i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6

Available online at www.sciencedirect.com

ScienceDirect

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

Comparative performance assessment of three

configurations of magnesiumechlorine cycle

Hasan Ozcan a,*, Ibrahim Dincer b

a
Clean Energy Research Laboratory (CERL), Faculty of Engineering and Applied Science, University of Ontario
Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario, L1H 7K4, Canada
b
Center of Research Excellence in Renewable Energy Research Institute, King Fahd University of Petroleum and
Minerals, Dhahran 31261, Saudi Arabia

article info abstract

Article history: The magnesiumechlorine cycle is proposed to be one of the promising thermo-
Received 13 August 2015 electrochemical (hybrid) hydrogen production cycles. The main steps of the cycle include
Received in revised form hydrolysis of MgCl2, chlorination of MgO, and electrolysis of HCl. Based on the information
21 October 2015 from the high temperature hydrolysis of common salts (i.e., KCl, NaCl, CaCl2, CuCl2), a
Accepted 22 October 2015 higher steam/solid ratio is utilized for better conversion of the reactants. Thus, it is ex-
Available online 18 December 2015 pected that HCl would be in mixture with steam after the hydrolysis process, which yields
to a possible aqueous electrolytic process, resulting in a higher power consumption in the
Keywords: electrolyzer. This study aims to develop an optimum configuration of the MgeCl cycle in
MgeCl cycle order to decrease the electrical energy requirement of the electrolysis step, as well as to
Hydrogen production overcome the problems, such as oxygen evolution and chlorine solubility in water by
Efficiency substituting the aqueous electrolysis step to anhydrous electrolysis. Two configurations of
Energy the three-step cycle are comparatively assessed in terms of their efficiencies and energy
Exergy requirements. A novel fourth step is then introduced for anhydrous HCl production with an
additional decomposition step. Energy requirements of all individual cycles are presented
and comparatively discussed. The newly developed four-step cycle shows higher ther-
modynamic performances than those of three-step options with decreased electrical work
consumption. The calculated energy and exergy efficiencies of the four-step MgeCl cycle
are 43.7% and 52%, respectively, resulting in a lower electrical work consumption than that
of water electrolysis with a maximum temperature of 723 K.
Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.

carbon-free nature, environmental friendliness, high effi-

Introduction
Hydrogen is an energy carrier which is not readily available in
nature, but can be produced through disassociation of water.
Hydrogen is recognized as an ideal energy carrier due to its
ciency, reliability, sustainability, etc. [1]. There are several
methods to produce hydrogen. These methods can be cate-
gorized into six classes, namely electrochemical, thermo-
chemical, photochemical, radiochemical, biochemical and
hybrid (involving two or more of first five methods together).

* Corresponding author.
E-mail addresses: Hasan.Ozcan@uoit.ca (H. Ozcan), Ibrahim.Dincer@uoit.ca (I. Dincer).
http://dx.doi.org/10.1016/j.ijhydene.2015.10.078
0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
846 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6
Renewable energy resources, such as solar, hydro, wind, tidal,
biomass, geothermal and nuclear power and process heat are
used for sustainable hydrogen production [2].
Thermochemical splitting of water is an alternative to
electrolytic splitting of water requiring more than one chem-
ical reaction and sums thermolysis of water at elevated tem-
peratures. Water splitting at high temperature (thermolysis) is
possible at very high temperatures (>2000 K) and results in
production of oxygen and hydrogen as one output which re-
quires a very challenging separation process [3]. Multi-step
thermochemical cycles work at lower maximum tempera-
tures than the one step thermolysis through a cyclic process of
splitting water. These steps vary from 2 to 6 steps in relation
with the chemistry of substances where a higher number of
steps tends to work at lower maximum temperatures [3].
While the majority of hybrid methods shows lower maximum
temperatures than pure thermochemical processes, the
UeEueBr cycle shows 573 K maximum temperature and is a
purely thermochemical cycle [4]. However, technical issues,
such as very low reaction rates of exothermic reactions, are
required to be considered for a more feasible cycle.
Hybrid thermochemical cycles can be arranged as thermo-
electrochemical, thermo-photo-electrochemical, and thermo-
radiochemical cycles. Thermo-electrochemical cycles utilize
electricity at least one of the steps through the cycle. Thermo-
photo-electrochemical cycles need a photochemical reaction
to complete the cycle, and a radiochemical reaction is needed
for thermo-radiochemical cycles [3].
Most thermo-electrochemical cycles have been proposed
in order to decrease the Reverse Deacon Reaction temperature
by using electrical work lower than the water electrolysis to
make the cycle a feasible one. Lower temperature cycles tend
to consume more electrical work to bring down the Gibbs free
energy to zero. Low maximum temperatures are one of the
most important advantages of hybrid cycles to utilize heat
from relatively low temperature heat sources such as nuclear
and solar energy. The chlorine family cycles as well as other
hybrid cycles under study can be listed as follows [5e8]:

Hybrid Sulfur cycle (HyS) (Ispra Mark 11)

Hybrid Chlorine cycle (Hallet air products)

CueCl cycle (3, 4 and 5 steps)

HBr cycle (ANL)

