The last crusade for 0.1 mg iridium

Two mil­li­grams. That’s all it takes for con­ven­tion­al PEM elec­tro­lyz­ers to con­vert water into hydro­gen gas. Two mil­li­grams per square cen­ti­meter of electrolyzer’s mem­brane, that is. That doesn’t sound like much. But it costs more than is eco­nom­ic­ally jus­ti­fi­able, and it’s more than is avail­able if the prom­ises of politi­cians and self-pro­claimed experts about green hydro­gen are to be ful­filled. This is why research­ers are pulling out their last gray hairs in pure frus­tra­tion as they fever­ishly try to reduce the iridi­um require­ment by 95% – to 0.1 mg/​cm². Achiev­ing this goal is the holy grail of the hydro­gen industry. You could read about it in our art­icle Unlock­ing the green hydro­gen eco­nomy. 

But how is it going? We’re going to find out. But first, a quick primer on how an elec­tro­lyz­er works.

Proton exchange membrane (PEM)

PEM elec­tro­lyz­ers are so named because they have a pro­ton exchange mem­brane (PEM) at the cen­ter of each elec­tro­lys­is cell. As the name implies, the mem­brane allows pro­tons to pass through. As you may remem­ber from high school, a free pro­ton is noth­ing more than a hydro­gen ion. Hydro­gen is made up of a pro­ton and an elec­tron. Remove the elec­tron and you get a hydro­gen ion, which is a pro­ton. The mem­brane allows these to pass freely from one side to the oth­er while block­ing the elec­trons. This is the key to how PEM elec­tro­lyz­ers work.

On one side of the mem­brane is an elec­trode that is con­nec­ted to the pos­it­ive ter­min­al of a power source. This side is called the anode. On the oth­er side of the mem­brane is anoth­er elec­trode con­nec­ted to the neg­at­ive ter­min­al of the same power source. This side is called the cath­ode.

Porous transport layer (PTL)

Both the anode and cath­ode are made of a por­ous, elec­tric­ally con­duct­ive material.

They must be elec­tric­ally con­duct­ive to allow the power source and the elec­tro­lyz­er to form an elec­tric­al cir­cuit through which cur­rent can flow.

They must also be por­ous to allow water to flow past them and the hydro­gen and oxy­gen gas formed to escape.

Both are referred to as gas dif­fu­sion lay­er (GDL) or por­ous trans­port lay­er (PTL). We prefer the lat­ter name, since they not only allow gas to dif­fuse, but also water to flow.

On the cath­ode side, where hydro­gen is formed when hydro­gen ions com­bine with elec­trons from the power source to form hydro­gen gas (H₂), a lay­er made of car­bon fibers serves as the PTL.

On the anode side, where oxy­gen is formed, a plat­in­ized titani­um sintered lay­er is used. Car­bon can­not be used because it would imme­di­ately react with the oxy­gen to form car­bon diox­ide. And we don’t want car­bon­ated water! Titani­um is more res­ist­ant to the aggress­ive envir­on­ment, and the plat­in­um pro­tects it and provides bet­ter elec­tric­al conductivity.

Electrolysis cell

The mem­brane is placed between the two trans­port lay­ers. When cur­rent is applied across the two elec­trodes, water flow­ing past the anode is split into hydro­gen ions and oxy­gen. The oxy­gen escapes through the PTL of the anode, while the hydro­gen ions, which are pro­tons, migrate through the mem­brane to the cath­ode side, where they com­bine with elec­trons (cur­rent is free elec­trons) to form hydro­gen gas, which escapes through the PTL of the cathode.

To dis­trib­ute water and elec­tri­city and to evac­u­ate oxy­gen and hydro­gen gas, a bipolar plate (BPL) with chan­nels for water and gas trans­port and elec­tric­al con­nec­tions is moun­ted dir­ectly on the out­side of each PTL.

Togeth­er, the mem­brane, both PTLs and both BPLs form an elec­tro­lys­is cell.

Two cells can be con­nec­ted in series by using one bipolar plate for both. It serves as the anode side for one and the cath­ode side for the oth­er. That’s why it’s called bipolar; it’s the pos­it­ive ter­min­al for one side and the neg­at­ive ter­min­al for the oth­er. In this way, sev­er­al elec­tro­lyz­er cells can be con­nec­ted in series to form an elec­tro­lyz­er stack.

An elec­tro­lyz­er plant typ­ic­ally con­sists of sev­er­al elec­tro­lyz­er stacks.

Iridium to the rescue

Is that all?

No.

We need a dash of iridi­um as well. Without the nobel met­al, the reac­tion is unbear­ably slow. Only small amounts of hydro­gen and oxy­gen gas are produced.

