Chemical Engineering Journal 431 (2022) 134310

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

High purity, self-sustained, pressurized hydrogen production from

ammonia in a catalytic membrane reactor
Jose L. Cerrillo a, 1, Natalia Morlanés a, 1, Shekhar R. Kulkarni a, Natalia Realpe a,
Adrian Ramírez a, Sai P. Katikaneni b, Stephen N. Paglieri b, Kunho Lee b, Aadesh Harale b,
Bandar Solami b, Aqil Jamal b, S. Mani Sarathy c, Pedro Castaño a, Jorge Gascon a, *
a
KAUST Catalysis Center (KCC), King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia
b
Carbon Management R&D Division, Research and Development Center, Saudi Aramco, Dhahran 31311, Saudi Arabia
c
Clean Combustion Research Center, King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Keywords: The combination of catalytic decomposition of ammonia and in situ separation of hydrogen holds great promise
Membrane reactor for the use of ammonia as a clean energy carrier. However, finding the optimal catalyst – membrane pair and
NH3 decomposition operation conditions have proved challenging. Here, we demonstrate that cobalt-based catalysts for ammonia
Pd-Au membrane
decomposition can be efficiently used together with a Pd-Au based membrane to produce high purity hydrogen at
High purity hydrogen
Pressurized hydrogen production
elevated pressure. Compared to a conventional packed bed reactor, the membrane reactor offers several oper­
ational advantages that result in energetic and economic benefits. The robustness and durability of the combined
system has been demonstrated for>1000 h on stream, yielding a very pure hydrogen stream (>99.97 % H2) and
recovery (>90 %). When considering the required hydrogen compression for storage/utilization and environ­
mental issues, the combined system offers the additional advantage of production of hydrogen at moderate
pressures along with full ammonia conversion. Altogether, our results demonstrate the possibility of deploying
high pressure (350 bar) hydrogen generators from ammonia with H2 efficiencies of circa 75% without any
external energy input and/or derived CO2 emissions.

capacity [12,13]. Already since its discovery by J. Priestley at 1774

1. Introduction
Over the last decades, the term ‘hydrogen economy’ has gained
increasing significance [1–3 4,5]. Hydrogen is an environmentally
friendly, COx-free, energy vector [6] with a high gravimetric energy
storage density [7]. However, other properties such as its low energy
volumetric density (1.5 kWh/L for liquid H2 and 0.8 kWh/L for gas H2)
or its high flammability are not desirable characteristics for a fuel
[8–10]. Overall, the scientific community agrees that hydrogen will be
key for the decarbonization of our current energy system, although very
important challenges related to production, distribution and storage
need to be addressed [11].
Ammonia is a promising energy carrier in the hydrogen economy.
This chemical offers a couple of advantages in terms of storage potential:
it has a volumetric energy density of 3 kWh/L and a 17.8 wt% hydrogen
[14,15] and its following industrial synthetic production in 1908 by the
famed Haber-Bosch process [16,17], ammonia is one of the most pro­
duced chemicals in the world in terms of volume [18,19]. This is because
NH3 is an important raw material for the manufacture of different
products such as fertilizers, pharmaceuticals, plastics, refrigerants, nitric
acid or explosives, among others [9,18,20]. But not only its production is
mature, the distribution and transport of ammonia through pipelines is
already available in countries such as USA, Russia or Ukraine, while
trucks, trains and ships are used worldwide for the distribution of this
important commodity. For instance, the global maritime trade is 18⋅106
tones per year [21]. Furthermore, liquefaction of ammonia is cost-
affordable and possible through compression to 10 bar at room tem­
perature or by cooling to –33 ◦ C at atmospheric pressure. Additionally,
working with liquified ammonia provides a volumetric hydrogen den­
sity 45% higher than liquified hydrogen [8,22,23]. Lastly, this chemical

Abbreviations: WHSV, Weight hourly space velocity.

* Corresponding author at: King Abdullah University of Science and Technology, KAUST Catalysis Center (KCC), Advanced Catalytic Materials, Thuwal 23955-
6900, Saudi Arabia.
E-mail address: Jorge.gascon@kaust.edu.sa (J. Gascon).
1
These authors contribute equally to this work.

