforced the reactor seals and slowed the speed of the table we were ready to start.
Over the next 48 hours we collected the required data, showing how the oxygen content of the material changed as it underwent repeated oxidation and reduction. The ability to carry out operando XRD on a working reactor allowed us to determine how the oxygen chemical potential of the carrier within the reactor changed and shifted with reaction. The results confirm the idea that each end of the reactor retains a memory of the feed gas it is exposed to. This supplies the information needed to the subsequent product stream that exits from that end of the bed. Without this transfer of memory the reactor would simply not work. Please see the full paper and a high-resolution video explaining the principle of operation.
We are immensely grateful to the staff at ID22 for their help and patience as we worked with them to carry out this work. We are also grateful to ESRF, EPSRC and the ERC for funding.
200 kg of equipment placed on the high precision positioner table