MgeCl cycle (ANL)
This study represents a comprehensive thermodynamic
comparison of various options of the MgeCl cycle. Aspen Plus
simulations are conducted to determine thermophysical and
thermochemical data of reactors and streams throughout all
options, and a thermodynamic assessment is conducted to
evaluate performances of all cycles under study.
Configurations of MgeCl cycle
The MgeCl cycle can be simulated by considering all individual
chemical reactions throughout the cycle using the hydrolysis
step as the main influencing reaction. Aspen Plus software is
used to simulate all options of the MgeCl cycle, and reactions
are simulated using RGibbs reactor, where heaters are used for
temperature manipulations through streams. The RGibbs
reactor aims to minimize the Gibbs free energy of a reaction
which is a promising built-in reactor type to determine reaction
conditions. However, majority of the reactions are already
known with their existing kinetics, thus, these data are also
taken into consideration for the yields. Here, RYield type re-
actors are used for this case for known conversions such as
chlorination reactor. There are several exothermic reactions
throughout the cycle. These type of reactors can be conveniently
used to determine the thermodynamics of a reaction with
known conversions at specific reactor conditions. These re-
actions are assumed to be isothermal for the sake of preventing
any interruption of the reaction rates. Thus, RYield type
reactors are adapted considering existing experimental data
from the literature. Initially three options are evaluated with
respect to their heat requirements. The first option is taken to be
high temperature hydrolysis of MgCl2 into MgO and HCl gas
(MgeCl-A). The second option is relatively lower temperature
hydrolysis of MgCl2 into MgOHCl and HCl gas, where stoichi-
ometry of MgCl2 should be doubled in order to obtain same
amount of HCl gas as in the first option (MgeCl-B). Finally, a
novel fourth step is introduced in the third option with a
decomposition step of the MgOHCl into MgO and HCl gas
(MgeCl-C). All configurations are designed to produce
3600 kmol/h of hydrogen, and the net heat and work re-
quirements are determined.
MgCl2eMgO (MgeCl-A) cycle
The Hybrid MgeCl cycle is initially proposed as a three-step
thermo-electrochemical cycle utilizing heat and electrical
work. Main steps of the cycle are; hydrolysis of MgCl2, chlori-
nation of MgO, and electrolysis of HCl gas. It can be run with two
different options, namely MgCl2eMgO, and MgCl2eMgOHCl
cycle as reported by Ref. [9]. Feasible chemical reactions,
mature electrolysis technology, and relatively low maximum
temperature of the cycle are promising in terms of integrating
this cycle with nuclear and solar energy sources. Aspen flow-
sheet of this cycle is represented in Fig. 1. Argonne National
Laboratory (ANL) studied the hydrolysis step of this cycle with
additives and obtained promising results for yields of the re-
actants at desired temperatures [6]. Individual reactions of the
three step MgeCl cycle has also been studied and adapted into
integrated systems and can be found elsewhere [10e12].
MgCl2eMgOHCl (MgeCl-B) cycle
The MgCl2eMgOHCl cycle is also one of the alternative options
for the MgeCl cycle where the hydrolysis reaction is exothermic
and commences at 513 K with a full conversion at 573 K, and in
order to produce same amount of hydrogen as in MgCl2eMgO
cycle, the stoichiometry of this reaction should be doubled. Fig. 2
shows the Aspen flowsheet of this cycle. Chlorination reaction
produces a mixture of steam and oxygen, where possible
problems may occur during the separation process due to sol-
ubility of oxygen in water. The excess steam from the chlori-
nation reactor is then recovered back in the hydrolysis reactor.
One of the main issues with three-step MgeCl cycle con-
figurations is possible high steam/HCl ratios after the hydro-
lysis step which results in aqueous HCl production. The
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6 847

Fig. 1 e Aspen flowsheet of the MgeCl-A cycle.

Fig. 2 e Aspen flowsheet of the MgeCl-B cycle.

aqueous HCl electrolysis is a mature process where 1.8 V per
mole of hydrogen is required with several solubility issues of
oxygen and chlorine gas in water, which makes the electrol-
ysis step less advantageous than the conventional water
electrolysis even though the theoretical requirement is lower
(0.99 V). The anhydrous (dry) HCl production is one of the
crucial issues throughout the cycle in order to consume less
electrical work than water electrolysis. Fig. 3 represents
practical voltage comparison of some electrolysis processes
for hydrogen production [13e17].
Four step MgeCl (MgeCl-C) cycle
As the last alternative, a fourth step is introduced to decompose
MgOHCl in order to recover the remaining HCl in anhydrous
form, and this step is introduced for the following reasons:

It is possible to obtain dry HCl gas by decomposing
MgOHCl, and considering the practical voltage
requirements, this option directly reduces the electrical
work consumption by 11.2% compared to three-step
configurations.

The stoichiometry to produce same amount of hydrogen as
in the MgeCl-B option can be reduced from two to one, by
capturing HCl gas in the decomposition step.

Produced MgO from the high temperature hydrolysis is less
reactive with the chlorine gas and has less surface area for
proper reaction than the hydrolysis of Mg(OH)2. However,
experimental studies showed that fine MgO can be pro-
duced from decomposition of MgOHCl which would
enhance the reactivity of MgO with the chlorine gas [18].

The only endothermic reaction is the decomposition of
MgOHCl, and maximum temperature of this reaction can
be reduced to 723 K from ~800 K by using an inert gas to
remove HCl gas from the surface of MgO particles. This also
leads to a very fast decomposition of MgOHCl into desired
products. The lower maximum temperature provides the
option to link this cycle to low-temperature heat sources.
848 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6

Fig. 3 e Practical Voltage comparisons of some selected electrolysis methods for hydrogen production.

