Hydrogen Embrittlement of CoCrFeMnNi High-Entropy Alloy Compared with 304 and IN718 Alloys

Abstract

The hydrogen embrittlement (HE) behaviors of a CoCrFeMnNi high-entropy alloy (HEA), 304 stainless steel (304SS) and IN718 alloys were studied and compared via electrochemical hydrogen pre-charging, slow strain rate tensile tests, and fracture surface analysis. The results demonstrate that the HEA exhibited the greatest HE-resistance, followed by 304SS and then IN718 alloy, when the alloys were charged at 1.79 mA cm⁻² for 24 h and 48 h, and 179 mA cm⁻² for 2 h. Hydrogen-induced reduction in ductility was observed for 304SS and IN718 alloys, whereas the hydrogen-affected fracture strain of the HEA was dependent on the hydrogen charging time. The resistance to HE was improved at a short hydrogen charging time (24 h), but reduced at a long charging time (48 h). This is attributed to the competing mechanisms between hydrogen-enhanced twin formation and HEDE (hydrogen-enhanced decohesion).

1. Introduction

In 2004, Ye et al. [1] proposed a novel design to produce multi-component alloys formed by mixing equal or high proportions of five or more elements. These are termed high-entropy alloys (HEAs) and regarded as a breakthrough in alloy material development. Unlike traditional alloys, HEAs comprise 5–13 principal elements, with each of element having an atomic fraction ranging from 5–35% [2]. Due to the presence of multiple principal elements in the base alloy, these materials will typically exhibit four effects [2,3]. These include high-entropy, severe lattice distortion, sluggish diffusion, and cocktail effect. These are accompanied with excellent mechanical properties, such increased hardness [4,5] and strength [6,7], wear [8,9], corrosion [10,11] and hydrogen embrittlement (HE) resistance [12,13], tolerance to radiation [14,15], and brittleness at low temperatures [16,17]. As a result, HEAs have potential application to aerospace, petroleum, nuclear power, and other specialist fields or industries.

Hydrogen embrittlement (HE) of alloys occurs when alloys are exposed to hydrogen-containing environments [18], such as high-pressure hydrogen gas and corrosion. This will result in premature failure of the alloy. Since 2014, various studies concentrating on HE behavior and characteristics of HEAs have been reported [19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34]. These studies can be broadly divided into two categories: the mechanical properties of HEAs in the presence of hydrogen and hydrogen-assisted cracking mechanism. On the one hand, hydrogen can increase or decrease yield strength and ductility of HEAs, depending on the hydrogen concentration. At low hydrogen concentrations (8.01–33.25 wppm), Luo et al. [35] reported that both the strength and ductility of CoCrFeMnNi alloy were improved after hydrogen charging, which was attributed to the formation of twins. Zhao et al. [12] reported that hydrogen (76.5 wppm) had a negligible influence on yield strength and fracture strain of CoCrFeMnNi alloy when compared to uncharged samples. However, under severe hydrogen charging condition (129 wppm), Koyama et al. [20] reported that the fracture strain of CoCrFeMnNi alloy decreased from 40% to 20% after hydrogen charging. Nygren et al. [29] reported that as hydrogen concentration was 146.9 wppm, fracture strain of CoCrFeMnNi alloy reduced from 67% to 34%. On the other hand, a hydrogen-enhanced-plasticity mediated decohesion mechanism has widely used to explain hydrogen-assisted cracking. This includes intergranular fracture in CoCrFeMnNi alloy [26], cracking at dual-phase interfaces of metastable Co₁₀Cr₁₀Fe₅₀Mn₃₀ alloy [32] and cracking at the interfaces between Co_19.9Cr_19.9Fe_19.9Mn_19.9Ni_19.9C_0.5-matrix and carbides [19]. In contrast, Xie et al. [33] reported hydrogen-retarded spallation of CoCrFeMnNi alloy under plate impact loading, which was proposed to be related nucleation of microvoids. This was thought to be caused by high migration energies of hydrogen-vacancy complexes, and a reduction in the growth rate of micro-voids due to the formation of nano twins.

