Post on 10-Apr-2022
transcript
1
Hydrogen permeation in dual phase (DP) and quenched and partitioned
(Q&P) advanced high strength steels (AHSS) under cathodic charging and
simulated service conditions
Qinglong Liu
1,*, Jeffrey Venezuela
1, Mingxing Zhang
1, Qingjun Zhou
2, Andrej Atrens
1
1 The University of Queensland, Division of Materials, School of Mining and Mechanical
Engineering, St. Lucia, 4072 Australia * Corresponding author, andrejs.atrens@uq.edu.au
2 Baoshan Iron & Steel Co., Ltd, Research Institute, Shanghai, 201900, China
Abstract: Hydrogen permeation through DP and Q&P advanced high strength steels (AHSS) was
investigated (i) for cathodic charging at different potentials in 0.1 M NaOH, (ii) for immersion in 3wt%
NaCl solution at the free potential simulating the corrosion of steel in service and at zinc potential for
corrosion of the galvanized steel during service, and (iii) for immersion in 0.1 M HCl solution at the
free potential. For cathodic charging, hydrogen permeation and hydrogen concentration decreased
with less negative charging potentials. Hydrogen permeation and hydrogen concentration under
simulated service conditions in 3wt% NaCl were lower than those under the least negative cathodic
charging potential of -1.100 VHg/HgO in 0.1 M NaOH. For galvanized steels, once the Zn coating was
corroded, more hydrogen would be introduced into and permeate through the steel. The hydrogen
concentration was also measured by a hot extraction analyser, providing consistent results with those
from the permeability experiments. Key words: steel; hydrogen permeation; advanced high strength steel
1. Introduction
In recent decades, advanced high strength steels (AHSS) were developed, and have
been adopted, for auto components to meet the demands for light weight and increased
vehicle safety [1, 2]. Among AHSS, DP and TRIP steels exhibit a good combination of
strength and ductility, and good energy absorption ability. Quenching and partitioning (Q&P)
is a new heat treatment to produce TRIP steel. This process was designed with the desire for
producing steels with a combination of better strength and ductility, and high strain hardening
capacity, achieved by a microstructure containing more retained austenite and a greater TRIP
effect. Q&P steels are thus third generation AHSS with high strength and high ductility [3].
Hydrogen can be introduced into the steels, during steel making, or during auto
construction processes such as painting, or during corrosion practical service. The hydrogen
evolution reactions are as follows [4]:
H3O+ + M + e MHads + H2O (acid) (1)
H2O + M + e MHads+ OH- (neutral, alkaline) (2)
where M represents the metal surface, and MHads represents the adsorbed hydrogen on the
metal surface. The atomic hydrogen may be desorbed by the desorption Eq. (3), through
which two adsorbed hydrogen atoms combine to form a molecule of hydrogen that leaves the
metal surface, or by the electrochemical desorption reaction given by Eq. (4) or (5).
2MHads →H2 + 2M (3)
MHads + H3O+ + e H2 + H2O + M (acid) (4)
MHads + H2O + e H2 + OH- + M (neutral, alkaline) (5)
Some of the adsorbed hydrogen enters the metal, MHabs, by the following equilibrium
reaction:
2
MHads ↔ MHabs (6)
This hydrogen, in combination with an applied stress, interacts with the steel and can
cause the degradation of the mechanical properties of the steel and even catastrophic failure.
This phenomenon of hydrogen embrittlement (HE) and is a possible concern for AHSS [5, 6].
Thus it is important to understand how much hydrogen can enter into and permeate through
the AHSS in various service environments. This research is part of a wider research program
that seeks to understand the influence of hydrogen on the mechanical behaviour of AHSS,
and to provide information for their applications in auto industry.
The permeability experiments based on the method of Devanathan and Stachurski [7]
have been widely used for studying hydrogen transportation in metals.
This paper builds on existing research that studied hydrogen embrittlement (HE) of
AHSS under increasingly severe hydrogen charging conditions [8-10], from inert, moderate,
to severe, and summarises our prior studies on hydrogen permeation in DP and Q&P steels
under various conditions [11, 12], including (i) cathodic charging in alkaline solution, (ii)
simulated service corrosion and (iii) free corrosion in 0.1 M HCl solution.
2. Experimental
2. 1. Materials
The DP and Q&P steels were rolled sheets, and were designated as 980 DP, 1200 DP-
GI and 980 QP. “GI” referred to galvanization of the steel, which involving applying a Zn
corrosion-protection coating on the steel. The average as-received sheet thicknesses were 1.35
m for 980 DP and 1200 DP-GI, and 1.93 mm for 980 QP. The specimens were cut to 30 × 30
mm with the as-received thickness. The chemical composition, mechanical properties and
microstructural composition of the steels are presented in Table 1.
Table 1 Chemical composition (in wt %), mechanical properties and microstructural
composition of the studied steels. Steel
designation
C Si Mn Yield
stress,
MPa
Tensile
stress,
MPa
Elongation
at fracture,
ef, %
ferrite
%
martensite
or banite
%
retained
austenite
%
980 DP 0.085 0.276 2.255 592 930 8.3 40 60 0
1200 DP-GI 0.121 0.244 2.425 896 1198 4.7 26 74 0
980 QP 0.209 1.386 1.876 682 1020 11.3 39 53 8
2.2. Permeability experiments
(i) Cathodic charging
A double-cell arrangement was used for permeability experiments based on that of
Devanathan and Stachurski [7]. Each cell contained 0.1 M NaOH solution, and a three
electrode system with the steel as the working electrode, a Pt wire as the counter electrode,
and an Hg/HgO, KOH (20%) reference electrode connected to a Luggin capillary.
