X-ray diffraction study of multiphase reverse reaction with molten CuCl and oxygen

Article abstract of DOI:10.1016/j.tca.2011.07.001

The thermochemical copper-chlorine (Cu-Cl) cycle for hydrogen production includes three chemical reactions of hydrolysis, decomposition and electrolysis. The decomposition of copper oxychloride for oxygen production establishes the high-temperature limit of the cycle. At 450-530 °C, copper oxychloride (Cu₂OCl₂) decomposes to produce a molten salt of cuprous chloride (CuCl, copper I chloride) and oxygen gas. Minimization of the reverse reaction and undesirable products is critical for the proper operation of the Cu-Cl cycle. This paper examines the operating conditions that disfavor the reverse reaction of the oxygen production, and the parameters that maximize the extent of the forward reaction. Experiments were designed to disperse oxygen gas into a molten CuCl bath to study its reaction at 450-500 °C. The composition of the products was quantified with X-ray diffraction measurements. Experimental results indicate that a high decomposition extent of copper oxychloride is obtained at equilibrium when the temperature is higher than 500 °C, and the oxygen pressure is below 2 bar. The thermochemistry data of the reactants and products were also determined and reported. These thermodynamic data provide a key missing gap in the understanding of the Cu-Cl cycle of thermochemical hydrogen production. The data includes the standard formation entropy, enthalpy and Gibbs free energy at different temperatures. Also, in this paper, a thermodynamic analysis is performed to investigate the reverse reaction from the aspects of spontaneity and optimization of the operating parameters. It is found that the optimal operating parameters for minimizing the reverse reaction lie in the pressure range of 1-2 bar and a temperature range of 500-525 °C.

Full text of DOI:10.1016/j.tca.2011.07.001

Thermochimica Acta 524 (2011) 109–116
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
X-ray diffraction study of multiphase reverse reaction with molten CuCl and
oxygen
∗
G.D. Marin, Z. Wang, G.F. Naterer , K. Gabriel
Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario, Canada L1H 7K4
a r t i c l e i n f o
a b s t r a c t
Article history:
The thermochemical copper–chlorine (Cu–Cl) cycle for hydrogen production includes three chemical
reactions of hydrolysis, decomposition and electrolysis. The decomposition of copper oxychloride for
oxygen production establishes the high-temperature limit of the cycle. At 450–530 C, copper oxychloride
Received 8 March 2011
Received in revised form 30 June 2011
Accepted 1 July 2011
◦
(Cu2OCl2) decomposes to produce a molten salt of cuprous chloride (CuCl, copper I chloride) and oxygen
Available online 7 July 2011
gas. Minimization of the reverse reaction and undesirable products is critical for the proper operation of
the Cu–Cl cycle. This paper examines the operating conditions that disfavor the reverse reaction of the
oxygen production, and the parameters that maximize the extent of the forward reaction. Experiments
Keywords:
Hydrogen production
Reverse reaction
Molten salt
◦
were designed to disperse oxygen gas into a molten CuCl bath to study its reaction at 450–500 C. The
composition of the products was quantified with X-ray diffraction measurements. Experimental results
indicate that a high decomposition extent of copper oxychloride is obtained at equilibrium when the
X-ray diffraction
◦
temperature is higher than 500 C, and the oxygen pressure is below 2 bar. The thermochemistry data
of the reactants and products were also determined and reported. These thermodynamic data provide a
key missing gap in the understanding of the Cu–Cl cycle of thermochemical hydrogen production. The
data includes the standard formation entropy, enthalpy and Gibbs free energy at different temperatures.
Also, in this paper, a thermodynamic analysis is performed to investigate the reverse reaction from the
aspects of spontaneity and optimization of the operating parameters. It is found that the optimal operating
parameters for minimizing the reverse reaction lie in the pressure range of 1–2 bar and a temperature
◦
range of 500–525 C.
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
search in the 1970s and early 1980s resulted in the identification
of several promising thermochemical cycles including S–I, Cu–Cl,
Cu–SO , and Zn–SO cycles. Carty et al. [2] performed an analysis of
Hydrogen is a promising future clean fuel and also a necessity
4
4
for numerous existing industries such as petrochemicals, fertiliz-
ers and increasingly as a future transportation fuel. Conventional
hydrogen production technologies require the utilization of pri-
mary energy sources in the form of electricity, natural gas, coal, oil,
biomass or others that emit greenhouse gases. The predominant
existing process is steam methane reforming (SMR) [1].
Researchers have endeavored to develop alternate hydrogen
production methods that can split water by using clean energy
sources, and consequently lower greenhouse gas emissions. Ther-
mochemical hydrogen production is a technology that produces
hydrogen through a cycle of chemical and physical processes for
the overall splitting of water. McQuillan et al. [1] surveyed and
compared over 200 cycles for thermochemical hydrogen produc-
tion based on stoichiometric and energy requirements. A literature
processes of hydrogen production based on the criteria of reactions
whose free energy change lies within 10 kcal for a given tempera-
ture. Lewis and Taylor [3] applied these criteria, in addition to the
advantages of relatively low temperature (≤600 C) operation, high
◦
efficiency, and materials requirements, to recommend six cycles of
hydrogen production. Among the most promising cycles, the Cu–Cl
cycle has the lowest temperature requirement that can be supplied
by the temperatures of SCWR (Supercritical Water-cooled Reactor)
and current solar thermal plant technology. Furthermore, due to
the lower temperature requirement of the Cu–Cl cycle, its indus-
trialization would be anticipated to have fewer material challenges
than other higher temperature cycles [4,5].