https://doi.org/10.1016/j.cej.2021.134310
Received 28 October 2021; Received in revised form 17 December 2021; Accepted 20 December 2021
Available online 24 December 2021
1385-8947/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
J.L. Cerrillo et al.
Nomenclature
Da Damkohler number
Pe Peclet number
ΔHr ammonia decomposition reaction enthalpy (kWh/kg NH3)
ΔHvap ammonia evaporation enthalpy (kWh/kg NH3)
CpNH3 ammonia specific heat (kWh/K/kg NH3)
ΔT temperature difference (K)
LHVH2 hydrogen Lower heating value (kWh/kg H2)
LHVNH3 ammonia Lower heating value (kWh/kg H2)
MWH2 hydrogen molecular weight (kmol/kg H2)
MWNH3 ammonia molecular weight (kmol/kg NH3)
ηC ammonia cracker efficiency, (90–95%),
ηH2-rec hydrogen recovery (permeated hydrogen) from the
cracked ammonia, (85–95%),
ηheat-rec heat recovery from the retentate, (70–90%)
is less hazardous in terms of flammability and explosion limits than
current fuels or even hydrogen [11,24]. In contrast, the toxicity of
ammonia is higher, being an important barrier for its implementation in
mobile applications. Despite its toxicity, scientists argue that ammonia
risks are acceptable and similar to current fuels, [25] placing NH3 as one
of the most important candidates for the decarbonization of the current
energy system [12,26].
As stated above, ammonia can be directly used as a COx-free fuel,
either in fuel cells [8,27] or through conventional combustion [24,28].
Main issues with the direct utilization are the high temperatures
required by ammonia fuel cells and the potential NOx emissions during
combustion. On the other hand, it can be also used as a hydrogen carrier
to obtain carbon-free H2, which can be directly combusted or used in
fuel cells [29,30]. Nowadays, different techniques such as plasma
[31,32] or photo-driven catalysis [33,34] are investigated to obtain H2
from ammonia, although electrocatalytic and thermocatalytic processes
are the most studied. The electrolysis of ammonia is not viable presently
due to insufficient energy efficiency of the process, formation of NOx,
and NH3 slip during the electro-oxidation [23,35,36]. Hence, the ther­
mocatalytic decomposition of ammonia or ammonia cracking (Equation
(1))is the most promising method.
2NH3 ↔ 3H2 + N2
Ammonia decomposition is an endothermic reaction thermody­
namically limited at low temperatures. Accordingly, higher tempera­
tures than 400 ◦ C are needed to achieve equilibrium conversions above
99% at atmospheric pressure, and even higher temperatures if the
working pressure is higher (Fig. 1) [37,38].
When it comes to catalyst design, ruthenium is the most active metal
for the decomposition of ammonia [6,39–50]. However, the high cost
and availability of this metal make its industrial implementation
doubtful [51]. Besides, when targeting 100% conversion to avoid any
ammonia slip, thermodynamic limitations become more important than
kinetics, making the use of Ru less desired (still almost 500 ◦ C are
needed for full ammonia decomposition at 1 bar [38,52,53]). Other
authors have looked at more abundant and cost-effective elements and
their combinations (i.e. Co, Ni, Mn, Ir or Fe) supported in different
materials [22,51,54–63].
In general, next to thermodynamic limitations, the considerable
endothermicity of the reaction,[6] the inhibitory effect of the produced
hydrogen,[64] and the high purity H2 required for most applications
(higher than 99.97% [22]), they all point at the in situ removal of
hydrogen as the most feasible strategy to efficiently decompose
ammonia.[65–68] Specifically, the combination of membranes and
catalytic ammonia decomposition allows to separate high purity H2 in
situ. Various authors have already demonstrated that the combination of
Chemical Engineering Journal 431 (2022) 134310
Fig. 1. Thermodynamic ammonia conversion versus temperature at
different pressures.
H2-selective membranes with catalytic systems in packed-bed catalytic
membrane reactors (PBCMR) can lead to overcoming equilibrium con­
straints and speed up reaction kinetics [67,69,70]. The thin Pd layers
used as H2 selective membranes present restrictions in terms of working
temperature. At low temperature (T < 400 ◦ C), apart from the low
hydrogen permeation due to the limitation in the endothermic diffusion
of hydrogen through the membrane, the integrity of the membrane is
compromised due to hydrogen embrittlement. Meanwhile, at higher
temperatures T > 550 ◦ C, cracking at the ceramic–metal interface might
occur due to a mistmatch in thermal expansion. In constrast, PdAu
membranes offer improvements in terms of H2 permeation at low tem­
perature and higher resistance to embrittlement [71,72] In this spirit,
we selected PdAu alloy membranes as the most adequate to be combined
with our recently reported barium promoted cobalt-cerium catalyst.
[73]
Apart from purity requirements, hydrogen should be supplied or
stored at high pressure. Therefore, in order to avoid additional
compression costs [74–76] and derived emissions (estimates are about
6.0 kWh/kg for compression of H2 to 70 MPa, which leads to approxi­
(1) mately 1.3 kg of CO2/kg of hydrogen) [77], high pressure in the
permeate side of a PBCMR would be desirable.
Here, we demonstrate the very efficient production of hydrogen by
combining our recently developed Co-based catalyst and a state-of-the-
art H2-selective membrane. The membrane dramatically enhances con­
version over the catalyst, proving the ability of the PBCMR to both,
accelerate the kinetics and overcome equilibrium constraints. A full
process optimization with the objective of minimizing energy re­
quirements has been performed. The system offers the possibility of
utilizing the energy of the hydrogen in the retentate to cover the thermal
requirements for the process, developing in such a way a stand-alone
system with a H2 efficiency of ~ 78%. Furthermore, the possibility of
producing the H2-permeate stream at moderate pressure results in
enormous energy savings and prevents further efficiency losses, and also
decreases CO2 emissions associated to the compression step. Last but not
least, we demonstrate the industrial potential of the system by running
the membrane reactor for over 1000 h.
2. Result and discussion
2.1. Benchmarking the membrane reactor configuration
To elucidate the benefits of incorporating a H2-selective membrane
to a packed-bed catalytic reactor (PBCR), the catalytic activity was
2
J.L. Cerrillo et al.
firstly evaluated in the membrane module with the permeate stream
opened (PBCMR) and closed (PBCR). In the second case, the reaction
mixture goes exclusively through the retentate or catalytic chamber,
mimicking a PBCR. The evolution of ammonia conversion using both
approaches (PBCR and PBCMR) at different temperatures, contact times
and pressures was analyzed (Fig. 2). The comparison of both experi­
ments at different conditions reveals a significant improvement (10–30
%) in ammonia conversion when the PBCMR was employed, in agree­
ment with previous studies [69,70]. The presence of hydrogen has an
inhibitory effect on the kinetics of the reaction according to the Temkin-
Pyzhev mechanism [78,79]. Therefore, the removal of H2 from the
retentate chamber improves ammonia conversion. This effect can be
seen only at temperatures higher than 400 ◦ C, minimum temperature to
attain H2 transport through the Pd-Au membrane (Supplementary
Figure S5).
The purity of the recovered hydrogen in all the PBCMR experiments
was 99.97 ± 0.03 %, demonstrating the absence of defects in the
membrane (see Figures S6-S8). Minor N2 impurities can be observed at
pressures higher than 9 bar (Supplementary Table S2), in agreement
with previous studies [64,67]. Overall, ammonia conversion increases
when temperature (250 – 500 ◦ C), contact time (2.7 – 16.3 g h mol− 1)
and pressure (4 to 15 bar) are increased. A similar trend can be observed
in both reactor configurations in the case of temperature and contact
time. In contrast, higher pressures are only beneficial in case of the
PBCMR [40,70]. This negative effect in the PBCR, when pressure is
increased, is the result of the previously mentioned inhibitory effect of
hydrogen on the reaction kinetics and of the strong thermodynamic
limitations. [80 78 80 81] Finally, the PBCMR without catalyst did not
present any significant ammonia conversion, demonstrating no catalytic
activity of the membrane (Supplementary Table S1).
2.1.1. Optimizing operation in the PBCMR
Fig. 3 shows the effect of temperature in ammonia conversion,
hydrogen recovery, and composition of the permeate and retentate, as
well as the measured flows, when pressure in the retentate is fixed to 4
bar (ΔP = 3 bar). As described above, both ammonia conversion and
hydrogen recovery increase with the reaction temperature, reaching
total conversion and almost 80% H2 recovery at 500 ◦ C. Under these
conditions, pure hydrogen is obtained in the permeate. However, not all
produced hydrogen can permeate through the membrane after partial
pressures of this gas (PH2) are equalized at both sides of the membrane
(see maximum recovery, that would correspond to similar PH2 in both
permeate and retentate). Indeed, in membrane separation processes, the
flow of the permeable gas is a function of the driving force (expressed as
difference in chemical potential) across the membrane.
Matching the permeation properties of the membrane and the ac­
tivity of the catalyst is of primary importance to ensure the full
Fig. 2. Effect of the reaction conditions (T, W/F and P) using PBCR (blue) and PBCMR (red). 0.5%Ba-CoCe catalyst: 2 g (350–500 μm) diluted with 10 g of SiC ~ 680
μm. Fixed conditions are ammonia flow rate 100 N mL min− 1, T = 485 ◦ C , Pretentate = 4 bar and Ppermeate = 1 bar. In each graph one variable was varied: Temperature
250 – 500 ◦ C (left), ammonia flow rate 50–300 N mL min− 1 (center), and Pretentate = 4–15 bar (right). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
3
Chemical Engineering Journal 431 (2022) 134310
utilization of both components. In this line, we investigated the effect of
space velocity over membrane reactor performance (Fig. 4). The flow
rate of ammonia was modified in the range 50–300 N mL min− 1, varying
the weight hourly space velocity (WHSV) in the range 300–1800 ml gcat-
1 − 1
h . At low WHSV, full conversion and maximum hydrogen recovery
are achieved. Under these conditions, the membrane is capable of
permeating the produced H2 until similar PH2 are reached at both both
sides of the membrane. In contrast, at WSHVs higher than 1000 ml gcat-
1 − 1
h , the membrane is not able to permeate all produced hydrogen,
resulting in recovery rates below the maximum value. At higher WSHVs,
full conversion cannot be achieved and some NH3 slip is observed in the
retentate.
Additional experiments at increasing pressure in the retentate were
performed. As shown in Fig. 5, upon increasing the pressure in the
retentate chamber and consequently the driving force for permeation of
H2 through the membrane, full ammonia conversion with much higher
recovery rates can be achieved. This is the result of a much more effi­
cient use of the membrane that, in addition, strongly suppresses the
well-known inhibition of H2 on kinetics and further brings the reaction
far from equilibrium limitations [53,64,67,81]. The positive effect of
increasing the pressure on ammonia conversion and hydrogen re­
coveries is particularly interesting at higher space velocities (Supple­
mentary Figure S9).
2.2. Production of pressurized H2
A sets of experiments were conducted at 485 ◦ C and 200 ml NH3
min− 1, keeping constant the pressure in the permeate chamber at 1, 3, 5,
10 and 15 bar, and increasing the pressure in the retentate chamber from
1 to 50 bar. To the best of our knowledge, this is the first example in the
open literature reporting the production of hydrogen from ammonia at
pressures higher than atmospheric. Similarly, we could not find any
publication reporting experiments at >6 bar in the retentate.
As observed in Fig. 6, the pressure ratio retentate/permeate even­
tually determines hydrogen recovery and productivity. However,
reaching full ammonia conversion at higher permeate pressures be­
comes more difficult and may lead to undesirable ammonia slip (Sup­
plementary Figures S10-S12), together with the increase in the
impurities in the hydrogen permeate stream (Supplementary Table S3).
In addition, similar experiment were carried out at lower temperatures
(450 ◦ C) demonstrating similar trends at lower conversion and H2 re­
covery values (Supplementary Figure S14). These results however,
demonstrate the great potential of this configuration and offer a great
degree of flexibility for the optimization of the full process (vide infra).
J.L. Cerrillo et al. Chemical Engineering Journal 431 (2022) 134310

Fig. 3. Effect of the reaction temperature in the ammonia conversion and hydrogen recovery (left), the molar composition and retentate flow (right-up) and molar
composition and permeate flow (right-down). Reaction conditions: 0.5%Ba-CoCe catalyst: 10 g (350–500 mm) diluted with 20 g of SiC ~ 680 mm. Ammonia flow
rate = 200 N mL min− 1, WHSV = 1200 ml gcat-1h− 1, T = 400–500 ◦ C, Pretentate = 4 bar; Ppermeate = 1 bar.