Both stability and decomposition kinetics of MgOHCl sub-
stance have been studied by [18] and it was concluded that the
removal of HCl gas with Argon gas increases the rate of the
reaction significantly. This reaction occurs from 649 K, and a
minimum of 806 K is needed to fully decompose into HCl and
MgO. However lower reaction temperatures have not been
considered with a higher inert gas flowrate. The Aspen flow
sheet for this configuration is represented in Fig. 4.
The hydrolysis step is set to 553 K to form MgOHCl where the
reactor temperature for decomposition step is set to 723 K. The
required steam for the hydrolysis step is same as the MgeCl-B
cycle and it is set to 11:1 [9]. A preheating process is considered
to bring the MgOHCl temperature to the lower limit of decom-
position (649 K). Argon gas is used as the inert gas to remove the
HCl gas from MgO surface and adapted in the simulation with
varying flow rates. The produced HCl gas is utilized in the dry
HCl electrolysis process, where the steam/HCl mixture from the
hydrolysis step is cooled to the electrolysis temperature for
aqueous HCl electrolysis. The temperature of decomposition
and chlorination is assumed to be same, where an additional
heat exchanger is required to cool down MgCl2 from chlorina-
tion step. Both hydrolysis and chlorination steps are
exothermic, and the only endothermic reaction is decomposi-
tion of MgOHCl. The chlorine gas from both electrolysis paths is
mixed and heated to the chlorination reaction temperature.
Thermodynamic assessment
The minimum electrical energy requirement of the electro-
chemical step of all MgeCl cycle options can be calculated
using the Nernst equation, depending on the change in the
Gibbs free energy as follows:
DG ¼ nFE0 ¼ Wel (1)
0
where F denotes Faraday's constant, E is the standard cell
potential and n is the number of moles of electrons exchanged
in the electrochemical reaction. DG for decomposition

Fig. 4 e Aspen flowsheet of the MgeCl-C cycle.

 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6 849

reactions of hydrogen chloride to hydrogen and chlorine is

191.67 kJ/mol, hence standard potential can be calculated as
1.01 V. This value for direct water decomposition at 298 K
and 1 bar is 1.23 V. However, voltage requirements are taken
from practical applications for better representation of system
efficiencies. The efficiency of the MgeCl cycle is simply the
ratio of useful energy from, and required energy inputs to the
cycle, and defined as follows:

n_H2 LHVH2
hen;MgCl ¼ (2)
Q_ heat þ W_ elec

_ H
Ex
hex;MgCl ¼ 2
(3)
_ heat þ W
Ex _ elec

where the specific exergy content of the hydrogen is sum of its

chemical exergy (238.49 kJ/mol) and its physical exergy.
Assuming hydrogen is extracted from the cycle at ambient
conditions, its physical exergy content is almost zero. How-
ever, the physical exergy contents are included in the results.
For the energy efficiency calculations lower heating value
(LHV) of hydrogen is used (242 kJ/mol) [20]. The thermal exergy
rate of the heat input is then considered as the exergy input to
the cycle as follows:
 
_ heat ¼ Q_ heat 1  T0
Ex (4)
Tmax

where the T0 and Tmax refer to ambient and maximum cycle

temperatures. The total heat requirement is calculated by
providing 85% of the heat from exothermic components to
endothermic components, except for isothermally driven re-
actions. The definition for the Q_ heat is the written as Fig. 5 e Reaction temperature and steam/magnesium ratio
influence on mole flow of (a) solid products (b) gaseous
Q_ heat ¼ Q_ end þ Q_ ext *0:85 (5) products.

where subscripts end and ext refer to endothermic and

exothermic, respectively.

product in a separate reactor can lead to a less steam requiring

Results and discussion
Reactor simulations
The expected products from the hydrolysis reactions are
MgOHCl at lower reactor temperatures and pure MgO at
higher temperatures. The reactor simulation of the hydrolysis
step is shown in Fig. 5a, and mole flow rates of the gaseous
products for varying steam/magnesium ratios are plotted in
Fig. 5b. These results are in good agreement with previously
defined experimental results [18]. The MgOHCl production
shows the highest rate between 523 and 573 K and starts
converting into pure MgO particles slowly at 649 K. Complete
reaction can be obtained at a steam/magnesium ratio of 1.9,
for the MgOHCl production.
Previously the steam/magnesium ratio of 3.1 is specified
for full conversion of MgCl2 as reported in a previous work of
the authors [10]. In order to react 1 kmol/s of MgCl2 with full
conversion, 3.1 kmol/s water is heated up to 810 K. MgOHCl
production would require less amount of steam than that of
direct MgO production. Thus, keeping the hydrolysis step at
lower temperatures to produce MgOHCl and decomposing this
cycle. It has also been imposed that reactivity of MgO with
chlorine is dependent on how the MgO is produced [18]. The
decomposed MgOHCl results in fine MgO with superior parti-
cle size, which significantly increases the reaction surface of
the solid particles.
The simulations of the decomposition step leads to pro-
duction of MgCl2 instead of MgO due to the reaction of pro-
duced HCl gas back with MgO particles. Thus, it is crucial to
remove the produced HCl gas on the surface of the particles
with an inert gas. High temperature reaction leads to a com-
plete decomposition of MgOHCl particles. Inert gas effect on
the rate of this reaction has been studied by [17], and rate of
the reaction has been doubled by removing HCl gas from the
solid particle surface. The use of argon gas resulted in
completion of the decomposition at 723 K. The effects of the
decomposition reactor temperature and the inert gas flow rate
on the performance are parametrically studied and repre-
sented in Fig. 6a and b. At a constant 723 K reactor tempera-
ture gas flow rate effect on MgO production is plotted in
Fig. 6a. Argon gas is selected as the inert gas which is also
favorable as a carrier gas for the electrolysis step. All of the
MgOHCl is converted into fine MgO at gas/solid ratios higher
than 1.7. Considering a constant gas flow rate and keeping it at
850 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6