The choice of material used in hydrogen containing environments is critical to ensure the reliability and integrity of the components. To determine a suitable alloy for a given application, a comparison of HE behavior must be made between traditional and HEAs. For example, Zhao et al. [12] exposed 304, 316L, and CoCrFeMnNi alloys to high-pressure hydrogen gas (15 MPa, 300 °C, 72 h). This study demonstrated that plastic loss of the HEA was reduced by 5%, in comparison with 304 (61%) and 316L (27%) steels. Pu et al. [13] compared the susceptibility of CoCrFeMnNi, X80 pipeline steel and 316L stainless steel to HE at room temperature (298 K) and low temperature (77 K). The study revealed that the CoCrFeMnNi alloy exhibited the greatest resistance to HE than other steels, with the embrittlement indexes 13%, 25%, and 53% for HEA, 316L, and X80 steels, respectively. However, the reported data on HE comparison between traditional alloys and HEAs are limited. In addition, a comprehensive understanding of hydrogen diffusion behavior and deformation mechanisms between traditional alloys and HEAs is still lacking.

In this study, the HE and hydrogen-assisted failure mechanisms were investigated for the CoCrFeMnNi alloy, a classical type of HEAs. In particular, the microstructural features of hydrogen-assisted brittle zone were observed and compared to samples without hydrogen charging. Experiments were also conducted on two alloys, i.e., 304 stainless steel (304SS) and IN718 alloy, which both exhibit a face centered cubic (fcc)-microstructure and are often used in hydrogen-containing conditions. The susceptibility of these alloys to HE was evaluated by tensile testing. Finally, the hydrogen-tolerance mechanisms and hydrogen diffusivity behavior were investigated using a combination of fracture surface observation, hydrogen diffusion modelling and microstructural evolution analysis.

2. Experimental Materials and Methods

2.1. Materials

An equiatomic CoCrFeMnNi ingot was prepared by using magnetic levitation melting technology. Pure metals (>99.9%) were used during this process. To ensure homogeneity, the ingot was remelted five times before casting into a mold. The cast ingot was hot-rolled and then solution-annealed at 1000 °C. For comparison, typical solution-annealed and -aged 304SS [36] and IN718 alloy [37] were also prepared. Their chemical compositions are listed in

Table 1. Chemical compositions of CoCrFeMnNi HEA, 304SS and IN718 alloy (wt. %).

	Fe	Cr	Ni	Mn	Co	C	Mo	Si	S	P	Cu	Al	Ti	Nb
HEA	19.92	18.54	20.93	19.59	21.02
304SS	Bal.	19.06	8.15	1.01		0.045	0.016	0.5	0.012	0.031	0.019
IN718	19.2	18.75	Bal.			0.028	3	0.078				0.49	0.97	5.35

. Flat dog-bone-shaped samples for tensile testing were prepared from bulk specimens, having a gauge length of 20 mm, width of 5 mm, and a thickness of 2 mm (Figure 1).

2.2. Microstructural Characterization

Samples for microstructural characterization were prepared by undertaking coarse- and fine-grinding, followed by mechanical polishing. The samples were etched first and then observed by optical microscopy (OM). CoCrFeMnNi alloy samples were etched for 90 s in a solution consisting of alcohol (25 mL) and hydrochloric acid (25 mL) with copper sulfate pentahydrate added (5 g). A solution of hydrochloric acid (100 mL), alcohol (100 mL), and copper chloride (5 g) was used to etch IN718 alloy. For 304SS, electrolytic polishing was utilized, with the process performed in a solution of acetic acid (200 mL), perchloric acid (10 mL), and glycerin (10 mL). To obtain high quality polishing, electrolytic polishing was carried out at low temperature through the addition of liquid nitrogen. The voltage and time were 30 V and 150 s, respectively.

Transmission electron microscopy (TEM) was used to investigate the changes in the microstructure of hydrogen-charged and uncharged samples. Foils were extracted using a focused ion beam (Helios NanoLab 600 FIB, Lausanne, Switzerland) with the operational voltage and current 30 kV and 10 nA, respectively. The samples were observed using a high-resolution TEM (JEM-F200(HR), JEOL, Tokyo, Japan) operated at 200 kV.