The 980 DP, 1200 DP-GI and 980 QP specimens were ground to: 0.53, 0.70 and 0.48
mm, respectively. The zinc coatings on the top and bottom surfaces of the galvanized 1200
DP-GI steel were ground off. The area exposed to the solution on the hydrogen-exit side was
3.394, 2.997 and 2.930 cm2, respectively. The specimen side exposed to hydrogen charging
cell was polished to 3 µm, washed with distilled water and ethanol, and dried. The hydrogen-
exit side of the specimen was plated with palladium to prevent oxidation of the steel.
Hydrogen was produced by a negative potential applied with a MP 81 potentiostat. A
PARSTAT 2273 maintained a potential of + 300 mVHg/HgO to oxidise the emerging hydrogen
and measure the amount of hydrogen that permeated through. N2 was bubbled throughout the
whole experiment to remove oxygen, which could contribute to the oxidation current density.
The exit side background current density was lower than 0.2 μA cm-2
before cathodic
3
charging on the hydrogen-entry side. This background current was subtracted from the
permeation current.
An uninterrupted pre-charging at -1.600 VHg/HgO was conducted for 60 h on the
hydrogen-entry side of the specimen in order to stabilise the surface. During pre-charging
there was typically a steady increase in permeation current density until a steady state was
reached. Thereafter, successive transients were measured, such as those from -1.600 VHg/HgO
to -1.700 VHg/HgO, as well as a transient loop from -1.700 VHg/HgO to -1.100 VHg/HgO and back
to -1.700 VHg/HgO. The potentials for the transient loop were: -1.700 VHg/HgO, -1.600 VHg/HgO,
-1.400 VHg/HgO, -1.200 VHg/HgO, -1.100 VHg/HgO, and similarly back to -1.700 VHg/HgO. The
same experimental sequence was carried out for each steel. All the experiments were carried
out at room temperature, 23 ± 2 ℃ .
The permeation transients can be expressed by:
in = 𝑖p− 𝑖p
0
𝑖p∞− 𝑖p
0 = 2𝐿
√𝜋𝐷𝑡∑ exp( −
(2𝑛+1)2𝐿2
4𝐷𝑡)∞
𝑛=0 (Rise transients) (7)
in = 𝑖p− 𝑖p
∞
𝑖p0− 𝑖p
∞ = 1−2𝐿
√𝜋𝐷𝑡∑ exp( −
(2𝑛+1)2𝐿2
4𝐷𝑡)∞
𝑛=0 (Decay transients) (8)
where in is the normalized current, ip is the measured permeation current density at time t, 𝑖p0
is the initial steady-state permeation rate at time t = 0, 𝑖p∞ is the new steady-state permeation
current density, and L is the thickness of the steel sheet. For the first charging, 𝑖p0 = 0, and for
the complete decay 𝑖p∞= 0. The experimental permeation curves were fitted to Eq. (7) or (8),
to determine the effective diffusion coefficient, Deff, and then the total hydrogen
concentration, CH, at the sub-surface on the cathodic side, can be calculated from:
CH = 𝑖𝑝
∞ 𝐿
𝐹 𝐷eff (9)
where F is the Faraday constant, L is the thickness of the membrane, 𝑖p∞ is the steady-state
permeation rate, and Deff is the measured effective diffusion coefficient of hydrogen in the
steels.
(ii) Simulated service corrosion
These permeability experiments used the same set-up and experiment conditions with
3wt% NaCl solution with pH 7.3 on the entry side either (i) with the steel at the free corrosion
potential, or (ii) with the steel surface polarised to the free corrosion potential of zinc, in order
to simulate the maximum hydrogen concentration for steel corroding in contact with a
galvanised layer.
For the experiments at the free potential in 3wt% NaCl solution, the 980 DP, 1200
DP-GI and 980 QP specimens were ground to 0.71, 0.86 and 0.47 mm, respectively. The area
exposed to the solution in the hydrogen-exit-side cell was 3.309, 3.423 and 2.588 cm2,
respectively. The open circuit potential was measured to be -0.668 VAg/AgCl for the steels
freely corroded in the 3wt% NaCl solution. The free corrosion potential was constant and did
not change significantly with time.
The experiments at the Zn potential in 3wt% NaCl solution used an applied potential
of -0.950 VAg/AgCl, which was the measured open circuit potential of the Zn coating in 3wt%
NaCl solution. The reference electrode was Ag/AgCl, KCl (saturated) and the counter
electrode was a Pt wire. The 980 DP, 1200 DP-GI and 980 QP specimens were ground to
0.56, 0.90 and 0.55 mm, respectively, and exposed an area of 2.608, 2.912 and 3.243 cm2,
respectively, to the deaerated 0.1 M NaOH solution on the hydrogen-exit side.
