membranes

Article
Hydrogen Separation and Purification from Various Gas
Mixtures by Means of Electrochemical Membrane Technology
in the Temperature Range 100–160 ◦ C
Leandri Vermaak 1, * , Hein W. J. P. Neomagus 2 and Dmitri G. Bessarabov 1, *

1 HySA Infrastructure Centre of Competence, Faculty of Engineering, Potchefstroom Campus, North-West

University, Potchefstroom 2520, South Africa
2 Centre of Excellence in Carbon Based Fuels, Faculty of Engineering, School of Chemical and Minerals
Engineering, Potchefstroom Campus, North-West University, Potchefstroom 2520, South Africa;
[email protected]
* Correspondence: [email protected] (L.V.); [email protected] (D.G.B.)

Abstract: This paper reports on an experimental evaluation of the hydrogen separation performance
in a proton exchange membrane system with Pt-Co/C as the anode electrocatalyst. The recovery of
hydrogen from H2 /CO2 , H2 /CH4 , and H2 /NH3 gas mixtures were determined in the temperature
range of 100–160 ◦ C. The effects of both the impurity concentration and cell temperature on the
separation performance of the cell and membrane were further examined. The electrochemical
properties and performance of the cell were determined by means of polarization curves, limiting



current density, open-circuit voltage, hydrogen permeability, hydrogen selectivity, hydrogen purity,


and cell efficiencies (current, voltage, and power efficiencies) as performance parameters. High purity
Citation: Vermaak, L.; Neomagus, hydrogen (>99.9%) was obtained from a low purity feed (20% H2 ) after hydrogen was separated from
H.W.J.P.; Bessarabov, D.G. Hydrogen H2 /CH4 mixtures. Hydrogen purities of 98–99.5% and 96–99.5% were achieved for 10% and 50%
Separation and Purification from CO2 in the feed, respectively. Moreover, the use of proton exchange membranes for electrochemical
Various Gas Mixtures by Means of
hydrogen separation was unsuccessful in separating hydrogen-rich streams containing NH3 ; the
Electrochemical Membrane
membrane underwent irreversible damage.
Technology in the Temperature Range
100–160 ◦ C. Membranes 2021, 11, 282.
Keywords: electrochemical hydrogen separation/purification; high-temperature proton exchange
https://doi.org/10.3390/membranes
11040282
membrane (PEM); polarization curve

Academic Editor: Metin Uz

Received: 16 March 2021 1. Introduction

Accepted: 1 April 2021 Hydrogen shows great potential regarding future energy prospects, especially in
Published: 10 April 2021 the potential low-carbon energy system known as ‘the hydrogen economy’. Some of the
benefits that promote hydrogen as an energy carrier include the following. Hydrogen
Publisher’s Note: MDPI stays neutral is the most abundant element in the universe; it is the lightest element with the highest
with regard to jurisdictional claims in energy content among all the existing fuels [1,2]; it is a highly efficient and clean energy
published maps and institutional affil- carrier [3], with only water as a by-product at conversion to energy [2,3]; and, furthermore,
iations.
hydrogen lends itself to a variety of storage methods (e.g., gaseous, liquid [2,4]). Hydrogen
could increase renewable energy usage, to a large extent, due to its ability to allow flexible
storage of intermittent renewable energy [5]. For this reason, hydrogen is seen as a key
solution to the global climate change [2], with the global ambition to maintain the increase
Copyright: © 2021 by the authors. in the average global temperature to <2 ◦ C above pre-industrial levels, as stated by the
Licensee MDPI, Basel, Switzerland. Paris Agreement of 2015 [6].
This article is an open access article Among other technologies that use hydrogen, specifically for mobility applications,
distributed under the terms and proton exchange membrane (PEM) fuel cells (FCs) are attracting the most attention due to
conditions of the Creative Commons
their favorable operational characteristics, such as quiet operation, near-zero emissions,
Attribution (CC BY) license (https://
high power densities, high conversion efficiencies, etc. [7]. However, for effective operation,
creativecommons.org/licenses/by/
high-purity hydrogen (>99.97%) is required [8].
4.0/).

Membranes 2021, 11, 282. https://doi.org/10.3390/membranes11040282 https://www.mdpi.com/journal/membranes

Membranes 2021, 11, 282 2 of 24

Currently, large-scale hydrogen production is commonly carried out by means of

steam methane reforming, coal gasification, or partial oxidation of hydrocarbons [9–11].
Although not widely implemented, there are several small-scale hydrogen production
methods, including partial oxidation of biomass, electrolysis and auto-thermal processes,
and small-scale steam methanol reformers [9,12]. However, both large- and small-scale
hydrogen production methods produce hydrogen-rich streams commonly accompanied
by impurities such CO2 , CO, N2 , and CH4 , among others [13–17] (e.g., steam methane
reforming (CO, CO2 , CH4 ), dry shifted syngas (N2 , CO, CO2 , CH4 , H2 S), hot shifted syngas
(N2 , CO, H2 S) [18,19]).
There are now several technological approaches being used to extract and purify
hydrogen from gas mixtures, utilizing various characteristics of hydrogen under different
industrial conditions. Common approaches for hydrogen recovery include adsorbing
the impurities (pressure swing adsorption (PSA)), condensing the impurities (cryogenic
distillation) or by using hydrogen-selective membranes [13,16,20,21]. See Table 1.

Table 1. Properties of different hydrogen purification processes [22–24].

Properties PSA Membranes Cryogenic

Min. feed purity (vol.%) >40 >25 15–80
Max. product purity (vol.%) >99.9 >98 ~97
Max. hydrogen recovery (%) Up to 90 Up to 95 Up to 98
Inlet pressure (bar) 10–70 14–138 14–83
Outlet pressure (bar) Similar to feed Substantially less than feed Similar to feed

Although PSA and cryogenic distillation technologies are both commercial processes,
multiple units are required and, in some instances, supplementary wash columns are
required to remove CO and CO2 [13]. Pressure-driven membrane processes are considered
to be better candidates for hydrogen production because they are not as energy intensive
and they yield high-purity hydrogen [25]. However, even though membrane technologies
are advantageous over the other purification methods mentioned, they commonly depend
on high-pressure feed streams, and hydrogen embrittlement is often experienced [13].
As an alternative to pressure-driven membranes, electrochemical membrane tech-
nology based on PEMs has been utilized for hydrogen separation and it has been used
to effectively recover hydrogen from H2 /N2 , H2 /CO2 /CO, and H2 /CH4 hydrogen-rich
gas mixtures [26]. Electrochemical hydrogen separation (EHS) offers several advantages
over its competitors [27,28], including high-purity hydrogen from one-step operation [29],
simultaneous hydrogen compression and separation is possible, as well as hydrogen sepa-
ration is achieved at low cell voltages with a high separation efficiency [30]. An additional
benefit of this electrochemical method is that CO2 , if present in the feed, is concentrated,
and it can be captured and stored without the requirement for any further treatment, thus,
reducing greenhouse gas emissions [26,30].
Perfluorinated sulfonic acid (PFSA) membranes (mainly Nafion-based), typically
operated at 60–80 ◦ C, are mostly used as PEMs due to their high proton conductivity,
mechanical integrity, and chemical stability [13]. However, one major challenge of Nafion
membranes include the complex water management of the membrane [13]. Furthermore,
swelling-induced stresses could result from water adsorption of the membrane [31]. An-
other major disadvantage is deactivation of the electrocatalyst (mainly based on Pt) when
small amounts of CO (10–1000 ppm) are present [13,14], which results in a significant
performance drop of the electrochemical cell [14].
High-temperature (HT)-PEMs (operating temperatures: 100–200 ◦ C) doped with
phosphoric acid (PA) have been developed to overcome the limitations associated with
low-temperature (LT) PSFA membranes [32–34]. HT PA-doped PEMs conduct protons
well, even in anhydrous conditions, which eliminates the requirement of pre-humidifying
the feed stream [35]. Some other advantages include their fast electrode kinetics, enhanced
mass transport, higher tolerance to impurities such as CO (up to 3 vol.%), and better sustain-
Membranes 2021, 11, 282 3 of 24

