Effect of CO2on hydrogen absorption in Ti-Zr-Mn-Cr based AB2type alloys

Article abstract of DOI:10.1016/j.jallcom.2017.02.067

The effect of CO₂on hydrogen absorption has been investigated for AB₂type alloys of Ti_0.515Zr_0.485Mn_1.2Cr_0.8(Ti-Zr-Mn-Cr) and Ti_0.515Zr_0.485Mn_1.2Cr_0.8M_0.1(Ti-Zr-Mn-Cr-M, M?=?Fe, Co, or Ni) in order to develop metal hydrides for a hydrogen purification and storage system. A magnitude of CO₂poisoning tolerance was evaluated by comparing hydrogen absorption properties before and after CO₂exposure. As the results, an order of CO₂tolerance was Ti-Zr-Mn-Cr-Ni?2alloys have dependence of additive 3d-transition elements on CO₂tolerance. Ti-Zr-Mn-Cr-Fe has the highest CO₂tolerance among them. The additive of Fe and Co improves tolerance of CO₂poisoning for hydrogen absorption properties. On the other hand, Ni element in AB₂alloy decreases CO₂tolerance. The estimated enthalpy change between CO₂and the alloys surface was comparable to the formation of metal oxide. And the CO₂-exposed AB₂alloys after reacting with hydrogen desorbed methane gas. Therefore, CO₂would dissociate to CO and O on the surface of alloys, and then oxygen atom reacts with the alloys.

Full text of DOI:10.1016/j.jallcom.2017.02.067

Accepted Manuscript
Effect of CO on hydrogen absorption in Ti-Zr-Mn-Cr based AB type alloys
2
2
Nobuko Hanada, Hirotaka Asada, Tessui Nakagawa, Hiroki Higa, Masayoshi Ishida,
Daichi Heshiki, Tomohiro Toki, Itoko Saita, Kohta Asano, Yumiko Nakamura, Akitoshi
Fujisawa, Shinichi Miura
PII:
S0925-8388(17)30483-8
10.1016/j.jallcom.2017.02.067
JALCOM 40795
DOI:
Reference:
To appear in: Journal of Alloys and Compounds
Received Date: 21 October 2016
Revised Date: 4 February 2017
Accepted Date: 7 February 2017
Please cite this article as: N. Hanada, H. Asada, T. Nakagawa, H. Higa, M. Ishida, D. Heshiki, T. Toki, I.
Saita, K. Asano, Y. Nakamura, A. Fujisawa, S. Miura, Effect of CO on hydrogen absorption in Ti-Zr-Mn-
2
Cr based AB type alloys, Journal of Alloys and Compounds (2017), doi: 10.1016/j.jallcom.2017.02.067.
2
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
0.6
0.25
Magnitude of CO2 tolerance
0.5 for hydrogen absorption
Under CO2 partial pressure
0.2
0.15
0.1
AB2-Fe
0.4
0.3
0.2
0.1
0
AB2-Co
AB2
0.05
0
AB2-Ni
0
10
20
30
40
50
60
70
80
AB2-Ni
AB2-Fe
AB2-Co
AB2
Time (min)

ACCEPTED MANUSCRIPT
Effect of CO2 on hydrogen absorption in Ti-Zr-Mn-Cr
based AB2 type alloys
Nobuko Hanadaa,*, Hirotaka Asadaa, Tessui Nakagawab, Hiroki Higaa,
Masayoshi Ishidaa, Daichi Heshikic, Tomohiro Tokib, Itoko Saitad, Kohta Asanod,
Yumiko Nakamurad, Akitoshi Fujisawae, and Shinichi Miurae
a Graduate School of Systems and Information Engineering, University of Tsukuba, 1-1-1
Tennodai, Tsukuba, Ibaraki 305-8573, Japan
b Faculty of Science, University of Ryukyus, 1 Senbaru, Nishihara-cho, Nakagami-gun,
Okinawa 903-0213, Japan
c Graduate School of Engineering and Science, University of the Ryukyus, 1 Senbaru,
Nishihara-cho, Nakagami-gun, Okinawa 903-0213, Japan
d National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi,
Tsukuba, Ibraki 305-8565, Japan
e Kobe Steel Ltd., 1-5-5 Takatsukadai, Nishi-ku, Kobe, Hyogo 651-2271, Japan
Abstract
The effect of CO2 on hydrogen absorption has been investigated for AB2 type alloys
of Ti0.515Zr0.485Mn1.2Cr0.8 (Ti-Zr-Mn-Cr) and Ti0.515Zr0.485Mn1.2Cr0.8M0.1 (Ti-Zr-Mn-Cr-M, M
= Fe, Co, or Ni) in order to develop metal hydrides for a hydrogen purification and storage
system. A magnitude of CO2 poisoning tolerance was evaluated by comparing hydrogen
absorption properties before and after CO2 exposure. As the results, an order of CO2 tolerance
was Ti-Zr-Mn-Cr-Ni < Ti-Zr-Mn-Cr < Ti-Zr-Mn-Cr-Co < Ti-Zr-Mn-Cr-Fe. It suggests that
the AB2 alloys have dependence of additive 3d-transition elements on CO2 tolerance.
1

ACCEPTED MANUSCRIPT
Ti-Zr-Mn-Cr-Fe has the highest CO2 tolerance among them. The additive of Fe and Co
improves tolerance of CO2 poisoning for hydrogen absorption properties. On the other hand,
Ni element in AB2 alloy decreases CO2 tolerance. The estimated enthalpy change between
CO2 and the alloys surface was comparable to the formation of metal oxide. And the
CO2-exposed AB2 alloys after reacting with hydrogen desorbed methane gas. Therefore, CO2
would dissociate to CO and O on the surface of alloys, and then oxygen atom reacts with the
alloys.
Keywords: hydrogen absorbing materials, carbon dioxide, gas-solid reactions, poisoning
tolerance, kinetics
*Corresponding author: Tel&Fax: +81-29-853-5149, E-mail: hanada@kz.tsukuba.ac.jp
Address: Graduate School of Systems and Information Engineering, University of Tsukuba,
1-1-1 Tennodai, Tsukuba, Ibaraki 305-8573, Japan
2