(iii) 0.1 M HCl
Experiments at the free potential in 0.1 M HCl solution, which was much more severe
than any environment encountered in auto service, were carried out using 980 DP steel. These
experiments were carried out because this solution has been used to study the influence of
4
hydrogen on AHSS [13, 14]. The specimen was ground to 0.91 mm and exposed area of
2.608 cm2 in the hydrogen-exit-side cell. The open circuit potential was measured to be
-0.476 VAg/AgCl.
2.3. Diffusible hydrogen concentration
The diffusible hydrogen concentration was measured using a BRUKER G4
PHOENIX DH carrier gas hot extraction analyser for 980 DP, 1200 DP-GI and 980 QP steels
which were uncharged as well as were (i) under hydrogen chathodic charging in 0.1 M NaOH
solution at over-potential values of -0.257 V, -0.557 V and -0.857 V, (ii) at free corrosion in
3wt% NaCl solution, simulating free corrosion of the steel of the car body in service, (iii)
cathodically polarised to the zinc potential in 3wt% NaCl to simulate the maximum amount
of hydrogen produced for the steels protected by a zinc coating applied by galvanisation, and
(iv) at free corrosion potential in 0.1M HCl, all for 24 h. The sample dimensions were
10 mm × 50 mm with the as-received thicknesses. All sample surfaces were ground to 1200
grit using SiC paper.
After 24-hour hydrogen charging, the sample was cleaned with distilled water and
ethanol, dried with blowing air, weighed, inserted in the quartz extraction tube in the
temperature-programmable infrared heated furnace, and the furnace was heated by the
instrument to 400 °C, which was used because it was expected that all diffusible hydrogen
would be released but that irreversibly bound hydrogen would not be released. The diffusible
hydrogen was released into the carrier gas flow of pure nitrogen, and the amount of hydrogen
was determined from the increase in thermal conductivity of the carrier gas. There was a
constant time interval of 3 min between the end of hydrogen charging and the beginning of
the hydrogen measurement.
3. Results
3.1. Cathodic charging
Fig. 1(a) presents the hydrogen permeation current density versus time for the three
steels during the 60-h pre-charging in the 0.1 M NaOH solution at -1.600 VHg/HgO, which was
started after the background current density in the right-hand cell decreased to less than
0.2 μA cm-2
. Thereafter, successive transients as well as a transient loop were measured.
The permeation current density increased significantly after the break through, and
then reached a maximum value, as shown in Fig. 1(a). The maximum current density
indicated approximately steady-state hydrogen charging conditions, and was 32 μA cm-2
,
21 μA cm-2
and 7 μA cm-2
, for 980 DP, 1200 DP-GI and 980 QP steel, respectively. The
significant increase in permeation current density with cathodic pre-charging time was
consistent with the literature [9, 15], and is attributed to a stabilisation of the steel surface due
to (i) the reduction of some of the air-formed oxide, which increased the surface coverage of
hydrogen, and (ii) the weakening of the bonding force between the adsorbed hydrogen and
the steel, which facilitated hydrogen absorption.
Fig. 1(b) shows a typical transient loop from -1.700 VHg/HgO to -1.100 VHg/HgO and
back to -1.700 VHg/HgO in 0.1 M NaOH for 980 QP steel. Each transient was a partial transient,
which meant that there was always a significant charging potential applied to the specimen,
and the specimen was not allowed to completely discharge all hydrogen. Similar transient
loops were also carried out for the other DP and Q&P steel grades. The experimental
permeation transients, such as those from the transient loop shown in Fig. 1(b), were fitted to
Eqs. (7) and (8) by Matlab to determine the hydrogen diffusion coefficient values, Deff, and
the total hydrogen concentration, CH, at the sub-surface on the cathodic side, was calculated
by Eq. (9).
5
Fig. 1 (a) Hydrogen permeation curves for the steels during cathodic pre-charging at-1.600
VHg/HgO in NaOH solution and (b) hydrogen permeation transients at different cathodic
charging potentials for the 980 QP steel after the pre-charging (P1: - 1.700 VHg/HgO, P2:
- 1.600 VHg/HgO, P3: - 1.400 VHg/HgO, p4:- 1.200 VHg/HgO and P5:- 1.100 VHg/HgO).
For the DP steels studied in our research, the values for effective diffusion coefficient,
as were presented in our prior study [11], were in the range of 0.7 × 10-6
~ 2.1 × 10-6
cm2 s
-1,
consistent with those from literatures [16-19]. For 980 QP steel, the effective diffusion
coefficient varied from 3 × 10-7
cm2 s
-1 to 7 × 10
-7 cm
2 s
-1, in agreement with the range of
0.3 × 10-7
cm2 s
-1 to 6 × 10
-7 cm
2 s
-1 from other studies [17, 20, 21], regardless that different
methods were used to determine the diffusion coefficient, which would lead to a variance of
the values. For instance, Yang et al. [20] used two methods, the time lag method and the
breakthrough method, to determine the hydrogen diffusion coefficient of their Q&P steels.
The results obtained from both methods had a difference of least a factor of 2. In addition, the
values of the diffusion coefficient were similar for the first two rise transients from -1.600
VHg/HgO and -1.700 VHg/HgO after long time pre-charging, and the fitting of the permeation
curve to the equations were good, as indicated in our prior research [11, 12]. These values of
the diffusion coefficient were identified as the lattice diffusion coefficient, DL. The diffusion
coefficients measured for the subsequent decay transients gave decreasing values of Deff.
