Going with the flow

US Engineers have built a fuel cell with no membrane that mimics the flow of toothpaste being squeezed from a tube.

The device could prove more efficient than conventional fuel cell designs for use in consumer electronics and military devices, and cost less to produce, University of Illinois researchers believe.

The team threw away the membrane usually needed to separate the fuel and oxidant by using ‘laminar flow’ behaviour. The fluids flow through micro-scale channels, which induce different flow behaviour compared with large-scale pipes. As a result, the two liquids have no turbulence and so meet without mixing, passing over the anode and cathode in a simple Y-shaped channel.

Researcher Prof Paul Kenis said a key advantage of the fuel cell is that it can use more reactive alkaline chemistry, which is superior to conventional acidic systems in the same way that alkaline batteries perform better. ‘Alkaline chemistry is quicker and the reaction kinetics are better,’ he said. ‘But the majority of alkaline fuel cells are expensive and are used in space programmes. They can only run on ultra-pure hydrogen.’

If less pure fuels are used, CO2 reactions form carbonates that can quickly clog up the membrane and stop it working, he said. ‘In our system, however, the streams are flowing, so any carbonate forming is immediately transported away.’

The cell has no membrane, relying instead on 'laminar flow' behaviour (inset).

Fuel crossover is also a problem for many fuel cells, alkaline systems especially. This process occurs when the fuel that should be reacting with the anode diffuses through the membrane and reacts at the cathode instead. ‘We can control crossover by adjusting the flow rate, which if fast enough can prevent the fuel diffusing across,’ said Kenis. The technology also prevents anode dry-out and cathode flooding because the system is water based.

One challenge for the team before the fuel cell becomes available is to tackle the problem of low oxygen solubility in water in the cell. A fuel cell cycle needs plenty of oxygen, so conventional systems take oxygen out of the surrounding air. The team hopes to engineer its fuel cell to tackle the problem in a similar way, but declined to reveal further details.

Kenis said that laminar flow behaviour has been well known for over a century, but it has not been used in fuel cells before because the flow only occurs in micro-channels and cannot be repeated on a larger scale. This means multiple tiny channels in an array would be needed to provide sufficient power. ‘Instead of scaling up, what you do is scale out,’ he said. ‘The advantage is that if you have many units in an array, and a few fail, the whole system still works.’

The team is building a fully working system, due to be ready in around two years, with its Illinois spin-off, INI Powersystems.