The processes of the Cu–Cl cycle are shown in Fig. 1. The three
reactions and their key parameters are described in Table 1. The
oxygen production reaction involves the decomposition of Cu OCl2
2
to produce molten CuCl and oxygen gas in a forward reaction. Past
studies by Zamfirescu et al. [6], Daggupati et al. [7] and Haseli
et al. [8] have reported the characteristics of various copper chlo-
∗ Corresponding author. Tel.: +1 905 721 8668; fax: +1 905 721 3370.
E-mail address: greg.naterer@uoit.ca (G.F. Naterer).
0
040-6031/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2011.07.001

1
10
G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
sirable products is particularly critical to the integration of the
Cu–Cl cycle.
Studies on this potential reverse reaction of the oxygen produc-
tion process have not been reported in past literature. Although the
reaction of CuCl and oxygen to produce Cu OCl has been reported
Nomenclature
B
bias
Cp
d
specific heat capacity at constant pressure (J/mol K)
distance of atomic layers in the crystal (m)
Gibbs energy of species i (kJ/mol)
standard free Gibbs energy of reaction at tempera-
2
2
previously for the synthesis of a catalyst (CuCl +1/2O = Cu OCl ), it
2
2
2
Gi
was a theoretical study on the reaction mechanism and not verified
by experiments [9]. Copper oxychloride is not commercially avail-
able and it exists in nature in very few locations as a mineral called
melanothallite. In past reported studies, it was prepared in a labora-
tory through a lengthy process that utilized an equimolar mixture
of copper oxide (CuO) and cupric chloride (CuCl2 or copper II chlo-
T
ꢀf
G
ture T (J/mol)
H
enthalpy (J)
ꢀf
K
ko
m˙
H
enthalpy of formation (J/mol)
equilibrium constant
pre-exponential factor
molar rate (mol/s)
◦
ride) in a quartz tube vacuum heated for several days at 400 C to
synthesize Cu OCl2 [10]. The synthesis was performed by utilizing
2
ni
n
P
moles of species i (mol)
integer, multiplier of wavelengths
pressure (bar)
CuCl2 and CuO rather than CuCl and O . However, an experimen-
2
tal verification of the reverse reaction of the oxygen production
process has not been reported in the past. The extent of decompo-
Pm
precision
sition of Cu OCl2 was reported for a small sample of 0.01–0.02 mol
2
p
partial pressure of oxygen (bar)
reaction quotient
universal gas constant (J/mol K)
entropy (J/mol K)
◦
O2
of Cu OCl2 that was heated to 530 C [11,12]. The results indicated
2
Q
R
S
that the oxygen yield can exceed 85%.
A two-stage reaction mechanistic analysis assumed that the
overall oxygen generation reaction was limited by the chemical
Ti
initial temperature of reactants (K)
target temperature (K)
reactor temperature (K)
overall uncertainty
bond splitting of CuCl2 of the Cu OCl2 crystal [12]. But the reverse
2
Tsp
Tr
U
reaction was not measured because the sample amounts in the
experiments were so small as to be unable to measure and observe
the reverse reaction. This paper will present an experimental study
of the reverse reaction from the perspectives of optimization of
operating conditions and reaction extent, so as to minimize the
reverse reaction and undesirable products. The thermochemical
properties such as the standard formation entropy, enthalpy and
Chemical compounds
CuCl
cuprous chloride
CuCl2
CuO
cupric chloride
copper oxide
Gibbs free energy at different temperatures for Cu OCl2 and CuCl
2
CuOHCl copper hydroxide chloride
Cu OCl copper oxychloride
will be determined for the oxygen production process of the Cu–Cl
cycle. Then a thermodynamic analysis will be performed to study
the reverse reaction from the aspects of spontaneity and optimiza-
tion of the operating parameters.
2
2
HCl
hydrogen chloride
Greek symbols
ꢁ
ꢂ
wavelength of incident X-rays
angle of diffraction
2. Experimental configuration
The experimental loop for the reverse reaction is shown in Fig. 2.
The reactor contains a crucible that can withstand a very high tem-
perature and it is chemically inert to both solid and molten CuCl.
The crucible contains the sample and holds the probes for air injec-
tion and temperature measurement. It is located inside of a heating
chamber surrounded by the top heating zone of the furnace, a verti-
2 (g)
H O
Hydrolysis
o
H
2
O
(l)
HCl(g)
4
00 C
Heat Input from
Solar, Nuclear, or
Industrial Source
cal split tubular furnace consisting of three separate heating zones
that can provide a stable temperature in the range of 50–1000 ◦C
O
2(g)
Cu
2 2
OCl
by adjusting the input power to each zone. The inside diameter
of the crucible varies from 5 to 20 cm so as to study the influence
of the CuCl quantity and mixing. The heating chamber also pro-
vides secondary containment and it is used to purge the reaction
environment from humidity and oxygen by using nitrogen. The
entire assembly is placed inside the furnace. Nitrogen is injected
and allowed to flow in order to obtain the desirable oxygen fraction
for the reverse reaction.
Heat
Recovery
Decomposition
o
4
50 -550 C
CuCl(l)
O
2
CuCl(aq)
HCl(aq
The bottom temperature of the molten CuCl in the reactor was
measured by type K thermocouples with a 4–20 mA transmitter
Electrolysis
Dryer
H
2
o
CuCl2(aq)
CuCl2(s)
8
0 C
H
2
O
(l)
◦
programmed for 30–1000 C. The uncertainty of the output is ± 0.1%
◦
of the span that is equivalent to ± 0.97 C. A mixture of oxygen and
Fig. 1. Schematic of the thermochemical copper–chlorine cycle.
nitrogen is utilized as the oxygen source for the reactant of the
reverse reaction. In the experiments, solid CuCl is fed into the reac-
tor and heated to a molten state, and then air is bubbled through the
molten CuCl at various temperatures. The gases exiting the heat-
ing chamber are cooled down to ambient temperature via a heat
exchanger. The gases are then sampled continuously with an oxy-
gen analyzer for the instantaneous oxygen concentration analysis.
ride compounds in the Cu–Cl cycle, including molten and vapor
CuCl. There may exist a reverse reaction in the decomposition of
Cu OCl , for example, with molten CuCl reacting with oxygen to
2
2
produce Cu OCl . Minimization of this reverse reaction and unde-
2
2

G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
111
Table 1
Reactions in the thermochemical Cu–Cl cycle.