Thereafter the partial rise transients gave values of the effective diffusion coefficient, Deff,
which increased towards the value for the lattice diffusion coefficient.
Fig. 1(b) indicates that as the charging potential was changed from -1.700 VHg/HgO to
-1.100 VHg/HgO the steady-state current density and the hydrogen concentration at each
potential decreased, indicating less hydrogen entered and permeated through the steel at a less
negative charging potential. When the charging potential was increased from -1.100 VHg/HgO
to -1.700 VHg/HgO, there was more hydrogen entering and permeating through the specimen,
providing a higher steady-state current density and hydrogen concentration. This trend was
consistent with those from Liu et al. [9] and Venezuela et al. [10], where they found that
during cathodic charging in 0.1 M NaOH solution, an increasingly negative charging
potential lead to an increased hydrogen fugacity, and thus an increased hydrogen
concentration and permeation current density in the steel.
Fig. 2 presents the experimentally determined relationship between the total hydrogen
concentration at the input side of the permeation specimen, CH, and the over-potential, η, for
all the steels under various cathodic hydrogen charging potentials in the 0.1 M NaOH solution.
For clarity, fitting lines are shown only for 1200 DP-GI and 980 QP. There was a turning
point for each steel at an over-potential of about -0.5 V. The relationship between ln CH and η
was linear both below and above the turning point, similar to that found in other research [9,
22].
6
Fig. 2 The hydrogen concentration, CH vs. over-potential, η, for the studied steels for various
conditions.
3.2. Simulated service corrosion
Fig. 3(a) presents the hydrogen permeation current density versus time for all the
steels under simulated auto service conditions in 3wt% NaCl solutions at Ecorr and EZn. These
are not permeation transients, as is apparent by a comparison with Fig. 1.
Fig.3 Hydrogen permeation current density versus time (a) for the steels for simulated service
corrosion and (b) for 980 DP steel under simulated service corrosion as well as under free
corrosion in 0.1 M HCl solution.
Fig. 3(a) shows that for the steels charged at EZn, the permeation curves experienced a
fast rising at the beginning to reach a maximum, then declined somewhat after a few hours,
attributed to the formation of an oxide film on the cathodic polarized surface, and finally
slowly increased with time. The permeation current density reached a steady state after about
60-hour charging. The steady-state permeation current density could be higher than the
maximum value during the fast rising stage. For the steels charged at Ecorr, the curves all
consisted of two stages: a rising stage and a steady state stage. The increase in the permeation
current density over this long time is attributed to changes at the ingress side of the specimen.
The hydrogen permeation current density increased with time, due to an increased amount of
hydrogen atoms that formed and permeated through the steel during the corrosion reaction,
and reached a steady state when the ingress surface conditions had reached steady state [9].
After about 60 hour at Ecorr in 3wt% NaCl solution, most of the surface area of the studied
steels was covered by a brown corrosion layer, which easily chipped off after removal of the
specimen from the cell, revealing a black layer of magnetite [23].
7
Table 2 presents the values of (i) the steady state permeation current density, i∞, (ii)
the effective hydrogen diffusivity, Deff, and (iii) the total hydrogen concentration, CH, under
simulated service conditions.
Table 2 Permeability parameter values for the studied DP and QP steels under (i) simulated
service conditions in 3wt% NaCl solution and (ii) at free corrosion in 0.1 M HCl solution. Steel
designation
Condition Over-potential
(η, V)
i∞
(µA cm-2
)
Deff
(cm2 s
-1)
CH
(µg g-1
)
Cd
(µg g-1
)
980 DP
at Ecorr in 3wt% NaCl -0.039 0.095 4.39 × 10-7
0.020 0.031
at EZn in 3wt% NaCl -0.322 1.480 8.78 × 10-7
0.133 -
at Ecorr in 0.1M HCl -0.203 0.748 7.24 × 10-7
0.129 -
1200 DP-GI at Ecorr in 3wt% NaCl -0.039 0.090 4.62 × 10
-7 0.022 0.053
at EZn in 3wt% NaCl -0.322 0.582 7.03 × 10-7
0.098 0.090
980 QP at Ecorr in 3wt% NaCl -0.039 0.053 1.71 × 10-7
0.018 0.010
at EZn in 3wt% NaCl -0.322 0.438 3.03 × 10-7
0.109 0.100
Fig. 3(a) and Table 2 show that the steady state permeation current density, i∞,
decreased in the following order for the steels at both EZn and Ecorr: 980 DP, 1200 DP-GI, and
980 QP. In addition, the i∞ values were significantly higher when hydrogen was charged at
EZn than for those at Ecorr in the 3wt% NaCl solution, as was also the case for the values of the
total hydrogen concentration, CH. For example, the values of i∞ and CH for 980 DP steel were
0.020 µg g-1
and 0.095 µA cm-2
at Ecorr, and 0.133 µg g-1
and 1.480 µA cm-2
at EZn, increased
at least by about 7 times.
