Investigating electrochemical charging conditions equivalent to hydrogen gas exposure of X65 pipeline steel

In this study, we systematically investigate electrochemical hydrogen charging conditions equivalent to hydrogen gas pressures relevant for hydrogen transportation in X65 pipeline steel. By performing hydrogen gas permeation, a relationship for Sieverts' law was established, which was used in combination with electrochemical hydrogen permeation to determine the equivalent hydrogen pressure. The results revealed that cathodic protection simulated condition at –1050 mVAg/AgCl was equivalent to a hydrogen pressure of 12.3 bar. The addition of thiourea, a hydrogen recombination poison, and changing the applied potential in the cathodic direction increased the equivalent hydrogen pressure. In this way, an electrochemical charging condition equivalent to a potential hydrogen gas pressure for hydrogen transportation (200 bar) was determined.


| INTRODUCTION
With hydrogen emerging as a clean energy carrier, [1] the transportation and storage of pressurized hydrogen gas have gained recent interest, [2][3][4][5][6] for example, repurposing existing natural gas pipelines for hydrogen gas transportation. [7]Pipelines should withstand pressures in the range 150-200 bar, due to the low energy density of hydrogen gas. [8]Hydrogen can degrade a material's mechanical properties, a phenomenon known as hydrogen embrittlement (HE). [9]Because of safety concerns using pressurized hydrogen gas, electrochemical charging is sometimes used as a substitute for hydrogen gas exposure during investigations of HE susceptibility of a material. [10,11]e severity of HE effects is highly dependent on the hydrogen charging condition. [2,12]Relating electrochemical and gaseous charging, would enhance the comparability of HE investigations performed under different charging conditions.The equivalence between electrochemical and gaseous charging has previously been investigated by hydrogen permeation, [13][14][15][16] thermal desorption analysis, [17][18][19] and by measuring the hydrogen pressure developed in a hollow sensor during charging. [20]hile these studies investigated the pressure equivalency from common electrochemical charging conditions, a systematic approach to determine electrochemical charging conditions equivalent to pressures relevant for hydrogen gas transportation is lacking.
Chemical additives can enhance hydrogen absorption during electrochemical charging. [21,22]Several mechanisms explaining the increased hydrogen absorption have been proposed, which have been discussed in detail in Protopopoff and Marcus. [23]One prominent hypothesis is that the enhanced hydrogen absorption is caused by a chemisorbed element or species obstructing the recombination of adsorbed atomic hydrogen, hence, these additives are commonly referred to as hydrogen recombination poisons. [24]Increasing the cathodic current density or cathodic potential are other measures to enhance hydrogen uptake, [17,25] however, high cathodic current densities can cause unsteady surface conditions due to extensive bubble formation, changes in the local pH, or detachment of second phase particles. [26,27]Akiyama and Li investigated the application of hydrogen recombination poison and different cathodic current densities to optimize electrochemical charging, [22] however, its equivalence to gaseous hydrogen charging was not determined.
Herein, the objective is to optimize the electrochemical charging conditions to obtain a desired and predetermined equivalent hydrogen pressure, which simulates a potential hydrogen pressure level (200 bar) in a pipeline system aimed for hydrogen transportation.

| Material and sample preparation
The steel used in this study was API 5L X65 grade, from a quenched and tempered, seamless pipe.The wall thickness was 15.4 mm.It consisted of homogeneously distributed grains of ferrite and bainite.The phase volumes of ferrite and bainite were 89% and 11%, respectively.The average grain size was 7 µm.The chemical composition is given in Table 1.The samples were machined from a position close to the inner side of the pipe wall, parallel to the longitudinal direction.Plates with dimensions of 50 × 150 × 1.9 mm (width × length × thickness) were used for hydrogen gas charging.Disks with dimensions of 30 mm diameter and 1.1 mm thickness were used for electrochemical charging.The exposed areas during gaseous and electrochemical charging were 16 and 3.7 cm 2 , respectively.Both sides of all samples were prepared with SiC grinding papers to a final grade of #P1000.A Pd coating was electrodeposited on the detection side of all samples, following the method described by Husby et al. [28] based on the work of Bruzzoni et al. [29] and Castaño Rivera et al. [30] Samples used for gas charging were also coated with Pd on the charging side.Pd facilitates hydrogen oxidation on the detection side, [31] while on the charging side, Pd is applied to overcome the surface impedance which can be caused by an oxide layer during gaseous hydrogen charging. [3,13,16]