◦
Reaction
Process
T ( C)
Enthalpy change (kJ/mol)
CuCl(aq) + HCl(aq) → CuCl2 (aq) + (1/2) H2(g)
Electrolysis (hydrogen production)
Hydrolysis (water splitting)
Decomposition (oxygen production)
80–100
340–400
450–550
93.7 electrolytic
116.7 endothermic
129.3 endothermic
2
CuCl2(s) + H2O(g) → Cu2OCl2(s) + 2HCl(g)
Cu2OCl2(s) ↔ (1/2) O2(g) + 2CuCl(molten)
Sum: H2O (g) = H2(g) + (1/2) O2(g)
Fig. 2. Experimental configuration for the reverse reaction of molten CuCl and O2.
The oxygen analyzer, AMI-201RSP, has four volumetric mea-
surement spans of 0–1%, 0–5%, 0–10% and 0–100%. The uncertainty
is ± 1% of the selected span, or equivalent to ± 0.01% of the oxygen
volumetric fraction when the span is set at 0–1%. The concentra-
tion of oxygen from the reactor is also measured with the oxygen
analyzer before the experiments, allowing for measurement of the
oxygen concentration decrease that is caused by its reaction with
molten CuCl in the experiments. The total quantity of reacted oxy-
gen can be determined from integration of the concentration curve
and the flow rates of air and nitrogen. The flow rates of air and
nitrogen are controlled continuously, and the values are monitored
in a real-time mode. The flow controllers, Omega FVL2600, have
an uncertainty of ± 0.8% of the full scale. Three flow controllers
are utilized in order to approach various flow rates in the range
of 0–100 sccm, 0–2 slm, and 0–50 slm. The oxygen quantity was
calculated by assuming ideal behavior as follows:
sensitivity of coefficients based on the Kline and McClintock equa-
tion [13] for error propagation:
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
∂
m˙
∂m˙
∂T
∂m˙
∂A
∂m˙
2
m˙
2
2
T
2
AO
2
FN
P
=
P
+
P
+
P
+
P
(2)
p
∂p
2
∂FN2
2
O2
and
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
ꢂ
ꢃ
2
2
2
2
∂
m˙
∂m˙
∂T
∂m˙
∂AO2
∂m˙
∂FN2
2
m˙
2
2
T
2
A
2
N
B
=
B
+
B
+
B
+
B
p
F
O
∂p
2
2
(
3)
The overall uncertainty in the determination of the total oxygen
evolution is expressed by:
1
2
/2
2
U = (P + B )
(4)
m˙
m˙
ꢀ
ꢁ
where the subscripts p, T, AO2 , FN2 of precision (P) and bias (B)
represent pressure, temperature, percent analyzer reading and
flow meter, respectively; and U represents the overall uncertainty.
Uncertainty estimates were taken over a representative period of
the experiments. Experiments were repeated under similar condi-
tions of initial mass and molar ratios, initial weight, reactor pressure
and purge gas flow. A steady set of data was selected over a period
for the uncertainty calculations. The measured uncertainty of moles
of percent volume oxygen reading was determined as 1.87%, based
on precision (random) contributions to the uncertainty of m˙ , Pm
and a bias contribution to the uncertainty of m˙ , Bm. The fit of XRD
1
2.4
P T
1
2
m˙ O2 = 2
(pO2 )(FN2 )
(1)
P T
2
1
In this equation, P , T1 and P , T2 represent the pres-
1
2
sure/temperature at the inlet and outlet conditions, respectively.
Also, pO2 , FN2 and m˙ O2 represent the partial pressure of oxygen,
flow of nitrogen and mass flow of oxygen, respectively. The uncer-
tainty of the total mol evolution was determined by following a
combination of a precision (random) contribution to the uncer-
tainty of m˙ , P , and a bias contribution to the uncertainty of m˙ , B .
m˙
m˙
The two contributions can be evaluated separately in terms of the

1
12
G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
data compared to data from the International Centre for Diffraction
Data and inaccuracies introduced by the Rietveld method result in
a weighted R-factor of approximately 3% in the phase determina-
tion. The composition of the reverse reaction product is analyzed
by an X-ray diffraction (XRD) technique after solidification of CuCl.
The Bragg’s Law of XRD diffraction is represented as follows:
3. Thermodynamic analysis of the reverse reaction
The reverse reaction in the decomposition of copper oxychloride
occurs as follows:
2CuCl(l) + (1/2)O (g) = Cu OCl (s)
(6)
2
2
2
The spontaneity of the reaction in Eq. (6) can be predicted from
nꢁ = 2d sin(ꢂ)
(5)
T
the change of the Gibbs free energy of the reaction, ꢀ G. Any devi-
r
ation from standard conditions can influence the behavior of the
The equation describes how the cleavage face of crystals reflects
X-ray beams at certain angles of incidence. When the angle of
incidence ꢂ equals the angle of scattering, and the path-length dif-
ference equals an integer number of wavelengths n, then peaks of
scattered intensity can be observed [14–16]. In the equation, d rep-
resents the distance between atomic layers in the crystal. Lambda,
T
reaction. The reaction Gibbs free energy ꢀ G can be calculated as
r
follows:
ꢀ Tr G = ꢀ G + RT ln(Qr)
T
◦
(7)
r
T
◦
where ꢀ G is the change of Gibbs free energy of the reaction at
r
ꢁ
, is the known wavelength of incidence of the X-ray beams and n
standard pressure and temperature, and Qr is the reaction quo-
tient defined by the stoichiometric ratio of the product and reactant
is an integer.
