CARRIERS,

RECOURCESSTORAGE
• SOLAR& TRANSFORMATION • HYDROGEN & FUEL CELLS

The Hydrogen Fuel Alternative

G.W. Crabtree (Argonne National Laboratory, USA)
M.S. Dresselhaus (Massachusetts Institute of Technology, USA)

Abstract
The cleanliness of hydrogen and the efficiency of fuel cells taken together offer an appealing alter-
native to fossil fuels. Implementing hydrogen-powered fuel cells on a significant scale, however,
requires major advances in hydrogen production, storage, and use. Splitting water renewably
offers the most plentiful and climate-friendly source of hydrogen and can be achieved through
electrolytic, photochemical, or biological means. Whereas presently available hydride compounds
cannot easily satisfy the competing requirements for on-board storage of hydrogen for transporta-
tion, nanoscience offers promising new approaches to this challenge. Fuel cells offer potentially
efficient production of electricity for transportation and grid distribution, if cost and performance
challenges of components can be overcome. Hydrogen offers a variety of routes for achieving a
transition to a mix of renewable fuels.

Introduction
Increased energy use is the universal driver for raising the As shown in Figure 1, the cost of the present generation of
quality of life in all societies, from developing to developed fuel cells is 100 times that of a gasoline engine; with mass pro-
countries. However, the present reliance on energy from fossil duction, the cost differential might fall to a factor of 10. In
fuels produces unwanted side effects: environmental pollution addition to cost, longevity, frequency of repair, and perfor-
that threatens human health, carbon dioxide emissions that accel- mance in cold weather are other areas in which fuel cells are not
erate global warming, and geopolitical tensions arising from the yet competitive with gasoline engines for widespread use in
nonuniform distribution of fossil resources throughout the world. transportation. Lowering cost and enhancing performance pres-
The challenge is to find highly efficient ways to produce, deliver, ent materials research challenges in terms of the cathodes, elec-
and use energy that enhance quality of life but do not threaten the trolyte membranes, and catalysts of fuel cells.
environment and climate or strain geopolitical relations. Hydrogen storage for transportation presents a major mate-
The energy carrier hydrogen is an alternative to fossil fuels rials research challenge, namely, to find a storage medium that
with the potential to achieve these goals.1–3 Hydrogen is abun- combines a hydrogen density greater than that of the liquid
dant in chemical compounds such as water and the organic with fast kinetics allowing rapid charging and discharging.
compounds of biomass, and its combustion produces only heat Many conventional bulk materials have been explored and
and water with no offensive pollutants or carbon dioxide. rejected as storage media because they do not meet these cri-
Hydrogen can be combined with oxygen in the electrochemical teria. However, nanoscience opens new opportunities for
reactions of a fuel cell to produce electricity, a clean, versatile addressing this challenge, with the potential for high surface
carrier of energy enabling many end uses including lighting, areas and hybrid structures that enable multifunctional perfor-
refrigeration, communication, information processing, and mance, such as low-energy dissociation of hydrogen molecules
transportation. The intimate connection between hydrogen and on the surface and rapid diffusion of atomic hydrogen to the
electricity through fuel cells makes hydrogen much more than interior.
simply a clean substitute for fossil fuel combustion. The challenge for production is finding a source of hydrogen
Achieving the promise of hydrogen as an efficient, sustain- that can supply the needs of a full-fledged hydrogen economy
able, and environmentally friendly fuel requires widespread and that is not dependent on fossil fuels for feedstocks.
innovation and development of the means for its production, Producing hydrogen from natural gas, a widely used established
storage, and use. The energy-chain and technical challenges for process, in sufficient quantities to power the world’s cars and
creating a viable hydrogen economy are shown in Figure 1. light trucks would strain the world’s supply of conventional
The most effective use for hydrogen is conversion of its chemi- methane, making natural gas as geopolitically sensitive as oil.
cal energy to electrical energy in fuel cells. The high conversion Furthermore, the production of hydrogen from natural gas gen-
efficiency of fuel cells, up to 60%, makes them attractive com- erates as much pollution and carbon dioxide per unit of energy
pared to other electrical generation alternatives based on fossil output as burning the gas directly. Aside from fossil fuel,
fuels, which are about 34% efficient on average. The high effi- ­possible sources of hydrogen include reforming the carbohy-
ciency of electric motors (typically well above 90%) makes the drates of biomass or splitting water molecules thermally, elec-
fuel cell–motor combination attractive for transportation com- trolytically, or photochemically. Although these nonfossil
pared to gasoline engines, typically about 25% efficient. This routes to hydrogen production are attractive and potentially
potential for high-efficiency end use adds additional appeal to viable, they require breakthroughs in materials research to dis-
the environmental argument for hydrogen: not only is it free of cover effective, robust catalysts that lower the energy barriers
pollutants and greenhouse gases, but it also uses considerably to production; to achieve nanoscale integration of the sequen-
less primary energy for a given energy use. tial production steps; and to lower cost.

MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy 421
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS, STORAGE & TRANSFORMATION • HYDROGEN & FUEL CELLS

Hydrogen Production
solar / wind In a mature hydrogen econ-
H2O hydro omy, hydrogen could replace
automotive
electrolysis
Compressed fuel cells fossil fuel for transportation in
gas cars and light trucks, produce
solar/nuclear
consumer electricity for distribution
thermochemical
cycles
H2 Liquid H2 electronics through the grid, and provide
Solid portable electricity for personal
bio- and fossil fuel
hydrogen stationary electronics and other applica-
compounds
bio-inspired reforming
electricity/heat tions. Transportation now con-
generation
solar chemistry + sumes about 20% of the world’s
carbon capture energy, and electricity about
12%.5 About 600 Mt/year of
use hydrogen will be needed world-
production storage wide to power all cars and light
in fuel cells
trucks in 2030, compared to
50 M tons/year 4.4 MJ/L (Gas, 10,000 psi) about 50 Mton/year now pro-
global production $3000/kW
8.4 MJ/L (LH2) duced.2 About half of the global
$300/kW supply of hydrogen is produced
mass production
600 M tons/year $30/kW by reforming natural gas.2 The
(light trucks and cars 9.72 MJ/L (Internal Combustion Engine) reforming of natural gas, how-
(2015 FreedomCAR Target)
world-wide in 2030) ever, is not an attractive produc-
tion route for a mature hydrogen
Figure 1. The three links of the hydrogen energy chain: production, storage, and use in fuel cells. Each
link must connect seamlessly with the others to create an effective infrastructure, and each link has
economy, because the order-of-
fundamental challenges that must be solved, as indicated. magnitude increase in demand
would deplete our finite reserves
and the concentration of gas
reserves in a relatively few
regions of the world could lead
Progress toward a mature hydrogen economy depends on to geopolitical tension and unstable supplies. Environmental
breakthroughs in finding new materials and basic understand- impact is also a major concern, as reforming natural gas to
ing of the nanoscale phenomena that govern the interactions of hydrogen produces as much pollution and CO2 as burning the
hydrogen with materials. The ultimate evolution of a hydrogen natural gas directly.
economy, however, depends on much more than technical fea- Splitting water renewably, that is, using only renewable
sibility. As with all new technologies, comparisons of hydrogen energy inputs, is an attractive production route for hydrogen as
with a developing mix of alternatives for performance, cost, a fuel of the future. Water is abundant on the surface of the earth
efficiency, convenience, reliability, and safety will determine and more widely distributed geographically than fossil fuels.
its future course. The outcomes of these comparisons are them- The water–hydrogen cycle (see Figure 2) is closed, unlike the
selves developing, as the costs of fossil fuels and environmental fossil fuel energy chain that operates on a “once-through” basis,
mitigation of their use increase and as engineering and scien-
tific discoveries alter the mix of competing alternatives.
Although a mature hydrogen economy requires technical
and economical solutions for production, storage, and use, hydrogen
there is value in implementing any of these without full-blown
solutions for the others. Hydrogen fuel cells can be used in
large stationary installations for neighborhood heat and power
and for small personal electronics without solving the problem Energy Energy
of on-board storage for transportation, and hydrogen storage in out
can be used with intermittent sources of renewable energy such
as solar and wind without large-scale hydrogen production. As
a highly storable carrier of energy, hydrogen offers remarkable
versatility, enhanced by its compatibility with electricity, the
less-storable energy carrier that forms the backbone of our water
energy distribution system. Thus, although a full-fledged
Figure 2. Hydrogen–water cycle. Hydrogen is a green carrier
hydrogen economy encompassing production, storage, and use
of energy; water is a green carrier of hydrogen. Hydrogen is
could be the ultimate goal, a partial implementation of hydro-
created from water by injecting energy at the point of
gen as a storable energy carrier for use in stationary and per- production from any source—photons, electrons, or heat.
sonal fuel cell applications is a desirable outcome on its own Water is created from hydrogen when energy is released at
merits. the point of use, by electro-oxidation to electrons and heat in
This article presents the current status of and promising fuel cells, or combustion to heat in engines or turbines. Even
research opportunities to overcome the technical and economic though the hydrogen–water cycle is closed and sustainable,
challenges in hydrogen production, storage, and end use. the energy chain it links can be open and depletable, if, for
Although the focus is on recent developments and future direc- example, fossil energy is used to split water. Sustainability is
tions, there is a long and distinguished history of hydrogen maintained if renewable energy is used to split water and the
research that laid the foundations for the current surge of energy released is used only for work and heat.
research activity.4

422 MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS,
RECOURCESSTORAGE
• SOLAR& TRANSFORMATION • HYDROGEN & FUEL CELLS