| Hydrogen permeation tests
The hydrogen permeation experiments follow the principles first described by Devanathan and Stachurski. [32]A detailed description of the gas permeation set-up and electrochemical permeation set-up was described in a previous paper. [16]n the detection chamber, for both gas and electrochemical permeation, a 0.1 M NaOH solution (pH = 12.6) and an applied potential of +315 mV SCE was used to ensure hydrogen oxidation.For gaseous hydrogen charging, the applied hydrogen gas pressures were 10, 50, and 100 bar.In the charging chamber, a 3.5 wt% NaCl solution (pH = 6.6) with an applied potential of -1050 mV Ag/AgCl was used to simulate cathodic protection (CP) conditions.To enhance the uptake of hydrogen, thiourea (CH 4 N 2 S) with various concentrations (0.5-3 g L -1 ) was added and the cathodic potential was changed in steps of 75 mV in the cathodic direction.Two charging and discharging cycles were performed for all charging conditions.All tests were performed at room temperature (21 ± 1°C).
To determine the effective hydrogen diffusion coefficient, D eff , partial permeation transients were conducted in a 3.5 wt% NaCl solution with a 2 g L -1 thiourea addition.Performing partial permeation transients can reduce surface and trapping effects, such that the permeation transients can be fitted to solutions of Fick's second law. [33]Initially, a charging potential of -1050 mV Ag/AgCl was applied until a steady permeation current density was achieved.Subsequently, the applied charging potential was changed every 5500 s in the following order: -1200, -1350, -1200, and -1050 mV Ag/AgCl , such that two build-up transients and two decay transients were obtained.

| Analysis
The subsurface hydrogen concentration in lattice and reversible trap sites, C 0R , is proportional to the steady-state permeation current density, ∞ i p . [34]C 0R was calculated using Equation ( 1), where L denotes the thickness of the sample and F denotes the Faraday constant (96,485 A s mol -1 ).
D eff was determined by fitting the partial permeation transients to analytical solutions of Fick's second law to the build-up transient, Equation (2), and the decay transient, Equation (3). [33]i p denotes the permeation current density, i p 0 denotes the permeation current density at the time of changing the applied potential, t denotes the time after changing the applied potential, and D denotes the hydrogen diffusion coefficient.
According to Sieverts' law, the amount of hydrogen dissolved in steel is proportional to the square root of the hydrogen fugacity, f H 2 , where S denotes Sieverts' con- stant. [35]At steady state, the lattice hydrogen concentration will scale as the reversibly trapped hydrogen concentration, [36] hence, Sieverts' law can be expressed as shown in Equation (4).f H 2 is related to the hydrogen pressure, p H 2 , through the Abel-Noble equation of state, Equation (5), where b denotes a constant (1.584 × 10 −5 m 3 mol −1 ), T denotes the temperature in K, and R denotes the gas constant (8.314J K −1 mol −1 ). [37]The method to determine the equivalent hydrogen fugacity, f H eq 2 , is described in detail in a previous paper [16] and is based on work reported by Atrens and colleagues. [14,15,17]It is considered that charging conditions that produce the same C 0R in a steel are equivalent.In such matter, f H eq 2 was calculated by inserting C 0R determined from electrochemical charging into Equation ( 4) and solved for f H 2 .Then Equation ( 5) was solved iteratively using the Newton-Raphson method to determine the equivalent hydrogen pressure, p H eq 2 , from f H eq 2 . (5)