2.3. Electrochemical Hydrogen Pre-Charging

To minimize the impact of surface roughness on HE of alloys, all specimens were mechanically ground in sequence using #120, #240, #400, #600, #800, to #1000-grit SiC sandpapers. Hydrogen was introduced into the samples by electrochemical pre-charging at room temperature (298 K). Hydrogen charging was conducted in a H₂SO₄ (0.5 mol L⁻¹) solution, with the addition of CH₄N₂S (1 g L⁻¹) to prevent hydrogen escape [38]. The DC power supply was connected to the sample (cathode) and a Pt sheet acted as the anode. During hydrogen charging, the current density and time were 1.79 mA cm⁻² for 24 h and 48 h, respectively. Severe hydrogen charging condition, 179 mA cm⁻² for 2 h, was also conducted. It is well documented that the HE of alloys is sensitive to the stress concentration [39]. To avoid the fracture at regions with high stress concentration, both transition arc and root sections of the samples were sealed by a paraffin prior to hydrogen charging, as illustrated in shaded region in Figure 1.

2.4. Slow Strain Rate Tensile Test

After hydrogen pre-charging, samples were immediately tensioned on a universal electronic tensile machine (Instron1195). The HE characteristics of alloys are dependent on the loading rate [40]. A slow tensile strain rate should be adopted to properly ascertain the HE susceptibility of the alloys. However, more hydrogen will escape from the samples driven by hydrogen concentration gradient under slow tensile rate, resulting in a reduction in alloy HE. Considering both factors, a tensile cross-head rate of 0.06 mm min⁻¹ was utilized, and the corresponding strain rate was at 5 × 10⁻⁵ s⁻¹.

In according with GB/T 228.1-2010 standard, the yield strength and tensile strength of hydrogen-charged and uncharged samples were measured from the stress–strain curves. Based on the fractured samples, the reduction in the area was measured. As the effect of hydrogen on mechanical properties of alloys is primarily reflected on plastic loss, the hydrogen embrittlement index ($\mathrm{HEI}$) of the samples was assessed by the relative losses of reduction in area, which can be described as:

$$\mathrm{HEI}=\frac{{\mathsf{\psi }}_{\mathrm{HU}}-{\mathsf{\psi }}_{\mathrm{HC}}}{{\mathsf{\psi }}_{\mathrm{HU}}}\times 100\%$$

(1)

where ${\mathsf{\psi }}_{\mathrm{HU}}$ and ${\mathsf{\psi }}_{\mathrm{HC}}$ indicated the reduction in area of hydrogen-charged and uncharged samples, respectively.

2.5. Fracture Surface Observation

After tensile testing, samples with a height of 5–8 mm near the fracture surface were cut. Following this, they were ultrasonically cleaned in an alcohol solution for 10 min and dried. Samples were mounted on Al substrate and observed by scanning electron microscopy (Hitachi SU6600 SEM, Tokyo, Japan). The brittle zone area fraction ($\mathrm{BZAF}$) was quantified by the following equation:

$$\mathrm{BZAF}=\frac{2\mathrm{D}(\mathrm{L}+\mathrm{W})}{\mathrm{L}\times \mathrm{W}}\times 100\%$$

(2)

where $\mathrm{L}$ and $\mathrm{W}$ were the length and width of fracture surface. $\mathrm{D}$ was the brittle zone depth.

3. Results

3.1. Initial Microstructure

OM images of CoCrFeMnNi, 304SS and IN718 alloys are presented in Figure 2. The alloy samples all exhibit equiaxed grains (Figure 2a–c), with grain size histograms quantitatively analyzed (Figure 2a′′–c′′)). The grain sizes of the alloys follow a normal distribution, and the average grain sizes of the HEA, 304SS and IN718 alloy are 10 μm, 54 μm, and 15 μm, respectively. High magnification images indicate that the HEA shows a single-phase microstructure (Figure 2a′), whereas elongated carbides are presented within the grains and along grain boundaries for samples of 304SS (Figure 2b′) [36], and IN718 alloy has needle- and rod-shaped δ phases (Figure 2c′) in the alloy grains [37]. As these alloys exhibit a fcc-structure with low stacking fault energy (20 mJ m⁻² for CoCrFeMnNi [41], 20–25 mJ m⁻² for 304SS [42,43], and 50 mJ m⁻² for IN718 alloy [44,45]), a number of twins are observed, as marked by the white arrows.

