Argonne supercomputer helped Rice/Minnesota team identify materials to improve fuel production
Li-ion maker Electrovaya acquires Evonik Litarion and licenses SEPARION IP

Toyota pops the hood on the technology of the fuel cell Mirai at SAE World Congress

The new fuel cell stack in Mirai increases the current density by a factor of 2.4 compared to the conventional FC stack. Konno et al. Click to enlarge.

At SAE 2015 World Congress last week, Toyota presented a set of four technical papers describing some of the technology innovations used in its production fuel cell hybrid electric vehicle Mirai (earlier post). The papers provide technical details on the high performance fuel-cell (FC) stack; specific insights into FC separator, and stack manifold; the newly developed boost converter; and the new high-pressure hydrogen storage system with innovative carbon fiber windings.

The Toyota papers were part of a larger World Congress technical session on practical hydrogen fuel cell technology: PFL 720, Advances in Fuel Cell Vehicle Applications, chaired by Jesse Schneider of BMW.

Toyota has been developing fuel cell technology for more than 20 years, culminating with the introduction of Mirai in December 2014. An major step along the way was the release of the “FCHV-adv” in 2008 (earlier post), which delivered major improvements in efficiency, driving range (more than double its predecssor), durability, and cold start capability, compared to its predecessor. Despite those improvements, further further reductions in size and cost as well as enhanced performance were necessary for commercial adoption. A major step toward addressing those challenges was Toyota’s development of the new FC stack used in Mirai—the first FC stack without a humidifying system.

The Mirai stack features an innovative cell flow field structure and membrane electrode and gas diffusion layer assembly (MEGA), delivering a compact, high performance unit. Other cost reduction measures incorporated by the new FC stack include reducing the amount of platinum in the catalyst by two-thirds and adopting a carbon nano-coating for the separator surface treatment.

The Polymer Electrolyte Fuel Cells (PEFCs) used in automotive applications enable low temperature start-up and operation, fast reaction speeds, and high current densities. However, reducing proton conductivity resistance to enable high performance requires maintaining humidity within the electrolyte membrane and catalyst layer (CL). Conventionally, this is handled by recirculating water from the FC stack air outlet to the FC stack air inlet through a humidifier.

To eliminate this subsystem, Toyota engineers developed an innovative integrated cell flow field structure and membrane electrode and gas diffusion layer assembly (MEGA) that self-humidifies.

Unlike conventional strait channel and porous metal flow fields, the new Toyota fuel cell uses a 3D fine-mesh flow field in the cathode. The 3D micro-lattice directs air toward the MEGA and promotes O2 diffusion to the catalyst layer. The designers optimized the geometry and surface wettability of the flow field to draw water generated by the MEGA to the back surface of the 3D flow field.

Conventional flow field structure. Konno et al. Click to enlarge.   New 3D fine mesh flow field structure. Konno et al. Click to enlarge.

For the anode, Toyota engineers used an integrated channel-based fine-pitch flow field structure with H2 flow on the front and coolant flow on the back. This creates a 2-turn, 3-step cascade microstructure in which the H2 and air flow in counter directions on either side of the membrane electrode assembly (MEA).

This structure provides several important functions to enable self-humidification. First, generated water is recycled from the air outlet through back diffusion to humidify the H2 inlet. Second, the air inlet is humidified through back diffusion from water vapor in the anode. The H2 pump plays a critical support role by recirculating water emitted from the H2 outlet as vapor and transferring water generated by back diffusion from the H2 inlet to the outlet. Finally, the flow field structure is also designed to provide increased coolant flow and heat transfer.

—Konno et al.

Outline of self-humidification within cell surface. Konno et al. Click to enlarge.

To enhance performance without a humidifier, Toyota reduced the electrolyte membrane thickness by two-thirds, promoting the generation of water by back diffusion and increasing the proton conductivity by a factor of at least 3. Other measures restrict chemical deterioration of the electrolyte membrane.

In the catalyst layer, Toyota lowered the equivalent weight (EW) of the ionomers (ionomers are ion-containing polymers)—i.e., by increasing the functional groups—and also optimized the ratio of the ionomers to enhance proton conductivity and gas diffusion performance. Catalyst activity was boosted by a factor of 1.8 by optimizing the platinum (Pt)/cobalt (Co) alloy ratio and adopting an acid treatment.