| RESULTS AND DISCUSSION
The normalized partial permeation transients of the build-up and decay are presented in Figure 1a and 1b, respectively.The predicted solution of Fick's second law for lattice hydrogen diffusivity (D L = 7.27 × 10 -5 cm 2 s -1 ) [38] is included in the figures, as well as the average of the best fits to Equations ( 2) and ( 3).In alloys such as X65 pipeline steels, a significant trap occupancy can cause the normalized permeation transient to be steeper than predicted by Fick's second law. [39]The consequence can be a false interpretation of D eff , which would appear to be changing with time. [39]By using the method of partial permeation transients, the effect of changing trap occupancy on the transient is reduced, [33] and the partial permeation transients can be fitted to the analytical solutions of Fick's second law.The average determined D eff from the partial permeation transients is 2.3 × 10 -6 cm 2 s -1 , which was used for the determination of C 0R .
Figure 2a shows two hydrogen permeation transients performed at 10, 50, and 100 bar.∞ i p increases with increasing applied p H 2 , manifesting that the hydrogen flux through the samples increases with increasing p H 2 .C 0R determined from gas charging at 10, 50, and 100 bar is 0.031, 0.066, and 0.095 wppm, respectively.Figure 2b shows the hydrogen permeation transients obtained in CP-simulated conditions.The corresponding C 0R obtained is 0.032 wppm.As shown in Figure 2c, C 0R obtained from gas charging is proportional to f H 1/2 2 , which is consistent with Sieverts' law (Equation 4).The calculated S is 0.0095 wppm bar -1/2 , which is slightly lower than determined in a previous study for a hot rolled and welded X65 pipeline steel. [16]S is almost three times less than reported for a 3.5 NiCrMoV martensitic steel, [18] however, it is two to three times higher than for a martensitic advanced high strength-steel [17] and a dual phase steel. [19]It is emphasized that the surface condition during gas permeation differs significantly from the inside of a pipeline in operation, for example, presence of an oxide layer.However, pipelines are subjected to cyclic stresses caused by pressure fluctuations which can rupture an oxide layer and expose the bare steel surface.It is reported in the literature, that exposure to pressurized hydrogen gas reduces the fracture toughness and increases the fatigue crack growth rate of pipeline steels, [4][5][6] indicating that hydrogen uptake occurs locally.The linearity between C 0R and f H 1/2 2 and S being comparable to reported values suggests that the Pd-coated surface can resemble the hydrogen uptake of a bare steel surface after breakage of an oxide layer, however, this should be investigated further.In Figure 2c, the C 0R obtained at CP simulated conditions is superimposed onto the linear regression line determined for Sieverts' law, which corresponds to a f H eq 2 of 12.4 bar.Based on Equation ( 5), p H eq 2 is 12.3 bar, which is significantly lower than the desired hydrogen pressure of 200 bar.
Two methods to increase the hydrogen uptake are the addition of a hydrogen recombination poison, [22,26] or to modify the cathodic charging parameters (i.e., apply a more negative potential or higher cathodic current density). [15,25]ased on the previous study, [16]  (corresponding to a potential of about -2600 mV Ag/AgCl ).However, it can be difficult to maintain stable surface conditions at such high cathodic current density.Unsteady surface conditions can be a result of excessive hydrogen bubble formation or changes in local pH, leading to fluctuations in hydrogen flux or detachment of inclusions. [26,27,40]In addition, the increasing influence of electrochemical recombination at higher cathodic current densities [41] can reduce the effect of applied current density or potential on the p H eq 2 .Consequently, it was decided to use a hydrogen recombination poison as the first measure to increase the hydrogen absorption, before changing the applied cathodic potential.
Thiourea is a hydrogen recombination poison often used to enhance the hydrogen uptake. [21]Hydrogen permeation transients obtained at a cathodic potential of -1050 mV Ag/AgCl in a 3.5 wt% NaCl solution with 0-3 g L -1 thiourea, are shown in Figure 3a.An addition of 0.5 g L -1 thiourea increases ∞ i p by approximately 2.5 times compared to the pure 3.5 wt% NaCl solution.Further, the additions of 1.0 and 2.0 g L -1 thiourea about triples ∞ i p .The p H eq 2 values as a function of thiourea concentration are presented in Figure 3b.It can be observed that p H eq 2 increases initially with an increasing thiourea concentration until 2 g L -1 , where a decrease in p H eq 2 started.The reduced hydrogen uptake at high hydrogen recombination poison concentration could be attributed to the formation of deposits acting as an inhibiting layer to hydrogen uptake. [42]Because the addition of thiourea reached its limit contribution in p H eq 2 at 111.5 bar, it was necessary to apply more negative cathodic potential to obtain the desired p H eq 2 of 200 bar.
Hydrogen permeation transients obtained with different applied cathodic potentials in a 3.5 wt% NaCl solution with 2 g L -1 thiourea concentration are presented in Figure 4a.The ∞ i p increases when more negative cathodic potentials are applied.Compared with that obtained at -1050 mV Ag/AgCl, ∞ i p increases by 1.17 times at -1125 mV Ag/ AgCl , 1.35 times at −1200 mV Ag/AgCl , and 1.68 times at -1275 mV Ag/AgCl .4b.By using Equation ( 5), it is calculated that a 200 bar p H 2 corresponds to a f H 2 of 227.2 bar.The relationship between f H eq 2 and applied potential from Figure 4b was solved for 227.2 bar, and the determined corresponding potential was approximately -1225 mV Ag/ AgCl .Following the above deduction, a permeation transient was performed at a potential of -1225 mV Ag/AgCl in 3.5 wt% NaCl solution with 2 g L -1 thiourea concentration, and the results are included in Table 2.The corresponding C 0R is 0.137 wppm and the p H eq 2 is 201.9 bar.Thus, electrochemical charging at a potential -1225 mV Ag/AgCl in 3.5 wt% NaCl solution with 2 g L -1 thiourea concentration can represent approximately the same C 0R as a p H eq 2 of 200 bar.The determined equivalence in terms of hydrogen uptake was obtained with a Pd coating during gas permeation.However, in real service conditions, the equivalence can be affected by relevant environmental parameters (e.g., the presence of surface oxides, temperature, and local stress conditions), which should be carefully considered in the design of HE susceptibility investigations.In addition, the transferability of the determined equivalent hydrogen pressure to mechanical testing needs to be validated.This study established an equivalence between electrochemical and gaseous hydrogen charging by the addition of a hydrogen recombination poison in combination with different charging potentials.The results show that the subsurface hydrogen concentration was proportional to the square root of the hydrogen fugacity, in agreement with Sieverts' law, which formed the basis for determining the equivalent hydrogen pressure.Electrochemical charging at the condition simulating cathodic protection at -1050 mV Ag/AgCl was equivalent to a hydrogen pressure of 12.3 bar at room temperature.Higher equivalent pressures were achieved by adding thiourea to the 3.5 wt% NaCl solution, and by applying more negative cathodic potentials.The equivalent pressure as a function of thiourea concentration went through a maximum at 2 g L -1 .The desired equivalent pressure of about 200 bar, which is a potential hydrogen pressure during hydrogen gas transportation using pipeline steel, was obtained in a 3.5 wt% NaCl solution with 2 g L -1 thiourea and an applied potential of -1225 mV Ag/AgCl at room temperature.The equivalent hydrogen pressure was deduced using a Pd coating during gas permeation, and further evaluation must be conducted to elucidate the effect of surface conditions and transferability to mechanical testing.