Fig. 4. Effect of the weight hourly space velocity in the ammonia conversion and hydrogen recovery (left), the molar composition and retentate flow (right-up) and
molar composition and permeate flow (right-down). Reaction conditions: 0.5%Ba-CoCe catalyst: 10 g (350–500 μm) diluted with 20 g of SiC ~ 680 μm. Ammonia
flow rate = 50–310 N mL min− 1, WHSV = 300–1860 ml gcat-1h− 1, T = 485 ◦ C, Pretentate = 4 bar; Ppermeate = 1 bar.

2.3. Controlling regimes ( )

Ea
k0 exp − RT
P(a+b)
p W
In order to gain further insights into the controlling regimes during Da =
membrane operation, we made use of two well known dimensionless FNH3 ,0
numbers in the field of membrane reactors: the Damkohler (Da) and the
FNH3 ,0
Peclet (Pe) numbers. The definitons are equivalent to the ones proposed Pe =
JH2 APp
by Gomez-garcia et al. [82] who modelled a similar shell-tube type
membrane reactor for ammonia decomposition. The choice of catalyst Pf
mass in the Da definition arrives from the definition of k0 unlike the Pr =
Pp
original work, where volume based definition were employed. An
additional dimensionless number was introduced for the regimes eval­ Where k0 is the pre-exponential factor (8.2 × 107 mol kgcat− 1 s− 1);
uation at higher pressures. This number was derived as the ratio be­ Ea the activation energy (116.11 kJ mol− 1); a the ammonia reaction
tween the feed and the permeate pressures. order (0.73); b the hydrogen reaction order (-1.62); Pf and Pp are feed

4
J.L. Cerrillo et al. Chemical Engineering Journal 431 (2022) 134310

Fig. 5. Effect of the retentate pressure in the ammonia conversion and hydrogen recovery (left), the molar composition and retentate flow (center) and molar
composition and permeate flow (right). Reaction conditions: 0.5%Ba-CoCe catalyst: 10 g (350–500 μm) diluted with 20 g of SiC ~ 680 μm. Ammonia flow rate = 200
N mL min− 1, WHSV = 1200 ml gcat-1h− 1, T = 485 ◦ C, Pretentate = 4–9 bar, Ppermeate = 1 bar.

Fig. 6. Effect of the permeate pressure using different retentate pressures on conversion and H2-recovery. Reaction conditions: 0.5%Ba-CoCe catalyst: 10 g (350–500
μm) diluted with 10 g of SiC ~ 680 μm. Ammonia flow rate 200 N mL min− 1, T = 485 ◦ C , Pretentate = 4–50 bar, Ppermeate = 1–15 bar.

and permeate pressures respectively; W the catalyst mass (g); FNH3,0 is
the initial ammonia flow rate (mol s− 1); JH2 the hydrogen permeance
(1.17 × 10-7 mol m− 2 pa-1 s− 1) and A the membrane surface area
(0.00716 m2). The values of reaction parameters are estimated with
similar methodology followed by Sayas et al. [80]
Da compares the reaction rate against the convective mass transfer
rate, while the Pe number correlates the rates of two modes of mass
transfer: convection through the catalytic bed against diffusion through
the membrane. High Da values (obtained for either a high catalyst
loading or a low reactant feed flow) would indicate a fast reaction that is
close to equilibrium and will be controlled by the convective mass
transfer from the gas to the catalyst. Low Pe values would indicate
relatively faster diffusion through the membrane on account of high
permeability. As a consequence, a combination of high Da and low Pe

