Artificial photosynthesis - a difficult but necessary part of the energy revolution

Why we need to store energy  in liquids

May 28th. It’s a fairly typical early summer’s day in Germany. At noon, PV and wind are providing just over 50% of electricity supply. Solar output will fade sharply over the afternoon and cease completely by about 9.30pm. Although the wind will continue to blow, renewables won’t make much contribution to overnight electricity supply.

Despite this intermittency, we want solar to continue growing. In most parts of the world, though probably not Germany, its annualised costs are now no higher than electricity produced from fossil fuels. But however much we desire PV to succeed, it doesn’t deliver electricity for much of the day, it is unreliable in higher latitudes and it is still requires a huge capital investment to move from old and fully depreciated gas and coal plants to open fields of PV. Even those of us who think that solar is the world’s best hope need to acknowledge that it fits uncomfortably with the energy systems of advanced economies. However cheap solar gets, it doesn’t solve the problem of the need for seasonal storage.

The purpose of this post is to argue for a new focus on artificial photosynthesis and, in particular, investment in the development of low cost technologies to convert sunlight directly into carbon-bearing liquids. Put at its most simplistic, we need techniques that use photons of light to disassociate the hydrogen and oxygen in water and then use the hydrogen protons and electrons to provide the energy to mimic the natural photosynthetic reactions in plants - or, more probably, bacteria - that capture atmospheric CO2 and combine the gas with organic molecules to make sugars and more complex energy carrying chemicals. Artificial photosynthesis (AP) may (eventually) become cheap, be able to use non-potable water and require little land that could be otherwise used for agriculture. In other words, it avoids all the problems bedevilling today’s renewable energy technologies.

A tricky problem

But using solar energy directly to make fuels is an intensely difficult problem. One recent academic paper wrote

The scientific challenges for efficient and globally deployable AP are complex; requiring coupled breakthroughs in light harvesting, charge separation, catalysis, semiconductors, nanotechnology, modelling from synthetic biology and genetic engineering, photochemistry and photophysics, photoelectrochemistry, catalysis, reaction mechanisms and device engineering.

But the very next paragraph of this paper says

In favour of AP is the vast excess of available solar energy compared to present and projected human needs, its capacity to reduce the atmospheric concentration of greenhouse gases and address the problem of intermittent renewable energy (solar pv, wind and hydro) electricity supplies as well as the need for a zero-carbon source for transportation fuels.

 I’ve tried to write before about the importance of using surplus electricity for conversion into methane (natural gas) because gas can be stored easily within the existing infrastructure of pipes, gasometers and exhausted fossil gas fields. ‘Power to gas’ - the conversion of surplus electricity into hydrogen and possibly then on to methane – is an extraordinarily important technology in which the UK has no shown no interest whatsoever. It faces two obstacles in addition to policy neglect. First, neither the UK nor any other country has any significant investment in the infrastructure to store hydrogen. Second, hydrogen isn’t particularly good as an energy carrier, delivering relatively little power per unit weight or volume. But using the hydrogen  to make methane looks much more interesting, although this requires a dense source of CO2, such as flue gases.

In this article, I want to put forward the view that we should also be pushing for expansion of research into the direct capture of the photons of sunlight and their conversion not into gas, but into liquids. Liquids have high energy density (kWh per kilogramme and per litre) and are also easy to store within today’s economy. At this moment, for example, the UK has oil in store equivalent to about 80 days consumption. In other words, we have the pipes and tanks to take seasonal surpluses of electricity from PV and keep that energy in the form of energy-rich liquids.

Time to start sponsoring research

After decades of statis, recent progress in academic and commercial research in artificial photosynthesis is exciting. I guess we are about 15-20 years from cheap and easily deployable systems that be used everywhere around the world to provide abundant energy to supplement solar. I hear you immediately saying ‘but we don’t have 15-20 years’. Probably true, but we must not let that stop us engaging in urgent research. There is no conceivable alternative in high latitude countries if we want a modern economy and abundant and cheap energy available throughout the year. (I am disregarding nuclear power as an alternative because it is looking increasingly impossible in much of the world, even including the UK).

In European countries modern life requires the continuous supply of about 4 kilowatts of energy per person. (Not just electricity, which is less than half this, even including conversion losses turning fossil fuels into power). We can probably compress this by switching to electric cars (3 times as energy efficient as petrol) and possibly by using heat pumps (up to 3 times as much useful heat per unit of input electricity as a resistive heater). But we probably cannot get the total energy need much below 3 kilowatts a person in the near future. (That’s the equivalent of two electric kettle working continuously for each of us).

Provided we could store the surplus power when the sun is shining, we could get this electricity from installing about 30 kilowatts of PV per head. (The average British PV panel generates about 10% of its maximum capacity over the course of the year). The UK has currently about 0.12 kW per head, by the way, and Germany about 0.45 kW or between 0.4% and 1.5% of what is needed.