F I G U R E 1
Normalized partial permeation transients obtained in a 3.5 wt% NaCl solution with a 2 g L -1 thiourea addition, including the prediction by Fick's second law for lattice diffusion and the average of the best fits.(a) Step 1: -1050 to -1200 mV Ag/AgCl , Step 2: -1200 to -1350 mV Ag/AgCl .(b) Step 3: -1350 to -1200 mV Ag/AgCl , Step 4: -1200 to -1050 mV Ag/AgCl .[Color figure can be viewed at wileyonlinelibrary.com]

2 2
a p H eq 2 of 196.4 bar was obtained at a current density of -50 mA cm -2 F I G U R E 2 (a) Hydrogen permeation transients obtained at 10, 50, and 100 bar hydrogen gas charging pressures.(b) Hydrogen permeation transients were obtained at -1050 mV Ag/AgCl in a 3.5 wt% NaCl solution.(c) Subsurface hydrogen concentration, C 0R , versus the square root of hydrogen gas fugacity, f H 1/, with C 0R obtained at -1050 mV Ag/AgCl in a 3.5 wt% NaCl solution superimposed onto Sieverts' law regression line.[Color figure can be viewed at wileyonlinelibrary.com]

F
I G U R E 3 (a) Hydrogen permeation transients obtained at -1050 mV Ag/AgCl in a 3.5 wt% NaCl solution with different thiourea concentrations.(b) Equivalent hydrogen pressure, p H eq

F
I G U R E 4 (a) Hydrogen permeation transients performed at different cathodic potentials in a 3.5 wt% NaCl solution with 2 g L -1 thiourea.(b) Equivalent hydrogen fugacity, f H eq 2 , versus the applied potential.[Color figure can be viewed at wileyonlinelibrary.com]T A B L E 2 C 0R , f H eq 2 , and p H eq 2 obtained by charging at different cathodic potentials in a 3.5 wt% NaCl solution with 2 g L -1 thiourea.
The chemical composition of the API 5L X65 studied steel.
T A B L E 1

eq 2 .
Table 2 lists the determined C 0R , f H At -1200 and -1275 mV Ag/AgCl the p H