For years, scientists have used artificial photosynthesis to generate clean energy.

This process typically requires electricity or natural gas, but now researchers have developed a new device that's more than twice as efficient and runs only on water and sunlight.

Joining us via Google Hangout to discuss this breakthrough is Zetian Mi, professor of electrical engineering and computer science at the University of Michigan in Ann Arbor.

First of all, plants have been doing this for millions of years or hundreds of thousands of years, right?

How close have we come to being as efficient as Mother Nature?

Okay, that's right.

And nature has been doing this for millions of years, and the current efficiency is about .6%. So, it's actually a very low-efficiency process.

So, as electrical engineers, our job is to make this a much more efficient process.

So, the questions I challenge myself and my students are, 'Can we make this 10 times, 100 times more efficient?'

Actually, if you look at solar cells, currently the solar-cell energy efficiency can reach about 40%-plus.

That's already much more efficient than the natural photosynthesis process.

But the challenge for solar cells is you convert solar energy into electricity.

Electricity is not storable.

So, the cost associated with the battery is actually very high.

That limits the usage of solar cells.

In the artificial photosynthesis process, we convert solar energy into storable, chemical fields -- hydrogen, in this case.

So we can eliminate the cost and also the efficiency loss associated with batteries.

So, instead of having to store the energy from the wind and from the sun in some sort of a giant battery that can eventually be converted to electricity that we want for our car or whatever, you're just saying, convert the sunlight into hydrogen.

Exactly. Into hydrogen.

How kind of energy-intensive is the process?

Because we've had this ability before, but we've had to use fossil fuels, ironically, natural gas or coal-powered electricity, to make this process work.

That's right.

And today, in U.S. alone, we produce nearly 10 million metric tons of hydrogen.

And this is mostly produced from fossil fuels through a process known as steam reforming.

This indeed is a very energy-intensive process.

So, in the artificial photosynthesis process, as we do, so, we use only solar energy.

There's no carbon dioxide emission in this process.

Yeah.

What are the ingredients to get to the hydrogen?

Okay, that's a great question.

Actually, this is a very simple device.

I'm sure you are familiar with solar-cell panels.

Yeah.

So, our device is very similar to a solar-cell panel, except it's actually easier to make.

So, we put nanostructured semiconductors -- in this case, gallium nitride, which is also the material used in LED lighting.

We put them on the silicon wafer, and we tailor the size to be in the nanoscale about .001 of the width of human hair.

And we change the properties so that they become a very efficient photocatalyst.

And then we put this wafer, which could be only a few inch or larger, immerse that in water, and shine sunlight.

Then hydrogen and oxygen bubbles are produced.

You're putting this nanoscale substance on a wafer, sticking it in water, adding sunlight.

That's it.

That's it.

And, you know, gallium nitride has been known for decades.

Gallium nitride is one of the most produced semiconductors, next only to silicon.

But for decades, people cannot transform this material to be a very efficient photocatalyst.

So, what we have discovered is, by growing this using a standard process known as molecular beam epitaxy, we can tailor the material properties to bring the water molecules into hydrogen and oxygen very efficiently.

And we can achieve that by using the nanoscale materials.

If gallium nitride is so widely known and available, I'm assuming it is inexpensive to make this?

Well, so, gallium nitride is so well produced, the cost has been reduced dramatically over the years.

So, for example, use LED lighting as an example.

And we are using the same material, a very similar manufacturing process.

So, that's the most exciting part of our discovery -- the device we build is based on a very scalable manufacturing process.

It uses very well known, very well produced semiconductors.

So, that will make our technology scalable and potentially very low cost.

I'm assuming there's some sort of a container that's capturing the hydrogen and capturing the oxygen and there's no more water that's gonna be left, right?

In the real applications, we can envision that there will be a continuous flow of water and a continuous collection of the hydrogen fields generated, which will be used, you know, for fuel cell vehicles, as an example.

Mm-hmm.

So, basically, this would be the stuff that your hydrogen car can run on.

But as you said, there's also already so much hydrogen being manufactured in a very energy-intensive manner in the United States every year.

So you could actually switch the process to this and have a cleaner way of making hydrogen.

Exactly. Much cleaner way.

And I think the very exciting part of this is we can produce hydrogen directly on site.

We know that the current hydrogen-generation process is a very energy-intensive process, but it's also very challenging for hydrogen distribution and transportation.

Now let's imagine that we can generate hydrogen directly on site whenever, wherever we need that.

That will not only reduce the cost of hydrogen itself, but, more importantly, reduce the cost associated with hydrogen distribution and transportation.

So, give me an example of -- How much hydrogen and oxygen can you generate from, say, an ounce or a gallon of water?

Okay, so, you know, our preliminary calculations have shown that, for example, if we use a 1-by-1-meter-square photocatalyst wafer that we produce and using about 100-by concentrated sunlight, then we can produce a few kilogram of hydrogen per day.

And that's enough to power our fuel cell vehicles for many, many miles.

So [chuckles] you could have one of these 1-by-1-meter wafers in your backyard, add a gallon of water or two to it every couple of days, and have enough fuel to run your car.

I think that's what will happen in the very near future.

All right, Zetian Mi from the University of Michigan, thanks so much for joining us.

Thank you so much for the interview.

[ Computer keys clacking ]