Fig. 2 also presents the total hydrogen concentration at the input side of the steel
specimen, CH, for all the steels under simulated service corrosion, and provides a comparison
with the hydrogen concentrations produced by cathodic hydrogen charging.
3.3. In 0.1 M HCl solution
Fig. 3(b) presents the permeation current density versus time curves for 980 DP steel
at free corrosion in 0.1 M HCl solution and compares the data with those measured
previously under simulated auto service corrosion in 3wt% NaCl solutions at Ecorr and EZn, as
in Fig. 3(a). The permeation current density increased faster than that in 3wt% NaCl solutions
at EZn, reaching to a nearly steady-state after about 5 hours and remained steady thereafter.
Values of the steady state permeation current density, i∞, and the total hydrogen concentration,
CH, for 980 DP steel at Ecorr in 0.1 M HCl solution, were also included in Fig. 2 and Table 2,
and were found to be significantly higher than those for 980 DP at Ecorr in 3wt% NaCl
solution. The steel surface was observed to be fully covered by corrosion products in the 0.1
M HCl solution. The corroded area was measured after the experiments to be 3.002 cm2
and
2.014 cm2
for the steel at Ecorr in 0.1 M HCl and in the 3wt% NaCl solution, respectively.
3.4 Diffusible hydrogen content
Table 3 and Fig. 4 present the diffusible hydrogen concentration, Cd, for the studied
steels for (i) being uncharged, (ii) hydrogen cathodic charging in 0.1 M NaOH solution at
over-potential values of -0.257 V, -0.557 V and -0.857 V, (iii) free corrosion in 3wt% NaCl
solution, simulating free corrosion of the steel of the car body in service, (iv) cathodic
polarisation of the steel to the zinc potential in 3wt% NaCl to simulate the maximum amount
of hydrogen produced for the steels protected by a zinc coating applied by galvanisation, and
(iv) free corrosion in 0.1 M HCl solution.
The values of diffusible hydrogen concentration, Cd, for the uncharged samples were
not higher than 0.002 µg g-1
and were significantly lower than those at Ecorr in 3wt% NaCl.
The Cd values at Ecorr in 0.1 M HCl were much higher than those at Ecorr in NaCl. Comparing
the simulated auto service conditions in NaCl, the hydrogen concentrations at EZn were higher
than the corresponded values at Ecorr. For example, the diffusible hydrogen concentration for
8
980 QP at EZn in 3wt% NaCl was 0.100 µg g-1
, 10 times higher than 0.010 µg g-1
at Ecorr. For
those under cathodic charging potentials in NaOH solution, the hydrogen concentration
values were significantly higher than the values for simulated service conditions or at Ecorr in
0.1 M HCl solution. In addition, the hydrogen concentration increased with increasingly
negative charging potential for all the studied steels, such as for 980 DP, the hydrogen
concentration increased from 0.107 µg g-1
to 0.171 µg g-1
and up to 0.240 µg g-1
with the
charging over-potential changed from -0.257 V to -0.557 V and to -0.757 V. This trend was
consistent with the results from permeability experiments, as shown in Fig. 2.
Table 3 Diffusible hydrogen concentrations, Cd, in the studied steels under various hydrogen
conditions. Steel
designation
Conditions Charging potential
(VSCE)
Over-potential
(η, V)
Cd
(µg g-1
)
980 DP
uncharged - 0 0.002
at Ecorr in 3wt% NaCl - -0.039 0.031
in 0.1M NaOH -1.248 -0.257 0.107
in 0.1M NaOH -1.548 -0.557 0.171
in 0.1M NaOH -1.848 -0.857 0.240
1200 DP-GI
uncharged - 0 0.001
at Ecorr in 3wt% NaCl - -0.039 0.053
at EZn in 3wt% NaCl -0.998 -0.322 0.090
at Ecorr in 0.1M HCl - -0.203 0.070
in 0.1M NaOH -1.248 -0.257 0.120
in 0.1M NaOH -1.548 -0.557 0.218
in 0.1M NaOH -1.848 -0.857 0.300
980 QP
uncharged - 0 0.001
at Ecorr in 3wt% NaCl - -0.039 0.010
at EZn in 3wt% NaCl -0.998 -0.322 0.100
at Ecorr in 0.1M HCl - -0.203 0.085
in 0.1M NaOH -1.248 -0.257 0.197
in 0.1M NaOH -1.548 -0.557 0.252
in 0.1M NaOH -1.848 -0.857 0.347
Fig. 4 also provides a comparison between the measured diffusible hydrogen
concentration, Cd, and the prior measurements of the hydrogen concentration, Cd, from
permeability experiments, as presented in Fig. 2. Under simulated service conditions in 3%
NaCl solution, the diffusible hydrogen concentrations were similar to the values of total
hydrogen concentration determined by the permeability experiments, as was also the case for
that cathodically charged at over-potential of -0.257 V in NaOH solution. However, at more
negative charging over-potentials of -0.557 V and -0.857 V, the diffusible hydrogen
concentrations were lower than the total hydrogen concentration values obtained from our
permeability experiments.
9
Fig. 4 The measured diffusible hydrogen concentration, Cd, using the BRUKER G4
PHOENIX DH analyser in comparison with the hydrogen concentration, CH, determined
from permeability experiments.