To speed up the pro­cess, the water molecules need iridi­um, or more spe­cific­ally iridi­um oxide, to ”hold on to” as they split. The mere pres­ence of this pre­cious met­al speeds up the pro­cess con­sid­er­ably. It doesn’t con­sume itself, it’s only a cata­lyst for the reaction.

Plat­in­um is used on the cath­ode side for the same reason.

Triple phase boundary

For each iridi­um atom to per­form its magic trick, it must be in con­tact with the anode, the water, and the mem­brane. All at the same time. This is called hav­ing a triple phase bound­ary or three phase bound­ary. And achiev­ing this is one of the biggest chal­lenges in design­ing elec­trode struc­tures for PEM electrolyzers.

The con­ven­tion­al solu­tion is to mix iridi­um oxide with a solvent to form a slurry, which is then used to ”paint” the anode side of the membrane.

The iridi­um atoms on the out­er­most sur­face of the paint lay­er are act­ive as cata­lysts. But many more iridi­um atoms are inact­ive; they don’t get a triple phase bound­ary because they are inside the paint lay­er. They are pure waste of the extremely rare and expens­ive metal.

But you can’t paint too thin. The paint will crack and form islands of iridi­um atoms that don’t have elec­tric­al con­tact. The lay­er of ”iridi­um paint” must be 5–10 microns thick to work, and then it typ­ic­ally requires 2 mil­li­grams of iridi­um per square cen­ti­meter of membrane.

The big challenge for the hydrogen industry

But, as those of you who have read our art­icle Unlock­ing the green hydro­gen eco­nomy know, 2 mg/​cm² is not sus­tain­able either eco­nom­ic­ally or in terms of resources. A sus­tain­able level is 0.1 mg/​cm² – a 95 % reduc­tion from today’s level. Get­ting there is the big chal­lenge for the hydro­gen industry.

A chal­lenge that Smol­tek has taken on.

But before we take a closer look at our solu­tion, let’s exam­ine oth­er aven­ues that have been explored.

Composite anode

The com­pos­ite anode tech­nique is basic­ally the same idea as the one already used. How­ever, most of the iridi­um is replaced with plat­in­um black – a fine powder of plat­in­um. It doesn’t act as a cata­lyst, but as a con­duct­or between the islands of few iridi­um atoms on the surface.

This meth­od has shown the abil­ity to reduce iridi­um use by up to 80 %, but the fun­da­ment­al struc­ture makes it dif­fi­cult to go below 0.5 mg/​cm² without los­ing per­form­ance. In addi­tion, the increased use of oth­er pre­cious metals increases the over­all cost, mak­ing it dif­fi­cult to achieve an eco­nom­ic­ally sus­tain­able solution.

Core-shell structures

Core-shell struc­tures are based on cre­at­ing nan­o­particles with a core of a cheap­er mater­i­al, such as rutheni­um or nick­el, and a thin shell of iridi­um oxide. This design allows the amount of iridi­um to be drastic­ally reduced, because only the shell needs to be made of the expens­ive metal.

A triple phase bound­ary is cre­ated because the iridi­um oxide in the shell is in dir­ect con­tact with both the elec­tro­lyte and the elec­tric­ally con­duct­ive core mater­i­al. The most prom­ising attempts have achieved an 85% reduc­tion in iridi­um use.

Unfor­tu­nately, core-shell struc­tures have proven dif­fi­cult to man­u­fac­ture on an indus­tri­al scale. The pro­cess is com­plex and requires care­ful con­trol of the syn­thes­is to ensure uni­form dis­tri­bu­tion of the shell. In addi­tion, there are sta­bil­ity issues where the core mater­i­al can leak out over time in the acid­ic envir­on­ment. This means that des­pite prom­ising res­ults in the lab, the tech­no­logy is unlikely to reach the 0.1 mg/​cm² goal in com­mer­cial production.

Manganese-iridium composites

Man­ganese-iridi­um com­pos­ites are some­what like core-shell struc­tures, but now man­ganese oxide is used as the core, and there is no shell of iridi­um, but indi­vidu­al atoms scattered over the entire sur­face, max­im­iz­ing the util­iz­a­tion of each iridi­um atom while tak­ing advant­age of the sta­bil­ity of man­ganese oxide in acid­ic environments.

A triple phase bound­ary is cre­ated because the iridi­um atoms are per­fectly exposed on the man­ganese oxide sur­face, provid­ing dir­ect con­tact with both the elec­tro­lyte and the con­duct­ive substrate.

Although the tech­no­logy has shown impress­ive res­ults with up to 95% reduc­tion in iridi­um con­sump­tion in labor­at­ory set­tings, there are sig­ni­fic­ant chal­lenges to indus­tri­al imple­ment­a­tion. The com­plex syn­thes­is pro­cess is dif­fi­cult to scale up, and there are ques­tions about the long-term sta­bil­ity of these atom­ic­ally dis­persed cata­lysts under real­ist­ic oper­at­ing conditions.