1 kmol/s, reactor temperature effect is also plotted in Fig. 6b,

where a complete conversion of MgOHCl is accomplished at
480  C. Inert gas flow is a very important parameter for the
decomposition step to remove the HCl gas from the surface of
the MgO particles in order to prevent MgCl2 and steam for-
mation due to possible reaction of MgO and HCl at higher
temperatures. This variable is also a decision maker for the
maximum temperature selection of the cycle. Additional inert
gas compression energy and recovery of the inert gas from the
electrolysis step should be considered for the overall analysis
and assessment of this cycle, where it is included in perfor-
mance assessment as auxiliary work and heat.
The MgO produced from the hydrolysis reaction is reacted
with chlorine gas from the electrolysis step in the chlorination
reactor. This reaction is exothermic and spontaneous at
ambient temperature. Therefore, kinetics of this reaction Fig. 7 e MgCl2 production at varying chlorination reactor
should also be considered in order to determine the range of pressure and temperatures.
temperature for faster production of the desired substances.
The Gibbs minimization method is in good agreement with

the thermochemistry data of this reaction. Fig. 7 shows the

conversion of solid particles with the reaction temperature. It
is expected that the chlorine/magnesium ratio would be
required to completely convert the MgO particles into MgCl2,
however, production rate of these two agents are 1
(Cl2:MgO ¼ 1), and additional use of chlorine is not appro-
priate. As reported in the literature, chlorination reaction at
higher or elevated temperatures does not result in completion
of reaction, and hence, there has been several other additives
to complete the reaction with such as carbo-chlorination.
However, the reactions occur at very high temperatures
(>1300 K) with carbon containing substances [19]. Thus, this
reaction can be run at a fluidized bed by recirculating
unreacted MgO particles without additives.
The electrical energy required for the theoretical electrol-
ysis of HCl can save up to 18% electrical work compared to the
water electrolysis. However, practical aqueous HCl electrol-
ysis needs ~1.8 V at 5 kA/m2 current density, and it is a mature
technology. This amount can be reduced to 1.4 V by anhy-
drous HCl electrolysis process. The change in equilibrium cell
potential of the HCl electrolysis by HCl conversion ratio is
represented in Fig. 8. The cell voltage requirement increases

Fig. 6 e Effects of (a) Inert gas flow rate and (b)

decomposition reactor temperature on reactants and Fig. 8 e Equilibrium cell potential variation with
products. temperature and HCl conversion ratio.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6 851

up to 10% with full conversion, the changes are proportional
and slightly different at every 50 K increase. Electrolysis of the
aqueous HCl has some major issues in terms of solubility of
chlorine gas, and oxygen evolution. Thus, one should consider
generating pure HCl gas in order to make the MgeCl cycle a
feasible one. A more practical approach depicts that the
electrical work saving can be up to 13%, which is less than the
theoretical ratio. The four-step MgeCl cycle leads to produce
half of the HCl in anhydrous form with the decomposition
step, where the other half is still in mixture with steam. For
the sake of less work requirement, a separation process at
slightly elevated temperatures can be considered to increase
the amount of anhydrous HCl gas.
Performance assessment
The Aspen Plus results for the conventional MgeCl cycle are
summarized in Table 1a and b. The results are given for
1 kmol/s hydrogen production from the electrolysis step.
Excess water from electrolysis step is fed back to the hydro-
lysis reactor by mixing it with the external water. The main
inputs to the system are water, various grades of heat, and
electrical work for the electrolysis process. In the efficiency
assessment, energy content of water input and oxygen output
are not included due to their negligible contents compared to
heat and work requirements, and the main output of the
system is hydrogen at 298 K. Considering the excess steam
requirement of 17 for the conventional cycle (MgeCl-A), the
water boiling process for a large amount of water is the most
energy consuming process. However, the potential of recov-
ering this energy by internal heat recovery, and low grade heat
requirement of the water boiling process makes it easier to
provide the internally recovered energy. Table 1b represents
the energy requirement of the conventional MgeCl cycle.
The calculation of the energy need is based on internal
heat recovery from exothermic heat exchanging components
with a heat exchanger effectiveness of 85%. Here, the chlori-
nation reactor also shows an exothermic trend. The total heat
requirement of this cycle results in 310.85 MJ/kmol H2. Here, it
should also be noted that the heat requiring components at
high temperatures are only the hydrolysis reactor and the
chlorine heating exchanger. Even if Hex-1 shows the highest
heat requirement, this heat supply can be supplied around the
phase change temperature of water. The electrolysis power
requirement is calculated based on 1.8 V requirement and
results in 347.35 MW for 1 kmol/s H2 production.
The energy efficiency of the MgeCl-A cycle is calculated to be
36.8% where it is drastically lower than that of calculations made
theoretically under stoichiometric conditions. Note that the
previous studies reported up to 52% for the ideal cycle [10e12].