4. Discussion
4.1. Hydrogen evolution
Under cathodic polarization or at the free corrosion potential, hydrogen evolution
reaction proceeds through reaction given by Eq. (1) or (2), and acts as part of the cathodic
partial reaction, balancing the anodic reaction. Ootsuka et al. [24] indicated that one mol of
hydrogen was absorbed for every 1000 mol of steel corroded, indicating that the hydrogen
evolution reaction compromised 0.1% of the cathodic reaction.
For the tests under cathodic polarization in 0.1 M NaOH and 3wt% NaCl solutions,
the electrons for hydrogen reduction reaction were provided by the external potentiostat,
whereas for those at free potential in 0.1 M HCl and 3wt% NaCl solutions, the electrons for
hydrogen evolution reaction came from the anodic reaction as following:
Fe → Fe2+
+ 2e (10)
Since the cathodic partial reaction as in Eq. (2) and the oxygen partial reduction
reaction balance the anodic partial reaction as in Eq. (10) in the 3wt% NaCl solution, a
change of either reaction rate causes an adjustment to minimize the effect of this change.
Cathodic polarization causes increased hydrogen evolution by Eq. (2) and decreased
corrosion rate by the rate of Eq. (10). This explains why the amount of corrosion for steels
cathodically polarized at EZn is significantly less than that at Ecorr in 3wt% NaCl solution, as
observed in our permeability experiments. During cathodic polarization in the 0.1 M NaOH
solution, hydrogen is produced due to the continuously provided electrons. A more negative
over-potential increased the current of the hydrogen evolution reaction as in Eq. (2), and
resulted in more hydrogen being produced.
In our study, for the experiments under cathodic charging in the 0.1 M NaOH solution,
prolonged charging was conducted. Thus, it was expected that the surface oxides were
reduced to a stable state, and there was equilibrium between the adsorbed hydrogen on the
hydrogen-entry side of the steel surface and the hydrogen dissolved into the steel [9]. For the
experiments in the 3wt% NaCl and 0.1 M HCl solutions, the same equilibrium was also
expected, since water molecule or hydronium ion diffused through gaps in the oxide film to
the steel surface and was reduced, releasing hydrogen adsorbed on the steel surface and then
introducing the adsorbed hydrogen into the specimen. Therefore, both of cathodic charging
10
and corrosion conditions in our study provided hydrogen evolution and the adsorbed
hydrogen on the steel surface was in equilibrium with the hydrogen dissolved into the steel,
so that hydrogen could be produced, be absorbed into the steel, permeate through and be
oxidized by the applied positive potential on the other side.
4.2. Permeation current density
For permeability experiments in the 3wt% NaCl and 0.1 M HCl solutions, the
adsorption of aggressive ions such as Cl- on the metal surface facilitates the corrosion rate of
the material and further the iron dissolution during corrosion depends on H+ ion concentration
more than on the Cl- ion concentration [25, 26]. Therefore, corrosion occurred in these
solutions at the free potential in this study and the corrosion rate in the 0.1 M HCl solution
was higher, reflected by the steeper initial slope of the permeation current density as in Fig.
2(b). As a consequence, more hydrogen was produced, entered into and diffused through the
specimen, leading to higher permeation rates and higher hydrogen concentrations at free
corrosion in 0.1 M HCl than in 3wt% NaCl, as shown in Fig. 2(b) and Table 3
For the experiments at EZn in the 3wt% NaCl solution, the cathodic polarization
increased the hydrogen reduction reaction, and decreased the corrosion of the steel, as stated
above. This explained the highest permeation rate and hydrogen concentration at EZn in 3wt%
NaCl solution, compared with the other conditions simulated auto service. During actual auto
service, the steel could be corroded, introducing hydrogen, and influencing the steel
properties, so the permeation experiments at Ecorr in 3% NaCl solution indicated how much
hydrogen had permeated under these conditions. In addition, in order to prevent the steel from
corrosion, steel can be galvanized, by applying a zinc coating on the steel surface. The
permeation experiments at EZn in the 3wt% NaCl solution indicated that this Zn coating could
significantly reduce the corrosion of the steel, but more hydrogen could be introduced into the
steel in addition to the hydrogen from the corrosion on the open surface of the steel. Thus,
more hydrogen would be introduced into and permeate through the galvanized steel if there is
access of the steel surface to the corroding solution.
As presented in Figs. 1 and 3, the permeation rate, i∞, for experiments under simulated
service conditions, comparing with cathodic charging conditions in NaOH solution, was
significantly lower, even much lower than those at -1.100 VHg/HgO, due to the lower rate of
hydrogen evolution reaction under simulated service conditions in NaCl solutions, despite
that the presence of Cl- that could accelerate the corrosion reactions. Besides, in the 0.1 M
NaOH solution, the permeation rate, i∞, increased with an increasingly negative potential, due
to the higher rate of hydrogen reduction reaction and a less potent trapping effect, which was
also indicated by the lower hydrogen trap site density as in our prior study [11]. This trend is
consistent with that in other research [9, 22].