Nanostructured thin films

Nano­struc­tured Thin Films (NSTF) use a tech­nique in which iridi­um oxide is depos­ited in extremely thin lay­ers on crys­tal­line organ­ic whiskers. These whiskers, made of a spe­cial organ­ic pig­ment, cre­ate sub­micron sup­port struc­tures that allow for a very effi­cient cata­lyst arrange­ment in which almost all of the iridi­um atoms can par­ti­cip­ate in the reaction.

The triple phase bound­ary is optim­ized by the large avail­able sur­face area of the whisker struc­tures and the dir­ect con­tact between cata­lyst, elec­tro­lyte and con­duct­ive sub­strate, while the crys­tal­line nature of the whiskers provides excep­tion­al cor­ro­sion resistance.

While NSTF tech­no­logy is eleg­ant in the­ory, its prac­tic­al imple­ment­a­tion has proven prob­lem­at­ic. The man­u­fac­tur­ing pro­cess is del­ic­ate and dif­fi­cult to scale up. In addi­tion, the films have proven to be sens­it­ive to mech­an­ic­al stress and can delamin­ate dur­ing oper­a­tion. This means that des­pite its poten­tial, the tech­no­logy prob­ably won’t be able to deliv­er stable per­form­ance at 0.1 mg/​cm² on an indus­tri­al scale.

Nanoprinting technology

Nan­o­print­ing uses an advanced print­ing pro­cess to cre­ate extremely small (2 nm) iridi­um particles that are prin­ted dir­ectly onto the mem­brane sur­face. The tech­no­logy allows for optim­al iridi­um util­iz­a­tion by ensur­ing that each particle is avail­able for cata­lys­is while main­tain­ing dir­ect con­tact with the pro­ton-con­duct­ing membrane.

The triple phase bound­ary is optim­ized by the pre­cise place­ment of the particles on the mem­brane and their min­im­al size, ensur­ing each particle has both elec­tric­al con­tact and access to water.

While the tech­no­logy has shown prom­ising res­ults with a factor of 10 reduc­tion in iridi­um usage, there are still chal­lenges with scalab­il­ity and cost. The advanced equip­ment and pre­cise pro­cess para­met­ers make it dif­fi­cult to achieve cost-effect­ive mass pro­duc­tion of PEM elec­tro­lyz­ers at 0.1 mg/​cm².

Smoltek’s solution

Smoltek’s tech­no­logy uses ver­tic­ally ori­ented car­bon nan­ofibers coated with an ultra-thin lay­er of plat­in­um on which iridi­um atoms are placed. This cre­ates a unique struc­ture where each nan­ofiber acts as a ”stick” to which the cata­lyst can attach.

The triple phase bound­ary is optim­ized because the iridi­um is on the sur­face of the elec­tric­ally con­duct­ive nan­ofibers, which are per­fectly integ­rated into the membrane.

This approach has already demon­strated stable oper­a­tion with as little as 0.2 mg/​cm² of iridi­um and shows great poten­tial to reach the tar­get of 0.1 mg/​cm². The simple but eleg­ant solu­tion is also well suited for indus­tri­al pro­duc­tion. The nan­ofibers can be pro­duced using proven CVD tech­no­logy, and the sub­sequent met­al coat­ing is a well-known pro­cess. Smoltek’s suc­cess­ful 1,000 hour test shows that the tech­no­logy is ready for scale-up to indus­tri­al production.

The winning technology emerges

Look­ing at the dif­fer­ent approaches to redu­cing iridi­um con­tent in PEM elec­tro­lyz­ers, a clear pat­tern emerges. While sev­er­al tech­no­lo­gies show prom­ise in labor­at­ory set­tings, most face sig­ni­fic­ant hurdles in scal­ing up to indus­tri­al pro­duc­tion. The key to suc­cess lies not just in achiev­ing the 0.1 mg/​cm² tar­get, but in doing so with a tech­no­logy that can be reli­ably man­u­fac­tured at scale.

Smoltek’s solu­tion rep­res­ents one prom­ising path for­ward. Our approach using car­bon nan­ofibers has already demon­strated stable oper­a­tion at 0.2 mg/​cm² in demand­ing 1,000-hour tests. By basing our innov­a­tion on estab­lished man­u­fac­tur­ing meth­ods like PECVD, we’ve pri­or­it­ized indus­tri­al scalab­il­ity from the start – because a break­through tech­no­logy only mat­ters if it can reach the market.

And now, after two dec­ades of research and devel­op­ment, this cru­sade for the holy grail of green hydro­gen is draw­ing to a close. The 0.1 mg/​cm² mile­stone is with­in reach.