International Journal of Hydrogen Energy 53 (2024) 434–440

Contents lists available at ScienceDirect

International Journal of Hydrogen Energy

journal homepage: www.elsevier.com/locate/he

Hydrogen-enhanced entropy (HEENT): A concept for hydrogen

embrittlement prediction
Masoud Moshtaghi a, b, c, *, Mahdieh Safyari d, M.M. Khonsari b
a
Laboratory of Steel Structures, LUT University, P.O. Box 20, 53851, Lappeenranta, Finland
b
Mechanical and Industrial Engineering Department, Louisiana State University, Baton Rouge, LA, 70803, USA
c
Chair of General and Analytical Chemistry, Montanuniversität Leoben, Franz Josef-Straße 18, 8700, Leoben, Austria
d
LKR Light Metals Technologies Ranshofen, Austrian Institute of Technology, 5282, Ranshofen, Austria

A R T I C L E I N F O A B S T R A C T

Handling Editor: Ramazan Solmaz The concept of hydrogen-enhanced entropy (HEENT) for hydrogen embrittlement prediction was firstly intro­
duced and discussed through experimental validation methods, multiscale characterization, and trapping studies.
Keywords: The accumulated generated entropy reaches a constant value (fatigue fracture entropy (FFE)) regardless of
Hydrogen-enhanced entropy (HEENT) hydrogen content. Hydrogen-enhanced localized plasticity (HELP) is the dominant mechanism for hydrogen-
Hydrogen embrittlement
assisted cracking in the studied material, due to quasi-cleavage fracture pattern with serrated marking and H-
Hydrogen trapping
enhanced planar slip. Fatigue life is reduced due to increasing hydrogen uptake with increasing current density of
Quasi-cleavage fracture
Hydrogen-enhanced localized plasticity H charge, and the accumulated generated entropy reaches a constant FFE value. As H content increases, a
Fatigue fraction of weakly trapped hydrogen at interstitials and dislocations in the ferrite increases, and the possibility of
reaching hydrogen to crack tip and slip planes increases. Direct evidence of HELP mechanisms with a well-
developed dislocation substructure is reported for the first time in pearlitic steels with nanoscale observations
near the H-assisted crack path. The contribution of microplasticity due to the HELP mechanism to the total
entropy compensates for the total entropy reduction generated due to reduced fatigue life. This introduces the
hydrogen-enhanced entropy (HEENT) concept for hydrogen embrittlement prediction.

1. Introduction irreversible thermodynamics and the details of that for evaluating fa­
tigue degradation. The details of the formulation and implementation of
Consideration of the Low-cycle fatigue (LCF) life of the steel struc­ FFE can be found elsewhere [7]. Briefly, researchers have shown that
tures is vital for designing components for application under cyclic stress FFE can be viewed as a material parameter and is independent of the
conditions. In the LCF regime, the fatigue crack growth rate (FCGR) is an size, geometry, or type of loading [6,8–10]. It has been demonstrated [7,
indicator of the fatigue behaviour of the alloys. Several studies [1–3] 8] that the FFE can be estimated by integrating the following equation
have reported an increase in the FCGR when the material is exposed to a from t=0 to time of fracture (tf) in the LCF regime when the entropy
hydrogen-containing environment. Also reported is that the pearlitic generation owing to plastic deformation is dominant:
structure is less sensitive to hydrogen-assisted FCG (HA-FCG) compared ∫ tf
Wp
to the martensitic and ferritic steels [2]. In the range of carbon steels, the η= dt (1)
T
strength of the pearlitic steels can be comparable to the martensitic 0

steels, suggesting that these alloys are good candidates for hydrogen
where Wp is the internal heat generation consisting of inelastic defor­
applications.
mation (there is a negligible elastic part due to the small temperature
While FCGR is a viable indicator of hydrogen embrittlement sensi­
fluctuations via thermoelastic effect), T is the measured temperature, η is
tivity, realistic prediction of the LCF fatigue life for components exposed
accumulated entropy generation.
to hydrogen environment remains challenging. In previous studies
In this study, we firstly apply the concept of FFE to analyze the
[4–6], one of the present authors introduced the concept of fracture
hydrogen-assisted fatigue fracture behaviour of pearlitic carbon steel
fatigue entropy (FFE) as a general framework based on the principles of
and examine the extent of damage at different levels of hydrogen

* Corresponding author. Laboratory of Steel Structures, LUT University, P.O. Box 20, 53851, Lappeenranta, Finland.
E-mail address: [email protected] (M. Moshtaghi).