ability [13,36–38]. Recent advances in other HT membranes, specifically for fuel cells, have
been made [39–41]. These include, for example, phosphonated poly(pentafluorostyrene)
for fuel cells and fuel cells based on quaternary ammonium-biphosphate ion pairs. A
comparison of the properties of typical HT and LT PEMs can be found in Reference [32]. In
terms of EHS, limited information is currently available on HT PEMs separation compared
to LT separation (Nafion membranes) [42].
In the present work, hydrogen was separated from various concentrations of H2 /CH4 ,
H2 /CO2 and H2 /NH3 mixtures at higher temperatures of 120–160 ◦ C. The gas mixtures
were chosen based on their industrial relevance in processes related to syngas, the water–
gas shift reaction, the power-to-gas (P2G) concept, and ammonia production. A HT-PEM
(TPS® -based) was selected as the commercially available membrane choice due to the
higher operating temperatures of these membranes compared to the PBI membranes of
the same supplier (Advent technologies, Inc., Cambridge, MA, USA). To understand, and
evaluate, the characteristics of the membrane, polarization curves and electrochemical
impedance spectroscopy (EIS) were used. An in-line gas chromatograph (GC) was used
to evaluate the separation performance of the membrane. This was done by measuring
the composition of the permeate stream and comparing it with the composition of the feed
stream, using the hydrogen selectivity parameter.

2. Working Principle
An electrochemical hydrogen separator (EHS), illustrated in Figure 1, is principally
based on the following mechanism: a molecular gaseous hydrogen-containing gas stream
is fed, at low pressure, at the anode side of the PEM electrochemical cell. When the
hydrogen makes contact with the anode electrode, hydrogen is oxidized into protons and
electrons, facilitated by a Pt-based catalyst. Hydrogen is then selectively transported, in
the form of protons, through the PEM to the cathode compartment and the electrons then
complete an electrical circuit. Finally, the protons and electrons are combined at the cathode
compartment to form gaseous hydrogen. This reaction can be carried out at high pressure
to improve the specific volumetric energy density of hydrogen, for storage purposes.

Figure 1. Working principal of an ideal electrochemical hydrogen separator [42,43].

Membranes 2021, 11, 282 4 of 24

The hydrogen oxidation reaction does not occur spontaneously and an external load
(DC power source) is required to drive the electrochemical reaction (electrolytic mode) [44].
However, minimal power is required to operate the cell, as the oxidation and reduction
reactions of hydrogen are facile and nearly Nernstian/Faradic in their electrochemical
behavior. Unlike conventional membrane purification systems that rely on pressure or
concentration differentials to bring about separation, an EHS generates hydrogen at a rate
dependent on the applied current, I (A). This is defined by Faraday’s law:

. MI
m= (1)
nF

where ṁ depicts the hydrogen flow rate, normally in g s−1 or g min−1 , n is the number of
electrons, F depicts the Faraday constant (96,485 C mol−1 ) and M is the molecular weight
(g mol−1 ).

3. Materials and Methods

3.1. Experimental Set-Up: Cell and Membrane Electrode Assembly
Experiments were carried out using an off-the-shelf electrochemical cell with a square
active area of 25 cm2 (Fuel Cell Technologies, Inc., Albuquerque, NM, USA). See Figure 2.
The cell consisted of a catalyst coated proton-conducting membrane, which was placed
between two pieces of carbon cloth, acting as gas diffusion layers (GDLs), and Ti flow
fields. The cell was sealed with HT PFA gaskets (used to compress the membrane electrode
assembly (MEA)). The cell was assembled with aluminium back plates and compressed
using a torque wrench to ensure a uniform pressure distribution over the compressed MEA
and to minimize the electrochemical and thermal resistance of the GDLs. The assembly
torque for the 8 bolts was 6 Nm. The MEA consisted of carbon supported platinum (Pt/C)
as the cathode catalyst and a bimetallic carbon supported platinum-cobalt (Pt-Co/C) as the
anode catalyst, and a TPS® -based PEM (supplied by Advent Technologies, Inc., Cambridge,
MA, USA). The total catalyst loading was 1.8 mg cm−2 (specified by Advent Technologies,
Inc., Cambridge, MA, USA).

Figure 2. Cross section of the electrochemical cell.

Membranes 2021, 11, 282 5 of 24

The properties of the TPS-based membranes used in the experiments can be found in
Table 2.

Table 2. Properties of the TPS® -based membrane (as received from the supplier).

Membrane Type TPS® -Based

Membrane thickness 60–65 µm
Catalyst used anode Pt-Co/C
Atomic ratio 1:1 (Pt:Co)
Catalyst used cathode Pt/C
Total catalyst loading 1.8 mg cm−2
Membrane area 25 cm2
Temperature range 120–200 ◦ C
Proton conductivity 8 × 10−2 S cm−1

An external DC power supply was connected to the bus plates of the electrochemical
cell and the system was operated in constant current mode. The current and voltage were
recorded with an electronic load, attached to a computer, making use of LabVIEW software.
The cell temperature was controlled by two FIREROD® (Watlow, St. Louis, MO, USA)
heating rods, one at the anode side and one at the cathode side, with incorporation of a
K-type thermocouple (anode side) and PID controller (DCL-33A). Furthermore, the cell
was enclosed in a heat insulation jacket to reduce heat losses. The temperature at the
centre of the anode back plate was taken as the cell temperature. The permeate volumetric
flow rate was measured using a soap flow meter and the composition of the permeate
was determined with an in-line gas chromatograph (GC) (SRI 8610C-GC; SRI Instruments,
Torrance, CA, USA). See GC and bubble flow meter set-up in Figure 3.

Figure 3. Experimental set-up used for electrochemical hydrogen separation.

3.2. Characterization Methods Used

3.2.1. Polarization Curve Measurements
The hydrogen flow rate was kept constant at 100 mLn /min throughout all the experi-
ments (as described below). Polarization curves were drawn, starting at 0 A cm− 2 , and
then increasing the current density with a step change of 0.02 A cm− 2 every 10 min. The
voltage output for each current density setting was recorded every 5 Sections (performed
in constant current mode). This was done until a limiting current was reached or until the
Membranes 2021, 11, 282 6 of 24

voltage limit (1.1 V) was reached, after which the run was terminated. The volumetric flow
rate of the permeate was recorded for all current density values, making use of a soap flow
meter, in order to determine the current efficiency.

3.2.2. Electrochemical Impedance Spectroscopy Measurements

To determine the resistance of the MEA, in situ EIS was performed in constant cur-
rent/galvanostatic mode (1 A), frequency range 100 kHz to 0.1 Hz, at 120–160 ◦ C in
hydrogen pumping mode. For feed streams containing impurities, the hydrogen flow rate
was kept constant at 100 mLn /min H2 . The AC current amplitude was a standard 10% of
the applied direct current [14]. A Nyquist plot was generated of the impedance spectrum.

3.3. Experimental Procedure

All experiments were carried out at atmospheric pressure (~86 kPa), where the cell
operating temperature was varied between 120 and 160 ◦ C. All experiments were con-
ducted in hysteretic order as per assessment of the performance deterioration during
experimentation. All the flow rates herein reported are given in normal millilitres per
minute (mLn /min), with 0 ◦ C and 1 bar taken as reference.
Preliminary tests performed with pure hydrogen showed that the thermal cycles
(heating and cooling of the cell), which lead to the expansion and contraction of the cell
components, could cause the assembly bolts to loosen. Therefore, once a thermal cycle is
complete and the cell is cooled down, the assembly torque on the bolts was reinsured. This
procedure was applied after each thermal cycle.
The experimental tests were organized as follow. First, as a basic characterization step,
the cell was operated using pure hydrogen. The effect of temperature on the performance
of the cell was investigated by varying the temperature between 120 and 160 ◦ C, in 20 ◦ C
steps. Polarization curves and EIS graphs were drawn for data at each temperature. After
the pure hydrogen tests, the separation experiments were performed. A constant hydrogen
flow rate of 100 mLn /min was maintained throughout. The gas mixtures were tested in
the following order: H2 /CH4 , H2 /CO2 , and then the H2 /NH3 gas mixture, starting at the
lowest impurity mixtures and moving towards the higher impurity concentrations.