Table 1 e State point information (a) and energy requirement (b) of the MgeCl-A cycle.
(a)
State Stream n_ T h s exph extot _
Ex
kmol/s ( C) kJ/kmol kJ/kmolK kJ/kmol kJ/kmol MW
1 H2O 1 25.00 285,820 163.14 241,790 3120 3.09
2 H2O 17 67.36 282,640 153.15 238,610 3120 55.91
3 H2O 17 537.00 223,430 8.78 223,430 11,710 331.78
4 MgCl2 1 537.00 601120 88.55 840,490 151,860 168.92
5 Mix 19 537.00 226,700 3.81 e
6 H2O/HCl 18 537.00 207,180 0.51 e 388.63
7 H2O/HCl 18 70.00 259,290 128.19 289.47
8 Mix 18 70.00 249,290 129.34 e
9 H2/H2O 17 70.00 264,620 139.88
10 H2O 16 70.00 282,440 152.57 238,410 3120 53.05
11 H2 1 70.00 1302 4.17 240,698 238,490 238.55
12 H2 1 25.00 1 0.11 241,999 238,490 238.46
13 MgO 1 537.00 577,990 63.35 695,680 59,170 69.26
14 Cl2 1 70.00 1485 4.82 160,225 117,520 117.60
15 Cl2 1 537.00 18,473 35.94 143,237 117,520 125.31
16 Mix 1.5 537.00 395350 48.60 e
17 O2 0.5 537.00 16,173 31.29 16,173 3970 5.41
18 O2 0.5 25.00 8 0.09 8 3970 1.97
(b)
Component Process Temperature DH W_

C MJ/kmol H2 MW
Hydrolysis MgCl2 hydrolysis 537 92.21 e
Chlorination MgO Chlorination 537 33.5 e
Electrolysis HCl(aq) electrolysis 70 e 347.35
Hex-1 Water superheating 537 1006.4 e
Hex-2 HCl cooling 70 937.9 e
Hex-3 Cl2 heating 537 16.98 e
Hex-4 H2 cooling 25 1.3 e
Hex-5 O2 cooling 25 8.1 e
Total 310.85 347.35
852 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6

Use of excess water and practical voltage for the electrolysis cell
are main efficiency reducing factors. Exergy efficiency of the
cycle is also calculated by considering the exergy rate of state 12,
in Table 1a. Here, the exergy contents of water input and oxygen
output are also included in the calculations, where it is calcu-
lated as 44.1%. Compared to exergy efficiency values in the
previous studies for stoichiometry (68%), almost 35% reduction
is observed for the cycle for practical assumptions. Another
important issue to be pointed out is that the energy consumption
of this cycle is higher than that of conventional water electrol-
ysis. Since most thermo-electrochemical cycles are proposed to
be better performing and less power consuming alternatives
than the water electrolysis, MgCl-A cycle is not a feasible alter-
native considering practical applications.
The literature studies show that another configuration for
the MgeCl cycle can be simulated by using a low temperature
hydrolysis reaction. However, for the same amount of
hydrogen production, stoichiometry should be doubled, where
the required steam has to be doubled as well. Simulation re-
sults of the MgeCl-B cycle are represented in Table 2a and b.
The main difference between the option A and B are the hy-
drolysis step of the cycle. The hydrolysis reaction for option B
requires lower temperatures than that of direct MgCl2 hydro-
lysis as in option A, and this reaction is exothermic.
The chlorination reaction for MgOHCl chlorination is
highly endothermic, and results in steam production at high
temperature in mixture with oxygen. The separation process
of these substances might be very challenging and can pre-
vent the reuse of steam in the cycle. Thus, an additional
drying process for water separation from oxygen would be
required, in order to recirculate the water content after the
chlorination. For the reaction chain inside the reactor, one can
predict that a possible MgOHCl decomposition is initially
required to liberate HCl gas first. This may cause mixture of
HCl in the outlet stream. However, previous information for
reaction kinetics of MgO and HCl shows that this reaction is
more feasible and faster than MgO chlorination, with a very
good conversion of MgO into MgCl2 [21].
The energy calculations of the cycle shows relatively
higher energy load mainly due to energy intensive phase

Table 2 e State point information (a) and energy requirement (b) of the MgeCl-B cycle.
(a)
State Stream n_ T h s exph extot _
Ex

kmol/s C kJ/kmol kJ/kmolK kJ/kmol kJ/kmol MW
1 H2O 1 25 285,820 163.142 33.773 3086.227 3.086
2 H2O 22 67.8 282,600 153.04 175.890 3295.89 72.509
3 H2O 22 280 233,040 22.9991 2433.74 14143.74 311.162
4 MgCl2 2 280 621940 119.425 5438.501 157298.5 314.597
5 Mix 23 280 268,400 31.5061
6 H2O/HCl 21 280 218,920 15.5103 14322.06 300.763
7 H2O/HCl 21 70 262,630 131.78 13810.47 290.019
8 Mix 21 70 254,090 132.826
9 H2/H2O 20 70 267,290 141.779
10 H2O 19 70 282,440 152.569 195.502 3315.502 62.994
11 H2 1 70 1301.857 4.171019 57.3860 238547.4 238.547
12 H2 1 25 0.987862 0.107765 32.633 238457.4 238.457
13 Cl2 1 70 1484.522 4.822405 569247.4 628417.4 628.417
14 Cl2 1 537 18473.2 35.93761 7792.224 125312.2 125.312
15 MgOHCl 2 280 787,930 199.462 3209.706 148329.7 296.659
16 MgOHCl 2 537 776,180 182.004 9757.192 154877.2 309.754
17 Mix 3.5 537 405,020 46.3679
18 O2 0.5 537 16173.08 31.29298 16174.96 20144.96 10.072
19 O2 0.5 25 7.95734 0.088981 32.592 3937.408 1.968
20 Mix 3 537 475,220 61.9581
21 MgCl2 2 537 601,120 88.5468 17056.94 168916.9 337.833
22 H2O 1 537 223,430 8.7809 7806.704 19516.7 19.5167
(b)
Component Process Temperature DH W_

