CATL to build battery production base in Guangdong; US$1.9B, 25 GWh in first phase, 150 GWh by end of 2030
MIT study finds Uber & Lyft increase congestion, decrease transit ridership and don’t affect vehicle ownership (updated w/ Uber response)

Fraunhofer suggests e-scooters as application for its magnesium hydride paste hydrogen storage technology

Researchers at Germany’s Fraunhofer Institute for Manufacturing Technology and Advanced Materials IFAM in Dresden have developed an ultra-high-capacity hydrogen storage substance for PEM fuel cell applications based on solid magnesium hydride. Fraunhofer’s POWERPASTE releases hydrogen on contact with water. It has a hydrogen capacity of about 10 mass-% (i.e. 10 kg POWERPASTE → 1 kg hydrogen). This is a specific energy of 1.6 kWh/kg and an energy density of 1.9 kWh/liter—about 10 times the capacity of Li-Ion batteries.


Specific energies and energy densities including conversion losses. Source: Fraunhofer IFAM

The award-winning POWERPASTE technology is patented (EU, US) and offers advantages over other energy storage technologies, in particular in the power range from 100 W to 10 kW, Fraunhofer researchers said.

Now, Fraunhofer is suggesting the use of POWERPASTE for use with e-scooters. Hydrogen is not currently an option for small vehicles such as electric scooters and motorcycles, since the pressure surge during refilling would be too great. POWERPASTE, on the other hand, the researchers suggest, would be ideal.


POWERPASTE (left); POWERPASTE cartridge (middle); portable 100 W power supply unit (right). Source: Fraunhofer IFAM

POWERPASTE stores hydrogen in a chemical form at room temperature and atmospheric pressure to be then released on demand, said Dr. Marcus Vogt, research associate at Fraunhofer IFAM. POWERPASTE only begins to decompose at temperatures of around 250 °C, so it remains safe even when an e-scooter stands in the sun for hours.

Refueling would be simple; riders merely have to replace an empty cartridge with a new one and then refill a tank with water. This can be done either at home or underway.

To create POWERPASTE, magnesium powder is combined with hydrogen to form magnesium hydride in a process conducted at 350 °C and five to six times atmospheric pressure. An ester and a metal salt are then added in order to form the finished product.

Onboard the vehicle, the POWERPASTE is released from a cartridge by means of a plunger. When water is added from an onboard tank, the ensuing reaction generates hydrogen gas in a quantity dynamically adjusted to the actual requirements of the fuel cell. Only half of the hydrogen originates from the POWERPASTE; the rest comes from the added water.


POWERPASTE thus has a huge energy storage density. It is substantially higher than that of a 700 bar high-pressure tank. And compared to batteries, it has ten times the energy storage density.

—Marcus Vogt

This means that POWERPASTE can offers a range comparable to—or even greater than—gasoline. It also provides a higher range than compressed hydrogen at a pressure of 700 bar.

With its huge energy storage density, POWERPASTE is also an interesting option for cars, delivery vehicles and range extenders in battery-powered electric vehicles, Fraunhofer suggests. Similarly, it could also significantly extend the flight time of large drones, which would thereby be able to fly for several hours rather than a mere 20 minutes. This would be especially useful for survey work, such as the inspection of forestry or power lines. In another kind of application, campers might also use POWERPASTE in a fuel cell to generate electricity to power a coffeemaker or toaster.

Unlike gaseous hydrogen, POWERPASTE does not require a costly infrastructure. This makes it suited for areas lacking such an infrastructure. In places where there are no hydrogen stations, regular filling stations could sell POWERPASTE in cartridges or canisters instead. The paste is fluid and pumpable. It can therefore be supplied by a standard filling line, using relatively inexpensive equipment.

Initially, Fraunhofer suggests, filling stations could supply smaller quantities of POWERPASTE—from a metal drum, for example—and then expand in line with demand. This would require capital expenditure of several tens of thousands of euros. By way of comparison, a filling station to pump hydrogen at high pressure currently costs between one and two million euros for each fuel pump. POWERPASTE is also cheap to transport, since no costly high-pressure tanks are involved nor the use of extremely cold liquid hydrogen.

Fraunhofer IFAM is currently building a production plant for POWERPASTE at the Fraunhofer Project Center for Energy Storage and Systems ZESS. Scheduled to go into operation in 2021, this new facility will be able to produce up to four tons of POWERPASTE a year.


Erik Tigerstedt

Sounds like this is the solution to most of our energy problems.

interesting of course how much energy it takes to make the paste and can the MgO be recycled.

Could the paste be made locally?

Why the 100-10kW range?