4. Results and Discussion

4.1. Pure Hydrogen Experiments: Membrane Characterization
The performance of the MEA was first investigated using a pure hydrogen feed at
operating temperatures between 120 and 160 ◦ C and a hydrogen flow rate of 100 mLn /min.
The resulting voltage–current characteristics are shown in Figure 4. It is evident that the cell
performance is sensitive to small changes in temperature (20 ◦ C). A clear increase in the cell
performance was observed as the voltage values decrease, with an increase in the applied
current density. This results in higher overall power efficiencies. Moreover, the limiting
current density values (vertical portion of the curves) also increased with temperature. The
highest achievable current density was found to be 0.40 A cm−2 at 160 ◦ C.
The linear nature of the initial portion of the curves indicates that the ohmic losses
(IR-losses) dominate the voltage losses in this region. This behavior is common in electro-
chemical cells and suggests that the contact resistance and the membrane resistance play a
crucial role in the general performance of the cell [15,28,45]. It can therefore be concluded
that the cell performance can be enhanced by minimizing both the contact and membrane
resistance. In addition, the curved nature of the final portion of the curves indicates that
the mass transfer limitations dominate in the high current density region. This is explained
by the depletion of reactants at the surface of the electrode.
Membranes 2021, 11, 282 7 of 24

Figure 4. Polarization curve for pure hydrogen (100 mLn /min, 1 atm).

In an ideal cell, the hydrogen permeation flux is given as a function of current, accord-
ing to Faraday’s law (see Equation (1)), which is derived by assuming that the hydrogen
is an ideal gas. However, the real value is always smaller than that depicted by an ideal
process, due to losses in the system. In this particular case, the losses can be determined by
the current efficiency, which is defined as the ratio of the actual hydrogen permeation rate
and the theoretical hydrogen permeation rate:
.
m H ( Experimental )
εi = . 2 (2)
m H2 ( Theoretical )
. .
where (m H2 ( Experimental )) depicts actual hydrogen flow from the cathode and m H2 ( Theoretical )
depicts the theoretical hydrogen flow rate, determined from Faraday’s law. A visual
correlation between the theoretical and actual hydrogen permeation rates is plotted in
Figure 5. The hydrogen permeability is also shown on the same graph (left y-axis). In
general, the measured data show good correlation with the theoretical values.

Figure 5. Theoretical and measured hydrogen permeate flow rates and hydrogen permeability
(hydrogen flow rate = 100 mLn /min, 1 atm).
Membranes 2021, 11, 282 8 of 24

The calculated current efficiencies, together with the voltage and power efficiencies
are illustrated in Figure 6. The voltage efficiency is defined in terms of the thermal energy
of the hydrogen generated (Equation (3)). This is calculated with the voltage equivalent of
the heat of combustion of hydrogen (1.484 V), whereas the power efficiency can be defined
as the product of both the voltage and current efficiencies (Equation (4)).

V
εv = 1 − (3)
1.484
ε p = εi × εv (4)
At low current densities (<0.08 A cm−2 ) the current efficiencies were relatively low
(~80%) for all temperatures considered here. However, as the current density was increased,
and therefore also the hydrogen flux, the current efficiency increased to >95%. Due to the
high current efficiencies observed at the higher current densities, it was concluded that the
lower efficiencies at lower current densities were caused by back diffusion of hydrogen.
This can be explained by the low hydrogen fluxes at low current densities. Overall, it can be
concluded that the current efficiency increased with current density at constant temperature,
whilst temperature seems to play a negligible role in the current efficiency. Also, the voltage
efficiency increased with an increase in temperature and decreased with an increase in
current density. Lastly, higher power efficiencies were reached when the temperature was
increased. The maximum power efficiencies at the respective temperatures were ~91%
(120 ◦ C), ~93% (140 ◦ C) and ~95% (160 ◦ C) at 0.16 A cm−2 . Therefore, 0.16 A cm−2 was
considered the optimum operating current density with regards to the power consumption
at 120–160 ◦ C.
In order to obtain a better understanding of the temperature effect on the performance
of the EHS, EIS was used to diagnose the cell. Figure 7 shows the in situ AC impedance
spectrum of the EHS operated at 0.1 A cm−2 with variations in the temperature and the
hydrogen flow rate. Similar spectra have been widely reported in the literature for both
FCs [14,46] and EHSs [47,48]. The intercept of the real axis (x-axis) represents the ohmic
resistance, which is defined as the sum of the contact resistance, electrical resistance and the
proton conduction resistance in the membrane). Figure 7 shows that the impedance spectra
move towards the left as the temperature is increased. The leftwards movement indicates
that the ohmic resistance of the cell decreases with temperature. This is in agreement
with what is reported in the literature, as the membrane can be considered as an electric
insulator, which means that theoretically its resistance decreases with temperature [49].
The diameter of the impedance arc is the sum of the charge transfer and mass transfer
resistance. The reduction in the arc diameter at higher temperatures indicates that there is a
slight improvement in mass transport and high frequency charge transfer [49]. According to
Su et al. [50], an increase in temperature enhances the cell performance due to the increase
in the mass transfer rate, and a lower cell resistance, as confirmed in this work. Furthermore,
the overall performance enhancement with increasing temperatures (polarization curves)
can be attributed to the decrease in ohmic resistance and the improvement in mass transport
and high frequency charge transfer observed in the EIS data [49].
Membranes 2021, 11, 282 9 of 24

Figure 6. Current-(εi ), voltage-(εv ), and power (εp ) efficiencies as a function of current density at (a)
120 ◦ C, (b) 140 ◦ C and (c) 160 ◦ C (hydrogen flow rate = 100 mLn /min, 1 atm).

Figure 7. Electrochemical impedance spectroscopy (EIS) data showing the change in ohmic resistance
(x-intercept) with temperature. Hydrogen flow rate = 100 mLn /min, 1 atm.
Membranes 2021, 11, 282 10 of 24

4.2. Hydrogen Separation from H2 /CH4 Gas Mixtures

The EHS was evaluated for the electrochemical hydrogen separation from H2 /CH4
gas mixtures (10% CH4 , 50% CH4 , and 80% CH4 , with the balance hydrogen), where the
hydrogen flow rate was kept constant at 100 mLn /min. The voltage–current characteristics
of the H2 /CH4 mixtures at different cell temperatures and feed concentrations are shown
in Figure 8 and were compared to that of pure hydrogen (refer to Figure 4).

Figure 8. Polarization curves for (a) 10%, (b) 50%, and (c) 80% methane, balance hydrogen (hydrogen
flow rate = 100 mLn /min, 1 atm).

Generally, higher voltages were reached with the H2 /CH4 gas mixtures, compared
to that of pure hydrogen. Here it is noteworthy that the open-current voltage (OCV),
determined by the Nernst equation, is not zero, as is the case with pure hydrogen. This is
due to the differences in partial pressures [13]. The OCVs were calculated for the different
feed compositions and temperatures at a fixed current density of 0.1 A cm−2 . See Table 3.
These values were compared to the cell voltages corresponding to pure hydrogen at the
same current density: 0.071 V (120 ◦ C), 0.042 V (140 ◦ C) and 0.021 (160 ◦ C), respectively.
Based on Kim et al. [47], the actual operating voltage for the gas separation can be estimated
by subtracting the OCV from the measured cell voltage. However, from Table 3, it is
evident that these values are still much higher than the cell voltages corresponding to
pure hydrogen.
Membranes 2021, 11, 282 11 of 24

Table 3. Hydrogen purity and cell voltage values as a function of temperature for the H2 /CH4 feed
mixtures, measured at 0.1 A cm−2 .