Bragg’s Law conditions occur at a different d-spacing in materi-
als with a polycrystalline structure. A pattern can then be obtained
by plotting the angular positions and intensities of the resultant
diffracted peaks of radiation. The diffracted peaks are a character-
istic that depends exclusively on the crystalline structure of the
sample. If the sample contains a mixture of different phases, the
pattern shows the reflections of individual phase characteristics at
different angular positions. Powder diffraction XRD was applied to
a 1 mm thick sample in a 2 cm2 bed in order to characterize the
crystallographic structure. The measurements used a Philips PW
fugacity. Since CuCl and Cu OCl2 in Eq. (6) exist in the form of a
condensed state, the reaction quotient can be simplified by:
2
1
Qr =
(8)
1/2
o
(
P /P )
O
ꢂ
where Qr represents the reaction quotient, P is the standard pres-
sure (100 kPa) and PO2 is the partial pressure of O2 under actual
reaction conditions. When the reaction reaches equilibrium, the
quotient is equal to the reaction equilibrium constant K as defined
by:
1
830HT generator with a PW1050 goniometer, PW3710 control,
X-pert software (2007 version; International Centre for Diffraction
Data) and Rietveld refinement software. In the XRD analysis of the
T
r
◦
CuCl samples, the angle 2-ꢂ was allowed to vary within a range of
ꢀ G = −RT ln K
(9)
◦
5
–60 .
Two purity grades of CuCl were used for the reactant, i.e., analyt-
and K is then determined by:
ical grade for thermodynamic analysis, and industrial (assay 99%)
for simulating the effects of trace amounts of impurities on the
reverse reaction. Impurities such as moisture originally present in
the reactants may promote the formation of hydroxychlorides. The
pale green color of the CuCl reactants indicated the possible pres-
ence of impurities. The grade of reactants was expected to better
reflect conditions during regular operation and it was preferred
for sets of experiments over an analytical grade of CuCl. The results
indicated that the presence of hydroxychlorides was out of the XRD
composition analysis detection range.
1
K = Q = (
(10)
PO2 /Po)1/2
T
ꢂ
Here ꢀ G in Eq. (7) can be calculated from the molar Gibbs free
energy of formation of the product and reactants:
r
ꢀ Tr G = ꢀ G
◦
T
f
◦
− [2ꢀ G
T
◦
+ 1/2ꢀ G (O2)]
T
f
◦
(11)
(Cu2OCl2)
f
(CuCl)
T
◦
T
f
◦
and ꢀ G
(CuCl)
The forward reaction for oxygen production is endothermic, so
its reverse reaction is exothermic. However, this does not imply
that no external heat is required for the reactor because the reac-
where ꢀ G
are the molar Gibbs free energy
(
Cu2OCl2
)
f
of formation for CuCl and Cu OCl , respectively. The Gibbs free
energy of formation of each reactant and product can be calculated
from their enthalpy and entropy of formation:
2
2
tants, i.e., CuCl and O , must be heated to the reaction temperature.
2
Heat losses to the environment must be offset to maintain accu-
racy of the reaction temperature. Moreover, if the extent of the
reverse reaction is not significant, the released heat from the
reaction may not be sufficient to sustain the reaction tempera-
ture. In the experiments in this paper, the heating profiles of the
reactor were obtained and studied to characterize the reaction
temperature.
In the experiments, molten CuCl in the reactor is operated in a
batch mode where no species are fed or removed during the time of
the experiment. By comparison, oxygen flows continuously as bub-
bles through the molten CuCl. Due to the large density difference
of molten CuCl and oxygen, the species would remain separated
from each other if oxygen and CuCl are confined within the reac-
tor with no feeding/removal during the experiment. The feeding
and removal rates of oxygen are balanced, so the oxygen quantity
in the reactor is kept at a constant value. If there is no product
observed within the period of the experiment, the reverse reaction
is insignificant.
ꢀ Tf G = ꢀ H − Tꢀ S
◦
T
◦
T
◦
(12)
f
f
T
f
◦
T
f
◦
where ꢀ H and ꢀ S are the standard enthalpy and entropy of
T
f
◦
formation at temperature T. Unfortunately few data of ꢀ H and
T
f
◦
ꢀ S have been reported for CuCl and Cu OCl in the temperature
2
2
◦
range of 400–550 C. CuCl exists in the solid phase up to the
temperature of 412 C. At this temperature, it remains solid with
a different microstructure and thermophysical properties. This
phase is identifiable between the temperatures of 412 and 423 C.
Beyond this temperature, CuCl remains liquid in the range of
temperatures of interest with significant vaporization beyond the
temperature of 550 C. The thermophysical properties of Cu OCl
are calculated from the properties of compounds and elements.