ACCEPTED MANUSCRIPT
1. Introduction
Polymer electrolyte fuel cells (PEFC) have advantages like high efficiency, low operating
temperature, e.g. 80 °C, and quick response for electric load demand. It is important for
PEFCs to provide pure hydrogen since the cell voltage drops when impurities exist in
hydrogen gas [1]. When hydrogen is produced by steam reforming and following water-gas
shift reaction [2], it involves 20-25% CO2 and 0.5-1% CO as by-products. A CO Adsorbent
and Metal hydride Intermediate-Buffer (COA-MIB) system has been proposed to purify from
methane reforming gas and store hydrogen [3-9]. This system can supply pure hydrogen to
PEFCs corresponding to electric load demand. The CO adsorbent used in this system
decreases CO less than 1 ppm. The metal hydride stores hydrogen from the CO2-mixed gas,
and it desorbs pure hydrogen when required. The metal hydride used currently is LaNi5-based
AB5 type alloy which contains Misch Metal in A site and mainly Ni in B site because of
excellent tolerance of CO2 poisoning [8]. Here, CO2 poisoning is defined as the decrease of
hydrogen absorption kinetics and hydrogen absorption amount. It is important for practical
application of the COA-MIB system to reduce cost and volume of the metal hydride tank by
replacing with one of low-cost and large hydrogen capacity. For designing the high functional
metal hydrides with high tolerance of CO2 poisoning, we investigated the CO2 poisoning
factors of constituent elements in AB5 type alloys of LaNi5, CaNi5, and LaCo5 [10]. As the
result, CaNi5 showed the best tolerance of CO2 poisoning among them. This means that Ni in
B site has more important role to maintain high CO2 tolerance rather than Co in LaCo5 and
the rare-earth metal of La in LaNi5. It has been reported that surface modified AB5-type
alloys have higher poisoning tolerance and separates hydrogen from CO2 and CO containing
gas mixtures [11-13]. For other type of metal hydrides, TiFe alloys have been investigated the
effects of CO2 as the impurity gas on hydrogen sorption properties [14-16]. The hydrogen
3

ACCEPTED MANUSCRIPT
absorption rate and capacity decreased even under 0.1% CO2 concentration [15].
We focused on AB2 type laves phase alloys because of higher hydrogen capacity than AB5,
rare earth free, and easy activation [17]. There have been a number of publications reporting
hydrogen sorption properties for AB2 type alloys [17-19]. We chose C14 laves phase alloys of
Ti-Zr-Mn-Cr [18] as possible candidate for COA-MIB system because it is composed of low
cost elements such as Zr and Mn.
In this study, an effect of CO2 on hydrogen absorption has been investigated for
Ti0.515Zr0.485Mn1.2Cr0.8 which has a moderate equilibrium pressure of 0.1 MPa at 25 °C for the
COA-MIB system. In order to improve tolerance of CO2 poisoning, 3d-transition elements of
Fe, Co, or Ni, which potentially have a catalytic effect on hydrogen molecule dissociation,
have been added as Ti0.515Zr0.485Mn1.2Cr0.8M0.1 (M = Fe, Co, or Ni) and investigated the effect
of CO2 on hydrogen absorption. Furthermore, mechanism of the reaction between CO2 and
the alloys was discussed.
2ꢀExperimental
Å Alloy samples of AB2 (Ti0.515Zr0.485Mn1.2Cr0.8) and AB2-M (Ti0.515Zr0.485Mn1.2Cr0.8M0.1; M
= Fe, Co, Ni) were synthesized by arc melting (NEV-AD 03, Nisshin Giken Co.) in an argon
atmosphere. Starting materials of Zr (99.7%), Mn (99.99%), Cr (99.99%), Fe (99.9%), Co
(99.9%), and Ni (99.9%) were purchased from Furuuchi Chemical Co. Titanium (99.5%) was
purchased from The Nilaco Corporation. As-prepared ingots of alloys were annealed at
1100 °C for 24 h in the argon atmosphere for homogenization. The cross sections of the
annealed ingots were polished and observed by Scanning Electron Microscope (SEM,
S-3400N, Hitachi High-Technologies). Elemental analysis was performed by
Energy-Dispersive X-ray Spectroscopy (EDX, APOLLO XP, AMETEK) equipped with SEM.
The
ingots
were
ground
into
powders
of
150-500
ꢀm
diameters.
4