Hydrogen Purity (%) Measured Measured-OCV

Temperature OCV (V)
Inlet Permeate Voltage (V) (V)
20 99.98 0.027 0.564 0.537
120 ◦ C 50 99.98 0.012 0.284 0.272
90 99.99 0.002 0.077 0.075
20 99.98 0.029 0.352 0.323
140 ◦ C 50 99.98 0.012 0.219 0.207
90 99.99 0.002 0.075 0.074
20 99.86 0.030 0.253 0.233
160 ◦ C 50 99.95 0.013 0.083 0.070
90 99.99 0.002 0.074 0.072

General conclusions can be drawn from the performance curves (Figure 8). Firstly, at
constant inlet composition, the general performance of the cell increased with temperature.
This is explained by the enhanced reaction activity (mass transfer rate) of the electrodes
at higher temperatures. Similar results have been reported by Lee et al. [15]. However,
at constant temperature, an increase in methane concentration leads to a decrease in the
performance of the cell. This performance decrease shows that the partial pressure of the
hydrogen influences the general performance of the cell. Secondly, the limiting current
density values, indicating where most of the hydrogen in the H2 /CH4 mixtures has been
consumed and where the hydrogen partial pressure at the anode becomes small, were seen
to increase with temperature and decrease with methane concentration.
The current, voltage, and power efficiencies of the H2 /CH4 gas mixtures were also
considered and compared to those of pure hydrogen. The experimental results obtained for
10%, 50%, and 80% CH4 are illustrated in Figure 9. In general, the initial current efficiencies
were significantly lower than that obtained with pure hydrogen. The current efficiencies
for H2 /CH4 were in the range ~47–51% at 0.02 A cm−2 , whereas those for pure hydrogen
were ~70%. The current efficiencies of the H2 /CH4 mixtures also increased with current
density, as was the case with pure hydrogen. The voltage efficiencies, on the other hand,
decreased with an increase in current density, which correlates well with the results of
pure hydrogen experiments. Moreover, the voltage efficiencies seemingly increased with
temperature, observed with the 50 and 80% CH4 inlet mixtures. However, in the case of
the 10% CH4 , this trend was only visible at current densities > 0.16 A cm−2 .
The power efficiencies had a maximum value at 0.12 A cm−2 with 10% CH4 , 0.08–0.1 A
cm−2 with 50% CH4 and 0.04–0.08 A cm−2 with 80% CH4 . This depicts the optimum current
density for hydrogen separation from the H2 /CH4 mixtures at the respective operating
conditions, with regards to the power consumption. Furthermore, the power efficiencies
decreased with an increase in methane concentration and increased with temperature. The
highest power efficiency was, therefore, observed with the 10% CH4 mixture at 160 ◦ C:
~89%. Overall, the power efficiencies obtained with the respective H2 /CH4 mixtures were
much lower than with pure hydrogen, especially when the hydrogen-to-methane ratio of
the inlet decreased.
In terms of separation, a high purity hydrogen (>99.9%) was achieved with all the
methane inlet concentrations (10%, 50%, and 80% CH4 ). This implies that a high hydrogen
purity can be generated from H2 /CH4 gas mixtures, in a single-stage process, regardless of
the inlet methane concentration. Moreover, the hydrogen purity was slightly enhanced
with an increase in temperature. Besides the hydrogen purity, the hydrogen selectivity
values were also calculated and reported [51]:
yi
xi
αij = yj (5)
xj
Membranes 2021, 11, 282 12 of 24

where yi and yj are the molar fractions of gas species i and j on the permeate side, while xi
and xj are the molar fractions of gas species i and j on the feed side.

Figure 9. Current-(εi ), voltage-(εv ), and power (εp ) efficiencies as a function of current density at
120–160 ◦ C for (a) 10%, (b) 50%, and (c) 80% methane, balance hydrogen (hydrogen flow rate =
100 mLn /min, 1 atm).

The relationship between the hydrogen selectivity and current density, as a function
of methane concentration, is illustrated in Figure 10. At fixed temperature and impurity
concentration, the hydrogen purity of the permeate stream increased with an increase
in current density. This is because, according to Faraday’s law, the proton flux through
the membrane increases with current density, while the impurities (in this case CH4 ) are
unaffected by the current density. The permeation of the impure gas species (CH4 ) through
the membrane was proportional to/dependent on the change in partial pressure. When
the hydrogen partial pressure at the anode side decreased, the selectivity and the hydrogen
purity of the permeate (cathode outlet) increased.
Membranes 2021, 11, 282 13 of 24

Figure 10. H2 /CH4 selectivity, in terms of methane concentration, versus current density: (a) 10% (b) 50% and (c) 80%
methane, balance hydrogen (hydrogen flow rate = 100 mLn /min, 1 atm).

According to Rowe et al. [52], the separation performance of a membrane can be

improved by reducing the temperature, as explained by the Arrhenius equation. Our
findings were in agreement with this. According to Figure 10, much higher hydrogen selec-
tivity values were achieved at lower temperatures. For the 10% methane feed stream, the
hydrogen selectivity decreased from ~13,000 to ~12,750 and then to ~1980 as temperature
was increased from 120 to 140 ◦ C and then to 160 ◦ C. For the 50% methane feed stream, the
selectivity decreased from ~9800 (120 ◦ C) to ~6250 (140 ◦ C) and then to ~4800 (160 ◦ C). The
highest hydrogen selectivity (~22,200) was achieved with the 80% methane feed stream
at 140 ◦ C. It was therefore evident that the selectivity values increase with both methane
concentration and current density, and decrease when the temperature is increased.
Furthermore, the temperature effect on the hydrogen selectivity of the PEM was
more pronounced at higher temperatures. Compared to the hydrogen selectivity values of
other membrane technologies, for example, dense metallic (>1000), dense ceramic (>1000),
microporous ceramic (5–139), porous carbon (4–20), and dense polymeric (low), it is evident
that in terms of H2 /CH4 separation (selectivity) the EHS far exceeds its competitors [18,25].
Membranes 2021, 11, 282 14 of 24

4.3. Hydrogen Separation from H2 /CO2 Gas Mixtures

The voltage–current characteristics of electrochemical hydrogen separation from the
H2 /CO2 gas mixtures, as a function of temperature, are shown in Figure 11. The curved
profile of the polarization curves suggests that the electrode reaction dominates the over-
potential. Here the cell showed significantly larger voltages than in the case of the perfor-
mance with pure hydrogen. For the 10% CO2 (balance hydrogen) separation mixture, the
maximum achievable current density at 160 ◦ C was 0.16 A cm−2 . This is a large decline
from the 0.4 A cm−2 that was reached with pure hydrogen. Furthermore, at 120 and
140 ◦ C, the limiting current densities decreased from 0.28 and 0.32 A cm−2 , obtained with
pure hydrogen, to 0.12 and 0.16 A cm−2 , respectively. For the 50% CO2 inlet stream, the
limiting current densities were 0.14, 0.14, and 0.12 A cm−2 , respectively at 160, 140, and
120 ◦ C. A slight performance decline, in the order of 0.02 A cm−2 , was seen when the CO2
concentration was changed from 10% (Figure 11a) to 50% (Figure 11b).

Figure 11. Polarization curves for (a) 10%, (b) 50% carbon dioxide, balance hydrogen (hydrogen flow
rate = 100 mLn /min, 1 atm).

The change in the overpotential characteristics can be explained by the hydrogen to

impurity ratio of the inlet stream. Since the impurity (CO2 ) makes up a large portion of
the gas mixture, many of the active sites are covered by impurities and, consequently, the
reaction sites are ‘blocked’ by the impurities. This was also observed with the H2 /CH4
gas mixtures. However, the performance decay was more severe in the presence of CO2
than with CH4 , possibly caused by the electrocatalytic reduction/hydrogenation of CO2 .
Previous studies, where similar experiments were performed, reported the reduction of
CO2 to CO by the reverse water–gas shift (RWGS) [28,35,53]:

H2 + CO2 → CO + H2 O (6)

Here, the adsorbed hydrogen reduces CO2 to CO on the electrocatalyst surface [54,55].
The formed CO strongly bind to the Pt sites, consequently causing the Pt to be poisoned [56].
Therefore, Pt–H binding is hindered [54,57]; hence, a voltage increase result.
Theoretically, the electrode tolerance to CO should increase with temperature, which
is confirmed by the present work; the performance increased when the temperature was
raised (Figure 11), with lower voltages seen at higher temperatures. However, no CO was
Membranes 2021, 11, 282 15 of 24

detected with the in-line GC, situated downstream from the cathode outlet. Rather, trace
amounts of methane were found on ppm level. This could be explained by the methanation
of CO2 to CH4 , believed to be driven by the presence of Co in the electrocatalyst. Co is
known to have a high activity, similar to that of Ni [58,59] (Activity: Ru > Fe > Ni > Co >
Mo [60]) and high selectivity (Selectivity: Ni > Co > Fe > Ru [60]) towards the methanation
of CO2 . The possible reaction network for the reduction of CO2 can be described by
Equation (6) together with Equations (7) and (8) [61,62].