4.3. Service versus cathodic charging
Fig. 2 presents the relationship between the total hydrogen concentration at the input
side of the steel specimen, CH, and the over-potential, η, for all the steels under simulated
service conditions, and under cathodic charging in the 0.1 M NaOH solution, determined
through permeability experiments. This allows a direct comparison between the hydrogen
concentrations produced under service conditions with those under cathodic charging
conditions. A turning point at over-potential of around -0.5 V was observed in the
relationship between CH and η for each steel under cathodic charging, as was also reported by
Bockris et al. [27] and Liu et al. [9]. This turning point was attributed to the change of
hydrogen evolution mechanisms from coupled discharge-recombination mechanism at lower
over-potentials to a slow discharge-fast electrochemical mechanism at higher over-potentials,
meaning that the electrochemical discharge of adsorbed hydrogen became more likely.
11
Fig. 2 shows that, for each steel, the hydrogen concentration at the input side of the
steel, CH, was significantly lower under simulated auto service conditions in 3wt% NaCl or in
0.1 M HCl solution than that under cathodic charging in 0.1M NaOH solution, even lower
than that at the least negative charging potential of -1.100 VHg/HgO, as was also the case for
hydrogen permeation current density. And further, extrapolating the experimental data
obtained in the 0.1 M NaOH solution, the CH value for all the steels in 0.1 M NaOH solution,
at each over-potential corresponding to that under simulated service corrosion conditions, was
also higher than that experimentally determined under simulated auto service conditions. This
indicates that at the same over-potential, more hydrogen entered in and permeated through the
steel specimen in the 0.1 M NaOH solution than that in the 3wt% NaCl solution. This is
considered attributed to the lower rate of hydrogen evolution reaction under simulated service
conditions in NaCl solutions, as also stated above in section 4.2.
4.4. Diffusible and total hydrogen content
Fig. 4 and Table 3 show that the amount of diffusible hydrogen in the as-received
steels was quite low, indicating that either there was little hydrogen in the steels due to steel
production, or that the hydrogen had effused out of the steels by the time that these steel
samples were tested in this work.
Fig. 4 also compares the amount of diffusible hydrogen, Cd, measured by hot
extraction analyser, with the total amount of hydrogen, CH, as measured by the permeability
experiments. There was good agreement between Cd and CH for the conditions of (i) free
corrosion in 3wt% NaCl solution, (ii) free corrosion in 0.1 M HCl, (iii) cathodic polarisation
of the steel to the zinc potential in 3wt% NaCl, and (iv) cathodic polarisation at the lowest
over-potential value of -1.100 VHg/HgO in the 0.1M NaOH solution. This good agreement
provides validation of using Eqs. (7) to (9) to analyse the permeation curves and estimate the
total hydrogen concentration, CH. Besides, Fig. 4 shows that, at the larger values of over-
potential, such as -0.557 V and -0.857 V, for cathodic hydrogen charging, the measured
amount of diffusible hydrogen, Cd, was somewhat less than the total amount of hydrogen, CH,
attributed to the loss of some diffusible hydrogen during the time between the end of the
cathodic charging and the start of measuring the amount of diffusible hydrogen. This delay
was inevitable since the specimen needed to be washed, dried, inserted into the measurement
apparatus and the commencement of heating for hydrogen extraction. The loss of diffusible
hydrogen by diffusing out of the sample is related to the amount of diffusible hydrogen and
was thus much larger at the higher hydrogen concentrations. In addition, as shown in prior
study [11], the effective diffusion coefficient tends to be larger at the higher concentrations
because of less trapping [9, 22], and consequently the diffusion loss is greater at the higher
hydrogen concentrations. In contrast, the total amount of hydrogen, CH, was measured by the
permeation technique under steady state conditions, and thus represented the total hydrogen
in the specimen.
5. Conclusions 1. The values of the effective diffusion coefficient measured in our work were consistent
with literature values.
2. The hydrogen permeation rates and hydrogen concentration increased with increasingly
negative charging potential in 0.1 M NaOH solution.
3. The hydrogen permeation rates and hydrogen concentrations for simulated auto service
conditions were significantly lower than the values at the least negative cathodic charging
potential of –1.100 VHg/HgO in 0.1 M NaOH solution.
4. The Zn coating on the steels could protect the substrate steel from corrosion, however,
once the Zn coating was corroded under actual service condition, more hydrogen could
enter and permeate through the steel.
12
ACKNOWLEDGEMENTS
This research is supported by the Baosteel-Australia Joint Research & Development Centre
(BAJC) Grant, BA13037, with linkage to Baoshan Iron and Steel Co., Ltd of China.