depleting reserves at the point of production and producing most abundant of fossil fuels and the most able to supply a large
unwanted CO2 and pollutants as byproducts at the point of use. quantity of hydrogen without depleting its reserves or inflating
In contrast, the hydrogen and oxygen produced from the split- its market price. Gasification of coal produces hydrogen and
ting of water are recombined at the point of use to release CO directly in syngas, and CO reacts with steam to produce
energy, replenishing the water taken for its production. This more hydrogen in the water–gas shift reaction. Several
closed chemical cycle is a prerequisite for a sustainable energy approaches are under development for producing hydrogen
system—no chemical compounds are created or destroyed once from coal with CO2 capture that maintain 60–90% of the origi-
the cycle is complete. There is, however, a net flow of energy nal energy content of the coal.16,17 The development of mem-
through the water–hydrogen cycle—in for water splitting and branes for separation of hydrogen from CO and CO2 is a key
out upon water formation. If fossil or nuclear energy is used to materials issue.18 Coal gasification and the water–gas shift are
split water, its once-through chemistry reappears, and the sus- commercial technologies ready to implement in integrated gas-
tainability of the closed water–hydrogen cycle is lost. Using a ification combined-cycle (IGCC) power plants for generating
source of renewable energy to split water, however, eliminates electricity from coal with CO2 capture at up to ~40% efficiency,
the once-through features and enables a sustainable, closed compared to an average efficiency of ~34% for existing elec-
cycle for energy use. tricity production from coal without CO2 capture. (See the
The efficiency of the water–hydrogen cycle is key to its article on coal by Powell and Morreale in this issue.) A rela-
value. Inevitably, more energy is needed to split water than is tively slight modification of this process enables coproduction
recovered on its recombination. Minimizing the energy barriers of hydrogen along with electricity and CO2 capture, and the
to water splitting and recombination are high-priority research fraction of production devoted to electricity and to hydrogen
challenges in catalysis and nanoscale science. Present technol- can be varied on demand with little or no sacrifice in effi-
ogy can split water with over 80% efficiency in large electrolyz- ciency.17 The switchability of production from electricity to
ers6 and recover electricity at 60% efficiency upon recombination hydrogen accommodates the intermittency of renewable
in combined-cycle turbines or in fuel cells. Although they dem- sources by producing hydrogen when solar and wind are active
onstrate the principle of the water–hydrogen cycle, these pro- and electricity when these sources are dormant. In this approach,
duction routes and efficiencies are only illustrative. Other the coal plant runs at full production capacity and high effi-
options for splitting water by thermal, photochemical, and elec- ciency continuously, and no storage of electricity produced by
trochemical means are now being explored in the laboratory. renewable sources is required. The hydrogen coproduced with
The proof of principle and most advanced incarnation of electricity can be transported for use off the production site or
solar water splitting is provided by photosynthesis, in which a used on-site to produce electricity at efficiencies approaching
complex protein assembly—Photosystem II—collects daylight, 60% in combined-cycle turbines or solid-oxide fuel cells for
channels its energy to a photocatalyst based on Mn3CaO4, and electrical grid distribution. Thus coproduction not only pro-
orchestrates the physical coordination of two water molecules vides a flexible, low-marginal-cost source of hydrogen for
and the catalyst to produce O2, protons, and electrons. Water transportation and other distributed uses, it encourages the
splitting is such a demanding process that nature has evolved development of hydrogen for electricity production and distri-
only one mechanism to accomplish it, used by all photosyn- bution through the grid.
thetic organisms for the past three billion years. Although Any source of heat can be used to split water through ther-
Photosystem II produces only protons and electrons separately, mochemical cycles that require several reaction steps operating
the enzyme hydrogenase can finish the job by joining protons at high temperature. Concentrated solar power can produce the
and electrons to produce H2.8 With high-resolution x-ray and high temperatures needed for thermochemical conversion, as
neutron diffraction, we know the structures and some of the can the heat from nuclear reactors. The challenges are discover-
working principles of Photosystem II and hydrogenase.9–11 The ing and developing the thermodynamics of chemical cycles that
challenge is to use this knowledge to develop genetically modi- split water and finding materials that can withstand the high
fied designer plants that produce hydrogen directly from sun- temperatures and often corrosive environments required by
light and water, and to assemble artificial structures inspired by these processes.19
nature that accomplish the same goal.8,10
A second route to solar hydrogen production uses semicon- Hydrogen Storage
ductors to produce electrons and holes that are, in turn, used to The use of hydrogen for transportation, personal electronics
split water with the aid of a catalyst. Wide-bandgap semicon- and other portable power applications requires an effective
ductors such as the oxides TiO2 and WO3 are favored substrates hydrogen storage medium. Existing technology for hydrogen
because they are inexpensive, are robust in water solutions, and storage is limited to compressed gas and liquefaction, both of
can be created in the form of nanotubes and other nanostruc- which are used now in demonstration vehicles. Compressed
tures with large surface areas to promote light harvesting.12 The gas, even at the highest practical pressure of 10,000 psi, is still
challenge is accessing the solar spectrum at energies below the a bulky way to store hydrogen that requires a significant frac-
wide bandgap of these oxides. One promising approach coats tion of the trunk space in a small car to enable a 500 km driving
the oxide surfaces with narrow-bandgap nanodot semiconduc- range. Liquid hydrogen takes up slightly more than half the
tors or dyes that transfer electrons excited by low-frequency volume of 10,000 psi compressed gas, but it loses 30–40% of
light to the oxide; another approach dopes the oxides to lower its energy in liquefaction. Although gas and liquid storage are
their bandgap.13,14 Catalysts are needed as well, to reduce the useful as temporary options in a provisional hydrogen econ-
energy barriers to water splitting.15 The ability to fabricate, omy, more compact and efficient storage media are needed for
characterize, and understand complex nanostructures is key to a mature hydrogen economy.
success with semiconductor-based water splitting, and impres- The most promising hydrogen storage routes are in solid
sive progress has been made recently in exploiting nanostruc- materials that chemically bind or physically adsorb hydrogen
tured materials for solar hydrogen production. at volume densities greater than that of liquid hydrogen. The
Hydrogen production from coal is an intermediate alterna- challenge is to find a storage material that satisfies three com-
tive to the present practice of natural gas reforming and the peting requirements: high hydrogen density, reversibility of
future development of splitting water renewably. Coal is the the release/charge cycle at moderate temperatures in the range

MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy 423
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS, STORAGE & TRANSFORMATION • HYDROGEN & FUEL CELLS

