This is exciting news, I love the idea of efficiently converting free energy into something we can use. I still wonder if charging batteries makes more sense.

@ guzmanchinky - "I still wonder if charging batteries makes more sense." Of *course* it does and by a factor of at least 5... which is why H2 will never be able to compete with ever-improving battery technology except for relatively niche markets. But that is only the beginning of why H2 is, effectively, a dead duck for transportation.
Some people might feel happy flying about in an aircraft with a tank full of massively compressed H2, one of the most reactive elements known to mankind, that also happens to be one of the smallest making it next to impossible to 100% contain (leaks?), bolted in somewhere. I am not!

It would be much more efficient to directly charge a battery from a solar array. The charging efficiency matches the efficiency of the solar cell:

The solar to battery charging efficiency was 8.5%, which was nearly the same as the solar cell efficiency, leading to potential loss-free energy transfer to the battery. ... This approach led to a high overall efficiency of 9.36% (average 8.52%) (Figure 2D) and storage efficiency of $77.2% at 0.5C discharge.

https://www.cell.com/joule/pdf/S2542-4351(18)30143-0.pdf

I'm very happy to read the process is "using relatively inexpensive semiconductors." Then puzzled to find platinum listed. As far as I know, platinum is neither inexpensive nor semiconductor. What am I missing?

some how i just can't see using up earths water supply to power out toys as being a smart idea.

Do I understand correctly:
this is really just a cell backed into a watter splitting chamber where the solar cell(s) energy forms one side and the path to ground generates the hydrogen (and oxygen on the other side?) how is this different from conducting the electricity over to a separate device to generate the hydrogen?

To be sure, there are other important advantages aside from the impressive 17.6 percent efficiency:
“Significant cost benefits could be achieved through the use of the solar-to-hydrogen approach as it avoids the need for added power and network infrastructure necessary when hydrogen is instead produced using an electrolyser. And by avoiding the need to convert solar power from DC to AC power and back again, in addition to avoiding power transmission losses, the direct conversion of solar energy into hydrogen can achieve a higher overall efficiency for the total process." -- Siva Karuturi

Charging batteries makes more sense for short term energy storage like transportation. Hydrogen, (H2) or Anhydrous Ammonia, (NH3) is a long term energy storage solution for the grid. Big picture problem is there is a shortage of wind and solar energy produced during the coldest months of the year, and an excess during the warmest. Storing electricity for 6 months with a battery is not a good solution. NH3, no problem. @Stegmeier water (H2O) is not lost, it is just separated into H2 and O2, to be recombined back into H2O when the H2 is oxidized or "burned".

Martin Winlow hit nail on head ... dude, think about it, hyped since auto electrolyzing kits hit big in 2005 ( never mind the “seven” auto computers immediately defeated the ignition settings) Plus nobody seems high on plugs for safe storage (metal hydride crystal in tanks). Lost cause, really. Solid state batteries have crossed or is crossing brittle, conduction, efficient anode/cathodes, And Soon ... Cost! Yea! Right!? H2 who’s kidding whom?

Loz - pretty good article. These developments are the baby steps to improved solar utilization, and one day the 20% efficiency will look like old wasteful technology. I've seen efficiency numbers go up with modalities are married - dualsun (french company) markets pv with hydrothermal backing layer that can provide hot water while mitigating the pv panel's temperature (pv is most efficient in at the 25 C standard test temperature and for each degree above or below, it drops by 0.1 to 0.3%. If the thermal gain tracks similarly on these panels, there should be no reason not to utilize the hydrothermal backing to mitigate temps and improve efficiency.

A shout out to martinwinlow: I've read of people who were so certain that innovations would never equal proven technologies. My favorite is when some Rockefeller was marketing lubricating oil and some new lamp oil - kerosene - at the same time automobiles were starting to appear in towns without massive wealth. A stock market tip of the day was "Invest in buggy whips, the noisy and dirty horseless carriages are a fad like penny farthings and will soon disappear!" I might speculate that they thought the kerosene would never rival whale oil for smokeless light production either!