The procedure utilizes known data of enthalpy and entropy at
ambient temperature. For the product Cu OCl2 in Eq. (6), the
◦
◦
◦
2
2
2
following schematic of a thermodynamic method is used for the

G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
113
T
f
◦
calculation of ꢀ H :
2
5
o
Δ H
f
o
o
o
o
2
Cu (s, 25 C) + ½ O
2
(g, 25 C) + Cl
2
(g, 25 C)
Cu
2
OCl
2
(s, 25 C)
==
T
o
T
o
T
o
T
o
ΔH (Cu)
ΔH (O ) ΔH (Cl )
ΔH (Cu OCl )
2
25
2
2
2
25
25
25
T
o
Δ H
f
2
Cu (s, T)
+
½ O
2
(g, T)
+
Cl
2
(g, T)
2 2
Cu OCl (s, T)
=
=
T
f
ꢂ
The method also applies to the calculation of ꢀ S , then
ꢀfT
H
◦
= [ꢀf H + ꢀH (Cu OCl ) ] − [ꢀH (Cu)
25
◦
◦
T
◦
T
0
200
400
600
2
2
25
25
0
20
40
0
◦
T
25
◦
T
25
+
ꢀH (O ) + ꢀH (Cl )
]
(13)
(14)
2
2
-20
-
-
ΔSo reaction
ΔHo reaction
-
40
-60
80
-100
120
ꢀfT
S
◦
= [ꢀf S + ꢀS (Cu OCl ) ] − [ꢀS (Cu)
25
◦
◦
T
◦
T
25
-60
80
2
2
25
-
-
◦
T
25
◦
T
+
ꢀS (O ) + ꢀS (Cl )
]
2
2
25
-
-
100
120
-
where the enthalpy and entropy changes for each component are
calculated based on the following equations:
-140
-160
-180
-200
ꢄ
-140
-160
T
T
◦
ꢀ
5H
=
CPdT
(15)
(16)
2
2
5
-180
ꢄ
T
Temperature, ºC
CPdT
T
T
◦
ꢀ
5S
=
2
T
ꢂ
T
ꢂ
2
5
Fig. 3. Standard enthalpy (ꢀ H ) and entropy (ꢀ S ) of the reverse reaction at
r r
various temperatures.
A similar methodology is adopted for the reactants CuCl and O
in Eq. (6) for the calculation of ꢀ H and ꢀ S . The transition of
2
T
f
◦
T
f
◦
CuCl, correlations of CP and the values of ꢀ25H for Cu, O , Cl and
◦
10
0
2
2
f
CuCl reported by NIST [17] are adopted in this paper. The reactant
CuCl experiences a phase change from solid to liquid at the temper-
ature of 430 C (803.15 K). The entropy change ꢀS at the melting
-10
◦
-20
30
40
-
-
point is calculated with the following equation:
◦
ꢃH
◦
melting
-50
60
ꢀmelting
S
=
(17)
8
03.15
-
where ꢀH represents the enthalpy of phase change. For Cu OCl ,
-70
-80
2
2
only the experimental data of specific heat capacity at constant
◦
pressure CP in the temperature range of 25–410 C was reported
18]. In this paper, the following equation is proposed to correlate
0
200
400
600
[
Temperature, ºC
the data:
T
r
ꢂ
Fig. 4. Standard Gibbs free energy (ꢀ G ) of the reverse reaction at various tem-
CP(Cu OCl ) = 99.23243 + 0.021622 T
(18)
peratures.
2
2
where CP and T have units of J/mol K and K, respectively. The rel-
ative standard deviation of Eq. (18) in the temperature range of
As indicated in Fig. 4, the transitional point from negative to
◦
T
◦
◦
2
5–410 C is within 3%, so it will be assumed to be extended to the
positive Gibbs free energy ꢀ G occurs close to 520 C, which indi-
r
◦
temperature of 550 C.
cates that the reverse reaction is spontaneous if the temperature
The thermodynamic properties for the formation of Cu OCl
is below 520 ◦C at standard pressure, or the reaction will not occur
2
2
T
◦
T
◦
T
◦
and CuCl, i.e., values of ꢀ H , ꢀ S and ꢀ G , calculated from Eqs.
when the temperature is higher than 520 ◦C at standard pressure.
f
f
f
(
11), (13) and (14), follow the thermodynamic method described
Fig. 5 shows the calculation results for the operating pressure
−
10
above (shown in Table 2). Based on these thermodynamic prop-
erties, the changes of standard enthalpy ꢀ H , entropy ꢀ S , and
Gibbs free energy ꢀ G for the reaction in Eq. (6) can be calculated,
as shown in Figs. 3–5. The negative values of the reaction enthalpy
in the temperature range of reactions as shown in Fig. 3 indicate
range of 1.0 × 10
to 100 bar to represent the range of vacuum to
T
◦
T
◦
highly pressurized reactors. An analysis of Fig. 5 indicates that the
reverse reaction may be minimized by either decreasing the oxygen
pressure or increasing the reaction temperature. When the oxy-
gen pressure is decreased to below 1 bar, a vacuum will be created.
Since the produced oxygen in the forward reaction is often pressur-
ized for further storage or utilization in other units that consume
oxygen, a vacuum oxygen production reactor may increase the
power consumption of the vacuum and pressurization for industrial
scale oxygen production. In addition, a vacuum is more likely than
pressurization to introduce other gases into the oxygen produc-
tion reactor. Furthermore, negative pressure may also increase the
r
r
T
◦
r
◦
that the reverse reaction is exothermic. The steep slope at 430 C is
explained by the phase change of CuCl at this temperature, which
also explains the steep slope for the standard entropy of reaction,
as shown in Fig. 3. This sudden change does not exist in Fig. 4 for
the standard Gibbs free energy ꢀ G because the phase change of
CuCl is close to a reversible thermodynamic process under isobaric
conditions.
T
◦
r

1
14
G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
Table 2
Thermodynamic properties for the formation of Cu2OCl2 and CuCl.