3.2. Tensile Mechanical Properties

Stress–strain curves of hydrogen-charged and uncharged CoCrFeMnNi, 304SS, and IN 718 alloy samples are shown in Figure 3. For uncharged samples, the stress–strain curve of the CoCrFeMnNi alloy is divided into elastic, uniform plastic deformation and necking stages. However, 304SS and IN718 alloys behave differently, with two stages observed, namely elastic and uniform deformation. The measured yield strength and fracture strain are 463 MPa and 39% for the CoCrFeMnNi alloy, 340 MPa and 104% for 304SS, and 1176 MPa and 17% for IN718 alloy, respectively. These results indicate that the HEA exhibits a trade-off between strength and ductility.

In the presence of hydrogen, the mechanical properties of the alloys are significantly different. The degree of plastic reduction is related to the types of alloys and hydrogen charging parameters. For HEA, both strength and ductility are simultaneously enhanced when the samples were charged at 1.79 mA cm⁻² for 24 h and at 179 mA cm⁻² for 2 h. With an increase in the hydrogen charging time (1.79 mA cm⁻² for 48 h), the yield strength still increases, but the fracture strain is reduced to 33%. In contrast, hydrogen has a negative effect on mechanical properties of 304SS and IN718 alloy, especially of the ductility. Regardless of the hydrogen charging parameters, the yield strength of both alloys slightly enhances after hydrogen charging due to solid solution strengthening [46]. However, the fracture strain is reduced to half that of uncharged samples. The yield strength, tensile strength, fracture strain, reduction in area, and HEI of the samples with and without hydrogen charging are presented in

Table 2. Mechanical properties of hydrogen-charged and uncharged samples of CoCrFeMnNi, 304SS, and IN718 alloy.

Material	Hydrogen Charging	Tensile Strength (MPa)	Yield Strength (MPa)	Fracture Strain (%)	Reduction in Area (%)	HEI (%)
CoCrFeMnNi	0	684	463	39	71
	1.79 mA cm−2—24 h	672	471	41	74	−3.86
	1.79 mA cm−2—48 h	691	468	33	68	4.38
	179 mA cm−2—2 h	709	493	39	72	−0.98
304SS	0	759	340	104	0.56
	1.79 mA cm−2—24 h	670	342	63	0.43	23.21
	1.79 mA cm−2—48 h	679	345	57	0.38	32.20
	179 mA cm−2—2 h	701	344	68	0.45	19.64
IN718	0	1481	1176	17	0.19
	1.79 mA cm−2—24 h	1344	1261	8	0.10	47.37
	1.79 mA cm−2—48 h	1476	1295	9	0.11	42.11
	179 mA cm−2—2 h	1464	1222	15	0.12	36.84

Note: negative HEI indicating that hydrogen increases the ductility of the alloy.

. Under same hydrogen charging conditions, the HEI of CoCrFeMnNi is the lowest, indicating that the HEA has superior HE-resistance in comparison to other alloys. It is worth noting that negative HEI values presented in Table 2 indicate that the reduction in area actually increases in the presence of hydrogen. This effect was also reported by Luo et al. [35] and Zhang et al. [47].

3.3. Fracture Surface

Fracture surfaces of hydrogen-charged and uncharged alloy samples are shown in Figure 4, Figure 5 and Figure 6, respectively. For uncharged samples, CoCrFeMnNi distinctly shows necking (Figure 4a), a typical feature of plastic deformation. No necking phenomena are observed in samples of 304SS (Figure 5a) and IN718 alloy (Figure 6a). These results are consistent with the features of the relevant stress–strain curves. The fracture surfaces are composed of three parts: a center zone, transition zone, and shear slip. Images at high magnification demonstrate that the center zone consists of large and deep dimples with various sizes. Small and shallow dimples are observed in the shear slip region, as shown in Figure 4a′–a′′′, Figure 5a′–a′′′, and Figure 6a′–a′′′ for CoCrFeMnNi, 304SS and IN718 alloys, respectively. After hydrogen charging, there are no significant changes in the fracture appearances, compared with hydrogen-uncharged samples (Figure 4b–d, Figure 5b–d, and Figure 6b–d). It is well documented that the hydrogen diffusion coefficients in fcc-microstructural alloys are typically very low. During electrochemical hydrogen charging, a hydrogen concentration gradient is generated near the sample′s surface [34,48], leading to a hydrogen-assisted brittle zone (Figure 4b′–d′, Figure 5b′–d′ and Figure 6b′–d′). As expected, these regions exhibit brittle fractures with mixed intergranular and quasi-cleavage modes for the CoCrFeMnNi (Figure 4b′′–d′′) and IN718 alloy (Figure 6b′′–d′′), with quasi-cleavage fracture observed in samples of 304SS (Figure 5b′′–d′′). Ductile dimple patterns cover the center zone of all three alloys due to the lack of hydrogen (Figure 4b′′′–d′′′, Figure 5b′′′–d′′′, and Figure 6b′′′–d′′′).