4H2 + CO2 → CH4 + 2H2 O (7)

3H2 + CO → CH4 + H2 O (8)

Only one article was found on the electrochemical reduction of CO2 in an electrolytic
cell based on polybenzimidazole (PBI) membranes doped with phosphoric acid (PA) [63].
Here, the electrocatalytic reduction of CO2 on Pt-Mo/C and Pt/C was examined and com-
pared [63]. Both CO and CH4 were detected as CO2 reduction products, although the Mo
addition into Pt in the Pt-Mo/C electrocatalyst significantly promoted CO formation, whilst
the formation of CH4 was suppressed. They confirmed that CO2 was reduced firstly to CO
by adsorbed hydrogen, and hereafter further reduced to CH4 by more adsorbed hydrogen.
Here, as was the case with the methane-containing mixtures, the OCV is not zero,
explained by the difference in partial pressures. See Table 4. The OCVs obtained were
much higher than those obtained with the H2 /CH4 mixtures at the same conditions. This
might be due to the fact that CH4 only acts as a hydrogen diluent [64], whilst CO2 actually
affects the electrode and its performance [47].

Table 4. Hydrogen purity and cell voltage values as a function of temperature for the H2 /CO2 feed
mixtures, measured at 0.1 A cm−2 .

Hydrogen Purity (%)

Measured Measured-OCV
Temperature Inlet Permeate OCV (V)
Voltage (V) (V)
(Anode) (Cathode)
50 99.39 0.012 0.454 0.442
120 ◦ C
90 99.56 0.002 0.300 0.298
50 99.29 0.012 0.132 0.120
140 ◦ C
90 99.44 0.002 0.170 0.168
50 99.34 0.013 0.125 0.112
160 ◦ C
90 99.63 0.002 0.138 0.136

In terms of hydrogen purity, 98–99.5% was reached with 10% CO2 in the feed and
96–99.5% with 50% CO2 in the feed stream. Therefore, a hydrogen purity of up to 99.5% can
be reached with a H2 /CO2 ratio of 1:1. No correlation was found between the hydrogen
purity of the permeate and the temperature, whereas the hydrogen purity did increase with
current density. Figure 12 shows the hydrogen selectivity values in terms of feed concentra-
tion. Reasonable separation was achieved from the 1:1 H2 /CO2 mixture, with selectivities
of up to ~200. This is, however, much lower than the hydrogen selectivity values obtained
with the H2 /CH4 gas mixtures (up to ~22,000). A similar trend was observed for both the
H2 /CH4 and H2 /CO2 mixtures in terms of current density and impurity concentration; the
hydrogen selectivity increased with current density and impurity concentration. However,
the trend of the selectivity in terms of temperature differed from that of the H2 /CH4 mix-
tures. The selectivity did not increase with a decrease in temperature; the best selectivity
was achieved at 160 ◦ C. This can be explained by the catalyst poisoning that occurs (CO
poisoning)—catalyst deactivation is alleviated at higher temperatures.
Membranes 2021, 11, 282 16 of 24

Figure 12. H2 /CO2 selectivity, in terms of carbon dioxide concentration, versus current density: (a) 10% and (b) 50% carbon
dioxide, balance hydrogen (hydrogen flow rate = 100 mLn /min, 1 atm).

Compared with other membrane technologies, the hydrogen selectivities here were
lower than achieved with dense metallic and dense ceramic membranes (>1000), but
higher than those obtained with microporous ceramic, porous carbon and dense polymeric
membranes [18,25].
The current, voltage, and power efficiencies of the H2 /CO2 gas mixtures are given
in Figure 13. In general, the current efficiencies increased with current density. The lower
current efficiencies at lower current densities can be attributed to hydrogen cross-over. The
current efficiencies at lower current densities were lower than in the case of pure hydrogen
(~70% at 0.02 A cm−2 ). In terms of voltage efficiencies, an increase was observed when the
temperature was increased, but a decrease with an increase in current density. A similar
trend was seen for both the pure hydrogen and the hydrogen–methane mixtures.
The power efficiencies showed a maximum value of 0.08–0.10 A cm−2 for both the 10
and 50% CO2 inlet mixtures. This depicts the optimum current density for hydrogen sepa-
ration from the H2 /CO2 mixtures at the respective operating conditions with regards to the
power consumption. Similar to the results recorded for the methane-containing mixtures,
the power efficiencies here decreased with an increase in carbon dioxide concentration
and increased with temperature. Moreover, nearly Faradic flow rates were achieved in all
experiments performed.
Membranes 2021, 11, 282 17 of 24

Figure 13. Current- (εi ), voltage- (εv ), and power (εp ) efficiencies as a function of current density at 120–160 ◦ C for (a) 10%
and (b) 50% carbon dioxide, balance hydrogen (hydrogen flow rate = 100 mLn /min, 1 atm).

4.4. Hydrogen Separation from H2 /NH3 Gas Mixtures

The voltage–current characteristic of the H2 /NH3 inlet mixtures are presented in
Figure 14. The cell performance is severely impaired in the presence of ammonia, even
in very low concentrations (ppm range), when compared to the performance with pure
hydrogen. The limiting current densities achieved were 0.10, 0.12, and 0.12 A cm−2 for 120,
140, and 160 ◦ C, respectively, when 1500 ppm ammonia was present. When the cell was
operated in the presence of 3000 ppm ammonia, the cell performance degraded completely,
even with an increase in temperature. The limiting current densities in the presence of
3000 ppm NH3 were 0.04, 0.04, and 0.02 A cm−2 for 120, 140, and 160 ◦ C, respectively. The
limiting current densities seemingly indicate that the performance of the cell decreases as
the temperature is raised. A reverse trend is, however, seen in the presence of 1500 ppm
NH3 . The results indicated that the performance of the cell deteriorates in the presence of
ammonia (continuous exposure to ammonia) rather than with temperature, as observed in
Figure 14b.
The selectivity graphs presented in Figure 15 show that very poor selectivities (<2) are
achieved for feed streams containing both 1500 and 3000 ppm ammonia. Ammonia and
hydrogen seem to compete with each other. In the cases where the selectivity values < 1,
ammonia transport through the membrane is favored, whereas hydrogen is favored at
selectivities > 1. Since hydrogen is split into its constituents, more specifically, protonated
to H+ ions, ammonia protonation results:

NH3 + H+ → NH4+ (9)

Membranes 2021, 11, 282 18 of 24

Figure 14. Polarization curves for (a) 1500 ppm, (b) 3000 ppm ammonia, balance hydrogen (hydrogen
flow rate = 100 mLn /min, 1 atm).

Figure 15. H2 /NH3 selectivity, in terms of ammonia concentration, versus current density:
(a) 1500 ppm and (b) 3000 ppm ammonia, balance hydrogen (hydrogen flow rate = 100 mLn /min,
1 atm).

NH3 + is then also transported through the membrane.