I also wondered why they thought that this technique would be limited to 100 W to 10 kW. Anyway, this would a even more energy inefficient power source than just using hydrogen in fuel cells as you not only need to generate the hydrogen, you also need to use energy to convert the magnesium oxide back to magnesium. This technique might be useful for some specialized applications where the cost of energy is not a major consideration. Maybe specialized military drone aircraft? However, I doubt that this technique will find a wide application.


What they have not specified is the power output, as opposed to the energy.

I would have thought low power density the likely reason for the range 100W-10KW.

It seems much of the energy to recycle magnesium oxide to magnesium can be concentrated solar, so at least from a GHG POV it would appear to be very attractive:


I would have thought e-scooters are OK as they are with LiIon batteries.
The "portable 100w" system looks rather large to stick in a scooter.
Military applications, perhaps.
I wonder how long the paste keeps ?
Backup power supply?
IMO to run a car on a motorway (at moderate speeds) you need ~15 kW.
+ campers can just use propane and not worry about the CO2 as the amounts are minuscule (in the grand scheme of things) (As long as you are not in the same campsite as Greta).
I am sure there are military applications as they are not so cost conscious.
Looks like a solution looking for a problem to me.

Account Deleted

Metal Hydrides look like the next step in Fuel Cell Tech, even though they have been working on this for over 40 years!
Power density should not be a concern since the discharge rate looks good enough to drive large fuel cells. It appears that the Fraunhofer Institute for Manufacturing Technology and Advanced Materials does not want to compete with Compressed H2 (read here: - use Google Chrome to translate).
One important note is that this is a Magnesium Hydride Hydolysis Reaction with half the hydrogen coming from water. The semi-solid POWERPASTE contains magnesium hydride and non noble metal salts as well as a non-toxic ester, there is more detail in this Whitepaper:

What I would like to see is no new Infrastructure. Something like Professor Kondo-Francois Aguey-Zinsou Hydrogen battery which also uses a metal salt catalyzed Magnesium Hydride H2 storage (read about it here: Now just use a reversible fuel cell to reduce costs.


Power density is the problem.


Thanks for the links, and the typically informed analysis, gryf

SJC - as gryf has shown, discharge rates and hence power density is fine.

The only think that I can add is that from the technical paper costs per kg are given as $2, and a 1kg has 1.6KWH of energy

Account Deleted

Hydrogen production or discharge rate should not be a problem. Higher rates can be achieved by multiple storage tubes for example. Also, Metal Hydrides have already been used in submarines (Howaldtswerke-Deutsche Werft GmbH U212) and grid applications (Toshiba H2One). However, these were low density Lanthanum MIschmetal Nickel alloy H2 storage. There have been other projects that have used Magnesium Hydride (Check: "Application of hydrides in hydrogen storage and compression: Achievements, outlook and perspectives", Volume 44, Issue 15, 22 March 2019, Pages 7780-7808 -

Account Deleted

Costs are probably the biggest concern (the LAVO system costs $35,000 AUD).
So low cost transportation BEV probably is the best approach. Long range applications. e.g. trucking, ships, air which also have expensive powerplants would be better.

Lars Metkemeyer

"Researchers at Germany’s Fraunhofer Institute for Manufacturing Technology and Advanced Materials IFAM in Dresden have developed an ultra-high-capacity hydrogen storage substance for PEM fuel cell applications based on solid magnesium hydride. Fraunhofer’s POWERPASTE releases hydrogen on contact with water. It has a hydrogen capacity of about 10 mass-% (i.e. 10 kg POWERPASTE → 1 kg hydrogen). This is a specific energy of 1.6 kWh/kg and an energy density of 1.9 kWh/liter—about 10 times the capacity of Li-Ion batteries."


The 100W PSU example is hardly adequate for a scooter application, given that eBikes typically have between 250-750W motors (higher for off-road applications). And that power supply unit looks too big to go on a scooter (the scale can only be guessed at from the plants in the background), even if its a prototype that could be shrunk to halve-size in a production unit.

Also their power density figure is for the paste itself and does not take into account the weight of the PEM fuel cell and the PSU to extract the H2 from the paste.

Per Davemart: "$2 for 1.6KWH". For the equivalent Li-ion battery the cost to charge would be 20 cents.


lightweight applications...
Power density is the problem.