The dependence of the brittle zone depth (BZD) on hydrogen charging times is quantified and listed in

Table 3. Brittle zone depth of hydrogen-charged CoCrFeMnNi, 304SS and IN718 alloy.

Hydrogen Charging Parameters	Brittle Zone Depth (μm)
	CoCrFeMnNi	304SS	IN718
1.79 mA cm−2—24 h	15.31 ± 1.0	31.74 ± 1.5	24.84 ± 1.0
1.79 mA cm−2—48 h	19.57 ± 0.5	36.77 ± 3.0	24.88 ± 0.6
179 mA cm−2—2 h	12.31 ± 0.8	29.40 ± 0.8	16.37 ± 0.4

. The results show that the BZD value gradually increases with time, with results of 15.31 μm and 19.57 μm for CoCrFeMnNi, 31.74 μm and 36.77 μm for 304SS, and 24.84 μm and 24.88 μm for IN718 alloy after charging at 24 h and 48 h, respectively. These results demonstrate that the BZD of 304SS is the greatest, followed by the IN718 alloy and then CoCrFeMnNi. According to Equation (2), the BZAF after hydrogen charging 24 h and 48 h is calculated to be 3.54% and 4.57% for CoCrFeMnNi, 5.93% and 6.39% for 304SS, and 3.86% and 3.91% for IN 718 alloy, respectively (

Table 4. Brittle zone area fraction of hydrogen-charged CoCrFeMnNi, 304SS and IN718 alloy.

Hydrogen Charging Parameters	Brittle Zone Area Fraction (%)
	CoCrFeMnNi	304SS	IN718
1.79 mA cm−2—24 h	3.54	5.93	3.86
1.79 mA cm−2—48 h	4.57	6.39	3.91
179 mA cm−2—2 h	3.71	5.12	2.56

). The results clearly show that an increase in charging time corresponds to deeper brittle zones.

3.4. Microstructural Comparison of Hydrogen-Uncharged and -Charged Samples of HEA

CoCrFeMnNi exhibits the highest resistance to HE in comparison to the 304SS and IN718 alloy (Table 2). Studies that investigate the deformation mechanism of novel HEAs in the presence of hydrogen have limited. Consequently, a focus was placed on the microstructural evolution of specific region in fractured hydrogen-charged and uncharged CoCrFeMnNi alloy (179 mA cm⁻² for 2 h). Taking into consideration the limited brittle zone depth (12.31 μm) on the fracture surface, TEM foil samples were prepared by the use of FIB. The samples near the fracture surface were extracted for TEM analysis. A representation of the preparation of a hydrogen-charged sample is follows: (1) Selection of a target region near the fracture surface (Figure 7a), (2) deposition of a Pt layer on the target region to protect the surface (Figure 7b), (3) cutting of a TEM sample with a depth of approximately 10 μm (less than the brittle zone depth, Figure 7c,d), (4) adhesion of the sample on a copper mesh (Figure 7e), and (5) Further thinning using the ion-beam until a nano-sized void appears (Figure 7f).