C MJ/kmol H2 MW
Hydrolysis MgCl2 hydrolysis 280 35.6 e
Chlorination MgOHCl Chlorination 537 116.3 e
Electrolysis HCl(aq) electrolysis 70 e 347.35
Hex-1 Water superheating 280 1040.8 e
Hex-2 HCl cooling 70 817.8 e
Hex-3 Cl2 heating 537 16.98 e
Hex-4 H2 cooling 25 1.3 e
Hex-5 O2 cooling 25 8.1 e
Hex-6 MgOHCl heating 537 23.5
Hex-7 MgCl2 cooling 280 41.7
Total 459.02 347.35
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6 853

change process for large amount of required water. The
endothermic reactor of this cycle also requires 20.5% more
heat than the hydrolysis reaction of MgeC-A at elevated
temperatures. Energy and exergy efficiency of this option is
30% and 37.7%, respectively. Considering the electrolysis
process, the practical voltage requirement is at the same
range of the MgeCl-A cycle and highly energy intensive.
Considering thermodynamic point of view, the MgeCl-B cycle
does not show any promising performance either for heat
requirement and power consumption. However, further
investigations can be made to enhance the cycle performance
and reduce the energy requirements by capturing HCl in dry
form, developing a less steam requiring hydrolysis reactor
configuration, and investigating the temperature dependence
and reaction characteristics of the MgOHCl chlorination.
The newly developed four-step MgeCl cycle is named as
MgeCl-C and the main idea behind the four-step option is to
capture HCl in dry form with an additional step to decompose
MgOHCl into solid and gas substances at elevated tempera-
tures. Table 3a and b represent the state point information

Table 3 e State point information (a) and energy requirement (b) of the MgeCl-C cycle.
(a)
State Stream n_ T h s exph extot _
Ex

kmol/s C kJ/kmol kJ/kmol K kJ/kmol kJ/kmol MW
1 H2O 1 25 285,820 163.142 33.773 3086.227 3.086
2 H2O 11 65.9 282,750 153.471 154.298 3274.298 36.017
3 H2O 11 280 233,040 22.9991 2433.744 14143.74 155.581
4 MgCl2 1 280 621,940 119.425 5438.501 157298.5 157.298
5 Mix 12 280 266,930 31.1178
6 H2O/HCl 11 280 219,570 15.8138 13762.5 151.387
7 H2O/HCl 11 70 263,540 132.758 13192 145.112
8 Mix 11 70 255,400 133.775
9 H2/H2O 10.5 70 268,010 142.293
10 H2O 10 70 282,440 152.569 195.502 3315.502 33.155
11 H2 0.5 70 1301.857 4.171019 57.386 238547.4 119.273
12 H2 0.5 25 0.987862 0.107765 32.633 238457.4 119.228
13 MgOHCl 1 280 787,930 199.462 3209.706 148329.7 148.329
14 MgOHCl 1 450 780,160 187.201 7326.017 152446 152.446
15 Mix 2 450 331,050 16.4064
16 HCl 1 450 79812.6 36.14458 4729.485 90679.49 90.679
17 HCl 1 70 91027.6 14.17917 60.197 86010.2 86.01
18 Mix 1 70 1406.357 10.28291
19 Cl2 0.5 70 1484.522 4.822405 75.877 117595.9 58.797
20 H2 0.5 70 1301.857 4.171019 57.386 238547.4 119.273
21 H2 0.5 25 0.987862 0.107765 32.633 238458.4 119.229
22 Cl2 0.5 70 1484.522 4.822405 75.877 117595.9 58.797
23 Cl2 1 70 1484.522 4.822405 75.877 117595.9 117.595
24 MgO 1 450 582,290 68.9573 7459.278 66629.28 66.629
25 Cl2 1 500 17097.69 34.19978 6934.588 124454.6 124.454
26 Mix 1.5 500 397,800 51.694
27 MgCl2 1 500 604,170 92.4003 15155.3 167015.3 167.015
28 O2 0.5 500 14926.99 29.71871 6072.696 10042.7 5.021
29 O2 0.5 25 7.95734 0.088981 32.592 3937.408 1.968
(b)
Component Process Temperature DH W_

C MJ/kmol H2 MW
Hydrolysis MgCl2 hydrolysis 280 17.82 e
Chlorination MgO Chlorination 500 31.51 e
Decomposition MgOHCl Decomposition 450 118.1 e
Electrolysis(aq) HCl(aq) electrolysis 70 e 173.67
Electrolysis(dry) HCl(dry) electrolysis 70 e 135.08
Hex-1 Water superheating 280 546.8 e
Hex-2 MgOHCl heating 450 7.8 e
Hex-3 HCl(aq) cooling 70 483.8 e
Hex-4 HCl(dry) cooling 70 11.2 e
Hex-5 H2 cooling 25 0.65 e
Hex-6 H2 cooling 25 0.65 e
Hex-7 Cl2 heating 500 15.61 e
Hex-8 O2 cooling 25 7.46 e
Hex-9 MgCl2 cooling 280 17.8 e
Total 244.98 308.75
854 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6