The anode reaction can be depicted as follows:

NH3 + 2H2 → NH4+ + H+ (10)

Membranes 2021, 11, 282 19 of 24

At lower current densities, however, the reaction seems to be the following, where
ammonia is favored and not hydrogen:

2NH3 + H2 → 2NH4+ (11)

The effect of ammonia on the performance of LT-FCs has been determined, and
reported in literature [65–67]. Some sources even report a decrease in FC catalytic activity
when only 1 ppm NH3 is present, over a short one-week period. Their tests were performed
with Nafion® (Wilmington, DE, USA) membranes. The loss in activity can be attributed
to the alkaline nature of ammonia. The formed NH4 + ions (Equation (11)) react with the
sulphonic groups (SO3 − ), which are characteristic of Nafion membranes. Consequently,
the water content and the conductivity of the membrane decreases, resulting in reduced
FC performances [66,67]. Likewise, as for the sulphonic groups of the Nafion® membranes,
so too for PA used in HT membranes (usually PBI-based membranes), which results in
phosphate groups (PO4 3− ). The performance losses observed in Figure 15 are therefore
attributed to the alkaline nature of ammonia, reacting with the PA on the membrane. More
specifically the formed NH4 + ions react with the PO4 3− groups, consequently resulting in a
loss of conductivity. Similar to the results reported here, Vassiliev et al. [68] reports on the
possible chemical incompatibility between the other organic fuels and the electrolyte in HT
membranes doped with PA. More specifically, it was concluded that the probable reasons
were that the catalyst deactivation and/or acid depletion within the electrode occurred.
Results of studies carried out with Nafion membranes to determine the effect of am-
monia on the performance of LT-FCs have revealed that although substantial deactivation
effects were observed in the presence of ammonia in the feed gas, the cell performance
was recoverable, with time (in the order of days), when neat hydrogen is introduced to
the anode side [67]. In the present study, however, this did not apply—the membranes
were deactivated beyond recovery; for example, membrane cracks occurred (see Figure 16).
Scanning electron microscopy (SEM) images of the microscopic morphological changes on
the surface of a membrane and GDLs are also shown in Figure 17. Small cracks are even
visible at the microscopic level.

Figure 16. Digital images of polybenzimidazole (PBI) membranes before and after ammonia exposure: (a) fresh membrane
(before) and (b) cracked membrane (after).
Membranes 2021, 11, 282 20 of 24

Figure 17. SEM images/surface morphologies of MEA components (before and after ammonia exposure): (a) membrane
before ammonia exposure, (b) affected membrane after ammonia exposure, (c) cathode GDL (Pt catalyst) before ammonia
exposure, (d) cathode GDL (Pt catalyst) after ammonia exposure, (e) anode GDL (Pt/Co catalyst) before ammonia exposure
and (f) anode GDL (Pt/Co catalyst) after ammonia exposure (MEA: membrane electrode assembly; GDL: gas diffusion layer).
Membranes 2021, 11, 282 21 of 24

4.5. Possible Future Work

As for a possible future work, one can consider using hydrocarbon-based PEM mem-
branes [69], PFSA-based materials with various modifications to address water manage-
ment challenges [70], as well as novel phosphonated materials [40].

5. Conclusions
For hydrogen feed, the voltage–current characteristics showed that the ohmic losses
dominated the overpotential. From this, it was concluded that the membrane- and contact
resistances of the cell play a major role in the efficacy of the EHS, and should, therefore, be
minimised. Overall, the general performance of the cell increased with minor temperature
changes (20 ◦ C). Moreover, the current efficiencies for the pure hydrogen inlet ranged from
~70% to tlevels >95%. The current efficiency increased with current density, whilst the
voltage efficiency increased with temperature and decreased with current density. This
resulted in higher power efficiencies as the temperature was raised. The optimal current
density with regards to power consumption was ~0.16 A cm−2 . The highest limiting
current density was 0.4 A cm−2 , which is well below a current density of practical interest,
1 A cm−2 .
Results revealed that hydrogen can be effectively recovered from H2 /CH4 mixtures,
yielding a hydrogen purity of >99.9%. From the polarization curves and selectivity graphs
of the H2 /CH4 mixtures, it was evident that the selectivity decreases with temperature,
whilst the overall cell efficiency (power efficiency) increases, therefore the selectivity and
cell efficiency behave antagonistically towards another. The highest selectivity achieved
with H2 /CH4 was ca. 22,000. Similar trends with regards to current, voltage, and power
efficiencies were visible when compared with trends of the pure hydrogen experiments.
Furthermore, performance of the cell was seen to decrease with an increase in methane
concentration. Therefore, the cell performance is sensitive to the partial pressure of the
hydrogen concentration in the feed.
In the H2 /CO2 experiments, the Pt/Co catalyst is severely influenced by CO2 , even at
elevated temperatures. Hydrogen purities of 98–99.5% can be achieved for feed streams
containing 10% CO2 and 96–99.5% for in the feed streams containing 50% CO2 . Overall,
poor selectivities (≤200) were reached for the CO2 -containing streams, with poor overall
cell performances due to catalyst deactivation caused by CO2 reduction. Similar to the
H2 /CH4 mixtures, the power efficiencies decreased with an increase in CO2 concentration
and increased with temperature. The power efficiencies were, however, lower than that
obtained by pure hydrogen and the CH4 -containing mixtures.
Ammonia has a noticeable effect on the membrane—ammonia reacts with the acid
embedded in the membrane. Even at very low concentrations (ppm) the membranes were
affected by ammonia beyond recovery and, hence, unsuitable for use in electrochemical
hydrogen separation. However, regarding the proton transport through the membrane,
ammonia and hydrogen competed; ammonia transport was favored at low current densities
and hydrogen transport was favored at higher current densities. The formation of NH4 +
demonstrates the possibility of electrochemical transport/pumping and compression of
ammonia with a PEM. In future research, alternative anion exchange membranes could be
considered to possibly facilitate this reaction.
In terms of the hydrogen selectivity, power efficiencies, and the effects on the cell
components, the performance of the cell based on the different gas mixtures were in the
order: H2 /CH4 > H2 /CO2 > H2 /NH3 .

Author Contributions: L.V.: Conceptualization, investigation, data curation, writing—original draft,

visualization and project administration. H.W.J.P.N.: supervision, writing—review and editing.
D.G.B.: conceptualization, resources, supervision, writing—review and editing and funding acquisi-
tion. All authors have read and agreed to the published version of the manuscript.
Funding: This work was supported by HySA Infrastructure (via KP5 program) at the North-West
University, Potchefstroom, South Africa.
Membranes 2021, 11, 282 22 of 24

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of 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.