Toyota switched from a hollow to a solid carbon support, allowing the Pt catalyst to be supported on the carbon surface, reducing the O2 diffusion resistance and increasing the effective utilization rate of Pt by approximately 2.0. Pt particle size and distribution were also optimized.

The redesigned cell flow field and MEGA structures result in:

  • Lower concentration overvoltage due to improved gas diffusion;
  • Lower resistance overvoltage due to improved proton conductivity; and
  • Lower active overvoltage due to improved catalytic activity.

As a result, the Mirai stack increases the current density by a factor of 2.4 compared to the conventional FC stack.

Also as a result of the new design and higher current density, the maximum power of the new stack increased from 90 kW to 114 kW (27% in total, a per cell increase of 36%). At the same time, cell thickness was reduced by 20%. As a result, cell volume was reduced by 24%.

Toyota also switched the separator material from stainless steel to titanium, reducing cell weight by 39%. Changing from a two-layer to a single-layer structure and halving the area of the compression plates reduced the size and weight of the stack compression structure. A stamped stack case was replaced by an aluminum casting and the number of compression parts was reduced by incorporating measures to increase structural strength of the case. As a result, the volume and weight of the case was reduced by 42% (from 64 to 37L and 108 to 56kg).

The Mirai FC system achieves a power density of 3.1 kW/L and 2.0 kW/kg, more than twice that of the conventional stack, and one of the highest in the world. With its compact dimensions, the FC stack fits under the seats.

Toyota achieved major cost reductions by reducing the amount of precious metals used (the amount of Pt in the catalyst was reduced by two-thirds and the Au surface treatment was eliminated). Furthermore, durability and reliability were improved by a factor of 3.

Boost converter. The boost converter—called the FC DC-DC converter (FDC)—steps up the voltage from the fuel cell to pass to the motor. Because of the high power requirement, a major challenge was developing countermeasures to limit component size. Toyota engineers achieved this by adopting a multi-phase converter and an innovative cooling structure.

In the 2008 FCHV-adv, the fuel cell and inverter are directly connected; the FC voltage is virtually the same as the motor voltage, and the inverter and motor have dedicated designs. However, Mirai uses the same inverter and motor as already adopted in mass-production hybrid electric vehicles. Because the maximum fuel cell voltage is lower than the maximum motor voltage (650V), the two components cannot be connected directly.

The new system implements the boost converter between the fuel cell and the inverter to step up the voltage from the fuel cell. By developing the FDC, it was possible to increase the voltage of the motor, reduce the number of fuel cell stack cells, and reduce the size and weight of the system. In addition, design innovations to the voltage-boost control and case structure provide quiet operation. The new system can be used with existing HV units, enhancing reliability and greatly reducing costs.

FCHV-adv hybrid system. Hasuka et al. Click to enlarge.   Mirai hybrid system. Hasuka et al. Click to enlarge.

The new efficient FDC features a simplified internal structure and innovative controls to reduce the size, weight, and cost of the system.

Structure of the FDC. Hasuka et al. Click to enlarge.
  • The phase drive control using the optimum number of phases in accordance with the power passing through FDC allows the FCV to be driven highly efficiently. Use of this control improved the loss by approximately 10% at 15kW.

  • TMC also reduced thermal resistance by approximately 50% compared with HEV reactors through the use of a new structure for reactor cooling and a dedicated filler.

  • Rubber in the body mounting structure helps prevent the transmission of vibration directly to the body, resulting in a reduction of 30 dB.

  • Nosie and vibration are also reduced by a carrier control that changes the switching frequency at random over time.

Toyota is continuing to develop and adopt new materials and further reduce the size of the FDC.

Hydrogen storage. Toyota reduced the weight, size, and cost of the high-pressure hydrogen storage system in the Mirai while improving fueling performance. The four 70 MPa tanks used on the 2008 Toyota FCHV-adv were reduced to two new larger diameter tanks; Toyota developed a new optimized laminated structure for tanks to reduce weight using a high-strength low-cost carbon fiber material. The shapes of the newly developed high-pressure hydrogen tanks were optimized to enable installation under the floor of the sedan. Further, Toyota reduced the size of the high-pressure valve and also modified a high-pressure sensor from a conventional vehicle for use in a high-pressure hydrogen atmosphere.