and energy requirements for the four-step cycle. This cycle

can be considered as a mixture of options A and B in terms of
reactions throughout the cycle. The hydrolysis reaction is
same as in the option B, and chlorination reaction is the same
as option A. The main differing part of the cycle is decompo-
sition of MgOHCl in a separate reactor. Electrolysis is made in
two different cells, one for aqueous and one for anhydrous
HCl, coming from different streams. Here the only endo-
thermic reaction is decomposition reaction with a relatively
high heat requirement (118.1 MJ/kmol H2). Reactivity of MgO
with chlorine has been previously reported to be low when it is
produced from direct hydrolysis of MgCl2. MgO from Mg(OH)2
and MgOHCl decomposition show significantly higher reac-
tivity than the previously mentioned chlorination method Fig. 10 e Efficiency comparison of the MgeCl cycle options.
[18]. Therefore, this cycle is also promising in terms of
achieving better reaction kinetics in the most challenging
process of the cycle. In order to validate this claim, a complete
experimental study to hydrolyze MgCl2 in to MgOHCl,
power requirement remains the lowest compared to practical
decompose MgOHCl into MgO and HCl, and chlorination of
applications of the cycle. Highest heat energy consumption
MgO can be conducted in a future work.
belongs to MgeCl-B cycle due to the double stoichiometry and
Since the required steam is lower than the direct hydro-
very high steam requirement. The energy requirement of the
lysis at elevated temperatures, energy requirement for water
four-step cycle is the lowest among all configurations with the
superheating is significantly lower than the previous options,
same internal heat recovery assumption. This energy
and recovery potential is also higher due to internal heat
requirement can be further decreased with a more efficient
exchanging. Using the same method for the previous options,
reactor configuration which would lead to a less steam
a total external heat requirement for this cycle is found to be
requirement to convert solid reactants into desired products.
244.98 MJ/kmol H2, which is 21.3% lower than option A, and
Lower steam requirement is also favorable for electrolysis of
almost 47% lower than that of option B. The electrical work
HCl at higher concentrations. However, optimum concentra-
requirement is also lower than previous two options due to
tion of HCl in the aqueous electrolysis is around 18e22%, and
lower voltage requirement of dry HCl capture. Electrical wok
it is a mature technology, which corresponds to a steam/HCl
consumption of this configuration is 11.2% lower than previ-
molar ratio of 10e11. If the electrolysis is desired to be made at
ous two options and it also shows lower voltage requirement
gaseous phase, higher concentration can be beneficial with
than that of water electrolysis by 3.1%, and calculated energy
higher utilization of HCl in more compact cells. However, this
and exergy efficiencies are 43.7% and 52%, respectively.
is not a well-known procedure and further studies should be
A simple comparison between all configurations and the
conducted.
theoretical calculation results can be made by matching heat
Energy and exergy efficiency comparisons show that the
and power loads of the cycles as well as their efficiencies. Fig. 9
four-step option is also promising in terms of system perfor-
shows the energy load comparison of all cycles. The theoret-
mances which also has the potential to be further enhanced
ical cycle is calculated using the Aspen results for required
by further studies as represented in Fig. 10. Since the power
excess steam and theoretical voltage requirement for the HCl
consumption of this cycle is lower than direct water elec-
electrolysis as in Ref. [10]. Previous study has been conducted
trolysis, further studies should be conducted for various
to optimize the maximum cycle temperature by manipulating
the excess steam rate into hydrolysis reactor. The heat and