The microstructural features of uncharged CoCrFeMnNi alloy are presented in Figure 8. The sample has severe plastic deformation with a number of microbands observed (Figure 8a). High dislocation density and considerable numbers of twins can be observed between these microbands (Figure 8b). The selected area electron diffraction (SAED) pattern (Figure 8c), demonstrates the <110> observation direction and the existence of twinning. To analyze the twins and stacking faults (SF) further, high-resolution TEM image was captured (Figure 8d). It can be concluded that the twin width is between 5 nm and 10 nm, with the thickness of SF bundles being 2–9 nm. Enlarged images of the twins and SFs (red rectangular region in Figure 8d) as confirmed by fast Fourier transform (FFT) patterns in Figure 8e′,f′, are presented in Figure 8e,f, respectively. These reveal that the SF bundles have a thickness that is dozens of atomic layers, whereas twin boundaries consist of several atomic layers. Similar to uncharged sample, the hydrogen-charged TEM foil exhibits a number of microbands and dislocations (Figure 9a). Nano twins, as demonstrated by the SAED pattern (Figure 9c), are also observed (Figure 9b). In comparison with uncharged sample, the distributions of the twin width and SF bundle thickness are more heterogeneous (Figure 9d,g). The width of the SF bundles varies from 2 nm to 14 nm, with the twin thickness between 3 nm and 10 nm. Micrographs of the twin boundary and SF bundle are presented in Figure 9e,h, respectively. These are confirmed by the embedded FFT patterns, indicating that the width of the SF bundles is greater than that of the twin boundaries. Edge dislocations are also analyzed, as indicated by the red color in Figure 8g for uncharged sample, and Figure 9f,i for hydrogen-charged sample.

To compare the effect of hydrogen on the microstructural features, statistical analysis of the microband gap, number of twins, and width of SF bundles in hydrogen-charged and uncharged CoCrFeMnNi samples was conducted (Figure 10). The average microband gap is determined to be 26.62 nm for the uncharged sample, whereas it is 46.96 nm for hydrogenated sample. Conversely, the number of twins and the width of SF bundles increases after hydrogen charging, with the average number of twins and average width of SF bundles calculated to be 2.57 and 5.02 nm for the uncharged sample and 2.71 and 5.77 nm for the hydrogen-charged sample. These results demonstrate that the presence of hydrogen promotes the formation of SFs and twinning.

4. Discussion

In comparison with the 304SS and IN718 alloys, CoCrFeMnNi demonstrates the lowest HEI under the same hydrogen charging conditions (Table 2). This result has similarities with the results reported by Zhao et al. [12] and Pu et al. [13], who demonstrated that CoCrFeMnNi exhibited a higher HE-resistance than 304SS. However, this study indicates that the HEA is the most resistant to HE among the three alloys tested. This result can be rationalized in terms of the hydrogen diffusivity and deformation mechanism, as discussed in detail below. Additionally, prior studies [12,13] have reported that hydrogen reduced the fracture strain of CoCrFeMnNi, whereas the current study reveals that the HE susceptibility of the CoCrFeMnNi is dependent on the hydrogen concentration. A short hydrogen charging time (1.79 mA cm⁻² for 24 h) can improve both strength and ductility, whereas the ductility is reduced at long hydrogen charging time (1.79 mA cm⁻² for 48 h). This observation is related to competition between hydrogen-enhanced twin formation and hydrogen-enhanced decohesion (HEDE).

4.1. Hydrogen Diffusivity of CoCrFeMnNi, 304SS, and IN718 Alloy

During hydrogen charging, the following reaction will occur on the sample surface:

$${\mathrm{H}}^{+}+\mathrm{e}\to \mathrm{H}$$

(3)

In this process, hydrogen atoms will combine to generate hydrogen gas and escape the surface, while some will diffuse into metal lattice as concentrations build up. The hydrogen diffusion into the lattice along the plate thickness direction follows Fick’s diffusion law: In this process, hydrogen atoms will combine to generate hydrogen gas and escape the surface, while some will diffuse into metal lattice as concentrations build up. The hydrogen diffusion into the lattice along the plate thickness direction follows Fick’s diffusion law:

$$\frac{\partial \mathrm{c}}{\partial \mathrm{t}}=\mathrm{D}\frac{{\partial }^2\mathrm{c}}{\partial {\mathrm{x}}^2}$$

(4)

where c is the hydrogen concentration. t is the time. x is the position. D is the diffusion coefficient. As hydrogen charging is conducted, it is assumed that hydrogen concentration at the surface is constant. Based on fracture surface observations, hydrogen does not reach the center of the samples (Figure 4, Figure 5 and Figure 6). Thus, the corresponding boundary conditions can be described as:

$$\mathrm{c}(0,\mathrm{t})={\mathrm{c}}_{\mathrm{s}};\mathrm{c}(\infty ,\mathrm{t})={\mathrm{c}}_0$$

(5)

$$\mathrm{c}(\mathrm{x},0)={\mathrm{c}}_0$$

(6)