of 70–100°C to be compatible with the present generation of Table I: Hydrogen Storage Capacity of Amine Borohydrides
fuel cells, and fast release/charge kinetics with minimum NHnBHn.
energy barriers to hydrogen release and charge. The first Reaction Wt % H2 T (°C)
requires strong chemical bonds and close atomic packing; the
second requires weak bonds that are breakable at moderate NH4BH4 ⇒ NH3BH3 + H2 6.1 <25
temperature; and the third requires loose atomic packing to NH3BH3 ⇒ NH2BH2 + H2 6.5 <120
facilitate fast diffusion of hydrogen between the bulk and the
NH2BH2 ⇒ NHBH + H2 6.9 ~155
surface, as well as adequate thermal conductivity to prevent
decomposition by the heat released upon hydriding. Although NHBH ⇒ BN + H2 7.3 >500
several materials have been found that satisfy one or more of
the requirements, none has proven to satisfy all three. In addi-
tion to these basic technical criteria, viable storage media storage under expected operating conditions. The remaining
must satisfy cost, weight, lifetime, and safety requirements hydrogens come off at <120°C, ~155°C, and 500°C. Thermal
as well.20 decomposition of NH3BH3 to NHBH removes the second and
Two recent developments in materials science hold promise third hydrogens and releases 12% hydrogen by mass below
for meeting the difficult hydrogen storage challenge. The first 155°C, making it attractive for portable storage. Confining
is the rapid progress in nanoscience in the past five years. The NH3BH3 in a nanoporous scaffold lowers its decomposition
small dimensions of nanoscale materials minimize the diffusion temperature, speeds its release kinetics, and suppresses undesir-
length and time for hydrogen atoms to travel from the interior able decomposition products. A path for reversibly recharging
to the surface. The large relative surface area provides a plat- NHBH has not yet been found and remains a subject of intense
form for dissociation of molecular hydrogen to atomic hydro- research.
gen, a prerequisite for diffusion and for chemical bonding with Ammonia (NH3) is a high-capacity storage medium with
the host. The surface area can be tailored with a monolayer of 17 wt% hydrogen. Its stability is too great to release hydrogen
catalyst to promote dissociation, and surface curvature can be easily, and furthermore, NH3 is dangerous to the environment.
adjusted through the size of the nanoparticles to create unbonded Ammonia can be combined with other compounds that allow
orbitals that promote reactivity with hydrogen. NH3 to be reversibly released and absorbed as a molecule,
The second promising development for hydrogen storage effectively eliminating the toxic hazard during storage. For
materials is the growing ability of density functional theory to Mg(NH3)6Cl2, the temperature for decomposition to MgCl2 and
numerically simulate material behavior.21–24 Density functional NH3 is 150°C.31 The ammine salt Mg(NH3)6Cl2 can be formed
theory implemented on computer clusters is now in widespread into nanoporous pellets as shown in Figure 3, which could
use for the calculation of the electronic structures, crystal struc- function as a potential carrier for ammonia that would remove
tures, bond strengths, and heats of reaction for many multi- its environmental danger if handled by trained personnel.
­element compounds. The number of compounds that can be Although decomposition of ammonia to hydrogen occurs at too
simulated and the level of comprehensive information about high a temperature to be viable for low-temperature fuel cells
their structures and stability far exceed what can be determined using polymer electrolytes, ammonia can be used directly in
experimentally by discovery synthesis in the laboratory on the solid-oxide fuel cells, where the high operating temperature can
same time scale. This potentially powerful tool for numerically be used to decompose it without an energy penalty.
screening materials is now being applied to hydrogen storage An alternative approach to the thermal release of hydrogen
compounds. A recent study examined 300 candidate compounds from a storage compound is destabilization of the storage com-
for their structures, hydrogen storage capacities, and hydrogen pound by reaction with a second compound. An example is
decomposition temperatures.23 LiNH2 + LiH ® Li2NH + H2, which releases hydrogen at
Hydrogen storage materials employ two complementary
strategies for releasing hydrogen for use: thermalization and
destabilization. In thermalization, hydrogen is released from
the storage media by heating to the decomposition temperature,
where some or all of the hydrogen is driven off. This traditional
approach emphasizes hydrides with light elements from the
first and second rows of the periodic table, to maximize the
mass percentage of hydrogen.25 The ternary and quaternary
hydrides of these elements have high storage capacities, nota-
bly the borohydrides M+BH4− (where M is Li, Na, or K and B
can be replaced by Al),26 and the boranes NHnBHn, where
n ranges from 1 to 4.27–29 The borohydrides have significant
storage capacities, up to 19% of the mass of the molecule for
LiBH4, but they suffer from high decomposition temperatures
and large activation barriers to rehydrogenation. Catalysts such
as Ti reduce the barriers for both the decomposition and rehy-
drogenation of borohydrides and alanates, offering a practical
route to their use as hydrogen storage materials.30 NH4BH4 sub-
stitutes the ammonium ion for a simple metal cation in the boro- Figure 3. Pellets of Mg(NH3)6Cl2, a carrier for ammonia. The
hydride structure, packing four more hydrogens into the pellets hold NH3 at nearly its density in liquid ammonia, but at
molecule. The hydrogen mass ratio is an impressive 24%, but a vapor pressure of only 2 mbar; there is no significant release
not all of the hydrogen can be easily removed thermally. The of ammonia to the environment and the pellets can be
hydrogen comes off in stages, with about 6% of the mass handled safely. The pellets are charged at room temperature
released for each decrease of n by one, as shown in Table I. The and release ammonia above ~100°C. After Reference 31.
first hydrogen comes off at <25°C, making it too unstable for

424 MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS,
RECOURCESSTORAGE
• SOLAR& TRANSFORMATION • HYDROGEN & FUEL CELLS

