[subnaught]

That's an apples-and-oranges comparison. The system you cite differs from the one under discussion in several ways; the most important of which are:

1. The cited system does not include a light-harvesting component. It merely postulates that the required energy could be generated from photovoltaics. This would introduce additional cost and complexity along with an efficiency hit.

2. The cited system comprises a bacterium in conjuction with an electrode-supported catalyst, whereas the system under discussion is solely an engineered bacterium.

Finally, it is not correct to refer to cadmium and cysteine as feedstocks. They are components of the catalyst, and they are not consumed during catalysis. The only feedstocks for both systems are CO2 and water.

[spqr0a1]

You're right that Liu's paper does use external electrodes. And as for Sakimoto's paper it's true that the cadmium should be reusable indefinitely, but if you read the second page it's clear than cysteine is consumed in stoichiometric amounts. I'd be excited to hear about more recent work which improves upon this and would be happy to be corrected.