In combination with Equations (4)–(6), the hydrogen concentration (c(x, t)) at position x and time t along the plate thickness direction can be calculated as [36]:

$$\mathrm{c}(\mathrm{x},\mathrm{t})={\mathrm{c}}_0+({\mathrm{c}}_{\mathrm{s}}-{\mathrm{c}}_0)\left[1-\mathrm{erf}(\frac{\mathrm{x}}{2\sqrt{\mathrm{D}\mathrm{t}}})\right]$$

(7)

where c₀ is the initial hydrogen concentration. c_s is the hydrogen concentration at the surface, which is dependent on hydrogen charging current density. D is the hydrogen diffusion coefficient. t is the time. erf is an error function. It is evident that as $\frac{\mathrm{x}}{2\sqrt{\mathrm{D}\mathrm{t}}}\ge 2$, $\mathrm{c}(\mathrm{x},\mathrm{t})\to 0$ [36]. Consequently, the maximum depth of hydrogen diffusion (x_max) can be calculated by the following equation:

$${\mathrm{x}}_{\max }=4\sqrt{\mathrm{D}\mathrm{t}}$$

(8)

With respect to Equation (8), as the values for x_max and t are given, the value of D can be estimated. In this study, the brittle zone depth is assumed to be x_max to simplify the calculation. By inserting the values of x_max and t (Table 3) into Equation (8), the values of D at various hydrogen charging times can be calculated. These calculated values are listed in

Table 5. Calculated hydrogen diffusion coefficient of CoCrFeMnNi, 304SS, and IN718 alloy.

Hydrogen Charging Parameters	Hydrogen Diffusion Coefficient (10−16 m2 s−1)
	CoCrFeMnNi	304SS	IN718
1.79 mA cm−2—24 h	1.70	7.29	4.46
1.79 mA cm−2—48 h	1.39	4.89	2.24
Ref	3.70 [53]	8.00 [54]	0.68 [55]

. Under identical hydrogen charging parameters, the hydrogen diffusion coefficient is CoCrFeMnNi < IN718 alloy < 304SS. It is well-known that HE susceptibility of alloys correlates with the hydrogen diffusion behavior. A high hydrogen diffusion coefficient facilitates hydrogen accumulation in the local micro-environment, in favor of hydrogen-assisted cracking [49,50,51]. Thus, the low hydrogen diffusivity of CoCrFeMnNi is responsible for the increased resistance to HE. From Table 4, a low hydrogen diffusion coefficient correlates with small brittle zone area fraction, indicating an increase in effective load-bearing area. This factor also contributes to the low HE susceptibility of the CoCrFeMnNi alloy. In addition, the CoCrFeMnNi alloy exhibits the lowest grain size (10 μm) compared with 304SS (54 μm) and IN718 alloy (15 μm). For an equivalent size area, the HEA will contain the most grain boundaries acted as hydrogen traps [18], which will reduce local hydrogen accumulation, also improving the HE resistance of the HEA. It should be noticed that the better resistance of the HEA against HE is most probably linked to the lower diffusion rate of hydrogen. If the charging time is longer or the charging current is higher, more hydrogen will be introduced in the samples and HEAs will exhibit higher HE susceptibility [52]. Meanwhile, the question that whether the HEAs maintain the best HE resistance among the alloys needs to be further clarified. Therefore, the obtained results in this study apply for the particular conditions of the experiments.

The calculated hydrogen diffusion coefficients of each of the samples after charging for 48 h is smaller than that of charging for 24 h (Table 5). During diffusion, surface hydrogen traps are filled first and then internal trap sites are occupied due to the hydrogen concentration gradient [34]. As hydrogen has a small atomic radius, they will dissolve in interstitial sites, and migrate through interstitial manner in the alloys. According to Equation (8), a long charging period correlates with deep hydrogen diffusion zone. In this case, the channels for hydrogen diffusion will be hindered, increasing the diffusion barrier and reducing hydrogen diffusivity for the sample charged 48 h.