◦
T ( C)
Cu2OCl2
CuCl
ꢀ Tf Hꢂ (kJ/kmol)
ꢀ S (kJ/kmol K)
T
f
ꢂ
ꢀ G (kJ/kmol)
T
f
ꢂ
ꢀ H (kJ/kmol)
T
f
ꢂ
ꢀ S (kJ/kmol K)
T
f
ꢂ
ꢀ G (kJ/kmol)
T
f
ꢂ
2
2
7
5
7
7
−381,286
−381,284
−380,897
−380,539
−380,200
−379,872
−379,550
−379,227
−378,897
−378,520.4
−378,520.4
−378,393
−378,027
−377,798
−377,640
−237.453
−237.444
−236.251
−235.293
−234.494
−233.804
−233.190
−232.627
−232.100
−231.735
−231.735
−231.364
−230.878
−230.591
−230.402
−310,486
−310,047
−298,205
−286,417
−274,672
−262,965
−251,289
−239,644
−228,025
−215,703
−215,703
−211,074
−199,515
−192,589
−187,976
−138,069
−138,055
−137,578
−136,946
−136,224
−135,446
−134,634
−133,799
−132,948
−132,025
−121,814
−121,392
−120,332
−119,702
−119,285
−57.665
−57.617
−56.155
−54.472
−52.772
−51.134
−49.586
−48.133
−46.771
−45.406
−30.873
−30.282
−28.863
−28.064
−27.552
−120,885
−120,770
−117,924
−115,158
−112,477
−109,879
−107,362
−104,919
−102,547
−100,098
−100.106
−99,494
1
1
2
2
3
3
4
4
4
5
5
5
27
77
27
77
27
77
30
30
50
00
30
50
a
−98,016
−97,162
−96,606
a
◦
CuCl changes from solid to liquid at 430 C.
◦
1
1
20
00
previous studies [18–22]. The temperature must be below 550 C
in order to maintain the vapor pressure of CuCl below 1000 Pa
(
1
.0 E-10 (M. A. Lewis, 2003)
0.01 bar
.1 bar
1 bar
bar
10 bar
00 bar
0.01 bar). For CuCl vapor minimization and the simultaneous
impact of pressure and temperature on the Gibbs free energy, the
optimum operating pressure is suggested in the range of 1–2 bar
and a temperature in the range of 525–550 C. Although it is desir-
able to decompose Cu OCl2 at lower temperatures, experimental
results of decomposition of copper oxychloride indicate that the
conversion rate is very low below 500 C [12]. As shown in Fig. 5,
the Gibbs free energy of the reverse reaction is slightly above zero
0
8
6
4
2
0
0
0
0
0
2
◦
1
2
◦
◦
at temperatures in the range of 525–550 C.
Fig. 5 also indicates that the reverse reaction is spontaneous
when the operating temperature is below 450 C, and the oxygen
-
-
-
-
20
40
60
80
◦
pressure is higher than 0.01 bar. This implies the reverse reaction
may occur in the process of reactant preheating from a temperature
◦
◦
significantly lower than 450 C to a temperature higher than 450 C,
as indicated by the temperature difference of the hydrolysis step
and its downstream oxygen production step shown in Table 1.
0
200
400
600
Reaction temperature, ºC
4. Results and discussion
Fig. 5. Gibbs free energy of the reverse reaction at various operating conditions.
Table 3 summarizes the experimental results and reaction con-
ditions. In order to have significant mixing and enhance the contact
of O2 and CuCl, 2–3 times of oxygen quantity in excess of the stoi-
chiometric balance was adopted in the experiments. The operating
time of each experiment was at least one hour. Two temperatures
generation of CuCl vapor. Therefore, a vacuum oxygen production
reactor is not suggested for the Cu–Cl cycle.
A method of minimizing the reverse reaction is to increase the
◦
operating temperature. If the temperature is higher than 530 C,
◦
were examined: 450 C, just above the melting point of CuCl, and
significant amounts of CuCl vapour may be generated, as observed
in the reactor and oxygen outlet in the experiments. Fig. 6 shows
the vapor pressure of CuCl based on experimental data reported in
◦
5
00 C, close to the transitional point of the Gibbs free energy of
the reaction. The implications of different temperatures will be dis-
cussed based on results of X-ray powder diffraction analysis (XRD).
Figs. 7–10 show the XRD results that correspond to experiments No.
5
4
4
3
3
2
2
1
1
000
500
000
500
000
500
000
500
000
1
–4 of Table 3. It can be found that no Cu OCl2 was produced in all
2
◦
experiments that operated in the temperature range of 450–500 C,
although experiment No. 3 resulted in some produced CuO. As to
the formation of CuO in experiment No. 3, it is believed that CuO
may have been generated from moisture in the reactant CuCl dur-
ing the preparation process because the purity of reactant CuCl is
larger than 99%, but the CuO detected in the product is 5%. The
composition of the reactant CuCl for experiment No. 3 was ana-
lyzed to determine the level of impurity. As shown in Fig. 10, the
only impurity in CuCl was found to be CuOHCl, which may generate
CuO through the following reaction [23–25]:
5
00
0
4
00 450 500 550 600 650 700 750 800 850
Temperature, ºC
2
CuOHCl = CuO + CuCl + H O
(19)
2
2
CuCl is then decomposed to CuCl and oxygen. It is believed that
2
Fig. 6. Vapor pressure of CuCl at various different temperatures.
the hydrogen element in CuOHCl occurred from moisture due to the

G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
115
Table 3
Experimental conditions and results for the reaction of CuCl and O2.
◦
No.