7. References
[1] WorldAutoSteel, Advanced high-strength steels application guidelines version 5.0, Available from http://www.worldautosteel.org/, Access on 2014. [2] X Zhu, Z Ma and L Wang, Current Status of Advanced High Strength Steel for Auto-making and its
Development in Baosteel, Rep Baosteel Research Institute. 201900 (2007) 1-8. [3] DK Matlock and JG Speer, "Third Generation of AHSS: Microstructure Design Concepts", in: Microstructure and Texture in Steels, Eds. A. Haldar, S. Suwas, D. Bhattacharjee, Springer London (2009) pp. 185-205. [4] A Lasia and D Grégoire, General model of electrochemical hydrogen absorption into metals, J
Electrochem Soc. 142 (1995) 3393-3399. [5] G Lovicu, M Bottazzi, F D’Aiuto, M De Sanctis, A Dimatteo, C Santus and R Valentini, Hydrogen
Embrittlement of Automotive Advanced High-Strength Steels, Metall Mater Trans A. 43 (2012) 4075-4087. [6] J Ronevich, B De Cooman, J Speer, E De Moor and D Matlock, Hydrogen effects in prestrained
transformation induced plasticity steel, Metall Mater Trans A. 43 (2012) 2293-2301. [7] M Devanathan and Z Stachurski, The adsorption and diffusion of electrolytic hydrogen in
palladium, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences. 270 (1962) 90-102. [8] Q Liu, B Irwanto and A Atrens, The influence of hydrogen on 3.5NiCrMoV steel studied using the
linearly increasing stress test, Corros Sci. 67 (2013) 193-203. [9] Q Liu, AD Atrens, Z Shi, K Verbeken and A Atrens, Determination of the hydrogen fugacity during
electrolytic charging of steel, Corros Sci. 87 (2014) 239-258. [10] J Venezuela, Q Liu, M Zhang, Q Zhou and A Atrens, The influence of hydrogen on the mechanical and fracture properties of some martensitic advanced high strength steels studied using the linearly
increasing stress test, Corros Sci. 99 (2015) 98-117. [11] Q Liu, J Venezuela, M Zhang, Q Zhou and A Atrens, Hydrogen trapping in some advanced high
strength steels, Corros Sci. Accepted June 01, 2016, DOI: 10.1016/j.corsci.2016.05.046. [12] Q Liu, Q Zhou, J Venezuela, M Zhang and A Atrens, Hydrogen concentration in dual phase (DP) and quenched and partitioned (Q&P) advanced high strength steels (AHSS) under simulated service
conditions compared with cathodic charging conditions, Adv Eng Mater. Accepted in May 2016, DOI: 10.1002/adem.201600217. [13] S Takagi, Y Toji, M Yoshino and K Hasegawa, Hydrogen embrittlement resistance evaluation of
ultra high strength steel sheets for automobiles, ISIJ Int. 52 (2012) 316-322.
[14] Y Toji, S Takagi, M Yoshino, K Hasegawa and Y Tanaka, Evaluation of hydrogen embrittlement for high strength steel sheets, Proceedings of Mater Sci Forum, 2010, 3537-3542. [15] J Flis, T Zakroczymski, V Kleshnya, T Kobiela and R Duś, Changes in hydrogen entry rate and in
surface of iron during cathodic polarisation in alkaline solutions, Electrochim Acta. 44 (1999) 3989-3997.
[16] J Malina, A Begić Hadžipašić and Š Nižnik, Electrochemical corrosion and hydrogen diffusivity in dual-phase steel, Proceedings of 2nd International Conference Corrosion and Material Protection, EFC Event, 2010, [17] A Begić Hadžipašić, J Malina and M Malina, The influence of microstructure on hydrogen diffusion and embrittlement of multiphase fine-grained steels with increased plasticity and strength,
Chem Biochem Eng Q. 25 (2011) 159-169.
13
[18] T Depover, E Wallaert and K Verbeken, Fractographic analysis of the role of hydrogen diffusion
on the hydrogen embrittlement susceptibility of DP steel, Mater Sci Eng A. 649 (2016) 201-208. [19] M Koyama, CC Tasan, E Akiyama, K Tsuzaki and D Raabe, Hydrogen-assisted decohesion and
localized plasticity in dual-phase steel, Acta Mater. 70 (2014) 174-187. [20] J Yang, F Huang, Z Guo, Y Rong and N Chen, Effect of retained austenite on the hydrogen embrittlement of a medium carbon quenching and partitioning steel with refined microstructure,
Mater Sci Eng A. 665 (2016) 76-85. [21] X Zhu, K Zhang, W Li and X Jin, Effect of retained austenite stability and morphology on the
hydrogen embrittlement susceptibility in quenching and partitioning treated steels, Mater Sci Eng A. 658 (2016) 400-408. [22] Q Liu and A Atrens, Reversible hydrogen trapping in a 3.5NiCrMoV medium strength steel, Corros
Sci. 96 (2015) 112-120. [23] H Moller, E Boshoff and H Froneman, The corrosion behaviour of a low carbon steel in natural
and synthetic seawaters, JOURNAL-SOUTH AFRICAN INSTITUTE OF MINING AND METALLURGY. 106 (2006) 585. [24] S Ootsuka, S Fujita, E Tada, A Nishikata and T Tsuru, Evaluation of hydrogen absorption into steel
in automobile moving environments, Corros Sci. 98 (2015) 430-437. [25] EA Noor and AH Al-Moubaraki, Corrosion behavior of mild steel in hydrochloric acid solutions, Int.
J. Electrochem. Sci. 3 (2008) 806-818. [26] M Sun, K Xiao, C Dong and X Li, Electrochemical corrosion behavior of 300M ultra high strength
steel in chloride containing environment, Acta Metall. Sin.(Engl. Lett.). 23 (2010) 301-311. [27] JM Bockris, J McBreen and L Nanis, The Hydrogen Evolution Kinetics and Hydrogen Entry into
[alpha]-Iron, J Electrochem Soc. 112 (1965) 1025-1031.