Batteries will be useful. But the size and cost will be enormous. Just to store the surplus power of a sunny day like today will require over 100 kWh of lithium ion battery per person. This is more than in the most powerful electric cars on the road today. Batteries like this have poor energy density, needing nearly a hundred times as much weight as petrol to store the same amount of energy and almost twenty times the space. The unfortunate truth is that batteries are fine for overnight electricity storage so that you can run the washing machine at 10pm from power harvested at noon that day. However they will never have any role, absent quite unexpected technological developments, in storing Britain’s summer photons for a dull December. That’s where power to gas and artificial photosynthesis come in.

Here’s brief details of three new developments in artificial photosynthesis, using contrasting approaches and with very different strengths and weaknesses.


The nearest entity to commercial production of fuel from sunlight is probably Joule Unlimited, a company headquartered near Boston and with production facilities in New Mexico. Joule’s technology uses transparent tubes of brackish water into which CO2 is flushed. The tubes are filled with genetically engineered microbes that take energy from sunlight to capture the CO2 and then excrete a liquid fuel.

The advantages of this approach include very high levels of productivity. The company claims it can produce 40,000 litres of ethanol per hectare a year. This implies a very high degree of photosynthetic efficiency – far greater than conventional green plants – and approaches the energy yield of the same area full of solar panels. The process is cheap and isn’t limited by having to use scarce water or minerals. Some openly doubt whether Joule’s technology will ever work but the company has just raised another $40m to add to the $120m already invested. This money will help build the first commercial plant. Rumours suggest that the recent fund-raising round included GE and other major international corporations, including Audi, an earlier backer.

The disadvantage of the Joule approach is that it needs a relatively dense source of CO2, such as the flue gas from a cement works or coal-fired power station. In other words, it can be argued that the technology requires the generation of CO2 in order to operate. Unless the CO2 from burning Joule’s ethanol is itself captured, unlike natural photosynthesis this process doesn’t reduce atmospheric carbon dioxide levels. But, if it works, Joule’s approach is inexpensive and effective.

Daniel Nocera at Harvard

Nocera and his colleagues published a paper in February 2015 that demonstrated a technique for using electricity – perhaps generated by solar panels – to split water into hydrogen and oxygen. A genetically modified bacterium uses the H2 to begin the conventional (and chemically complex) process of photosynthesis. Instead of producing biomass, as in conventional bacterial growth, the bacterium generates isopropanol, a potential carbon fuel.


Nocera’s team have made huge advances. The splitting of water is done using cheap electrodes and at a low voltage. Yields are still low, at least by comparison to the generation of electricity from PV, but already at least as good as in plant photosynthesis, which is usually no better than 1 or 2%. So, if this approach can be commercially exploited, more energy will be generated than could ever be generated by making biofuels from fermentation processes. The capital costs will be relatively low, but the disadvantage of this approach is that needs the delivery of electricity, rather than the direct use of sunlight, as well as a feed of concentrated CO2. Even if the Nocera process can be made 5% efficient at converting energy into carbon-bearing molecules, it will still require huge areas of solar panels to provide the fuel for each person’s energy needs.

Yang’s team at Berkeley

Professor Yang’s group published a paper last month that showed a mesh of interlocked microscopic nanowires made of silicon and titanium oxide which is populated by bacteria. (A different genus to Nocera’s bacterium). The nanowires absorb photons from sunlight and generate electrons. These  are captured by the bacteria and used to turn the CO2 being bubbled though the solution surrounding the mesh into more complex carbon-bearing elements.

The Yang team are excited by these results. One comment was ‘We believe our system is a revolutionary leap forward in the field of artificial photosynthesis’. The technology seems potentially inexpensive, robust and may have an energy yield that is as much as 10% of the energy initially provided by sunlight. But like the other two technologies, it needs sources of CO2 more concentrated than in the atmosphere.


There’s a long way to travel before artificial photosynthesis becomes an economically competitive means for storing surplus solar (or wind) energy. We eventually find it is better to focus on using surplus electricity in summer to turn into hydrogen and then into methane.

Power to gas, perhaps using the technology pioneered by Electrochaea may be able to convert more than 50% of the energy supplied as electricity into the chemical energy of methane. This is far better than the conversion efficiencies implied in the two academic results published this year.

Nevertheless, the arguments for spending research money on artificial photosynthesis remain strong. Without a means of energy storage for months at a time, the renewables revolution will stall. Most importantly, we need to be able to extract CO2 from air in order to reduce atmospheric concentrations.  Artificial photosynthesis is a hugely complicated problem to solve but the UK science base needs to get engaged in this vital area. At present, most of the research work is going on in the US (usually paid for by Federal money).It needs to be a similar priority here.