ACCEPTED MANUSCRIPT
Pressure-Composition-Temperature (PCT) curves were measured by using Sieverts-type
apparatus. Measurement conditions were at pressure of 0.05~1.0 MPa H2 and temperatures of
20, 30, 40, and 50 °C. The waiting time for each hydrogen absorption/desorption step was at
least 20 min. Equilibrium pressures were determined as a point that pressure change during
the waiting time became smaller than 0.002 MPa. A van’t Hoff plot was obtained using the
average of absorption and desorption pressures at a half of maximum amount of hydrogen in
PCT curves at each temperature. Powder X-ray diffraction (XRD) with Cu-Kα radiation
(RINT-2100, Rigaku) was performed to obtain crystallographic information. The powders
less than 150 ꢀm diameter was used for the measurements. The lattice constants of the
samples were calculated from peak positions.
Measurements for hydrogenation and dehydrogenation, and CO2 exposure were conducted
using semi-continuous flow-type reactor. Each sample of 1 g was loaded into a stainless steel
vessel of 14 ml volume with 4.6 mm in inner diameter. The sample vessel was covered with a
coolant tube for water cooling, a ribbon heater for electric heating, and a heat insulating
material on those. The sample temperature was monitored by a thermocouple located inside
the vessel in good contact with the sample powder. The flow rate gas inlet was controlled
using a needle valve and measured using a mass flow meter placed in the flow path of gas
inlet. More details of the apparatus were reported in our previous paper [10]. For an initial
activation process, the sample was heated up to 150 °C under vacuum by a rotary pump. Then
H2 of 0.9 MPa was introduced and kept flowing through the vessel for 1 h with 500 scc/min
at 150 °C. After that, the sample was cooled down to 80 °C and vacuumed at 80 °C as a
hydrogen desorption. Hydrogen of 0.9 MPa was introduced again at 80 °C as a hydrogen
absorption. The hydrogen absorption and desorption was repeated 3 times. Finally, the sample
was evacuated (~10 Pa) at 80 °C and then cooled to 20 °C.
For the hydrogenation test, H2 was introduced into the vessel up to 0.9 MPa at 20 °C with
5

ACCEPTED MANUSCRIPT
flow rate of approximately 5 scc/min (1 scc/min is flow rate at 20 °C under 0.1013 MPa). The
test was stopped after 80 min or when the flow rate decreased to 0.1 scc/min. A hydrogen
absorption rate Rhy was calculated by Rhy = Rin − Rcell, where Rin is a hydrogen flow rate, and
Rcell is rate of remained gas amount in the cell calculated by PV/RT. The amount of hydrogen
absorption was calculated by the integration of Rhy. More details were reported in Ref. [10].
For the dehydrogenation test, the sample after hydrogenation was simply evacuated for 30
min at 80 °C.
The CO2 exposure was performed by the following procedures. First, CO2 was introduced
into the vessel up to 0.9 MPa after the dehydrogenation process. Second, this pressurized
condition was kept for 30 min at 20 °C and then CO2 was evacuated for 30 min. After CO2
exposure, the hydrogenation and dehydrogenation was cycled for 3 times. The observed
properties were discussed comparing to the properties before the CO2 exposure.
After the CO2 exposure and hydrogenation property measurement, hydrogenation under
CO2 partial pressure was performed, CO2 of 0.1 MPa was firstly supplied to the
dehydrogenated state of alloys. Then H2 was supplied with flow rate of approximately 5
scc/min at 20 °C to increase the pressure from 0.1 to 0.9 MPa. This test was stopped after 600
min or when the gas flow rate decreases to 0.01 scc/min. Hydrogen of 7N purity and CO2 of
99.97% purity were used for all experiments.
Gas chromatography-Flame Ionization Detector (GC-FID; GC-2014, Shimadzu Co.)
analysis was performed in order to detect hydrocarbon in the sample gas which was prepared
by following procedures: (1) after the initial activation, three times of hydrogenation at RT
(~25 ºC) and 0.9 MPa H2 for 80 min and dehydrogenation at 80 ºC under vacuum condition
for 30 min; (2) CO2 was introduced at RT from vacuum condition to 0.9 MPa within 5 min
and then kept for 30 min; (3) after removing CO2 by evacuation at RT for 30 min, hydrogen
was pressurized to be 0.9 MPa; and finally (4) the sample was heated up to 200 °C for 3 days.
6

ACCEPTED MANUSCRIPT
The gas in the sample vessel was measured by GC-FID. All sample scale was 1 g and vessel
volume was ~55 mL. Blank data of the gas without CO2 exposure (procedure 2) were also
measured. Amount of detected gas was calculated by subtracting blank data from sample
data.
3. Results and discussion
3.1. Composition and crystal structure
Figure 1 and Table 1 show back scattering electron (BSE) images and the elements
compositions by SEM and EDX analysis for Ti0.515Zr0.485Mn1.2Cr0.8 (AB2),
Ti0.515Zr0.485Mn1.2Cr0.8Fe0.1
(AB2-Fe),
Ti0.515Zr0.485Mn1.2Cr0.8Co0.1
(AB2-Co),
and
Ti0.515Zr0.485Mn1.2Cr0.8Ni0.1 (AB2-Ni), respectively. The observed compositions in all area of
the samples almost corresponded to the target compositions. Since a vapor pressure at
melting point of Mn is the lowest among contained elements, the Mn ratio in all samples is
slightly less than the target one due to evaporation during arc melting. The spot 1 in AB2 was
a Zr-rich part, which was higher Zr concentration and lower Ti concentration than the target,
while the spot 2 was a Ti-rich part. The spots 3–8 in AB2-Fe, AB2-Co, and AB2-Ni were also
similar tendency with regard to Ti and Zr ratio. In particular, the BSE image of AB2-Co in Fig.
1 showed light gray (the Zr-rich part) and dark gray (the Ti-rich part) area in spot 5 and 6,
respectively. These results indicate that the AB2 alloys (AB2, AB2-Fe, AB2-Co, and AB2-Ni)
contain Ti-rich and Zr-rich components.
Figure 2 shows XRD profiles of the AB2 alloys. All samples were determined as single
phase of C14 Laves phase. No reflections from impurity phase was observed. Therefore, all
of Ti-rich and Zr-rich parts in BSE images are corresponding to C14 Laves structure. All
peaks were broad, and it can be explained by the variation of compositions as observed by
BSE and EDX. Lattice constants and lattice volumes of the AB2 phases were evaluated using
7