T ( C)
Reactants
puritya (%)
Runtime (h)
Ratio of total
quantity
Pressure (bar)
Inlet oxygen
concentration
(mole fraction)
Oxygen
decrease in
reaction (mole)
Composition of
product (by
XRD) % (± 3%)
b
(
O2/CuCl)
1
2
3
450
500
500
99%
99%
99%
1
1
1
1.3
1.3
1.3
1.2
1.2
1.2
5%
5%
5%
0
0
0
CuCl
CuCl
5 wt.% CuO
9
5 wt.% CuCl
97% CuCl
% Cu(OH)Cl
4c
500
99%
1
–
1.2
–
–
3
a
All sets of experiments used 0.5 mol of reactant and continuous air injection at 2 SLPM.
The stoichiometric ratio of O2 to CuCl is 0.5.
Baseline experiments with continuous nitrogen flow at 2 SLPM.
b
c
◦
Fig. 7. XRD results of products of CuCl + O2 at 450 C, 1-h test (experiment #1 in
Table 2).
◦
Fig. 9. XRD results of CuCl and O2 at 500 C, 1-h test (experiment #3 in Table 2).
The standard Gibbs free energy of the reverse reaction shown in
Fig. 4 suggests the spontaneity of the reaction in the temperature
range. This does not conflict with the experimental results because
the oxygen partial pressure in the experiments was below the stan-
dard value. As shown in Table 3, the partial pressure of O2 in the
experiments was far below 1 bar, which may significantly lower
the possibility of a reverse reaction based on the Gibbs free energy.
This observation also suggests that the reverse reaction may be a
function of the oxygen pressure in the reactor. When the operat-
ing pressure of oxygen is different from the standard pressure, the
Gibbs free energy of the reverse reaction at various pressures can
be calculated from Eq. (7).
presence of humidity when preparing the CuCl sample. It is unclear
how the CuOHCl was formed in the CuCl during the process. This
will be studied in future investigations.
From the experimental results, it can be concluded that the
reverse reaction as described in Eq. (6) has little likelihood to occur
under the conditions reported in Table 3. The low likelihood of the
reverse reaction to occur can be the result of two factors. First, the
reverse reaction is not thermodynamically spontaneous, and sec-
ond, the reverse reaction is spontaneous, but the runtime for the
reverse reaction is not long enough to produce significant amounts
of Cu OCl . If the reverse reaction is spontaneous, then the run time
2
2
must be very long, or the equipment size must be very large for the
reverse reaction to produce measurable amounts of Cu OCl .
The reverse reaction was not observed in the preheating of CuCl
and O2, in the experiments, although Fig. 5 shows negative values
of the Gibbs free energy when the operating pressure (0.06 bar)
is much higher than 0.01 bar. The XRD reflections (see Figs. 7–10)
2
2
◦
Fig. 8. XRD results of products of CuCl + O2 at 500 C, 1 h test (experiment #2 in
Table 2).
◦
Fig. 10. Impurities in the CuCl reactant, molten at 500 C (experiment #4 in Table 2).

1
16
G.D. Marin et al. / Thermochimica Acta 524 (2011) 109–116
did not show any trace of Cu OCl . Several conditions can lead
References
2
2
◦
to these results, for example, the reverse reaction below 450 C
is too slow to be measured. The sensitivity of the oxygen sensor
and XRD may not be sufficient to detect the composition change.
Also, the reverse reaction did take place, as suggested by the neg-
ative value of Gibbs free energy, and Cu OCl was produced, but
[
1] B.W. McQuillan, L.C. Brown, G.E. Besenbruch, R. Tolman, T. Cramer, B.E. Russ,
B.A. Vermillion, High efficiency generation of hydrogen fuels using solar ther-
mochemical splitting of water, General Atomics, Report GA-A24972.
[2] R.H. Carty, M.M. Mazumder, J.D. Schreiber, J.B. Pangborn, Thermochemical
Hydrogen Production , Institute of Gas Technology, 1981, GRI-80-0023.
2
2
[
3] M.A. Lewis, A Taylor, High temperature thermochemical processes, DOE Hydro-
gen Program, Annual Progress Report (2006) 182–185.
◦
after heating to 450–500 C, the Cu OCl decomposed back to CuCl
2
2
and O , as indicated in Fig. 5 by the positive value of Gibbs energy
[4] M. Sakurai, H. Nakajima, R. Amir, K. Onuki, S. Shimizu, Experimental study
on side-reaction occurrence condition in the iodine-sulfur thermochemical
hydrogen production process , International Journal of Hydrogen Energy (2000)
2
◦
within the temperature range. Under 430 C, CuCl exists in the
form of solid, and the mixing of oxygen and CuCl is not suffi-
cient to facilitate the reverse reaction. This condition may also
prevail in the mixture of molten CuCl and oxygen. Under normal
613–619.
[5] K. Schultz, Technical Report for the Stanford Global Climate and Energy Project,
General Atomics, Thermochemical production of hydrogen from solar and
nuclear energy, Technical Report for the Stanford Global Climate and Energy
Project, General Atomics (2003).
◦
operating conditions of 1–2 bar and 525–550 C, and within the
reaction time of interest, the reverse reaction can be neglected
in the design and operation of the oxygen production reactor of
the Cu–Cl cycle.
[
[
[
6] C. Zamfirescu, G.F. Naterer, I. Dincer, Vapor compression CuCl heat pump inte-
grated with a thermochemical water splitting plant , Thermochimica Acta 512
(
2011) 40–48.
7] V.N. Daggupati, G.F. Naterer, I. Dincer, Convective heat transfer and solid con-
version of reacting particles in a copper (II) chloride fluidized bed , Chemical
Engineering Science 66 (2011) 460–468.
8] Y. Haseli, I. Dincer, G.F. Naterer, Hydrodynamic gas–solid model of cupric chlo-
ride particles reacting with superheated steam for thermochemical hydrogen
production , Chemical Engineering Science 63 (2008) 4596–4604.