Account Deleted

Maybe a real Hydrogen EBike may help you visualize what can be accomplished with this H2 Storage.
Professor Kondo-Francois Aguey-Zinsou built a hydrogen powered eBike in 2014 using Hydride storage. (Read here: "Australia's first fuel cell bicycle", This eBike Specs were: Battery: 518 Wh Lithium-ion battery, Fuel cell power: 100W, Canister: 738 Wh capacity.
Fraunhofer has in a test environment used 300 W Fuel Cells which could charge a 1kW battery (PowerDensity). Larger Fuel Cells, e.g., 750 W could charge a 2.25 kW battery/electric motor system.
The problem is where do you get a POWERPASTE canister outside Fraunhofer?
Professor Kondo-Francois Aguey-Zinsou solves this with a "Hydrogen Battery" (the LAVO system has a FC and electrolyzer). A Unitized PEM Regenerative Fuel Cell would be a better solution. You can read some recent research here:
"A Novel Stack Approach to Enable High Round Trip Efficiencies in Unitized PEM Regenerative Fuel Cells"


For commercial aviation the Powerpaste 6X higher "fuel" weight will limit range to roughly 3000km unless speed is reduced and propfans are utilized, allowing perhaps 4000km range, but this is still much better than Lithium batteries. How much will it cost compared to other carbon neutral options?

Erik Tigerstedt

Ammonia as a source for hydrogen is actually cheaper. But the semi solid form has its advantages.


@GdB, if you could make a plane that could fly 3000km on H2, you would be a hero.
This is enough for most intra-Europe flights.
If you wanted to fly to the USA, you could either use traditional fuels or stop in Newfoundland to refuel as they used to do.
Also, for very long flights, you want to go at about M 0.82 - 0.85, so you need lots of power.
In terms of cost, it is really competing with biofuel based aviation fuels, which I would guess are expensive - you are certainly not competing with mains electricity for aircraft (!)
IMO, E-bikes and e-scooters are good enough as they are, and will only get better as LiIon batteries improve.
(As are push bikes, BTW)


Where scooters are heavily used include many, many people in high rise apartments in densely populated cities.

You don't want to be lugging a battery up and down stairs to charge it.

Hence snap in cannisters are a good solution.

Since the actual energy usage is so low, cost etc are very secondary consideration.

That is why scooters are a chosen early application.

Account Deleted

Long range commercial aviation can use low NOx, Drop-in Renewable JP8 fuels.
Fraunhofer is not going to compete with Toyota, Mercedes Benz, and the German government on mature tech Class 8 FCEV Compressed H2 trucks.

So here are some specs for a Scooter/Moto based on the very good Zero ZF7.2 eMoto. Current Specs: 34 kW, 7.2 kWh battery(weight around 40 kg), Range: 91 miles, Cost:$11k.

POWERPASTE eMoto: 8 kW Fuel Cell (4 kg weight), 1.2 kg POWERPASTE canisters (12.5 kg), Guoxuan High Tech JTM LiFePo battery (cheap as PbA, 14 kg). Total Energy store weight= 30.5 kg. Range (based on 19.2 kWh H2) = 243 miles.



I am finding it confusing what weight contribution the water makes, which provides much of the hydrogen, and also to a lesser extent the cost and energy figures as the boundaries are not always quite clear.

How much would the water cannister for the Scooter/Moto weigh?

Account Deleted

Have not seen a specific reference to component weights (MgH2 and H2O).
Did notice a comment on the Fraunhofer web site ( that states:"If water is available, gravimetric energy storage densities of more than 2.3 kWh / kg can be achieved - including all conversion losses - which corresponds to a gravimetric hydrogen storage capacity of approx. 15% by weight ."
Will continue to research your question.

Account Deleted

Two more references for POWERPASTE.
1. Explains the 2.3kWh/kg (recycling FC H2O output), "Hydrolysis of Mgh2 For ultra high energy applications",
2. PowerPaste for infrastructure-independent hydrogen and energy solutions,
Probably, looks like 1600 Wh/kg is combined MgH2 + H2O.


I think there is an error in the gasoline energy density stated by Fraunhofer.
It is not 1600 wh/Kg = 5.75 MJ/kg but it is 46.4 MJ/Kg.


@Waltersteffe1 The gasoline energy density figure they use is a 'funny money' one, which they have messed around by putting in assumed efficiencies of 17% percent to from the ICE to electric, or the drivetrain or something or other.

Account Deleted

You can read in my earlier post on the Fraunhofer Whitepaper (like Davemart says) that this is based on: "At an efficiency of ~ 0.17. Conversion with an internal combustion engine of a 1 kW system under realistic load changes."
Check "Engine Efficiency" on Wikipedia or other sources, and the average internal combustion engine efficiency is around 20%, subtracting mechanical losses and 17% is reasonable.
Forget 46.4 MJ/Kg, the only way to get that much energy is to light a match!


Ok, it makes sense. But it would have been better to compare directly energy densities of fuels. ICE efficiency depends on the kind of engine. In example Atkinson engine used in hybrid cars is more efficient. I think that it is closer to 30 % than 20 %.
Mechanical losses are the same for all solutions and there is no need to include them in the comparison.

The comments to this entry are closed.