ACCEPTED MANUSCRIPT
XRD peaks as shown in Table 2. For the samples added the transition element, lattice
constants a and c, and lattice volumes increased (AB2-Ni < AB2-Co < AB2-Fe) with the atom
radius of additive metal (Ni < Co < Fe).
3.2 Hydrogen absorption properties
PCT curves of AB2 alloys at 20, 30, 40, and 50 °C for hydrogen absorption and desorption
are shown in Figure 3. The plateau regions for all samples leaned possibly due to the
composition gradient. The equilibrium pressures of AB2-M (M = Fe, Co, and Ni) were
slightly larger than that of AB2. Table 3 shows maximum hydrogen absorption amounts at
20 °C and 1 MPa of all samples, which were 0.92-0.99 H/M(1.6-1.7 mass%). Their van’t
Hoff plots are shown in Figure 4. The enthalpy and entropy of hydride formation are shown
in Table 3. The absolute value of enthalpy −33 kJ/mol H2 of AB2 was slightly larger than that
of AB2-Fe (−30 kJ/mol H2), AB2-Co (−29), and AB2-Ni (−31). The entropies of all AB2 alloys
were almost similar value of −100 J /mol·K H2.
Figure 5 shows hydrogen absorption properties for all AB2 alloys at 20 °C. The maximum
amounts of hydrogen absorption H/M were 1.02, 0.93, 0.92 and 1.02 for AB2, AB2-Fe,
AB2-Co, and AB2-Ni, respectively. These values were slightly larger than PCT results shown
in Table 3. The deviations were probably caused due to measurement error of the mass flow
meter, which was nominally ±1% of full scale. The hydrogen absorption rates of samples
were almost constant at 5–6 scc/min during initial 5 to 20 min, which corresponds to the
controlled hydrogen flow rate. After 20 min, these rates gradually decreased and then their
hydrogen absorption finished within 60 min. The vessel pressure gradually increased until 60
min for all AB2 alloys according to the plateau region of hydrogen equilibrium pressure.
3.3. Hydrogen absorption properties after CO2 exposure
8

ACCEPTED MANUSCRIPT
Figure 6 shows hydrogen absorption properties of the AB2 alloys at 20 °C after CO2
exposure. The hydrogen absorption properties before CO2 exposure were also shown for
comparison. Amount of hydrogen absorption was normalized using maximum amount of
hydrogen absorption before CO2 exposure. The hydrogenated fraction at 80 min for AB2 and
AB2-Co gradually decreased to approximately 0.9 within 3 cycles after CO2 exposure. In the
case of AB2-Ni, the hydrogenated fraction decreased to 0.83 even at the 1st cycle and to 0.77
at the 3rd cycle. On the other hand, the fraction of AB2-Fe in all 3 cycles were almost identical.
The hydrogen absorption rates of AB2, AB2-Fe, and AB2-Co at the 1st cycle were slower than
those before CO2 exposure in the measurement beginning. When the pressure increased more
than 0.3 MPa at about 10 min, the rates suddenly increased up to 15–20 scc/min at maximum.
It indicates that those AB2 alloys require hydrogen pressure higher than equilibrium pressure
probably due to kinetic barrier by CO2 chemisorption. In the 2nd and 3rd cycles, the hydrogen
absorption rates from 5 to 20 min were almost constant at ~5 scc/min and then gradually
decreased. This behavior was almost same to that before CO2 exposure. In the case of AB2-Ni,
the hydrogen absorption rate in the 1st cycle suddenly increased at 20 min, where the
maximum absorption rate was 23 scc/min at 0.6 MPa hydrogen. In the 2nd cycle, a similar
peak to 1st cycle appeared at 10 min. In the 3rd cycle, that peak finally disappeared. The
chemisorbed CO2 would remain on AB2-Ni even after the hydrogen desorption at 80 °C.
These results suggest that the CO2 bonding with AB2-Ni is stronger than that with the other
AB2 alloys.
The hydrogen absorption properties of AB2 alloys in the 3rd cycle are summarized in Figure
7. Weak CO2–metal bonding was supposed to be removed at 3rd dehydrogenation treatment,
the strong CO2–metal bonding, which causes irreversible poisoning such as strong
chemisorption or surface reaction, would degrade hydrogen absorption properties of AB2-M.
The order of total hydrogenated fraction in 3rd cycle was:
9

ACCEPTED MANUSCRIPT
AB2-Ni < AB2 ≈ AB2-Co < AB2-Fe.
On the other hand, the order of hydrogen absorption rate at 5-20 min was AB2-Ni ≈
AB2-Co < AB2-Fe < AB2. It is difficult to directly compare their hydrogen absorption rate
regarding to CO2 poisoning tolerance with each other, because the hydrogen equilibrium
pressure effects the hydrogen absorption rate. Therefore, a magnitude of CO2 poisoning
tolerance regarding kinetics was defined by the ratio of the reaction rate constant after CO2
exposure to that before CO2 exposure. The following equation for the reaction rate of
hydrogen absorption was employed [20],
P(C)− P (C,T)
dC
dt
eq
v =
=
kf
(C)
,
(1)
P
0
where ν is the reaction rate, C is the amount of absorbed hydrogen (H/M), P(C) is hydrogen
pressure (MPa), Peq(C,T) is hydrogen equilibrium pressure (MPa), P0 is atmospheric pressure
(MPa), k is reaction rate constant (s−1) corresponding to temperature T (°C), and f(C) is
function of C regarding to reaction process (H/M). The reaction rate before and after CO2
exposure with elimination of the effect of hydrogen equilibrium pressure are defined as:
P (C) − P (C,T)
dC
dt




bef
eq
v'bef
=
= kbef
f
(
C
)
,
(2)
(3)
P

bef
0
P (C) − P (C,T)
dC
dt




aft
eq
v' =
= kaft
f
( )
C
.
aft
P

aft
0
The reaction rate (dC/dt) and P(C) was obtained by hydrogen absorption properties before
CO2 exposure and 3rd cycle, respectively. Peq(C,T) was chosen from a fitting curve of the PCT
plots in Fig. 3. If f(C) do not change before and after CO2 exposure, a ratio of reaction rate
constant regarding magnitude of CO2 poisoning tolerance χ is estimated as:
kaft
v'aft
=
= χ
.
(4)
v'bef kbef
Figure 8(a) shows magnitude of CO2 poisoning tolerance (χ) at C = 0.3, 0.4, 0.5, 0.6, and 0.7
for all samples. The values of χ were less than 0.5 and decreased with increasing amount of
10