~150°C, significantly lower than the release temperature for A recent breakthrough increasing the catalytic activity of Pt
either LiNH2 (lithium amide, 200°C) or LiH (500°C).32,33 The by a factor 10 for the oxygen reduction reaction at the fuel cell
storage capacity of 6.5% for this reaction is substantial, although cathode reveals a promising new research direction for tuning
not high enough for the ultimate storage solution. If Li2NH catalytic activity, as shown in Figure 5.38,39 The approach is to
could be destabilized to release the remaining H, the storage alter the electronic structure of the subsurface layers of Pt by
capacity would be 9.6%, high enough for hydrogen-powered substitution with a second element, leaving the surface layer
transportation. LiBH4 and MgH2 are another destabilization intact. The “Pt skin” on the surface maintains the atomic con-
couple, with a storage capacity of 11.5%.34 The promise of figuration of pure Pt that promotes breaking O–O bonds and
destabilized hydrides for storage is the much larger number of forming O–H bonds, while substitution in the first subsurface
paired (or multiple) hydrides than single hydrides might pro- layer adjusts the electronic structure and bonding strength of
vide effective storage. The advantage of the larger number of the skin by adding or subtracting electrons. Density functional
hydride pairs is balanced by the challenge of searching this theory39 plays a key role in subsurface manipulation of catalytic
enormous set. Guiding principles are needed to choose likely activity, describing and predicting its outcomes and providing
hydride pairs for which the reaction kinetics and destabilization a quantitative scientific framework for this novel approach.
temperature promise to be favorable. The screening potential of The membranes for polymer electrolyte fuel cells are based
density functional theory simulations plays a crucial role here, on C–F polymer backbones with sulfonic acid side chains.
potentially short circuiting years of experimental phase dia- Proton conductivity is achieved by hopping between the hydro-
gram work. gen sites of neighboring side chains. The hopping process
Nanostructured materials open broad new horizons for the requires hydration of the polymer with water, limiting the oper-
hydrogen storage challenge. The structural and compositional ating temperature of the membrane to about 80°C and requiring
complexity of nanostructured materials and their diverse sur- considerable attention to water management. New membrane
face morphologies enable control not only of the static but also materials that eliminate the hydration requirement and enable
of the dynamic features of the charge/release phenomena. hydrogen ion conductivity at higher temperatures are needed
Core–shell composites, for example, enable free exchange of and are being intensively pursued.40,41 One promising direction
protons between interior environments tailored for bonding to is the formation of nanocomposite membranes by addition of
the host and surface environments designed to make and break hydroscopic constituents that retain a high water content at high
the molecular H2 bond. Metal–organic frameworks are open temperatures, thereby extending the temperature range of con-
structures that physisorb molecular hydrogen at specific sites ventional proton conduction mechanisms. Another approach is
within the framework, yet allow easy diffusion to the surface to replace the C–F polymer backbone with a more thermally
and release hydrogen at low temperature; the richness of their resistant structure, but retain the same side chain sulfonation to
structure–property relationships is just beginning to be appreci- enable proton conductivity. Membranes based on some of these
ated.35,36 Nanoporous materials that divide and confine storage alternate polymers can operate above 120°C. A third creative
media such as NH3BH3 in nanoscale units can lower the decom- approach is to replace water with another solvent that assists
position temperature and activation barriers for hydrogen proton mobility but has a higher boiling point, such as phos-
release and alter the decomposition pathways and end prod- phoric acid. Some of these membranes conduct protons at tem-
ucts.28 Mg(NH3)6Cl2 is naturally nanoporous, accounting for its peratures up to 200°C in the absence of water.42 Key issues in
rapid NH3 charge/release times.31 Combined with the blossom- all membrane materials include acid loading to provide addi-
ing power of density functional theory to predict reaction path- tional sites for proton transfer, polymer crosslinking to improve
ways and materials properties, nanoscale science represents a thermal and mechanical stability at high temperature, and
new era of hydrogen storage possibilities. enhanced surface area to enable effective electrical and chemi-

Hydrogen Use in Fuel Cells

Fuel cells converting hydrogen and oxygen to electricity e−
and water are an appealing alternative to fossil fuel combus-
tion engines for their efficiency, versatility, and environmental O2
friendliness. The basic operation of a fuel cell is depicted in + +
Figure 4. Fuel cells produce electricity with a potential effi- H2 +
ciency of 60%; the electricity can be used directly or con- +
verted to motion, light, or heat. In contrast, gasoline engines +
operate with ~25% efficiency and are used almost exclusively +
+ +
for producing motion. Analysis of results from the first two H2O
years of the U.S. Department of Energy’s fuel cell learning Pt catalyst
demonstration based on a fleet of 77 fuel cell vehicles showed
that fuel cell efficiency at one-quarter power as tested on a H2 → 2H+ + 2e− 4H+ + 4e− + O2 → 2H2O
dynamometer reached 52.5–58.1%, close to the target effi- hydrogen oxidation oxygen reduction
ciency of 60%.37
Figure 4. Basic operation of a fuel cell. Hydrogen enters at
Polymer electrolyte membrane (PEM) fuel cells for trans-
the anode (red), where it is catalytically dissociated and
portation rely on dispersed Pt nanoparticle catalysts supported ionized to protons and electrons. The protons enter the
on carbon substrates for promoting the reaction of protons, polymer electrolyte membrane (green) where they migrate by
electrons, and oxygen molecules to water. Although Pt is the ionic conduction to the cathode (blue). Electrons travel
best-known catalyst for this reaction, it cannot meet the demands through an external circuit where they do electrical work
of a mature hydrogen economy because of its high cost and rel- before arriving at the cathode. The cathode reacts electrons,
ative scarcity. Orders-of-magnitude increases in its catalytic protons, and oxygen molecules to produce water. The net
activity are needed to reduce the required quantities, or it must input of the fuel cell is hydrogen and oxygen; its net output is
be replaced by an alternate catalyst that is active, abundant, and water, electricity, and heat.
inexpensive.

MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy 425
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS, STORAGE & TRANSFORMATION • HYDROGEN & FUEL CELLS

charge-deficient substitutions on the cation sites. The electrodes

a facing the electrolyte must have significant ion and electron
Pure Pt-skin conductivities at low temperatures, a key research objective.
Pt48Ni52 Promising anodes include rare-earth-doped ceria, such as
Pt87Ni13 SmxCe1-xO2-d, together with a Ni catalyst,45 and cathodes
include doped CoO3, such as Ba0.5Sr0.5Co0.8Fe0.2O3-d and
Pt75Ni25
Ba0.5Sr0.5CoO3-d. High performance of the electrolytes and
Pt3Ni crystal Pure Pt crystal electrodes depends not only on composition, but also on inter-
nal microstructure, a feature that has remained largely hidden
b until recently. Three-dimensional imaging of fuel cell compo-
Pt3Ni
nents using transmission electron microscopy is beginning to
Specific Activity reveal the inner microstructure of the three-phase boundary at
the intersection of the gaseous environment, the electrode, and
Ik kinetic current density

for the oxygen reduction the polymer membrane where molecular gases, electrons, and
reaction ions meet and exchange. Figure 6 shows the rich detail that 3D
−
+
1/2 O2+ 2 H + 2e ⇒ H2O imaging reveals about this three-phase boundary in a yttrium-
stabilized zirconia anode of a solid-oxide fuel cell.46 Such
images yield quantitative information about the volume frac-
tions, surface areas, and connectivities of the various phases
and will soon be applied to the electrodes and membranes of
Pt many established and emerging fuel cell types.

Evolution of a Hydrogen Economy

(111) (100) (110) Widespread hydrogen-powered transportation requires the
Crystal Surface simultaneous solution of challenges in production, storage, and
use, a much more difficult task than solving any of the three
Figure 5. Factor-of-10 improvement in catalytic activity of Pt alone. We are far from practical solutions in fossil-free produc-
with subsurface alloying. (a) Near-surface structure of single- tion and in solid-state storage, and the path to these outcomes
crystal alloys of nominal composition Pt3Ni produced by
will not be based on incremental innovations in today’s tech-
surface segregation. The top layer is a pure Pt-skin covering
nologies of reforming natural gas or coal, or storage in com-
a Ni-rich subsurface layer and Pt-rich layers beneath. The Pt-
skin composition and crystal structure are identical to those
pressed gases and liquids. Fuel cells are qualitatively different,
of a pure Pt crystal. (b) The specific activity of the Pt-skin is
however: a competitive technology can arise from innovation
enhanced over the pure Pt surface for the oxygen reduction
reaction. For the (111) Pt-skin surface, the enhancement is a
factor of 10, as a result of subtle changes in the Pt-skin
electronic structure. Theory plays a major role in subsurface
alloying, predicting and describing experimental outcomes.
After References 38 and 39.

cal exchange with the electrodes. Chemically crosslinking 5

membranes from liquid precursors at low temperature delivers
improvements in all of these properties.43 4
Beyond transportation, hydrogen and fuel cells can make a
substantial contribution to clean and efficient energy use for 3
stationary power generation for grid distribution, as battery
y (µm)

replacements for personal electronics, and as stationary or por- 3

table emergency power. For large-scale stationary applications, 2
solid-oxide fuel cells operating continuously in the range of 2
800–1000°C are the favored technology for their economies of
)

1
µm

1
scale. Here, O2− ions are the mobile species, traveling through
z(

a solid-oxide membrane to react with H+ ions to produce water. 0

0 4 5 6
The high operating temperatures required to produce sufficient 1 2 3
0
oxygen ion mobility in the perovskite membrane limit the life- x (µm)
time of components, create fatigue from thermal cycling, and
Figure 6. Three-dimensional reconstruction of yttria-stabilized zirconia
require long starting and stopping times. Reducing the operat- (YSZ)/Ni anode in a solid-oxide fuel cell by dual beam focused ion
ing temperature of solid-oxide fuel cells to 400–600°C is a beam scanning electron microscopy. The reconstruction reveals the
major materials research goal requiring new solid electrolytes inner microstructure of the anode at the 100 nm scale, a fundamental
with high oxygen ionic conductivities, active catalysts for feature that has been largely invisible until now. Three phases can be
hydrogen oxidation at the anode and oxygen reduction at the seen: the Ni catalyst and electron conductor (green), the YSZ ion
cathode, and electrode materials with suitable ionic and elec- conductor (translucent/grey), and the pores for gas conduction (blue).
tronic conductivities that are compatible with the electrolyte.44 From images like this, the length and geometry of the three-phase
Promising results have been achieved using Ce1-xMxO2-d (where boundaries enabling mutual contact of gas, catalyst, electrons, and
M is Gd, Sm, Ca, or Mg) and (La1-xSrx)(Ga1-yMgy)O3-d as the ions can be extracted. After Reference 46.
electrolytes, in which the oxygen vacancies are controlled by

426 MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS,
RECOURCESSTORAGE
• SOLAR& TRANSFORMATION • HYDROGEN & FUEL CELLS