5
J.L. Cerrillo et al.
Energy of the hydrogen recovered in the permeate − Net Energy requirements
η=
Energy of the hydrogen recovered in the permeate
maximizes the potential of membrane reactors against their packed bed
counterparts. On the contrary, low Da along with high Pe values are
obtained for membrane reactors operating close to the performance of
packed bed reactors.
The Da and Pe numbers have been calculated using the equations
described in the supplementary information and the values of reaction
parameters have been estimated following a similar methodology as that
previously reported by Sayas et al. [80]. Fig. 7 collectively compares Da
and Pe values for various experiments performed in the packed bed
catalytic membrane reactor. In two instances where different catalyst
loadings have been used (2 g represented by rhombuses and 10 g by
stars). At high temperatures, Da values are sufficiently high to reach
almost full ammonia conversion, as noted experimentally in Fig. 2.
Simultaneously, the Pe values are constant and sufficiently low to pro­
mote hydrogen permeation, resulting in high H2-recoveries on the
permeate side. Even at higher Pe numbers, such as in the case of 10 g
catalyst loading, the ammonia conversion and H2-recovery results are
still promising, due to the Da number values increasing roughly by an
order of magnitude when the catalyst amount is increased.
The effect of WHSV is shown in Fig. 7. At 485˚C and 2 g of catalyst
loading, when WHSV varies from 1800 to 9000 ml gcat-1h− 1 the Da and
Pe numbers change (represented by squares). The evaluated operating
conditions cover the low-to-moderate Pe range, and show conversions
higher than 80% and high (>80%) recovery as Pe decreases; in good
agreement with the experimental data reported earlier in Fig. 2. Simi­
larly, at the same temperature, 10 g of catalyst loading and WHSV
ranging from 360 to 1800 ml gcat-1h− 1 (represented by circles) the re­
sults cover the same Pe range but higher Da numbers; obtaining higher
conversions and higher recoveries, as expected (Fig. 4). In any case, both
experiments follow the same trend, where conversion and H2-recovery
increase as the ratio between Da and Pe increases. Lastly, the effect of
pressure ratio studied in Fig. 6 is further explored in terms of dimen­
sionless numbers, symbolized by triangles. Both, Da and Pe numbers are
kept constant at different pressure ratios. This effect ban be better
( )
MWH2
Net energy requirements = ΔHvap + CpNH3 × ΔT + ΔHr × ηC − LHVH2 ×
MWNH3
6
Chemical Engineering Journal 431 (2022) 134310
(2)
observed in Figure S15, where ammonia conversion and H2-recovery are
represented in terms of different pressure ratios. Once again, a signifi­
cant rise in Da values overcomes the slight increase in resistance to
permeation due to increased Pe numbers and allows for a higher re­
covery of hydrogen on the permeate side of the reactor. Overall, this
analysis shows that, under the studied reaction conditions, optimal
usage of both catalyst and membrane is possible and highlights the
importance of a good matching between both reactor components.
2.3.1. Efficiency of the process
Energy efficiency of the process is an issue of the utmost importance
in the field energy storage. The overall efficiency in this case was
calculated using Equation (2).[19,83,84 85]
The energy requirements include: (i) ammonia evaporation (ΔHvap),
(ii) energy required to reach the reaction temperature (400–500 ◦ C,
CpNH3 × ΔT), and (iii) the energy to convert ammonia into hydrogen in
the catalytic bed (ΔHr). The thermal energy losses in the system, or the
thermal energy that could be recovered from the reaction products
during the cooling step from the reaction temperature (400–500 ◦ C) to
room temperature are not considered here. [19]
The heat required for the ammonia decomposition in order to
maintain hydrogen production can be supplied by gas turbine, high
temperature fuel cell, or diesel power plants. However, for standalone
systems, a fraction of the hydrogen produced could be used to cover the
energetic demands of the whole process, for instance using the hydrogen
which is not recovered in the permeate. Thermodynamically, ammonia
decomposition requires around 15–20 % of the hydrogen calorific value
[84,86]. Therefore, with H2 recoveries around 80–85%, the energy
content of the retentate will be sufficient to satisfy the endothermic
requirement for NH3 decomposition. In such a stand-alone system, the
net energy requirements can be calculated according to Equation (3).
× ηC × (1 − ηH2− rec ) × ηHeat− rec − LHVNH3 × (1 − ηC ) × ηHeat− rec (3)
Fig. 7. Comparison of (a) ammonia conversion
and (b) hydrogen recovery as a function of Da
and Pe for various experimental conditions uti­
lized in Figs. 2-4. (■) 2 g catalyst, 485 ◦ C, 1800
to 9000 ml gcat-1h− 1, Ppermeate = 1 bar and Pre­
tentate = 4 bar; (●) 10 g catalyst, 485 C, 360 to
◦
1800 ml gcat-1h− 1, Ppermeate = 1 bar and Pretentate
= 4 bar; (◆) 2 g catalyst, 400 to 500 ◦ C, 3000 ml
gcat-1h− 1, Ppermeate = 1 bar and Pretentate = 4 bar;
(★) 10 g catalyst, 400 to 500 ◦ C, 1800 ml gcat-
1 − 1
h , Ppermeate = 1 bar and Pretentate = 4 bar; (▴)
10 g catalyst, 485 ◦ C, 1200 ml gcat-1h− 1, Ppermeate
= 1 to 45 bar and Pretentate = 1 to 15 bar.
J.L. Cerrillo et al.
η in terms of the available options to cover the energetic demand for the
Fig. 8. Energetic analysis of the hydrogen production and purification process
by ammonia decomposition conducted in a membrane reactor.
Where ηC is the ammonia cracker efficiency, (90–95%), ηH2-rec is the
hydrogen recovered (permeated hydrogen) from the cracked ammonia,
(85–95%) and ηheat-rec is the heat recovered from the retentate,
(70–90%). LHVi and MWi represent the lower heating value and the
molecular weight of the components, respectively. All the energetic
terms in Equations 2–3 are expressed in kWh per kg of ammonia. The
energy of the hydrogen recovered in the permeate is given by Equation
(4)
Energy of the hydrogen recovered in the permeate = LHV H2 ×
( )
MW NH3 × ηC × ηH2− rec (4)
MW H2
Fig. 8 summarizes the parameters considered to determine the effi­
ciency of the process. According to equations 2–4, the overall efficiency
for the hydrogen production-purification system would be 77.8 %, when
considering the use of the H2 in the retentate (and eventually the non-
converted NH3) to provide a large part of the energy needs; for an
85% H2-recovery and 95% the efficiency of the cracker. It has to be
stressed that this overall efficiency considers that the rest of the required
energy to drive the process is provided by waste streams, something that
may not be realistic in case of stand-alone systems. Indeed, the optimum
H2-recovery for such stand-alone units would depend on their location
Fig. 9. Energy requirements for hydrogen compression from 1 to 20 bar in the permeate up to 700 bar (left), and the energy savings resulting from the production of
a pressurized permeate streams in the range 5–20 bar (right).
7
Chemical Engineering Journal 431 (2022) 134310
whole process. If no external energy inputs are available, a maximum
recovery of ~ 70–75% would be the optimum value for an energy
neutral process.
Finally, an extra consideration could be added to the overall effi­
ciency of the process by including the the energy requirements for the
compression of hydrogen produced in the permeate. Here we take 3
different examples: fuel cells (up to 10 bar), vehicles (up to 350 bar), and
hydrogen storage (up to 700 bar). Energy requirements for compression
of hydrogen for these three scenarios are depicted in Fig. 9 considering
hydrogen in the permeate is directly produced in a range between 1 and
20 bar. As it can be observed, there is a considerable difference in the
required energy as a function of the starting pressure of H2. This trans­
lates into important energy savings and highlights the importance of
operating membrane reactors at elevated pressures, not only from a
productivity point of view. For example, in the case of producing
hydrogen in the permeate at 5 bar, this will reduce the energetic re­
quirements to compress this hydrogen up to 350 bar, by circa a 30%
(compared to the case in which the compression was done from atmo­
spheric pressure). The savings could reach up to almost 60% if hydrogen
was produced at 20 bar prior to compression at 350 bar.
Fig. 10 shows the efficiency of the process considering the energy
needed for compression. For example, if the hydrogen produced is
required to be compressed up to 350 bar, the efficiency would drop from
77.7% to ~ 68–73%, when hydrogen in the permeate is produced at