4.2. Failure Mechanisms of Hydrogenated CoCrFeMnNi, 304SS and IN718 Alloy

The hydrogen-assisted failure mechanisms of the 304SS and IN718 alloy have been widely investigated. The HE cracking mechanism is primarily correlated with the phase transformation from austenite to martensite for 304SS [36,56,57,58], with the interactions between slip bands or δ phase and hydrogen for IN718 alloy [37,59,60,61,62]. As expected, both 304SS and IN718 alloy are embrittled after hydrogen charging (Figure 3 and Table 2). To date, the HE mechanism of the CoCrFeMnNi HEA is still being determined, with preliminary results demonstrating the hydrogen-enhanced formation of twins. Based on first-principle calculations, Xie et al. [63] reported that the presence of hydrogen reduced the stacking fault energy and promoted the formation of twins. Luo et al. [35,64] demonstrated experimentally that hydrogen accelerated the formation of twins, and the number of twins depended on the hydrogen concentration. In previous studies [24,35,47,64], hydrogen-enhanced twin activity was qualitatively analyzed via microstructural observation. However, in this study, the number of twins and the width of SF bundles between hydrogen-charged and uncharged samples were quantified and compared (Figure 10). The results indicate that hydrogen accelerates the formation of SFs and twins. In comparison with grain boundaries, the twin boundaries often exhibit higher resistance to hydrogen-assisted cracking, primarily due to the high surface separation energy and low hydrogen solubility [52]. Thus, the HEA shows the lowest susceptibility to HE among the alloys tested.

As the hydrogen charging time is increased, the HEI of CoCrFeMnNi first decreases and then increases (Table 2). This can be explained by the competition between hydrogen-enhanced twin formation and the HEDE mechanism, which is supported by the intergranular fracture observed in Figure 4 and Figure 6. During tensile deformation, the formation of nano twins in the hydrogen-assisted brittle zone will refine the grains via the introduction of new twin interfaces. This will reduce the dislocation mean free path and strengthen alloy [35]. In addition, twin deformation is also a type of plastic deformation mechanism. In the presence of hydrogen, the number of twins and the width of SF bundles increases, leading to an increase in material’s ductility. Conversely, brittle IG and QC fractures are detected in terms of fracture surface observation, supporting the HEDE mechanism for reducing the ductility [18]. As hydrogen charging is 179 mA cm⁻² for 2 h and 1.79 mA cm⁻² for 24 h, hydrogen-enhanced twin formation mechanism dominates, improving the strength and ductility of the alloy. As the hydrogen charging time is prolonged, the HEDE mechanism becomes dominant, resulting in the reduction in the ductility of the alloy.

5. Conclusions

This study focused on the comparison of the HE of CoCrFeMnNi, 304SS and IN718 alloys. The microstructure of these alloys was analyzed by OM and high-resolution TEM, with the fracture mechanism characterized by SEM. The primary conclusions from this study are as follows:

• CoCrFeMnNi exhibits the lowest susceptibility to HE, followed by 304SS and then IN718 alloy, when hydrogen charging parameters are 1.79 mA cm⁻² for 24 h and 48 h, and 179 mA cm⁻² for 2 h. The result indicates that HEAs can potentially be regarded as a HE-resistance alloy for particular hydrogen-containing conditions.
• Regardless of the hydrogen charging conditions in this study, the fracture strain of 304SS and IN718 alloy decreases in the presence of hydrogen, whereas the effect of hydrogen on ductility of HEA depends on hydrogen charging time. At short charging time (24 h), a hydrogen-enhanced twin formation mechanism dominates, improving the material ductility. Meanwhile, the HEDE mechanism is responsible for the reduction in the fracture strain at long charging time (48 h).
• Based on the combination of brittle zone observations and the hydrogen diffusion modeling, the hydrogen diffusion coefficients are estimated to be 1.54 × 10⁻¹⁶ m² s⁻¹ for CoCrFeMnNi, 3.35 × 10⁻¹⁶ m² s⁻¹ for IN718 alloy and 6.09 × 10⁻¹⁶ m² s⁻¹ for 304SS, respectively.
• The number of nano twins (2.71) and the width of SF bundles (5.77 nm) in the hydrogen-assisted brittle zone is greater than in uncharged samples (2.57 and 5.02 nm). It clearly demonstrates that the presence of hydrogen promotes the formation of stacking faults and twins in CoCrFeMnNi alloy.

It is recommended that during future studies the mechanism of HE resistance behavior for HEAs will be explored based on multiscale and multi-physics modeling [65,66,67]. With this information, a method for designing materials that are hydrogen tolerant can be established based on the competing mechanisms. This will ultimately benefit multiple industries that are reliant on improved materials.