Fig. 9 e Energy load comparison of MgeCl cycle options (th: Fig. 11 e Effect of cycle maximum temperature and
theoretical calculations with Aspen yields). ambient temperature on exergy efficiency of cycles.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6
configurations of the four-step cycle. The four-step cycle
carries the potential to be more than 55% efficient with a
complete HCl electrolysis at the desired voltage values. Since
one of the major energy consuming devices is the aqueous
electrolysis step, a suitable dry HCl capturing process can be a
promising option for reduced electrical work consumption of
the cycle. The HCl capturing process can be another energy
intensive process which may contribute to the total thermal
energy requirement of the cycle.
Exergy efficiency values for all configurations are calcu-
lated based on thermal exergy input from total heat input.
Therefore, the heat exchanger effectiveness and the
maximum cycle temperature are two main parameters influ-
encing cycle efficiencies. The thermal exergy input to the cy-
cles is strongly influenced by the sink and source
temperatures. The higher reference and the lower maximum
temperatures result in better exergetic efficiency values.
However, manipulation of both parameters is not suggested
due to their uncontrollable nature. Chemical reactions occur
at a specific temperature range for the best possible yield, and
changing this parameter would only vary the thermal exergy
input, but changes in energy content of the streams cannot be
kept constant. Variation of these parameters and their effect
on exergy efficiency is represented in Fig. 11.
Since all calculations are made based on a heat exchanger
effectiveness assumption, effect of this parameter is also
represented in Fig. 12. As an expected outcome, increase of
this factor has a significant effect on cycle efficiencies. In a
perfect case with full internal heat recovery within all cycles,
up to 36% increase in energy efficiency and 25% increase in
exergy efficiency can be observed. MgeCl-B cycle is the one
effected the most from heat exchanger effectiveness factor,
which is due to relatively higher heat requirement of this
configuration.
Conclusion
The three configurations of the MgeCl cycle are comparatively
studied by considering the individual reaction behaviors and
practical electrical energy and steam requirements. The direct
electrolysis of aqueous HCl results in a higher electrical work
Fig. 12 e Effect of heat exchanger effectiveness on cycles
efficiencies.
855
requirement than the conventional water electrolysis for the
same current density. Therefore, a four-step MgeCl cycle is
proposed to generate dry HCl gas for the sake of lower elec-
trical work requirement. The four-step cycle can work at 723 K
maximum temperature with ~3% lower electrical energy
consumption than the water electrolysis. The present studies
show a separation process can enhance this saving up to 13%.
The present four-step cycle shows 43.7% energy efficiency and
up to 52% exergy efficiency. A small scale experiment is under
study in Clean Energy Research Lab (CERL) of UOIT to capture
HCl in dry form, and with a successful HCl capture, the four-
step MgeCl cycle can be a good candidate amongst other
hybrid thermochemical cycles with feasible reactions
throughout the cycle and mature electrolysis technology.
Nomenclature
ex specific exergy, kJ kg1
ex molar specific exergy, kJ mol1
F Faraday's constant, 96,485 C mol1
G Gibbs free energy, kJ kmol1
h specific enthalpy, kJ kg1
h molar specific enthalpy, kJ mol1
Ke equilibrium constant
LHV lower heating value, kJ mol1
N number of moles, mole, kmol
R universal gas constant, kJ mol1 K1
s specific entropy, kJ kg1 K1
s molar specific entropy, kJ mol1 K1
Greek letters
h efficiency ()
Superscripts
0 reference
Subscripts
end endothermic
eq equilibrium
ext exothermic
F formation
references
[1] Veziroglu TN, Barbir F. Hydrogen: the wonder fuel. Int J
Hydrogen Energy 1992;17:391e404.
[2] Dincer I, Zamfirescu C. Sustainable hydrogen production
options and role of AIHE. Int J Hydrogen Energy
2012;37:16266e86.
[3] Naterer GF, Dincer I, Zamfirescu C. Hydrogen production
from nuclear energy. London: Springer; 2013.
[4] Petri MC, Yildiz B, Klickman AE. US work on technical and
economic aspects of electrolytic, thermochemical, and
hybrid processes for hydrogen production at temperatures
below 550 C. Int J Nucl Hydrogen Prod Appl 2006;1(1):79e91.
[5] Yan XL, Hino R. Nuclear hydrogen production handbook.
Boca Raton FL: CRC Press; 2011. p. 50e4.
[6] Lewis MA, Masin JG, O'Hare PA. Evaluation of alternative
thermochemical cycles. Part I: the methodology. Int J
Hydrogen Energy 2009;34:4115e24.
[7] Simpson MF, Hermann SD, Boyle BD. A hybrid
thermochemical electrolytic process for hydrogen
856 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 8 4 5 e8 5 6
production based on the reverse Deacon reaction. Int J
Hydrogen Energy 2006;31:1241e6.
[8] Gooding CH. Analysis of alternative flow sheets for the
hybrid chlorine cycle. Int J Hydrogen Energy 2009;34:4168e78.
[9] Hesson RN. Kinetics of the chlorination of magnesium oxide.
1979. No. IS-T-823. Ames Lab., IA (USA).
[10] Ozcan H, Dincer I. Performance investigation of
magnesiumechloride hybrid thermochemical cycle for
hydrogen production. Int J Hydrogen Energy 2014;39:76e85.
[11] Ozcan H, Dincer I. Energy and exergy analyses of a solar
driven MgeCl hybrid thermochemical cycle for co-
production of power and hydrogen. Int J Hydrogen Energy
2014;39:15330e41.
[12] Ozcan H, Dincer I. Thermodynamic and environmental
impact assessment of steam methane reforming and
magnesiumechlorine cycle-based multigeneration systems.
Int J Energy Res 2015;39:1778e89.
[13] Eames DJ, Newman J. Electrochemical conversion of
anhydrous HCl to Cl2 using a solid-polymer-electrolyte
electrolysis cell. J Electrochem Soc 1995;142:3619e25.
[14] Motupally S, Becker AJ, Weidner JW. Water transport in
polymer electrolyte membrane electrolyzers used to recycle
anhydrous HCl I. Characterization of diffusion and electro-
osmotic drag. J Electrochem Soc 2002;149:D63e71.
[15] Bartling J, Winnick J. Chlorine recovery from anhydrous
hydrogen chloride in a molten salt electrolyte membrane
cell. J Electrochem Soc 2003;150:D99e107.
[16] Sivasubramanian P, Ramasamy RP, Freire FJ, Holland CE,
Weidner JW. Electrochemical hydrogen production from
thermochemical cycles using a proton exchange membrane
electrolyzer. Int J Hydrogen Energy 2007;32:463e8.
[17] Carmo M, Fritz DL, Mergel J, Stolten D. A comprehensive
review on PEM water electrolysis. Int J Hydrogen Energy
2013;38:4901e34.
[18] Kashani-Nejad S, Ng K-W, Harris R. MgOHCl thermal
decomposition kinetics. Metall Mater Trans B 2005;36:153e7.
[19] Kipouros GJ, Sadoway DR. A thermochemical analysis of the
production of anhydrous MgCl2. J Light Metals 2001;1:111e7.
[20] Kotas TJ. The exergy method of thermal plant analysis.
Elsevier; 2013.
[21] Kashani-Nejad S, Ng K-W, Harris R. Preparation of MgOHCl
by controlled dehydration of MgCl2$6H2O. Metall Mater Trans
B 2004;35:405e6.