1–20 bar, respectively. Fig. 10 clearly reveals the benefits of producing
the hydrogen permeate stream a certain value above the atmospheric
pressure, if the hydrogen produced requires further compression. The
drop in efficiency is smaller when increasing the pressure in the
permeate (for all the scenarios considered here of compression up to 25,
350 or 700 bar). Only a 5% of efficiency loss is observed when the
permeate is produced at 20 bar and requires further compression to 350
bar, compared to the 12% loss in the case of producing the permeate at 1
bar.
Another benefit of producing hydrogen in the permeate at certain
pressure above the atmospheric pressure is an important decrease in CO2
emissions associated to the compression step (considering that these
compressors are ran by non renewable energy sources). Fig. 11 shows
smaller CO2 emissions when hydrogen permeate is directly produced at
increasing pressures in the permeate stream. A decrease in the CO2
emissions in the range of 30–60% would be achieved when hydrogen is
compressed up to 350 bar from 5 to 20 bar, with respect to the case in
which the hydrogen was produced at atmospheric pressure.
In the previous section we have demonstrated the possibility of
producing the permeate at several pressures by simply increasing further
the pressure in the retentate (Fig. 6). It is important to highlight that
hardly any compression work is required to conduct the reaction at
J.L. Cerrillo et al.
Fig. 10. Efficiency and efficiency loss when including the energetic compression requirements up to higher pressures in the range 25–700 bar, for the cases in which
the permeate is produced at 1–20 bar (80% H2-recovery was used for the calculation).
Fig. 11. CO2 emission associated to the compression step of the hydrogen produced in the permeate.
elevated pressure, because pressurizing liquid ammonia is simple and
cheap.
2.3.2. Hydrogen purification technologies: Membrane reactor versus
conventional pressure swing Adsorption.
The feasibility of the ammonia decomposition process to produce
high purity hydrogen has been additionally investigated by comparing
the membrane reactor to a conventional pressure swing adsorption
(PSA) technology using Aspen plus simulations. Here we compare the
energy consumption for the two hydrogen purification technologies
(Supplementary Table S4).
The conventional process uses an equilibrium reactor (GIBBS), fol­
lowed by an NH3 recovery unit, a PSA purification unit and a H2
compression unit (Supplementary Scheme S2). The GIBBS reactor
operates at 20 bar and 500 ◦ C in the reactor. The PSA adsorption is
simplified as a theoretical separator with 90% H2 recovery and 0.1% N2
slip [87]. The energy consumption of the vacuum pump (0.5 bar) and
the further recompression are considered the main energy inputs for this
technology [88]. On the other side, the membrane reactor is modeled as
a combination of a GIBBs equilibrium reactor and a theoretical separator
with 95% recovery (Supplementary Scheme S3). No PSA unit is needed
in this case. The reaction conditions for the membrane reactor are
selected from experimental data, in particular 20 bar of feed, 3 bar of H2
permeate, and 95% H2 recovery with 99.97% purity (see Fig. 6).
The hydrogen purification technology coupled to the NH3 decom­
position was successfully modeled with Aspen plus for the membrane
reactor and the conventional PSA, delivering the target 99.9% H2 purity.
The initial case study has been considered at 20 bar and 500 ◦ C in the
reactor, with H2 delivery also at 20 bar. Interestingly, 51% electric
8
Chemical Engineering Journal 431 (2022) 134310
power savings in the membrane-based process are found due to the
absence of the PSA and the higher pressure of the final H2 stream after
the reactor (Supplementary Table S4). As a result, a 31% utility cost
savings could be achieved. Moreover, we also need to point out that, due
to the difficulty of achieving 100% conversion in standard reactors and
despite the NH3 recovery unit that works at − 100 ◦ C, close to 500 ppm of
NH3 slip is observed in the PSA scenario (stream S9, in supplementary
Scheme S2). This will most likely imply the need for an additional NH3
slip reaction unit (ASC) [89] as, apart from being one order of magni­
tude higher than the Permissible Exposure Limit (50 ppm), small traces
of NH3 can compromise the performance of the PSA and compression
units [90]. Furthermore, Fuel Cells require NH3 concentrations below
0.1 ppm[91]. As a result, extra energy requirements will be needed in
the conventional scenario for this ASC unit. On the other hand, in our
membrane system, the H2 stream directly comes out virtually free of
NH3 (see Fig. 3), revealing another important advantage of our system.
Moreover, the retentate stream still rich in H2, and could be used in a
cogeneration cycle to generate 3 MW of the electric power needed to run
the plant, reducing further the energy consumption and utility costs
(Supplementary Scheme S4).
Following, we evaluated the accuracy of the above estimations of
energy efficiency for an energy stand-alone membrane system by per­
forming additional simulations with 85% and 75% H2 recovery in the
membrane. While reducing the recovery will lessen the hydrogen pro­
duction rate, it will result in higher H2 content of the retentate stream for
the cogeneration cycle and, at the same time, higher pressures in the
permeate stream (see Fig. 6), both aspects being beneficial for the stand-
alone energetic operation of the plant. The results can be observed in the
Supporting Information at Table S4 and, indeed, lowering the H2
J.L. Cerrillo et al.
recovery in the membrane results in higher energy produced in the
coreneration cycle with lower electricity needed for the final compres­
sion stem. While in neither of the cases we generate enough electric
power for a stand-alone operation, we need to stress that the power cycle
of Scheme S4 (Supplementary Information) is solely designed for elec­
tric power generation and, in our particular case, we could omit the
second turbine with the water-steam cycle and used instead the heat of
the low pressure hot gas (Supplementary Scheme S4, 18.25 MW in
HEATX for a 75% H2 recovery) to provide the required energy for the
reactor (19.1 MW). In this case, the first turbine will provide more than
enough power (8.5 MW) for the rest of the utilities, effectively reaching
the desired energy and stand-alone operation.
2.3.3. Stability of the PBCMR
Last but not least, the stability of the system was further analyzed,
since it could become one of the limiting steps to the development of this
technology [83,92]. Fig. 12 shows the catalytic performance over the
reaction time for the long-term stability test conducted. No detrimental
effects were observed on ammonia conversion, hydrogen recovery,
composition or flow rates of the permeate and retentate. From the cat­
alytic point of view, no ammonia was detected in the chromatographic
analysis of the retentate, neither in the permeate stream. To uncover any
unreacted ammonia under the detection limit of the chromatographs
(detection limit 0.01–0.02 %vol.), a sulfuric acid trap (0.5 M) was
installed to capture ammonia at the ppm level (in the permeate and
retentate sides). Aliquots of the sulfuric acid trap were daily analyzed
following the indophenol blue protocol [93]. We can unequivocally
assure that no ammonia was detected in any of the exhausts of the
system, even at the ppm level.
In order to validate the robustness of the PBCMR, the system was set
at the maximum operating conditions specified by the supplier, 500 ◦ C
and 16 bar for 24 h. After that, the conditions were set back to the
previous long-term stability test for 24 extra hours. The results (Sup­
plementary Figure S16) were almost identical to the previous ones,
corroborating the stability and the robustness of the PBCMR. Finally,
9
Chemical Engineering Journal 431 (2022) 134310
Fig. 12. Composition and flow at
retentate (a), ammonia conversion and
H2 recovery (b) and composition and
flow at permeate (c) at long-term sta­
bility test with Pd-Au membrane module
Reaction conditions: 0.5%Ba-CoCe cata­
lyst: 10 g (350–500 μm) diluted with 20
g of SiC ~ 680 μm; Ammonia flow rate
= 170 N ml min− 1; T = 485 ◦ C, Pretentate
= 4 bar; Ppermeate = 1 bar. Ammonia
conversion, H2 recovery and H2 purity of
the permeate (d) using the same condi­
tions in cycles of heating/cooling.
several cooling /heating cycles were conducted. The membrane module
was purged with inert gas (Ar, 100 N mL min− 1 for 1 h), and kept under