5
. Conclusions
This paper presented an experimental study of a potential
reverse reaction of oxygen production in the Cu–Cl thermochemical
cycle. The operating conditions that disfavor the reverse reaction
were examined. The analysis of the products with XRD analysis
indicated that the reverse reaction did not occur at temperatures
in the range of 450–500 C, and oxygen pressures below 2 bar.
The presence of humidity should be avoided in the reaction sys-
tem so as to avoid undesirable products such as CuOHCl and CuO.
[9] C. Lamberti, C. Prestipino, L. Capello, S. Bordiga, A. Zec, G. Spotto, S. Diaz,
A. Marsella, B. Cremaschi, M. Garilli, S. Vidotto, G. Leofanti, The CuCl2/Al2O3
catalyst investigated in interaction with reagents , International Journal of
Molecular Science 2 (2001) 230–245.
[10] T. Parry, Thermodynamics and Magnetism of Cu2OCl2 , Department of Chem-
◦
istry and Biochemistry, Brigham Young University, 2008.
◦
[
11] M.A. Lewis, M. Serban, J.K. Basco, Hydrogen production at <550 C using a low
temperature thermochemical cycle , in: Proceedings of the Nuclear Production
of Hydrogen: Second Information Exchange Meeting, Argonne, IL, USA, 2–3
October, 2003, pp. 145–156.
The thermochemistry data of Cu OCl2 and CuCl at various tem-
2
[
12] M. Serban, M.A. Lewis, J.K. Basco, Kinetic study of the hydrogen and oxygen
production reactions , in: AIChE 2004 Spring National Meeting, 2004.
[13] ASME International, Policy on reporting uncertainties in experimental mea-
surements and results , Journal of Heat Transfer Policy 115 (1993) 5–6.
14] Introduction to X-ray difraction. PANalytical. [Online] Speciris, 2010. [Cited:
September 09, 2010.] http://www.panalytical.com/index.cfm?pid=135.
15] H.P. Alexander, L.E. Klug, X-ray Diffraction Procedures for Polycrystalline and
Amorphous Materials , John Wiley & Sons, 1975.
16] J. Leroux, D.H. Lennox, K. Kay, Direct quantitative X-ray analysis by diffraction-
absorption techniques, Journal of Analytical Chemistry 25 (1953) 740–743.
17] National Institute of Standards and Technology. Chemistry Web-
Book. [Online] NIST, February 1, 2010. [Cited: October 5, 2010.]
http://webbook.nist.gov/chemistry/.
18] L. Brewer, N.L. Lofgren, The thermodynamics of gaseous cuprous chloride,
monomer and trimer , Journal of the American Chemical Society 72 (1950)
3038.
19] M. Guido, G. Gigli, M. Spoliti, Mass spectrometric study of the vaporization of
cuprous chloride and the dissociation energy of Cu3Cl3, Cu4Cl4, and Cu5Cl5 ,
Journal of Chemical Physics 55 (1971) 4566.
peratures in the reaction system were calculated. Thermodynamic
methods were designed in order to utilize the data of standard
conditions, low temperatures, and elementary species. The data
included the standard entropy, enthalpy and Gibbs free energy
of formation at different temperatures. A thermodynamic anal-
ysis was also performed to study the spontaneity of the reverse
reaction under various operating conditions. The combined effects
of pressure and temperature on the Gibbs free energy of the
reaction indicated that the optimum operating conditions cor-
respond to pressures in the range of 1–2 bar, and temperatures
[
[
[
[
[
◦
within 525–550 C. In these temperature ranges, the Gibbs free
energy of the reverse reaction has positive values. Although a
vacuum or high temperature also disfavors the reverse reaction,
in this paper, they are not recommended as the operating con-
ditions in a scaled-up installation would be challenging due to
oxygen storage, as well as significant CuCl vapor under those
conditions, and the additional energy consumption implications.
It is concluded that the decomposition of copper oxychloride
occurs when operating at pressures of 1–2 bar and temperatures
[
[20] M. Guido, G. Gigli, G.1. Balducci, Dissociation energy of CuCl and Cu Cul3
3
gaseous molecules , Journal of Chemical Physics 57 (1972) 3731.
[
21] L.C. Wagner, P. Robert, Q. Grindstaff, 1. Grimley, Mass spectrometric study
of the fragmentation of the cuprous chloride vapor system , Journal of Mass
Spectrometry 15 (1974) 255.
[22] S. Mizuno, K. Nakayama, T. Kawachi, Y. Kojima, F. Watanabe, H. Matsuda, M.1.
Takada, Enhancement of chloride volatilization of Zn, Pb and Cu from molten
fly ash under reduced pressure , The Society of Chemical Engineers (Japan) 32
◦
of 525–550 C. Therefore, the reverse reaction of oxygen pro-
duction can be neglected in the design and operation of the
Cu–Cl cycle.
(
2006) 99–102.
[23] P. Ramamurtahnyde, A. Secco, Studies on metal hydroxyl compounds. IX. Kinet-
ics of thermal decomposition of twelve copper halide derivatives , Canadian
Journal of Chemistry 47 (1969) 3915–3918.
24] P. Ramamurtahnyde, A. Secco, Addendum: studies on metal hydroxyl com-
pounds, Heats of decomposition of twelve copper halide derivatives , Canadian
Journal of Chemistry 47 (1969) 2303–2304.
25] P. Ramamurtahnyde, A. Secco, Studies on metal hydroxyl compounds. VII. Ther-
mal analyses of copper derivatives, Canadian Journal of Chemistry 47 (1969)
2185–2190.
Acknowledgments
[
Financial support of this research from Atomic Energy of Canada
Limited and the Ontario Research Excellence Fund is appreciated,
as well as helpful input from Dr. Venkata Dagupatti.
[