References
1. Balat, M. Potential importance of hydrogen as a future solution to environmental and transportation problems. Int. J. Hydrogen
Energy 2008, 33, 4013–4029. [CrossRef]
2. Abe, J.; Popoola, A.; Ajenifuja, E.; Popoola, O. Hydrogen energy, economy and storage: Review and recommendation. Int. J.
Hydrogen Energy 2019, 44, 15072–15086. [CrossRef]
3. El-Shafie, M.; Kambara, S.; Hayakawa, Y. Hydrogen Production Technologies Overview. J. Power Energy Eng. 2019, 7, 107–154.
[CrossRef]
4. Rusman, N.; Dahari, M. A review on the current progress of metal hydrides material for solid-state hydrogen storage applications.
Int. J. Hydrogen Energy 2016, 41, 12108–12126. [CrossRef]
5. Chapman, A.; Itaoka, K.; Hirose, K.; Davidson, F.T.; Nagasawa, K.; Lloyd, A.C.; Webber, M.E.; Kurban, Z.; Managi, S.; Tamaki, T.;
et al. A review of four case studies assessing the potential for hydrogen penetration of the future energy system. Int. J. Hydrogen
Energy 2019, 44, 6371–6382. [CrossRef]
6. The Paris Agreement | UNFCCC. Available online: https://unfccc.int/process-and-meetings/the-paris-agreement/the-paris-
agreement (accessed on 23 January 2020).
7. Holladay, J.D.; Wang, Y.; Jones, E. Review of Developments in Portable Hydrogen Production Using Microreactor Technology.
Chem. Rev. 2004, 104, 4767–4790. [CrossRef] [PubMed]
8. SAE International-Hydrogen Fuel Quality for Fuel Cell Vehicles: J2719. 2015. Available online: https://www.sae.org/standards/
content/j2719_201109/ (accessed on 23 January 2020).
9. Thomassen, M.; Sheridan, E.; Kvello, J. Electrochemical hydrogen separation and compression using polybenzimidazole (PBI)
fuel cell technology. J. Nat. Gas Sci. Eng. 2010, 2, 229–234. [CrossRef]
10. Acar, C.; Dincer, I. Comparative assessment of hydrogen production methods from renewable and non-renewable sources. Int. J.
Hydrogen Energy 2014, 39, 1–12. [CrossRef]
11. Dincer, I.; Acar, C. Review and evaluation of hydrogen production methods for better sustainability. Int. J. Hydrogen Energy 2015,
40, 11094–11111. [CrossRef]
12. Baykara, S.Z. Hydrogen: A brief overview on its sources, production and environmental impact. Int. J. Hydrogen Energy 2018, 43,
10605–10614. [CrossRef]
13. Perry, K.A.; Eisman, G.A.; Benicewicz, B.C. Electrochemical hydrogen pumping using a high-temperature polybenzimidazole
(PBI) membrane. J. Power Sources 2008, 177, 478–484. [CrossRef]
14. Chen, C.-Y.; Lai, W.-H.; Chen, Y.-K.; Su, S.-S. Characteristic studies of a PBI/H3PO4 high temperature membrane PEMFC under
simulated reformate gases. Int. J. Hydrogen Energy 2014, 39, 13757–13762. [CrossRef]
15. Lee, H.; Choi, H.; Choi, K.; Park, J.; Lee, T. Hydrogen separation using electrochemical method. J. Power Sources 2004, 132, 92–98.
[CrossRef]
16. Grande, C.A.; Lopes, F.V.S.; Ribeiro, A.M.; Loureiro, J.M.; Rodrigues, A.E. Adsorption of Off-Gases from Steam Methane
Reforming (H2 , CO2 , CH4 , CO and N2 ) on Activated Carbon. Sep. Sci. Technol. 2008, 43, 1338–1364. [CrossRef]
17. Granite, E.J.; O’Brien, T. Review of novel methods for carbon dioxide separation from flue and fuel gases. Fuel Process. Technol.
2005, 86, 1423–1434. [CrossRef]
18. Ockwig, N.W.; Nenoff, T.M. Membranes for Hydrogen Separation. Chem. Rev. 2007, 107, 4078–4110. [CrossRef]
19. David, O.C. Membrane Technologies for Hydrogen and Carbon Monoxide Recovery from Residual Gas Streams; Universidad de Cantabria:
Santander, Spain, 2012; pp. 1–183.
20. Huang, C.; Ali, T. Analyses of one-step liquid hydrogen production from methane and landfill gas. J. Power Sources 2007, 173,
950–958. [CrossRef]
21. Peters, T.; Stange, M.; Klette, H.; Bredesen, R. High pressure performance of thin Pd–23%Ag/stainless steel composite membranes
in water gas shift gas mixtures; influence of dilution, mass transfer and surface effects on the hydrogen flux. J. Membr. Sci. 2008,
316, 119–127. [CrossRef]
22. Liu, F.; Zhang, N. Strategy of Purifier Selection and Integration in Hydrogen Networks. Chem. Eng. Res. Des. 2004, 82, 1315–1330.
[CrossRef]
23. Liemberger, W.; Groß, M.; Miltner, M.; Harasek, M. Experimental analysis of membrane and pressure swing adsorption (PSA) for
the hydrogen separation from natural gas. J. Clean. Prod. 2017, 167, 896–907. [CrossRef]
24. Ravanchi, M.T.; Kaghazchi, T.; Kargari, A. Application of membrane separation processes in petrochemical industry: A review.
Desalination 2009, 235, 199–244. [CrossRef]
25. Adhikari, S.; Fernando, S. Hydrogen Membrane Separation Techniques. Ind. Eng. Chem. Res. 2006, 45, 875–881. [CrossRef]
Membranes 2021, 11, 282 23 of 24

26. Wu, X.; Benziger, J.; He, G. Comparison of Pt and Pd catalysts for hydrogen pump separation from reformate. J. Power Sources
2012, 218, 424–434. [CrossRef]
27. Sakai, T.; Matsumoto, H.; Kudo, T.; Yamamoto, R.; Niwa, E.; Okada, S.; Hashimoto, S.; Sasaki, K.; Ishihara, T. High performance
of electroless-plated platinum electrode for electrochemical hydrogen pumps using strontium-zirconate-based proton conductors.
Electrochim. Acta 2008, 53, 8172–8177. [CrossRef]
28. Bouwman, P. Fundamental of Electrochemical Hydrogen Compression. In PEM Electrolysis for Hydrogen Production: Principles and
Applications; Bessarabov, D., Wang, H., Li, H., Zhao, N., Eds.; CRC Press: Boca Raton, FL, USA, 2015; pp. 269–299.
29. Grigoriev, S.; Shtatniy, I.; Millet, P.; Porembsky, V.; Fateev, V. Description and characterization of an electrochemical hydrogen
compressor/concentrator based on solid polymer electrolyte technology. Int. J. Hydrogen Energy 2011, 36, 4148–4155. [CrossRef]
30. Gardner, C.; Ternan, M. Electrochemical separation of hydrogen from reformate using PEM fuel cell technology. J. Power Sources
2007, 171, 835–841. [CrossRef]
31. Kusoglu, A.; Tang, Y.; Lugo, M.; Karlsson, A.M.; Santare, M.H.; Cleghorn, S.; Johnson, W.B. Constitutive response and mechanical
properties of PFSA membranes in liquid water. J. Power Sources 2010, 195, 483–492. [CrossRef]
32. Haque, M.A.; Sulong, A.; Loh, K.; Majlan, E.H.; Husaini, T.; Rosli, R.E. Acid doped polybenzimidazoles based membrane
electrode assembly for high temperature proton exchange membrane fuel cell: A review. Int. J. Hydrogen Energy 2017, 42,
9156–9179. [CrossRef]
33. Huth, A.; Schaar, B.; Oekermann, T. A “proton pump” concept for the investigation of proton transport and anode kinetics in
proton exchange membrane fuel cells. Electrochim. Acta 2009, 54, 2774–2780. [CrossRef]
34. Kwon, K.; Park, J.O.; Yoo, D.Y.; Yi, J.S. Phosphoric acid distribution in the membrane electrode assembly of high temperature
proton exchange membrane fuel cells. Electrochim. Acta 2009, 54, 6570–6575. [CrossRef]
35. Araya, S.S.; Zhou, F.; Liso, V.; Sahlin, S.L.; Vang, J.R.; Thomas, S.; Gao, X.; Jeppesen, C.; Kær, S.K. A comprehensive review of
PBI-based high temperature PEM fuel cells. Int. J. Hydrogen Energy 2016, 41, 21310–21344. [CrossRef]
36. Alegre, C.; Álvarez-Manuel, L.; Mustata, R.; Valiño, L.; Lozano, A.; Barreras, F. Assessment of the durability of low-cost Al bipolar
plates for High Temperature PEM fuel cells. Int. J. Hydrogen Energy 2019, 44, 12748–12759. [CrossRef]
37. Kim, B.; Ma, S.; Jhong, H.-R.M.; Kenis, P.J. Influence of dilute feed and pH on electrochemical reduction of CO2 to CO on Ag in a
continuous flow electrolyzer. Electrochim. Acta 2015, 166, 271–276. [CrossRef]
38. Chippar, P.; Oh, K.; Kim, W.-G.; Ju, H. Numerical analysis of effects of gas crossover through membrane pinholes in high-
temperature proton exchange membrane fuel cells. Int. J. Hydrogen Energy 2014, 39, 2863–2871. [CrossRef]
39. Lee, K.-S.; Spendelow, J.S.; Choe, Y.-K.; Fujimoto, C.; Kim, Y.S. An operationally flexible fuel cell based on quaternary ammonium-
biphosphate ion pairs. Nat. Energy 2016, 1, 16120. [CrossRef]
40. Atanasov, V.; Lee, A.S.; Park, E.J.; Maurya, S.; Baca, E.D.; Fujimoto, C.; Hibbs, M.; Matanovic, I.; Kerres, J.; Kim, Y.S. Synergistically
integrated phosphonated poly(pentafluorostyrene) for fuel cells. Nat. Mater. 2021, 20, 370–377. [CrossRef]
41. Kim, Y.S. Polymer Electrolytes with High Ionic Concentration for Fuel Cells and Electrolyzers. ACS Appl. Polym. Mater. 2021, 3,
1250–1270. [CrossRef]
42. Vermaak, L.; Neomagus, H.W.J.P.; Bessarabov, D.G. Recent Advances in Membrane-Based Electrochemical Hydrogen Separation:
A Review. Membranes 2021, 11, 127. [CrossRef]
43. Bessarabov, D.; Human, G.; Kruger, A.J.; Chiuta, S.; Modisha, P.M.; Du Preez, S.P.; Oelofse, S.P.; Vincent, I.; Van Der Merwe, J.;
Langmi, H.W.; et al. South African hydrogen infrastructure (HySA infrastructure) for fuel cells and energy storage: Overview of
a projects portfolio. Int. J. Hydrogen Energy 2017, 42, 13568–13588. [CrossRef]
44. Bard, A.J.; Faulkner, L.R. Electrochemical Methods: Fundamentals and Applications; Wiley: Hoboken, NJ, USA, 2001; pp. 261–304.
45. Barbir, F.; Görgün, H. Electrochemical hydrogen pump for recirculation of hydrogen in a fuel cell stack. J. Appl. Electrochem. 2006,
37, 359–365. [CrossRef]
46. Zhang, J.; Tang, Y.; Song, C.; Zhang, J. Polybenzimidazole-membrane-based PEM fuel cell in the temperature range of 120–200 ◦ C.
J. Power Sources 2007, 172, 163–171. [CrossRef]
47. Kim, S.J.; Lee, B.-S.; Ahn, S.H.; Han, J.Y.; Park, H.Y.; Kim, S.H.; Yoo, S.J.; Kim, H.-J.; Cho, E.; Henkensmeier, D.; et al. Characteriza-
tions of polybenzimidazole based electrochemical hydrogen pumps with various Pt loadings for H2 /CO2 gas separation. Int. J.
Hydrogen Energy 2013, 38, 14816–14823. [CrossRef]
48. Nguyen, M.-T.; Grigoriev, S.A.; Kalinnikov, A.A.; Filippov, A.A.; Millet, P.; Fateev, V.N. Characterisation of a electrochemical
hydrogen pump using electrochemical impedance spectroscopy. J. Appl. Electrochem. 2011, 41, 1033–1042. [CrossRef]
49. Sahlin, S.L.; Araya, S.S.; Andreasen, S.J.; Kær, S.K. Electrochemical Impedance Spectroscopy (EIS) Characterization of Reformate-
operated High Temperature PEM Fuel Cell Stack. Int. J. Power Energy Res. 2017, 1, 20–40. [CrossRef]
50. Su, A.; Ferng, Y.; Hou, J.; Yu, T. Experimental and numerical investigations of the effects of PBI loading and operating temperature
on a high-temperature PEMFC. Int. J. Hydrogen Energy 2012, 37, 7710–7718. [CrossRef]
51. Li, P.; Wang, Z.; Qiao, Z.; Liu, Y.; Cao, X.; Li, W.; Wang, J.; Wang, S. Recent developments in membranes for efficient hydrogen
purification. J. Membr. Sci. 2015, 495, 130–168. [CrossRef]
52. Rowe, B.W.; Robeson, L.M.; Freeman, B.D.; Paul, D.R. Influence of temperature on the upper bound: Theoretical considerations
and comparison with experimental results. J. Membr. Sci. 2010, 360, 58–69. [CrossRef]
53. Onda, K.; Araki, T.; Ichihara, K.; Nagahama, M. Treatment of low concentration hydrogen by electrochemical pump or proton
exchange membrane fuel cell. J. Power Sources 2009, 188, 1–7. [CrossRef]
Membranes 2021, 11, 282 24 of 24