ACCEPTED MANUSCRIPT
hydrogen absorption. It indicates that effect of CO2 poisoning becomes strong at higher
hydrogen absorption amount. Average values of χ for 0.3-0.7 H/M (Fig. 8(b)) were described
in the following order:
AB2-Ni < AB2 < AB2-Co < AB2-Fe.
This order was the almost same as the total hydrogenated fraction. From above results,
AB2-Fe has the highest CO2 tolerance among them. The fact which AB2-Fe and AB2-Co have
larger χ than that of AB2 indicates that Fe and Co additive improve tolerance of CO2
poisoning for hydrogen absorption properties. In contrast, because AB2-Ni has smaller χ than
that of AB2, Ni additive degrades CO2 tolerance.
3.4. Hydrogen absorption properties under CO2 partial pressure
The hydrogen absorption properties of AB2 alloys under CO2 partial pressure at 0.1 MPa
were evaluated after the CO2 exposure test as shown in Figure 9. In this condition there are
possible effects of both weak CO2 bonding such as physisorption and strong CO2 bonding,
which is irreversible poisoning, such as chemisorption and surface reaction with alloys. The
amount of hydrogenated fractions at 80 min became 0.074, 0.22, 0.14, and 0.066 for AB2,
AB2-Fe, AB2-Co and AB2-Ni, respectively. These values are smaller than the results of
hydrogen absorption after CO2 exposure tests. It indicates that CO2 adsorption on the surface
such as physisorption hinders further the hydrogen absorption kinetics of the AB2 alloys. The
order of hydrogenated fractions at 600 min (not shown) was AB2-Ni (0.48) < AB2 (0.55) <
AB2-Co (0.75) < AB2-Fe (0.86). Moreover, the order of hydrogen absorption rate in Fig. 9
was AB2-Ni ≈ AB2 < AB2-Co < AB2-Fe. From these results, the order of magnitude of CO2
tolerance under CO2 partial pressure can be described as
AB2-Ni < AB2 < AB2-Co < AB2-Fe.
The dependency of the additive element is the same order as the CO2 exposure test.
11

ACCEPTED MANUSCRIPT
3.5. Analysis of bonding state between CO2 and alloy
The hydrogen absorption properties of all samples were not recovered from CO2 exposure
test even after 3 cycles as mentioned in section 3.3. The possible bonding types between CO2
and AB2 are strong chemical adsorption, chemical dissociation, surface reaction, and bulk
reaction with the alloys. Here, we investigated the bonding state by analyzing the structure
and surface condition.
Firstly, in order to investigate the bonding state between the bulk and CO2, the samples after
hydrogen absorption under 0.1 MPa of CO2 and subsequently hydrogen desorption were
measured by XRD for AB2, AB2-Fe, AB2-Co, and AB2-Ni as shown in Fig. 10. All samples
showed peaks of C14 laves phase. There were also hydride peaks of C14 laves phase in AB2,
AB2-Fe, and AB2-Co. Since hydrogen desorption was performed at the atmospheric pressure,
not vacuum condition, the hydride phase remained in the samples. The peaks of C14 laves
phase and the hydride phase were relatively broad comparing to those just after heat
treatment as shown in Fig. 2. It is indicated that the hydrogen absorption and desorption
cause the pulverization accompanying by the decrease of crystalline size and induction of
lattice strain. Other peaks that indicates the reaction between CO2 and AB2 alloys were not
observed. It is concluded that the reaction between CO2 and the AB2 alloys occurs only on the
surface.
Secondly, to investigate the CO2 bonding state on the surface, a reaction heat between CO2
gas and the surface was estimated. Figure 11(a) shows temperature changes for all samples
during the CO2 exposure. Just after CO2 introduction, the temperatures increased around
15 °C in 10-20 sec. The blank data in empty vessel was also shown in Fig. 11. Since the
temperature increased only 0.8 °C, the adiabatic effect by the pressure increase was enough
negligible. A reaction heat Q (J) of the sample based on the temperature change is estimated
12

ACCEPTED MANUSCRIPT
as:
Q =
∆Q(t
)
=
∆
Q
(t
)
+ ∆Q
(t)
.
∑
t
∑
t
∑
(5)
t
s
v
A heat flow rate into and out of the sample ꢀQs (t) is deduced as:
Ts t + − Ts
(
∆
t
)
(t)
,
∆
Qs
(
t = csms
)
(6)
∆
t
where t is time, cs is specific heat capacity of the sample (J/K
·
g), which is determined by the
Neumann-Kopp rule that molar heat capacity is approximately equal to the total piece of the
atomic heat capacity [21], ms is sample weight (g), Ts(t) is sample temperature (°C), and ꢀt is
interval of measurement (sec). The conductive heat transfer rate from inside to outside of the
sample vessel ꢀQv(t) is defined as follow.
Ts (t) −T
v
∆Qv
(
t = Aλ
)
.
(7)
L
where t is time, λ is a thermal conductivity of sample vessel (J/s·m·K), A is heat transfer area
(m2), and L is a thickness of vessel wall (m). Tv is regarded as the temperature of outside the
vessel, which is kept approximately at 20 ºC by water cooling. Here, the temperature
distribution of the sample in the vessel was assumed to be homogenous because the sample
amount of 1g is enough small. Heat transfer by convection and thermal radiation was
negligible in this case. The constant value of Aλ/L in Eq. (7) was optimized to 0.035 for all
samples which the reaction heat rate ꢀQ(t) did not be negative value in all measured time. Fig.
11 (b) shows the estimated ꢀQ (t), ꢀQs(t), and ꢀQv(t) for AB2 sample. The reaction heat Q of
AB2 was calculated to 9.7 J. When it is assumed that CO2 gas adsorbs on the surface,
∆HCO
2
enthalpy change for CO2 adsorption
(J/mol CO2) is estimated by dividing the reaction
nCO
2
heat Q by adsorbed CO2 amount of
(mol),
Q
Å
Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å Å
.
∆
HCO =
2
(8)
nCO
2
It is difficult to quantify adsorbed CO2 amount, because the amount is too little to detect as a
13