on today’s approaches and designs. Like the internal combus- gency power applications do not require solutions to the
tion engine of a century ago, today’s fuel cells can spawn a long production and storage challenges before deployment. Although
line of progeny based on successive innovation of their elec- the relative cost of electricity from portable fuel cells is higher
trodes, membranes, and catalysts. than for battery or fossil fuel alternatives, the differential cost
The requirement of disruptive innovation for the success of these applications is small enough that the advantages of
of fossil-free production and solid-state storage makes the size, operating time between charges, efficiency, and conve-
entrepreneurial approach difficult to launch for transportation. nience could become deciding factors.
Unlike earlier energy shifts from wood to coal and from horse The prospects for hydrogen as an energy carrier are inti-
to internal combustion engine, all of the pieces are not yet in mately connected to the course of energy developments over
place for hydrogen-powered vehicles to offer an obviously the next half-century. Global energy demand is expected to
cheaper and higher performing alternative to fossil fuel power. double by 2050, and electricity demand to triple, while concern
The disruptive innovations that will enable production and for environment and climate expands. This unique combination
storage for a full-fledged hydrogen economy can come only of trends creates many opportunities for hydrogen to enter the
from high-risk, high-payoff basic materials research, rather global energy mix as a complementary energy carrier and natu-
than incremental approaches. ral partner for electricity. Hydrogen and fuel cells operate effec-
The value of a hydrogen economy, however, extends well tively at all scales, spanning personal, automobile, neighborhood,
beyond transportation. Hydrogen is remarkably attractive in and utility applications. Hydrogen links chemical and electrical
many applications as an energy carrier. Once produced, it is energy with much more flexibility, efficiency, and cleanliness
environmentally friendly, and it is flexibly convertible to other than is possible for fossil fuels. In the transition to new energy
forms of chemical, thermal, or electrical energy at high effi- sources and uses at mid-century, these advantages will have
ciency. In these respects, hydrogen shares the appealing attri- growing importance and appeal.
butes of electricity, the world’s most versatile and fastest-growing Apart from technical and economic factors, safety plays an
energy carrier. Hydrogen and electricity are natural partners, as important role in the evolution of a hydrogen economy. As for
they can be interchanged at high efficiency through electro- gasoline and electricity, the high energy density of hydrogen
chemistry in fuel cells and electrolyzers. Their compatibility is creates the possibility of harm to humans and property. Codes
enhanced by their complementary storage characteristics: and standards need to be developed that enable hydrogen to be
hydrogen stores energy indefinitely and at high density in handled safely and reliably. As hydrogen becomes more com-
chemical form, whereas electricity is typically used at the monplace, these codes and standards need to be disseminated
moment it is produced and lacks a convenient high-density widely to professional hydrogen workers and to the public. This
form of storage. The intermittent character of renewable elec- requires a concerted effort in education about the dangers of
tricity and the diurnal and seasonal fluctuations in demand hydrogen and in safe procedures for its handling. Gasoline and
make local hydrogen storage and interconversion to electricity electricity carry similar dangers, but codes, standards, and edu-
a powerful and natural option. cation have been effective in reducing the incidence of injury
The intimate connection between hydrogen and electricity from these energy sources and carriers to an acceptable level. A
provides attractive routes for the development of a hydrogen similar and deliberate level of commitment to safety codes and
economy. Hydrogen in stationary solid-oxide fuel cells pro- standards will be needed for hydrogen.
duces electricity much more efficiently than do existing coal-
fired power plants. Hydrogen for electricity production is Acknowledgments
available in abundant amounts from the gasification of coal, Funding for this work was provided by the U.S. Department
where CO2 capture is easier and less expensive than in existing of Energy, Office of Basic Energy Science, under Contract DE-
power plants based on the combustion of pulverized coal. The FG02-05ER46241 at the Massachusetts Institute of Technology
hydrogen produced in this way can be burned at high efficiency and the U.S. Department of Energy, Office of Science laboratory,
in combined cycle turbines to produce electricity, converted to operated under Contract DE-AC02-06CH11357 at Argonne.
electricity at high efficiency in solid-oxide fuel cells, or trans-
ferred off the power plant site for other uses such as transporta-
tion. This diversity of deployable options allows the References
1. M.S. Dresselhaus, G.W. Crabtree, M.V. Buchanan, Eds., Basic Research
entrepreneurial spirit to operate with present technology, ready Needs for the Hydrogen Economy (Office of Basic Energy Sciences,
to incorporate new scientific discoveries and technological Department of Energy, Washington, DC, 2003; www.sc.doe.gov/bes/reports/
innovations and to respond to changing market opportunities. abstracts.html#NHE) (accessed January 2008).
Because this development route uses known technologies, it is 2. The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs
not dependent on disruptive innovation in production and stor- (The National Academy Press, Washington, DC, 2004; http://books.nap.edu/
catalog.php?record_id=10922) (accessed January 2008).
age. Furthermore, it requires no new infrastructure off the 3. G.W. Crabtree, M.S. Dresselhaus, M.V. Buchanan, Phys. Today 57(12), 39
power plant site. (2004).
Other attractive routes for hydrogen development are elec- 4. R.L Cohen, J.H. Wernick, Science 214, 1081 (1981).
tricity for personal electronics and emergency power. In both 5. International Energy Outlook 2006, U.S. Energy Information Administration;
cases, performance trumps cost as the top priority, and the por- www.eia.doe.gov/oiaf/archive/ieo06/index.html (accessed January 2008).
table electricity of fuel cells offers advantages over batteries 6. J. Ivy, “Summary of Electrolytic Hydrogen Production” (NREL/MP-560–
35948, 2004; www1.eere.energy.gov/hydrogenandfuelcells/hydrogen_
and fossil fuel generators. Fuel cells can operate for longer publications.html) (accessed January 2008).
times between refueling or recharging than batteries of the same 7. M.Z. Jacobson, W.G. Colella, D.M. Golden, Science 308, 1901 (2007).
weight and volume. For emergency power, hydrogen and fuel 8. B. Esper, A. Badura, M. Rögner, Trends Plant Sci. 11, 543 (2006).
cells produce electricity at much higher efficiency than small 9. K.N. Ferreira, T.M. Iverson, K. Maghlaoui, J. Barber, S. Iwata, Science 303,
fossil fuel generators. They have no moving parts that require 1831 (2004).
10. C. Tard, X. Liu, S.K. Ibrahim, M. Bruschi, L. De Gioia, S.C. Davies,
lubrication and maintenance when not in use. Large units can X. Yang, L.-S. Wang, G. Sawers, C.J. Pickett, Nature 433, 610 (2005).
be stored at critical sites such as hospitals and communications 11. S. Ogo, R. Kabe, K. Uehara, B. Kure, T. Nishimura, S.C. Menon, R. Harada,
centers or delivered by truck to neighborhoods. As with utility- S. Fukuzumi, Y. Higuchi, T. Ohhara, T. Tamada, R. Kuroki, A. Dinuclear, Science
scale power generation, these personal electronics and emer- 316, 585 (2007).

MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy 427
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
 CARRIERS, STORAGE & TRANSFORMATION • HYDROGEN & FUEL CELLS

12. P.V. Kamat, J. Phys. Chem. C 111, 2834 (2007). 31. C.H. Christensen, T. Johannessen, R.Z. Sørensen, J.K. Nørskov, Catal.
13. P. Yu, K. Zhu, A.G. Norman, S. Ferrere, A.J. Frank, A.J. Nozik, J. Phys. Today 111, 140 (2006).
Chem. B 110, 25451 (2006). 32. P. Chen, Z. Xiong, J. Luo, J. Lin, K.L. Tan, J. Phys. Chem. B 107, 10967
14. J.H. Park, S. Kim, A.J. Bard, Nano Lett. 6, 24 (2006). (2003).
15. M. Matsuoka, M. Kitano, M. Takeuchi, Koichiro, M. Anpo, J.M. Thomas, 33. J.F. Herbst, L.G. Hector, Jr., Phys. Rev. B 72, 125120 (2005).
Catal. Today 122, 51 (2007). 34. J.J. Vajo, G.L. Olson, Scripta Mater. 56, 829 (2007).
16. X. Xu, Y. Xiao, C. Qiao, Energy Fuels 21, 1688 (2007). 35. T.B. Lee, D. Kim, D.H. Jung, S.B. Choi, J.H. Yoon, J. Kim, K. Choi, S.-H.
17. P. Chiesa, S. Consonni, T. Kreutz, R. Williams, Int. J. Hydrogen Energy 30, Choi, Catal. Today 120, 330 (2007).
747 (2005). 36. J.L.C. Rowsell, O.M. Yaghi, Angew. Chem., Int. Ed. 44, 4670 (2005).
18. C. Perkins, A.W. Weimer, Int. J. Hydrogen Energy 29, 1587 (2004). 37. K. Wipke, S. Sprik, H. Thomas, J. Kurtz, “Learning Demonstration
19. T.M. Nenoff, R.J. Spontak, C.M. Aberg, MRS Bull. 31, 705 (2006). Interim Progress Report—Summer 2007” (NREL Technical Report 560–
20. S. Satyapal, J. Petrovic, C. Read, G. Thomas, G. Ordaza, Catal. Today 120, 41848, 2007; www.nrel.gov/docs/fy070sti/41848.pdf) (accessed January
246 (2007). 2008).
21. P. Geerlings, F. De Proft, W. Langenaeker, Chem. Rev. 103, 1793 (2003). 38. V.R. Stamenkovic, B. Fowler, B.S. Mun, G. Wang, P.N. Ross, C.A. Lucas,
22. B. Kiran, A.K. Kandalama, P. Jena, J. Chem. Phys. 124, 224703 (2006). N.M. Marković, Science 315, 497 (2007).
23. S.V. Alapati, J.K. Johnson, D.S. Sholl, Phys. Chem. Chem. Phys. 9, 1438 39. J. Greeley, M. Mavrikakis, Nat. Mater. 3, 810 (2004).
(2007). 40. J. Zhang, Z. Xie, J. Zhang, Y. Tang, C. Songa, T. Navessin, Z. Shi, D. Song,
24. J. Greeley, M. Mavrikakis, Catal. Today 111, 52 (2006). H. Wang, D.P. Wilkinson, Z.-S. Liu, S. Holdcroft, J. Power Sources 160, 872
25. W. Grochala, P.P. Edwards, Chem. Rev. 104, 1283 (2004). (2006).
26. B. Bogdanovic, M. Felderhoff, A. Pommerin, F. Schüth, N. Spielkamp, 41. Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Chem. Mater. 15, 4896 (2003).
Adv. Mater. 18, 1198 (2006). 42. J.A. Asensio, S. Borrós, P. Gómez-Romeroa, Electrochim. Acta 49, 4461
27. A. Gutowska, L. Li, Y. Shin, C.M. Wang, X.S. Li, J.C. Linehan, R.S. Smith, (2004).
B.D. Kay, B. Schmid, W. Shaw, M. Gutowski, T. Autrey, Angew. Chem., Int. Ed. 43. Z. Zhou, R.N. Dominey, J.P. Rolland, B.W. Maynor, A.A. Pandya, J.M.
44, 3578 (2005). DeSimone, J. Am. Chem. Soc. 128, 12963 (2006).
28. M.H. Matus, K.D. Anderson, D.M. Camaioni, S.T. Autrey, D.A. Dixon, 44. A. Lashtabeg, S.J. Skinner, J. Mater. Chem. 16, 3161 (2006).
J. Phys. Chem. 111, 4411 (2007). 45. X.C. Lu, J.H. Zhu, J. Power Sources 165, 678 (2007).
29. C.W. Yoon, L.G. Sneddon, J. Am. Chem. Soc. 128, 13992 (2006). 46. J. Wilson, W. Kobsiriphat, R. Mendoza, H.-Y. Chen, J. Hiller, D. Miller,
30. E. Muller, E. Sutter, P. Zahl, C.V. Ciobanu, P. Suttera, Appl. Phys. Lett. 90, K. Thornton, P. Voorhees, S. Adler, S. Barnett, Nat. Mater. 5, 541 (2006). 
151917 (2007).

428 MRS BULLETIN • VOLUME 33 • APRIL 2008 • www.mrs.org/bulletin • Harnessing Materials for Energy
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press
https://doi.org/10.1557/mrs2008.84 Published online by Cambridge University Press