54. Hoshi, N.; Mizumura, T.; Hori, Y. Significant difference of the reduction rates of carbon dioxide between Pt(111) and Pt(110)
single crystal electrodes. Electrochim. Acta 1995, 40, 883–887. [CrossRef]
55. Janssen, G. Modelling study of CO2 poisoning on PEMFC anodes. J. Power Sources 2004, 136, 45–54. [CrossRef]
56. Kortlever, R.; Shen, J.; Schouten, K.J.P.; Calle-Vallejo, F.; Koper, M.T.M. Catalysts and Reaction Pathways for the Electrochemical
Reduction of Carbon Dioxide. J. Phys. Chem. Lett. 2015, 6, 4073–4082. [CrossRef] [PubMed]
57. Ru, F.Y.; Zulkefli, N.N.; Yusra, N.; Yusuf, M.; Masdar, M.S. Effect of Operating Parameter on H2 /CO2 Gas Separation using
Electrochemical Cell. Int. J. Appl. Eng. Res. 2018, 13, 505–510.
58. Rönsch, S.; Schneider, J.; Matthischke, S.; Schlüter, M.; Götz, M.; Lefebvre, J.; Prabhakaran, P.; Bajohr, S. Review on methanation—
From fundamentals to current projects. Fuel 2016, 166, 276–296. [CrossRef]
59. Kok, E.; Scott, J.; Cant, N.; Trimm, D. The impact of ruthenium, lanthanum and activation conditions on the methanation activity
of alumina-supported cobalt catalysts. Catal. Today 2011, 164, 297–301. [CrossRef]
60. Mills, G.A.; Steffgen, F.W. Catalytic Methanation. Catal. Rev. 1974, 8, 159–210. [CrossRef]
61. Guerra, L.; Rossi, S.; Rodrigues, J.; Gomes, J.; Puna, J.; Santos, M. Methane production by a combined Sabatier reaction/water
electrolysis process. J. Environ. Chem. Eng. 2018, 6, 671–676. [CrossRef]
62. Wu, H.C.; Chang, Y.C.; Wu, J.H.; Lin, J.H.; Lin, I.K.; Chen, C.S. Methanation of CO2 and reverse water gas shift reactions on
Ni/SiO2 catalysts: The influence of particle size on selectivity and reaction pathway. Catal. Sci. Technol. 2015, 5, 4154–4163.
[CrossRef]
63. Gao, D.; Cai, F.; Xu, Q.; Wang, G.; Pan, X.; Bao, X. Gas-phase electrocatalytic reduction of carbon dioxide using electrolytic cell
based on phosphoric acid-doped polybenzimidazole membrane. J. Energy Chem. 2014, 23, 694–700. [CrossRef]
64. Ibeh, B.; Gardner, C.; Ternan, M. Separation of hydrogen from a hydrogen/methane mixture using a PEM fuel cell. Int. J. Hydrogen
Energy 2007, 32, 908–914. [CrossRef]
65. Halseid, R.; Vie, P.J.; Tunold, R. Effect of ammonia on the performance of polymer electrolyte membrane fuel cells. J. Power Sources
2006, 154, 343–350. [CrossRef]
66. Uribe, F.A.; Gottesfeld, S.; Zawodzinski, T.A. Effect of Ammonia as Potential Fuel Impurity on Proton Exchange Membrane Fuel
Cell Performance. J. Electrochem. Soc. 2002, 149, A293–A296. [CrossRef]
67. Halseid, R.; Vie, P.J.S.; Tunold, R. Influence of Ammonium on Conductivity and Water Content of Nafion 117 Membranes. J.
Electrochem. Soc. 2004, 151, A381–A388. [CrossRef]
68. Vassiliev, A.; Reumert, A.K.; Jensen, J.O.; Aili, D. Durability and degradation of vapor-fed direct dimethyl ether high temperature
polymer electrolyte membrane fuel cells. J. Power Sources 2019, 432, 30–37. [CrossRef]
69. Kim, A.R.; Gabunada, J.C.; Yoo, D.J. Amelioration in physicochemical properties and single cell performance of sulfonated
poly(ether ether ketone) block copolymer composite membrane using sulfonated carbon nanotubes for intermediate humidity
fuel cells. Int. J. Energy Res. 2019, 43, 2974–2989. [CrossRef]
70. Vinothkannan, M.; Hariprasad, R.; Ramakrishnan, S.; Kim, A.R.; Yoo, D.J. Potential Bifunctional Filler (CeO2–ACNTs) for Nafion
Matrix toward Extended Electrochemical Power Density and Durability in Proton-Exchange Membrane Fuel Cells Operating at
Reduced Relative Humidity. ACS Sustain. Chem. Eng. 2019, 7, 12847–12857. [CrossRef]