ACCEPTED MANUSCRIPT
nCO
2
pressure change. Therefore, adsorbed CO2 amount of
surface area S (m2/g) of the sample.
nCO
was calculated by using specific
Sms
=
,
(9)
2
NA ACO
2
where ACO2 is occupation area (m2) of one CO2 molecule on the surface and NA is Avogadro
constant. The S was estimated by SEM image of AB2 after all the experiments. Particle
boundary was separated by the lines indicated in Fig. S1. The large particles were
agglomerates of small particles. The average diameter of the particles was 1.3 ꢀm. The
calculated specific surface area was 0.68 m2/g when the particles were assumed sphere shape.
The enthalpy change for CO2 adsorption was estimated to be 1800 kJ/mol CO2 by assigning
the values to Eq. (8) and (9). The enthalpy of other alloys were also evaluated using the same
method as shown in Table 4. The estimated ꢀQ (t), ꢀQs(t), and ꢀQv(t) and SEM images are
also shown in Fig. S2 and S1, respectively for AB2-Fe, AB2-Co, and AB2-Ni. The enthalpy
values were much larger than the CO2 adsorption enthalpy on metals [23-24]. On the other
hand, those were comparable to formation enthalpy of titanium or zirconium metal oxide [25].
It indicates that oxidation of AB2 alloys surface occurs via CO2 dissociation to CO and O.
Therefore, the surface reaction is assumed as follows:
CO2 + Msurf → COads + Msurf O,
where Msurf is surface atom of AB2 alloys, COads is CO molecule adsorbed on the alloys,
MsurfO is metal oxide of surface atom. In order to prove CO adsorption, methanation (COads
(10)
+
3H2 → CH4 + H2O) was tested. AB2-Fe and AB2-Ni after CO2 exposure were kept under 0.9
MPa H2 at 200 ºC for 3 days. As a result, hydrocarbons especially methane were detected in
the gas of the sample vessel by GC-FID analysis as shown in Fig. 12. Blank data without CO2
exposure showed small hydrocarbon peaks, which could originate from unremoved CO2
species on the surface of AB2 alloys during activation. Table 5 showed amounts of
hydrocarbons by GC-FID analysis. Heavier hydrocarbons than methane such as ethane,
14

ACCEPTED MANUSCRIPT
propane, and butane were also detected, although their amounts were less than one tenth of
methane. These hydrocarbons were probably generated by the reaction between adsorbed CO
and hydrogen. The results are an evidence of CO adsorption on AB2 alloys. Moreover, the
amount of hydrocarbons of AB2-Fe was smaller than that of AB2-Ni. It indicates that the
amount of CO2 dissociation correlates with magnitude of CO2 tolerance.
From these results, CO2 would dissociate to CO and O on the surface of AB2 alloys, and
then oxygen atom reacts with AB2 alloys. The CO adsorption and generation of metal oxide
on the surface might decrease their hydrogen absorption kinetics. Therefore, additive element
in AB2 alloys, which effects on magnitude of CO2 tolerance, probably controls the amount of
CO2 dissociation.
4. Conclusion
The effect of CO2 on hydrogen absorption has been investigated for AB2 and AB2-M0.1 (M =
Fe, Co, or Ni). All samples were single phase of C14 laves structure and had no impurity
phase. The ratio of reaction rate constant of hydrogen absorption after CO2 exposure to that
before CO2 exposure was evaluated as the magnitude of CO2 poisoning tolerance. An order of
CO2 tolerance was obtained as,
AB2-Ni < AB2 < AB2-Co < AB2-Fe.
Moreover, the hydrogen absorption properties under CO2 partial pressure at 0.1 MPa showed
the same order of CO2 tolerance. The results suggest that AB2 alloy has dependence of
additive elements on CO2 tolerance. AB2-Fe has the highest CO2 tolerance among them. The
additive of Fe and Co improves tolerance of CO2 poisoning for hydrogen absorption
properties. On the other hand, Ni element in AB2 alloy decreases CO2 tolerance. The
CO2-exposed AB2 alloys after reacting with hydrogen desorbed methane gas. It is suggested
that CO adsorbs on the surface via CO2 dissociation. Moreover, the estimated enthalpy
15