As a result, the whole storage system weighs approximately 15% less than that in the Toyota FCHV-adv, while reducing the number of component parts by half and substantially reducing cost.

Basic configuration of high-pressure storage system. The high pressure from the two hydrogen tanks is reduced in two stages by the high-pressure regulator and injector before reaching the fuel cell (FC) stack. Yamashita et al..   Comparison of conventional and new lamination methods. Toyota made three critical changes to the lamination method, resulting in a thinner tank wall and reduced weight. Yamashita et al. Click to enlarge.

Toyota engineers also reduced the time required to fuel the FCV by chilling the filling gas temperature at the hydrogen filling station to −40°C (as per SAE J2601). Furthermore, the layout of the tank temperature sensor and other aspects of the design were adjusted to increase the State of Charge (SOC) determined by SAE J2799 IrDA communications between the vehicle and hydrogen station.

The tanks use a plastic liner at the innermost layer to seal in the hydrogen gas. The liner is surrounded by a strong CFRP layer capable of withstanding high pressures, which itself is surrounded by a glass fiber reinforced plastic (GFRP) layer with high impact resistance, and a protector. Aluminum bosses are provided at both ends of the plastic liner, with one side for the valve fitting. The weight of the newly developed tanks was reduced by improving the CFRP layer and reducing the amount of material used.

Conventional tanks combine three types of winding methods: hoop winding to strengthen the central region of the tank; low-angle helical winding to strengthen the dome regions (in the axial direction); and high-angle helical winding to reinforce the boundaries of these regions. By necessity, the high-angle helical winding required to strengthen the boundary regions is also wound over the central region.

Because the high-angle helical winding is wound around the central region of the tank at an angle of 70°, the reinforcement efficiency is reduced. Toyota, however, made three critical changes to the lamination method:

  • The sectional shape of the liner was flattened to enable lamination by hoop winding at the boundary regions.

  • The boundary regions were strengthened while forming the conventional liner shape by gradually retracting the end positions of the hoop winding.

  • Hoop winding lamination was concentrated in the inner layers.

These changes had the following two effects. First, the high-angle helical winding, which accounted for approximately 25% of the total laminated structure, was eliminated. Second, hoop winding, which is a highly effective way of strengthening the central region of the tank, was concentrated in the inner layers where the generated stress is highest. This allowed the strength of the fibers to be used more effectively. This dual effect enabled a 20 wt% reduction in CFRP compared to the conventional lamination method.

—Yamashita et al.

For the FCHV-adv, Toyota had used aerospace grade carbon fiber. For Mirai, Toyota worked with carbon fiber providers to improve the properties of less costly general purpose carbon fiber.

On the road toward its anticipated future full-scale commercialization of fuel cell vehicles, Toyota will work to continue further to reduce the size of the hydrogen storage system and advance the performance of the next-generation FCV.


  • Konno, N., Mizuno, S., Nakaji, H., and Ishikawa, Y. (2015) “Development of Compact and High-Performance Fuel Cell Stack,” SAE Int. J. Alt. Power. 4(1) doi: 10.4271/2015-01-1175

  • Nakagaki, N. (2015) “The Newly Developed Components for the Fuel Cell Vehicle, Mirai,” SAE Technical Paper 2015-01-1174 doi: 10.4271/2015-01-1174

  • Hasuka, Y., Sekine, H., Katano, K., and Nonobe, Y. (2015) “Development Of Boost Converter for Mirai,” SAE Technical Paper 2015-01-1170 doi: 10.4271/2015-01-1170

  • Yamashita, A., Kondo, M., Goto, S., and Ogami, N. (2015) “Development of High-Pressure Hydrogen Storage System for the Toyota “Mirai”,” SAE Technical Paper 2015-01-1169 doi: 10.4271/2015-01-1169



Sounds like a brilliant effort with techniques that could be useful with not only fuel cells.