Ar to protect the H2-selective membrane and to prevent the reoxidation
of the Co-based catalyst. Then, cooling and heating ramps were set at
1 ◦ C min− 1 to prevent the stress between membrane and ceramic sup­
port. The cooling /heating cycles were followed by the ammonia reac­
tion at the same operating conditions as in the long-term stability test,
for 24 h each cycle. The catalytic performance for six cycles (Fig. 12 and
Supplementary Figure S17,) clearly showed the stability of the mem­
brane, with no significant differences in conversion and recovery.
Overall, the total working time of the system under several reaction
conditions reached ~ 1000 h (when including the long term and addi­
tional experiments performed with the same membrane module and
catalyst).
3. Conclusions
The production of high purity H2 from NH3 using packed bed
membrane reactors holds great promise. The robustness and durability
of such system has been demonstrated for >1000 h on stream, yielding a
very pure hydrogen stream (>99.97 % H2) and H2 recovery (>90 %). In
situ separation of hydrogen results in faster cracking kinetics and helps
overcome equilibrium limitations. However, a good match between the
properties of catalyst and membrane needs to be achieved: Pd based
membranes are very effective in a relatively narrow operation window
(namely 400 to 550 ◦ C), implying that the use of catalysts active and
selective in this temperature range is mandatory and that productivity of
membrane an catalysts need to be well matched. This is the case for the
combination studied in this work.
When considering NH3 as potential H2 carrier, efficiency of the
cracking has to account for the final application of the produced H2,
including compression to the desired pressure and important environ­
mental aspects such as potential NH3 slip. Our work demonstrates that
nearly thermodynamic efficiencies can be achieved when the appro­
priate membrane reactor is employed. Indeed, using this technology, we
J.L. Cerrillo et al.
have demonstrated that stand alone systems with a H2 efficiency of circa
75% can be achieved in the production of H2 at 350 bar. We expect our
results to pave the way towards the the development of this tehcnology
in a scalable manner through parallelization, enabling a pathway for
distributed use of hydrogen from ammonia for a wide range of
applications.
4. Experimental details
The barium promoted cobalt-cerium catalyst (BaCoCe, 80/20 M ratio
Co/Ce) was synthesized by coprecipitation followed by the addition of
barium as promoter using incipient wetness impregnation (0.5 wt%. Ba).
Experimental details of the synthesis procedure and characterization of
the catalyst can be found in our previous work [73]. Prior to the ex­
periments, the catalyst was reduced under pure hydrogen (at atmo­
spheric pressure) at 485 ◦ C for 8 h.
The membrane reactor (PBCMR) was supplied by GaoQ Functional
Materials Co. (Nanjing, China). The reaction system consists of a Pd-Au
alloy membrane (i.d., 8 mm; o.d, 12 mm; effective membrane area, 70
cm2; thickness, 8 µm; Au content, 15 wt%) in a stainless-steel module.
The thin Pd-Au alloy layer is supported on the external surface of a
porous ceramic tube. The shell and tube sides are the retentate and
permeate chambers, respectively. The catalyst (2–10 g) is incorporated
in the annular space (retentate chamber) together with SiC in a catalyst:
SiC weight ratio of 1:2.
The ammonia decomposition tests were performed using a PID
Microactivity Reference system. The experimental setup is shown in
Scheme S1 in the Supplementary Information. The original furnace has
been replaced with a longer one to ensure isothermal operation in the
membrane reactor. The experiments were carried out at different tem­
peratures (300 to 600 ℃), pressures (1 to 25 bar) and WHSV (300–1800
gcat mL-1h− 1) using different amounts of catalyst (2 to 10 g) diluted with
SiC in diverse ratios (catalyst: SiC = 1:10 and 1:2). Each set of experi­
mental conditions was kept for two hours. The flow rates of NH3, H2, N2
and Ar were carefully fixed by mass flow controller in order to explore
those different reaction conditions. Gases from the two chambers were
analyzed online by two Micro-GC in an online gas chromatograph
(3000A Micro-GC gas analyzer, Agilent) equipped with thermal con­
ductivity detectors and using He as internal standard (2 ml min− 1). One
channel equipped with a PLOT-U precolumn, followed by a Molsieve
column using Ar as the carrier gas was used for detection of N2 and H2; a
second channel with a PLOT-U column and He as the carrier gas, was
used for NH3 detection.
The experimental setup is designed to feed liquified ammonia with
the possibility of performing reactions at high pressure. This feature of
the system enables high-pressure experiments and could potentially be
used to generate pressurized pure hydrogen on the permeate side.
Experimental verification of the concept was done at feed pressure up to
50 bar and permeate pressure up to 20 bar.
During heating, both the tube/retentate side and the shell/permeate
side of the module were fed with Ar to avoid hydrogen embrittlement of
Pd membrane at low temperatures. The temperature was carefully
raised with 1–2 ◦ C/min.
The maximum flow rate was limited by the system’ s ability to
maintain a constant catalyst bed temperature, as the high flow rate and
endothermic reaction may result in a significant temperature gradient
across the catalyst bed.
The use of a sweep gas could decrease the partial pressure of
hydrogen in the permeate side, thereby, increasing the driving force for
permeation, but for hydrogen-production applications, dilution by a
sweep gas is, in principle, not desired. Because of this, all experiments
were conducted without sweep gas to obtain pure hydrogen in the
permeate.
A set of experiments was carried out using 200 mg of pelletized
BaCoCe between 300 μm and 500 μm and diluted with 2 g of SiC,
mimicking the experimental conditions reported in our previous work.
10
Chemical Engineering Journal 431 (2022) 134310
Additional experiments were performed using 10 g of BaCoCe and SiC in
a 1:2 ratio. The effect of the operational conditions (P, T and WHSV) on
ammonia conversion, hydrogen recovery and productivity were evalu­
ated using the PBCMR. Long-term stability tests were carried out for
>1000 h, including heating/cooling cycles and diverse operational
conditions to determine the robustness of the membrane and the sta­
bility of the catalyst. Preliminary experiments to evaluate the hydrogen
permeability and capability of the membrane module (Supplementary
Figures S1-S4).
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgment
The authors gratefully acknowledge the financial support provided
by Saudi Aramco, and the resources and facilities provided by the King
Abdullah University of Science and Technology.
Author Contributions
N.M. and J.G. conceived this work. J.L.C and N.M. designed, char­
acterized and conducted synthesis and catalytic experiments. N.R. and S.
R.K. performed control regime studies. A.R. performed the technology
simulation comparison and provided experimental support. J.G., P.C.
and S.M.S. provided revision suggestions. J.L.C and N.M. wrote the
original draft of the manuscript. All authors have given the approval to
the final version of the manuscript.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.cej.2021.134310.
References
[1] C.H. Christensen, T. Johannessen, R.Z. Sørensen, J.K. Nørskov, Towards an
ammonia-mediated hydrogen economy? Catalysis Today 111 (1-2) (2006)
140–144.
[2] M. Yadav, Q. Xu, Liquid-phase chemical hydrogen storage materials, Energy &
Environmental Science 5 (12) (2012) 9698–9725.
[3] G.W. Crabtree, M.S. Dresselhaus, M.V. Buchanan, The hydrogen economy, Physics
today 57 (12) (2004) 39–44.
[4] J.O. Abe, A. Popoola, E. Ajenifuja, O. Popoola, Hydrogen energy, economy and
storage: review and recommendation, International Journal of Hydrogen Energy
44 (29) (2019) 15072–15086.
[5] G.E. Ioannatos, X.E. Verykios, H2 storage on single-and multi-walled carbon
nanotubes, International Journal of Hydrogen Energy 35 (2) (2010) 622–628.
[6] T. Bell, L. Torrente-Murciano, H 2 production via ammonia decomposition using
non-noble metal catalysts: A review, Topics in Catalysis 59 (15) (2016) 1438–1457.
[7] C.D. Demirhan, W.W. Tso, J.B. Powell, E.N. Pistikopoulos, Sustainable ammonia
production through process synthesis and global optimization, AIChE Journal 65
(7) (2019) e16498, https://doi.org/10.1002/aic.16498.
[8] A. Valera-Medina, F. Amer-Hatem, A.K. Azad, I.C. Dedoussi, M. de Joannon, R.
X. Fernandes, P. Glarborg, H. Hashemi, X. He, S. Mashruk, J. McGowan,