ACCEPTED MANUSCRIPT
change between CO2 and the alloys surface was comparable to the formation of metal oxide.
Therefore, CO2 would dissociate to CO and O on the surface of AB2 alloys, and then oxygen
atom reacts with AB2 alloys. Additive element in AB2 alloys probably controls the amount of
CO2 dissociation corresponding to magnitude of CO2 tolerance.
Acknowledgement
This work has been partially supported by Ministry of the Environment under “Research
and development program of countermeasures against global warming in 2011-2012”.
References
[1] X. Cheng, Z. Shi, N. Glass, L. Zhang, J. Zhang D. Song, Z. Liu, H. Wang, J. Shen, A
review of PEM hydrogen fuel cell contamination: Impacts, mechanisms, and mitigation, J.
Power Sources 165 (2007) 739-756.
[2] A. Boyano, A.M. Blanco-Marigorta, T. Morosuk, G. Tsatsaronis, Exergoenvironmental
analysis of a steam methane reforming process for hydrogen production, Energy 36
(2011) 2202-2214.
[3] M. Ishida, Passive load following method by purifying hydrogen from reformed gas with
metal hydride for fuel cell, T. IEE 121-B (2001) 1036-1043 (in Japanese).
[4] Y. Taniguchi, M. Ishida, Passive load following method for PEFC system with reformer
and its efficiency improvement evaluated by using the concept of exergy, IEEJ Trans. PE
124 (2004) 851-858 (in Japanese).
[5] R. Sasaki, M. Ishida, Effect of CO in reformed gas on fluorinated metal hydride, IEEJ
Trans. PE 124 (2004) 965-970 (in Japanese).
[6] Y. Taniguchi, M. Ishida, Hydrogen purification method from reformed gas containing
high concentration of CO by using metal hydride, IEEJ Trans. PE 126 (2006) 1267-1274
16

ACCEPTED MANUSCRIPT
(in Japanese).
[7] S. Miura, A. Fujisawa, M. Ishida, A hydrogen purification and storage system using metal
hydride, Int. J. Hydrogen Energy 37 (2012) 2794-2799.
[8] S. Miura, A. Fujisawa, S. Tomekawa, Y. Taniguchi, N. Hanada and M. Ishida, A hydrogen
purification and storage system using CO adsorbent and metal hydride, J. Alloys Comp.
580 (2013) S414-S417.
[9] A. Fujisawa, S. Miura, Y. Mitsutake, M. Monde, Simulation study of hydrogen
purification using metal hydride, J. Alloys Comp. 580 (2013) S423-426.
[10] N. Hanada, T. Nakagawa, H. Asada, M. Ishida, K. Takahashi, S. Isobe, I. Saita, K. Asano,
Y. Nakamura, A. Fujisawa and S. Miura, Dependence of constituent elements of AB5
type metal hydrides on hydrogenation degradation by CO2 poisoning, J. Alloys Comp.
647 (2015) 198-203
[11] M. V. Lototsky, M. Williams, V. A. Yartys, Ye. V. Klochko, V. M. Linkov,
Surface-modified advanced hydrogen storage alloys for hydrogen separation and
purification, J. Alloys Comp. 509 (2011) S555-S561
[12] M. Lototskyy, K. D. Modibane, M. Williams, Ye. Klochko, V. Linkov, B. G. Pollet,
Application of surface-modified metal hydrides for hydrogen separation from gas
mixtures containing carbon dioxide and monoxide, J. Alloys Comp. 580 (2013)
S382-S385
[13] K. D. Modibane, M. Williams, M. Lototskyy, M. W. Davids, Ye. Klochko, B. G. Pollet,
Poisoning-tolerant metal hydride materials and their application for hydrogen separation
from CO2/CO containing gas mixtures, Int. J. Hydrogen Energy 38 (2013) 9800-9810
[14] F. R. Block, H. –J. Bahs, Investigation of selective absorption of hydrogen by LaNi5 and
FeTi, J. Less-Common Met. 89 (1983) 77-84.
[15] G. D. Sandrock, P. D. Goodell, Cyclic life of metal hydrides with impure hydrogen:
17

ACCEPTED MANUSCRIPT
Overview and engineering considerations, J. Less-Common Met. 104 (1984) 159-173.
[16] T. Hirata, Hydrogen absorption and desorption properties of FeTi1.14O0.03 in impure
hydrogen containing CO, CO2 and oxygen, J. Less-Common Met. 107 (1985) 23-33.
[17] D. G. Ivey, R. I. Chittim, K. J. Chittim, D. O. Northwood, Metal hydrides for energy
storage, J. Materials for Energy Systems, 3 (1981) 3-19
[18] S. Fang, Z. Zhou, J. Zhang, M. Yao, F. Feng, S. O. Northwood, The application of
mathematical models to the calculation of selected hydrogen storage properties
(formation enthalpy and hysteresis) of AB2-type alloys, Int. J. Hydrogen Energy 25
(2000) 143-149
[19] X. Guo and E. Wu, Thermodynamics of hydrogenation for Ti1-xZrxMnCr laves phase
alloys, J. Alloys Comp. 455 (2008) 191-196
[20] A. Yoshida, Y. Naka, and T. Ohkita, Experimental study on thermophysical and kinetic
properties of the LaNi5-H2 system, Transactions of the Japan Society of Mechanical
Engineers Series B 56 (1990) 536-540 (in Japanese)
[21] H. Kopp, Investigations of the specific heat of solid bodies, Phil. Trans. R. Soc. Lond.
155 (1865) 71-202.
[22] E. Hahne, J. Kallweit, Int. J. Hydrogen energy, Thermal conductivity of metal hydride
materials for storage of hydrogen: experimental investigation, 23 (1998) 107-114
[23] C. Liu, T. R. Cundari, and A. K. Wilson, CO2 reduction on transition metal (Fe, Co, Ni,
and Cu) surfaces: in comparison with homogeneous catalysis, J. Phys. Chem. C 116
(2012) 5681-5688
[24] A. E. Aksoylu, A. N. Akin, S. G. Sunol, Z. İ, Önsan, The effect of metal loading on the
adsorption parameters of carbon dioxide on coprecipitated nickel-alumina catalysts, Tr. J.
of Chemistry 20 (1996) 88-94
[25] Handbook of electrochemistry Ver.6, The electrochemical society of Japan (Eds.),
18