Can it be significantly cheaper? Can H2 be home/cheaply sourced from someone besides natural gas/oil oil companies? etc ..


To achieve 2.4X current density (one order of magnitude better than EV batteries) in about 6 years is something to be proud of. If that level of development is maintained, FCs could be highly compact and competitive by 2020/2021.

Mobile and fixed units production could multiply.

Concurrently, lower cost H2 production and distribution will expand in many countries.

Account Deleted

The Toyota Mirai is an impressive fuel cell vehicle. However, we should look at a realistic comparison of this FCEV to a current BEV - the Tesla Model.
Based on the 2016_Toyota_Mirai_FCV_Product_Information.pdf (, the energy density of the FCEV vs the Tesla Model S is not an order of magnitude better.
The Toyota Mirai Fuel Cell components (Fuel Cell Stack, H2 tank weigh 143.8 kg, the battery would add 29 kg - based on a NiMH Prius battery, for a total of 173 kg). The only question is how much energy is in the H2 tank: 5 kg is 195 kWh, for H2 @ 39 kWh/kg. However, Peak FC efficiency at 60% would lower that to 117 kWh and average efficiency could be roughly 30-35% would be <70 kWh (the press release actually states 60kWh). The Tesla Model S 60 kWh battery weighs 511 kg or roughly 3 times the weight of the Toyota Mirai.


These results are truly an achievement. Do they make sense?
I really have to ask if the extra cost to develop the H2 infrastructure will make FCVs a viable alternative. If BEVs get a range of 200+ miles in the next three years they still provide a cleaner driving option than H2 generation facilities + FCVs . I would expect that it will take at least 3 years to develop a full hydrogen fueling network, and this will be dependent upon strong FCV sales across the country. To keep hydrogen at -40C at each filling station will also likely require additional costs over existing filling station hardware and consume energy just to maintain the refrigeration systems (if this becomes a requirement).
Then there is the question of the cost to purchase hydrogen. This cost will be likely higher than electricity, especially if BEV drivers charge at home with solar. Hydrogen sold at a fueling station would primarily come from methane gas suppliers (non clean option) or solar (at the expense of generation time). If it is generated from solar, why not use the electricity directly rather than lose efficiency in converting the energy to hydrogen.
FCV sales really need to take off soon to take hold of a significant market share and bring down their costs. The cost of solar is approaching grid parity soon, and will make electricity a more viable fuel source for transportation in the next few years.



Well, for a start the hydrogen is not held at -40C at the filling station.
It is reduced in temperature at the point of filling the vehicle.

I am not going through all the rest, and will simply say that an accurate critique starts from accurate analysis of the technologies throughout the chain.

If you make a series of adverse assumptions about anything then it looks bad.


To compare BEVs and FCEVs one has to know users requirements.

If the vehicle is to be used in extender cold weather and on frequent long trips, BEVs are not yet capable of doing it with ease while FCEVs can do 500+ Km while keeping the passengers warm and in acceptable comfort zone.

Most cold weather areas already have Hydro and Wind energy to supply H2 generators with clean low cost energy (about $0.02/kWh), specially during off peak demands periods (about 16 hrs/day).

Highway H2 stations every 200 Km could be enough to start with.


Fuel Cell cars, the strategic mistake of Toyota and Honda.

Yordan Georgiev

So clearly there would be at least 2 major platforms for cleaner transport - Tesla's li-on battery pack vs. Toyota's Hydrogen storage. It is interesting to note that both the company are trying to push their technology by open-sourcing part of their patents. How about in 10 years driving autonomous fuel cell H2 , battery car having supercapacitors for faster acceleration for about 30 000 ...



Mass producing $30K extended range (500+ Km) BEVs and FCEVs could become a reality in the post 2025 era.

Which technology will be best suited on a regional basis?

Producing H2 from clean electricity during low consumption hours @ 80% efficiency and storing it could be an advantage over electrons.


With an excess of cheap electricity, this would be ideal to replace the carbon based fuels used in long haul trucks.

But to really compete with BEV's the round-trip efficiency will have to be improved.

Of coarse a huge investment in cost competitive H supply and associated electricity is needed.

Seems more likely to be a niche product.

The comments to this entry are closed.