C. Mounaim-Rouselle, A. Ortiz-Prado, A. Ortiz-Valera, I. Rossetti, B. Shu, M. Yehia,
H. Xiao, M. Costa, Review on ammonia as a potential fuel: from synthesis to
economics, Energy & Fuels 35 (9) (2021) 6964–7029.
[9] M. Benés, G. Pozo, M. Abián, Á. Millera, R. Bilbao, M.U. Alzueta, Experimental
Study of the Pyrolysis of NH3 under Flow Reactor Conditions, Energy & Fuels 35
(9) (2021) 7193–7200.
[10] T. Sinigaglia, F. Lewiski, M.E. Santos Martins, J.C. Mairesse Siluk, Production,
storage, fuel stations of hydrogen and its utilization in automotive applications-a
review, International journal of hydrogen energy 42 (39) (2017) 24597–24611.
[11] A. Klerke, C.H. Christensen, J.K. Nørskov, T. Vegge, Ammonia for hydrogen
storage: challenges and opportunities, Journal of Materials Chemistry 18 (20)
(2008) 2304–2310.
[12] N. Morlanés, S.P. Katikaneni, S.N. Paglieri, A. Harale, B. Solami, S.M. Sarathy,
J. Gascon, A technological roadmap to the ammonia energy economy: Current state
and missing technologies, Chemical Engineering Journal 408 (2021), 127310.
J.L. Cerrillo et al.
[13] A. Valera-Medina, H. Xiao, M. Owen-Jones, W.I.F. David, P.J. Bowen, Ammonia for
power, Progress in Energy and combustion science 69 (2018) 63–102.
[14] M.A. Sutton, J.W. Erisman, F. Dentener, D. Möller, Ammonia in the environment:
From ancient times to the present, Environmental Pollution 156 (3) (2008)
583–604.
[15] R.E. Schofield, The enlightenment of Joseph Priestley: a study of his life and work
from 1733 to 1773, Penn State Press, 1997.
[16] V. Smil, Detonator of the population explosion, Nature 400 (6743) (1999) 415.
[17] J.W. Erisman, M.A. Sutton, J. Galloway, Z. Klimont, W. Winiwarter, How a century
of ammonia synthesis changed the world, Nature Geoscience 1 (10) (2008)
636–639.
[18] I. Matito-Martos, J. García-Reyes, A. Martin-Calvo, D. Dubbeldam, S. Calero,
Improving ammonia production using zeolites, The Journal of Physical Chemistry
C 123 (30) (2019) 18475–18481.
[19] S. Giddey, S.P.S. Badwal, C. Munnings, M. Dolan, Ammonia as a renewable energy
transportation media, ACS Sustainable Chemistry & Engineering 5 (11) (2017)
10231–10239.
[20] X. Wen, J. Guan, Recent advancement in the electrocatalytic synthesis of ammonia,
Nanoscale 12 (15) (2020) 8065–8094.
[21] M. Goff, Distribution of amonia as ana energy carrier, Ammonia Energy Coference
2020, Ammonia energy association (2020).
[22] K.E. Lamb, M.D. Dolan, D.F. Kennedy, Ammonia for hydrogen storage, A review of
catalytic ammonia decomposition and hydrogen separation and purification,
International Journal of Hydrogen Energy 44 (7) (2019) 3580–3593.
[23] N.M. Adli, H. Zhang, S. Mukherjee, G. Wu, ammonia oxidation electrocatalysis for
hydrogen generation and fuel cells, Journal of the Electrochemical Society 165
(15) (2018) J3130.
[24] C. Zamfirescu, I. Dincer, Using ammonia as a sustainable fuel, Journal of Power
Sources 185 (1) (2008) 459–465.
[25] N.J.M. Duijm, F.; Paulsen, J. L.;, Safety assessment of ammonia as a transport fuel -
Risoe-R No. 1504, Risø National Laboratory, Denmark, 2005.
[26] Z. Cesaro, M. Ives, R. Nayak-Luke, M. Mason, R. Bañares-Alcántara, Ammonia to
power: Forecasting the levelized cost of electricity from green ammonia in large-
scale power plants, Applied Energy 282 (2021), 116009.
[27] A. Afif, N. Radenahmad, Q. Cheok, S. Shams, J.H. Kim, A.K. Azad, Ammonia-fed
fuel cells: a comprehensive review, Renewable and Sustainable Energy Reviews 60
(2016) 822–835.
[28] J. Li, H. Huang, N. Kobayashi, Z. He, Y. Osaka, T. Zeng, Numerical study on effect
of oxygen content in combustion air on ammonia combustion, Energy 93 (2015)
2053–2068.
[29] I.P. Jain, Hydrogen the fuel for 21st century, International journal of hydrogen
energy 34 (17) (2009) 7368–7378.
[30] I. Staffell, D. Scamman, A. Velazquez Abad, P. Balcombe, P.E. Dodds, P. Ekins,
N. Shah, K.R. Ward, The role of hydrogen and fuel cells in the global energy
system, Energy & Environmental Science 12 (2) (2019) 463–491.
[31] L. Wang, Y. Zhao, C. Liu, W. Gong, H. Guo, Plasma driven ammonia decomposition
on a Fe-catalyst: eliminating surface nitrogen poisoning, Chemical communications
49 (36) (2013) 3787–3789.
[32] M. Akiyama, K. Aihara, T. Sawaguchi, M. Matsukata, M. Iwamoto, Ammonia
decomposition to clean hydrogen using non-thermal atmospheric-pressure plasma,
International Journal of Hydrogen Energy 43 (31) (2018) 14493–14497.
[33] H. Yuzawa, T. Mori, H. Itoh, H. Yoshida, Reaction mechanism of ammonia
decomposition to nitrogen and hydrogen over metal loaded titanium oxide
photocatalyst, The Journal of Physical Chemistry C 116 (6) (2012) 4126–4136.
[34] A. Utsunomiya, A. Okemoto, Y. Nishino, K. Kitagawa, H. Kobayashi, K. Taniya,
Y. Ichihashi, S. Nishiyama, Mechanistic study of reaction mechanism on ammonia
photodecomposition over Ni/TiO2 photocatalysts, Applied Catalysis B:
Environmental 206 (2017) 378–383.
[35] F. Chang, W. Gao, J. Guo, P. Chen, Emerging Materials and Methods toward
Ammonia-Based Energy Storage and Conversion, Adv. Mater. 33 (50) (2021)
2005721.
[36] C. Zhong, W.B. Hu, Y.F. Cheng, Recent advances in electrocatalysts for electro-
oxidation of ammonia, Journal of Materials Chemistry A 1 (10) (2013) 3216–3238.
[37] M. Feyen, C. Weidenthaler, R. Güttel, K. Schlichte, U. Holle, A.-H. Lu, F. Schüth,
High-Temperature Stable, Iron-Based Core-Shell Catalysts for Ammonia
Decomposition, Chemistry–A, European Journal 17 (2) (2011) 598–605.
[38] F. Schüth, R. Palkovits, R. Schlögl, D.S. Su, Ammonia as a possible element in an
energy infrastructure: catalysts for ammonia decomposition, Energy &
Environmental Science 5 (4) (2012) 6278–6289.
[39] A.K. Hill, L. Torrente-Murciano, Low temperature H2 production from ammonia
using ruthenium-based catalysts: Synergetic effect of promoter and support,
Applied Catalysis B: Environmental 172-173 (2015) 129–135.
[40] A. Di Carlo, L. Vecchione, Z. Del Prete, Ammonia decomposition over commercial
Ru/Al2O3 catalyst: An experimental evaluation at different operative pressures
and temperatures, international journal of hydrogen energy 39(2) (2014) 808-814.
[41] L. Li, Z. Zhu, G. Lu, Z. Yan, S. Qiao, Catalytic ammonia decomposition over CMK-3
supported Ru catalysts: Effects of surface treatments of supports, Carbon 45 (1)
(2007) 11–20.
[42] L. Li, Z. Zhu, Z. Yan, G. Lu, L. Rintoul, Catalytic ammonia decomposition over Ru/
carbon catalysts: The importance of the structure of carbon support, Applied
Catalysis A: General 320 (2007) 166–172.
[43] S.-F. Yin, B.-Q. Xu, C.-F. Ng, C.-T. Au, Nano Ru/CNTs: a highly active and stable
catalyst for the generation of COx-free hydrogen in ammonia decomposition,
Applied Catalysis B: Environmental 48 (4) (2004) 237–241.
[44] X. Ju, L. Liu, P. Yu, J. Guo, X. Zhang, T. He, G. Wu, P. Chen, Mesoporous Ru/MgO
prepared by a deposition-precipitation method as highly active catalyst for
11
Chemical Engineering Journal 431 (2022) 134310
producing COx-free hydrogen from ammonia decomposition, Applied Catalysis B:
Environmental 211 (2017) 167–175.
[45] X. Li, W. Ji, J. Zhao, S. Wang, C. Au, Ammonia decomposition over Ru and Ni
catalysts supported on fumed SiO2, MCM-41, and SBA-15, Journal of Catalysis 236
(2) (2005) 181–189.
[46] C. Huang, Y. Yu, J. Yang, Y. Yan, D. Wang, F. Hu, X. Wang, R. Zhang, G. Feng, Ru/
La2O3 catalyst for ammonia decomposition to hydrogen, Applied Surface Science
476 (2019) 928–936.
[47] I. Lucentini, A. Casanovas, J. Llorca, Catalytic ammonia decomposition for
hydrogen production on Ni, Ru and NiRu supported on CeO2, International Journal
of Hydrogen Energy 44 (25) (2019) 12693–12707.
[48] X. Ju, L. Liu, X. Zhang, J.i. Feng, T. He, P. Chen, Highly Efficient Ru/MgO Catalyst
with Surface-Enriched Basic Sites for Production of Hydrogen from Ammonia
Decomposition, ChemCatChem 11 (16) (2019) 4161–4170.
[49] W. Zheng, J. Zhang, B. Zhu, R. Blume, Y. Zhang, K. Schlichte, R. Schlögl, F. Schüth,
D.S. Su, Structure–function correlations for Ru/CNT in the catalytic decomposition
of ammonia, ChemSusChem: Chemistry & Sustainability Energy & Materials 3 (2)
(2010) 226–230.
[50] K. Nagaoka, T. Eboshi, N. Abe, S.-I. Miyahara, K. Honda, K. Sato, Influence of basic
dopants on the activity of Ru/Pr6O11 for hydrogen production by ammonia
decomposition, International journal of hydrogen energy 39 (35) (2014)
20731–20735.
[51] X. Duan, J. Ji, G. Qian, C. Fan, Y. Zhu, X. Zhou, D.e. Chen, W. Yuan, Ammonia
decomposition on Fe (1 1 0), Co (1 1 1) and Ni (1 1 1) surfaces: A density functional
theory study, Journal of Molecular Catalysis A: Chemical 357 (2012) 81–86.
[52] J.P. Collins, J.D. Way, Catalytic decomposition of ammonia in a membrane reactor,