ACCEPTED MANUSCRIPT
Maruzen, Tokyo, 2013, pp.32-48 (in Japanese)
19

ACCEPTED MANUSCRIPT
Fig. 1 Back scattering electron (BSE) images of Ti0.515Zr0.485Mn1.2Cr0.8 (AB2),
Ti0.515Zr0.485Mn1.2Cr0.8Fe0.1 (AB2-Fe), Ti0.515Zr0.485Mn1.2Cr0.8Co0.1 (AB2-Co) and
Ti0.515Zr0.485Mn1.2Cr0.8Ni0.1 (AB2-Ni). Cross marks in the images are the spots analyzed by
EDX.
20

ACCEPTED MANUSCRIPT
C14 Laves phase
AB2
AB2-Fe
AB2-Co
AB2-Ni
35
30
40
45
50
55
60
65
70
75
80
2θ
(degree)
Fig. 2 X-ray diffraction profiles of AB2, AB2-Fe, AB2-Co, and AB2-Ni.
21

ACCEPTED MANUSCRIPT
1
0.1
AB2
AB2-Fe
Absorption
0.01
20°C
20°C
30°C
Desorption
30°C
40°C
40°C
50°C
50°C
0.001
1
AB2-Ni
AB2-Co
0.1
0.01
0.001
0
0.2
0.4
0.6
0.8
1
0
0.2
0.4
0.6
0.8
1
Hydrogen amount (H/M)
Hydrogen amount (H/M)
Fig. 3 PCT curves of hydrogen absorption and desorption at 20, 30, 40 and 50 °C for AB2,
AB2-Fe, AB2-Co, and AB2-Ni.
22

ACCEPTED MANUSCRIPT
1.5
1
0.5
0
AB₂
-0.5
-1
AB₂-Fe
AB₂-Co
AB₂-Ni
-1.5
3.05
3.1
3.15
3.2
3.25
3.3
3.35
3.4
3.45
1000/T (K-1)
Fig. 4 Van’t Hoff plots for AB2, AB2-Fe, AB2-Co, and AB2-Ni.
23

ACCEPTED MANUSCRIPT
1
0.8
0.6
0.4
0.2
AB2
AB2-Ni
AB₂
AB2-Co
AB2-Fe
AB₂-Fe
AB₂-Co
AB₂-Ni
0
6.0
AB2
AB2-Ni
5.0
4.0
3.0
2.0
1.0
AB2-Fe
AB2-Co
0.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
AB2-Co
AB2-Fe
AB2-Ni
AB2
0
10
20
30
40
Time (min)
50
60
70
80
Fig. 5 Pure hydrogen absorption properties at 20 °C for AB2, AB2-Fe, AB2-Co, and AB2-Ni.
24

ACCEPTED MANUSCRIPT
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
3rd
Before
2nd
(a)AB2
3rd
Before
2nd
(b)AB2-Fe
Before CO₂ exposure
1st cycle after CO₂ exposure
2nd cycle
1st
1st
3rd cycle
20
15
10
5
1st
1st
3rd
Before
2nd
Before
2nd
3rd
0
0.9
3rd
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2nd
3rd
Before
1st
2nd
1st
Before
30
0
10
20
30
40
Time (min)
50
60
70
80
0
10
20
40
50
60
70
80
Time (min)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Before
Before
2nd
1st
3rd
2nd
3rd
(c)AB2-Co
(d)AB2-Ni
1st
1st
20
15
10
5
1st
Before
2nd
3rd
Before
2nd
3rd
0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2nd
Before
1st
3rd
1st
10
3rd
2nd
Before
20 30
0
10
40
50
60
70
80
0
20
30
40
50
60
70
80
Time (min)
Time (min)
Fig. 6 Hydrogen absorption properties of 1st, 2nd and 3rd cycles at 20 °C for (a)AB2,
(b)AB2-Fe, (c)AB2-Co, and (d)AB2-Ni after CO2 exposure.
25

ACCEPTED MANUSCRIPT
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
AB2-Fe
AB2-Co
AB2
AB2-Ni
AB₂
AB₂-Fe
AB₂-Co
AB₂-Ni
6.0
5.0
4.0
3.0
2.0
1.0
AB2
AB2-Fe
AB2-Ni
AB2-Co
0.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
AB2
AB2-Fe
AB2-Co
AB2-Ni
0
10
20
30
40
Time (min)
50
60
70
80
Fig. 7 Hydrogen absorption properties of 3rd cycle at 20 °C for AB2, AB2-Fe, AB2-Co, and
AB2-Ni after CO2 exposure.
26

ACCEPTED MANUSCRIPT
0.6
0.5
0.4
0.3
0.2
0.1
0.0
AB₂
AB₂-Fe
AB₂-Co
AB₂-Ni
(a)
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Amount of hydrogen absorption (H/M)
0.6
0.5
0.4
0.3
0.2
0.1
0
(b)
Average
AB2-Ni
AB2-Fe
AB2-Co
AB2
Fig. 8 Ratio of reaction rate constant after CO2 exposure to that before CO2 exposure (χ)
regarding magnitude of CO2 poisoning tolerance at (a) H/M= 0.3-0.7 and (b) average values
at H/M= 0.3-0.7 for AB2, AB2-Fe, AB2-Co, and AB2-Ni.
27

ACCEPTED MANUSCRIPT
0.25
0.2
AB
AB
AB
AB
₂
₂
₂
₂
AB2-Fe
AB2-Co
-Fe
-Co
-Ni
0.15
0.1
0.05
AB2-Ni
0
0.6
AB2-Fe
0.5
0.4
0.3
0.2
0.1
AB2-Co
AB2
AB2-Ni
0
0.9
AB2
AB2-Ni
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
AB2-Fe
AB2-Co
0
10
20
30
40
50
60
70
80
Time (min)
Fig. 9 Hydrogen absorption properties at 20 °C for AB2, AB2-Fe, AB2-Co, and AB2-Ni under
CO2 partial pressure at 0.1 MPa.
28