Journal of membrane science 96 (3) (1994) 259–274.
[53] F.R. García-García, Y.H. Ma, I. Rodríguez-Ramos, A. Guerrero-Ruiz, High purity
hydrogen production by low temperature catalytic ammonia decomposition in a
multifunctional membrane reactor, Catalysis Communications 9 (3) (2008)
482–486.
[54] G. Lanzani, K. Laasonen, NH3 adsorption and dissociation on a nanosized iron
cluster, International journal of hydrogen energy 35 (13) (2010) 6571–6577.
[55] H. Muroyama, C. Saburi, T. Matsui, K. Eguchi, Ammonia decomposition over Ni/
La2O3 catalyst for on-site generation of hydrogen, Applied Catalysis A: General
443-444 (2012) 119–124.
[56] C. Plana, S. Armenise, A. Monzón, E. García-Bordejé, Ni on alumina-coated
cordierite monoliths for in situ generation of CO-free H2 from ammonia, Journal of
Catalysis 275 (2) (2010) 228–235.
[57] F. García-García, A. Guerrero-Ruiz, I. Rodríguez-Ramos, A. Goguet, S. Shekhtman,
C. Hardacre, TAP studies of ammonia decomposition over Ru and Ir catalysts,
Physical Chemistry Chemical Physics 13 (28) (2011) 12892–12899.
[58] J. Polanski, P. Bartczak, W. Ambrozkiewicz, R. Sitko, T. Siudyga, A. Mianowski,
J. Szade, K. Balin, J. Lelątko, S. Bhargava, Ni-supported Pd nanoparticles with Ca
promoter: a new catalyst for low-temperature ammonia cracking, PloS one 10 (8)
(2015) e0136805.
[59] D.A. Hansgen, D.G. Vlachos, J.G. Chen, Using first principles to predict bimetallic
catalysts for the ammonia decomposition reaction, Nature chemistry 2 (6) (2010)
484–489.
[60] Z.-W. Wu, X. Li, Y.-H. Qin, L. Deng, C.-W. Wang, X. Jiang, Ammonia decomposition
over SiO2-supported Ni–Co bimetallic catalyst for COx-free hydrogen generation,
International Journal of Hydrogen Energy 45 (30) (2020) 15263–15269.
[61] A. Srifa, K. Okura, T. Okanishi, H. Muroyama, T. Matsui, K. Eguchi, Hydrogen
production by ammonia decomposition over Cs-modified Co3Mo3N catalysts,
Applied Catalysis B: Environmental 218 (2017) 1–8.
[62] Ö. Akarçay, S.F. Kurtoğlu, A. Uzun, Ammonia decomposition on a highly-dispersed
carbon-embedded iron catalyst derived from Fe-BTC: Stable and high performance
at relatively low temperatures, International Journal of Hydrogen Energy 45 (53)
(2020) 28664–28681.
[63] X. Duan, G. Qian, X. Zhou, Z. Sui, D. Chen, W. Yuan, Tuning the size and shape of
Fe nanoparticles on carbon nanofibers for catalytic ammonia decomposition,
Applied Catalysis B: Environmental 101 (3–4) (2011) 189–196.
[64] V. Cechetto, L. Di Felice, J.A. Medrano, C. Makhloufi, J. Zuniga, F. Gallucci, H2
production via ammonia decomposition in a catalytic membrane reactor, Fuel
Processing Technology 216 (2021), 106772.
[65] N. Sazali, W.N.W. Salleh, A.F. Ismail, Carbon tubular membranes from
nanocrystalline cellulose blended with P84 co-polyimide for H2 and He separation,
International Journal of Hydrogen Energy 42 (15) (2017) 9952–9957.
[66] C. Zhao, B. Sun, J. Jiang, W. Xu, H2 purification process with double layer bcc-
PdCu alloy membrane at ambient temperature, International Journal of Hydrogen
Energy 45 (35) (2020) 17540–17547.
[67] Z. Zhang, S. Liguori, T.F. Fuerst, J.D. Way, C.A. Wolden, Efficient ammonia
decomposition in a catalytic membrane reactor to enable hydrogen storage and
utilization, ACS Sustainable Chemistry & Engineering 7 (6) (2019) 5975–5985.
[68] N.W. Ockwig, T.M. Nenoff, Membranes for hydrogen separation, Chemical reviews
107 (10) (2007) 4078–4110.
[69] A. Di Carlo, A. Dell’Era, Z. Del Prete, 3D simulation of hydrogen production by
ammonia decomposition in a catalytic membrane reactor, International Journal of
Hydrogen Energy 36 (18) (2011) 11815–11824.
[70] S.H. Israni, B.K.R. Nair, M.P. Harold, Hydrogen generation and purification in a
composite Pd hollow fiber membrane reactor: Experiments and modeling, Catalysis
Today 139 (4) (2009) 299–311.
[71] N.S. Patki, S.-T. Lundin, J.D. Way, Apparent activation energy for hydrogen
permeation and its relation to the composition of homogeneous PdAu alloy thin-
film membranes, Sep. Purif. Technol. 191 (2018) 370–374.
J.L. Cerrillo et al.
[72] L. Shi, A. Goldbach, G. Zeng, H. Xu, Preparation and performance of thin-layered
PdAu/ceramic composite membranes, Int. J. Hydrogen Energy 35 (9) (2010)
4201–4208.
[73] N. Morlanés, S. Sayas, G. Shterk, S.P. Katikaneni, A. Harale, B. Solami, J. Gascon,
Development of a Ba–CoCe catalyst for the efficient and stable decomposition of
ammonia, Catal. Sci. Technol. 11 (9) (2021) 3014–3024.
[74] J.B. Von Colbe, J.-R. Ares, J. Barale, M. Baricco, C. Buckley, G. Capurso, N.
Gallandat, D.M. Grant, M.N. Guzik, I. Jacob, Application of hydrides in hydrogen
storage and compression: Achievements, outlook and perspectives, international
journal of hydrogen energy 44(15) (2019) 7780-7808.
[75] C. Deng, M. Zhu, Y. Zhou, X. Feng, Novel conceptual methodology for hydrogen
network design with minimum compression work, Energy 159 (2018) 203–215.
[76] G. Sdanghi, G. Maranzana, A. Celzard, V. Fierro, Review of the current
technologies and performances of hydrogen compression for stationary and
automotive applications, Renewable and Sustainable Energy Reviews 102 (2019)
150–170.
[77] J.B. Von Colbe, J.-R. Ares, J. Barale, M. Baricco, C. Buckley, G. Capurso, N.
Gallandat, D.M. Grant, M.N. Guzik, I. Jacob, Application of hydrides in hydrogen
storage and compression: Achievements, outlook and perspectives, international
journal of hydrogen energy 44(15) (2019) 7780-7808.
[78] S. Armenise, E. García-Bordejé, J. Valverde, E. Romeo, A. Monzón, A Langmuir-
Hinshelwood approach to the kinetic modelling of catalytic ammonia
decomposition in an integral reactor, Physical Chemistry Chemical Physics 15 (29)
(2013) 12104–12117.
[79] G. Djéga-Mariadassou, C.-H. Shin, G. Bugli, Tamaru’s model for ammonia
decomposition over titanium oxynitride, Journal of Molecular Catalysis A:
Chemical 141 (1–3) (1999) 263–267.
[80] S. Sayas, N. Morlanés, S.P. Katikaneni, A. Harale, B. Solami, J. Gascon, High
pressure ammonia decomposition on Ru–K/CaO catalysts, Catalysis Science &
Technology 10 (15) (2020) 5027–5035.
[81] G. Li, M. Kanezashi, H.R. Lee, M. Maeda, T. Yoshioka, T. Tsuru, Preparation of a
novel bimodal catalytic membrane reactor and its application to ammonia
decomposition for COx-free hydrogen production, International journal of
hydrogen energy 37 (17) (2012) 12105–12113.
12
Chemical Engineering Journal 431 (2022) 134310
[82] M.Á. Gómez-García, I. Dobrosz-Gómez, J. Fontalvo, J.M. Rynkowski, Membrane
reactor design guidelines for ammonia decomposition, Catalysis Today 191 (1)
(2012) 165–168.
[83] L. Lin, Y. Tian, W. Su, Y. Luo, C. Chen, L. Jiang, Techno-economic analysis and
comprehensive optimization of an on-site hydrogen refuelling station system using
ammonia: hybrid hydrogen purification with both high H 2 purity and high
recovery, Sustainable Energy & Fuels 4 (6) (2020) 3006–3017.
[84] J.L. Nieminen, Experimental investigation of hydrogen production from ammonia
decomposition, 2011.
[85] J.H. Kim, O.C. Kwon, A micro reforming system integrated with a heat-
recirculating micro-combustor to produce hydrogen from ammonia, International
journal of hydrogen energy 36 (3) (2011) 1974–1983.
[86] K.E. Lamb, D.M. Viano, M.J. Langley, S.S. Hla, M.D. Dolan, High-purity H2
produced from NH3 via a ruthenium-based decomposition catalyst and vanadium-
based membrane, Industrial & Engineering Chemistry Research 57 (23) (2018)
7811–7816.
[87] Air-Liquid, Pressure Swing Adsorption (PSA) - Hydrogen Purification, 2021. htt
ps://www.engineering-airliquide.com/pressure-swing-adsorption-psa-hydr
ogen-purification. (Accessed 08/2021).
[88] M.T. Ho, G.W. Allinson, D.E. Wiley, Reducing the cost of CO2 capture from flue
gases using pressure swing adsorption, Industrial & Engineering Chemistry
Research 47 (14) (2008) 4883–4890.
[89] Umicore, Ammonia Slip Catalyst (ASC). https://ac.umicore.com/en/technolo
gies/ammonia-slip-catalyst/. (Accessed 08-2021 2021).
[90] G.F.B. Brendel, J E; Rathbone, R F; Frey, Jr, R N, INVESTIGATION OF AMMONIA
ADSORPTION ON FLY ASH DUE TO INSTALLATION OF SELECTIVE CATALYTIC
REDUCTION SYSTEMS, 2000.
[91] Z. Du, C. Liu, J. Zhai, X. Guo, Y. Xiong, W. Su, G. He, A Review of Hydrogen
Purification Technologies for Fuel Cell Vehicles, Catalysts 11 (3) (2021) 393.
[92] S. Adhikari, S. Fernando, Hydrogen membrane separation techniques, Industrial &
Engineering Chemistry Research 45 (3) (2006) 875–881.
[93] M. Wang, S. Liu, T. Qian, J. Liu, J. Zhou, H. Ji, J. Xiong, J. Zhong, C. Yan, Over
56.55% Faradaic efficiency of ambient ammonia synthesis enabled by positively
shifting the reaction potential, Nature, Communications 10 (1) (2019) 341.