https://doi.org/10.1016/j.ijhydene.2023.12.068
Received 1 October 2023; Received in revised form 20 November 2023; Accepted 8 December 2023
Available online 16 December 2023
0360-3199/© 2023 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

content. We study the content of trapped hydrogen and the reversibility ( ( ))

of the trapped hydrogen by thermal desorption spectroscopy (TDS). In d ln TΦ2
addition, the microstructure and fracture surfaces are studied using a
m E
( ) =− d (2)
scanning electron microscope (SEM) integrated with an electron back­ d Tm1 R
scatter diffraction (EBSD) detector and transmission electron micro­
scopy (TEM). To assess the efficacy of the FFE concept in hydrogen-
where Ф, R, Tm are the heating rate, the gas constant, and desorption
assisted fatigue fracture, we measured the accumulation of the en­
peak temperature, respectively.
tropy during the fatigue test and determined its relationship with the
hydrogen uptake content. The findings will considerably increase the
possibility of implementing the thermodynamic fundamentals to predict 2.5. Crystallographic features
the hydrogen-assisted fatigue fracture in steels and reduce the unex­
pected hydrogen-assisted fatigue fracture cost. The crystallographic features and grain boundary types were
analyzed via a Zeiss Crossbeam 450 equipped with an EBSD Oxford
system with a step size of 50 nm and an energy of 20 kV. Following the
2. Experimental and theoretical procedures
surface preparation via grinding and polishing, Hitachi IM4000+ device
was employed to cross-section the specimen via an ion beam. A titanium
2.1. Materials
mask with a portion of the specimen outside the mask was employed to
direct the specimen toward the beam and smooth the surface [14,15].
A pearlitic steel with a composition of 0.61C, 0.15Si, 0.34Mn, 0.002P
and 0.003S (wt.%) and Fe in balance was used for this investigation. A
pearlitic steel with a pre-deformation of 40 % was used for this study. A 2.6. High-resolution transmission electron microscopy (HRTEM)
pre-deformed pearlitic steel is used to minimize the effusion of hydrogen
during the measurements while cooling via liquid nitrogen is also The specimens were characterized via a high-resolution transmission
implemented. electron microscope (JEOL-JEM -ARM200F) to perform typical bright
field TEM observations. For this purpose, the samples were first me­
chanically ground, and disks were punched out from ~70 μm thick foils
2.2. Electrochemical charging and electropolished in a twin-jet polishing machine using a mixed so­
lution of 1/5 perchloric acid (HClO4) and 4/5 methanol (CH3OH). The
An electrochemical hydrogen charging system was used with a 3.5 % twin-jet electropolished TEM disks were cleaned and further thinned
NaCl solution with 1 g/l CH4N2S and 1, 2, 5, and 10 mA cm-2 current using Gatan Precision Ion Polishing System II (PIPS II) with beam energy
density for 24 h; the steel as a cathode and platinum cylindrical mesh as 4 to 0.5 kV and an incident angle of ± 3◦ .
the anode. Electrochemically charged specimens were utilized for low-
cycle fatigue and TDS tests. 3. Results and discussion

3.1. Microstructure
2.3. Fatigue testing
Fig. 1 shows the crystallographic orientation, texture, and grain
Dog-bone fatigue specimens with a gauge length of 12 mm and
boundary characteristics of the pearlitic steel observed by SEM equipped
diameter of 5 mm were used for the cyclic testing. The specimens were
with EBSD and TEM. Via the technique reported in the ref. [16] and the
polished with 80–1200 emery papers to minimize surface effect. Strain-
line interception method, the average interlamellar spacing was ob­
controlled tension–compression axial fatigue tests were performed using
tained that is 152 nm, and the average sizes of pearlite colonies and
a servo-hydraulic universal testing machine (KAPPA 50 SS-CF) at room
blocks were measured that are 8.1 and 14 μm, respectively. Pearlite
temperature with a frequency of 1 Hz, a strain ratio, R = -1 and strain
colony refers to a region in which cementite lamellas have nearly the
amplitude, Δε/2 = 0.35% using a triangular waveform. The surface
same direction. Pearlite block corresponds to a region sharing nearly the
temperature of the specimen during cyclic loading was monitored by a
same crystallographic orientation of ferrite and is the final stage of a
temperature logger with an accuracy of ±0.1 ◦ C (Testo-176H1). The
pearlite nodule that nucleates at an austenite grain boundary.
stress–strain results of the specimen subjected to cycle loading of a
sufficient magnitude to result in plastic deformation can be described as
a hysteresis loop. The hysteresis loops of the cyclic tension–compression 3.2. Hydrogen trapping behaviour
fatigue test are used to measure the entropy resulting from the ther­
modynamic calculations. The total inelastic dissipation (Wp) till a Fig. 2a shows the H desorption rate coupled with temperature for the
certain number of cycles on an ongoing basis is calculated from the different hydrogen charging times. The results show that the desorbed
Ramberg–Osgood constitutive model by fitting to the hysteresis loops hydrogen increases with increasing the current density of hydrogen
and corresponding to measured damage. More details can be found charging (1.2, 2.05, 5.1, and 8.21 wt ppm, respectively). Fig. 2b shows
elsewhere [11]. the hydrogen trapping behaviour for the deformed pearlitic steel. The
TDS results of the sample were analyzed using the Gaussian function to
identify overlapping and partially obscured peaks [17,18], as shown in
2.4. Studying the H trap sites via TDS Fig. 2b. The activation energies were calculated and the activation en­
ergy of Peak1 is 30.5 kJ/mol, meaning relatively reversible trapping of
The electrochemically charged samples were subsequently dipped hydrogen. While some researchers [13,19,20] indicate that the origin of
into the liquid nitrogen lower the possibility of the hydrogen release Peak1 is hydrogen trapped at the ferrite/cementite interface, Chen et al.
from the sample. A temperature-programmed quadrupole mass spec­ [21] recently used atom probe tomography (APT) and micromechanical
troscope is used for the evaluation of the different trap sites and their testing and indicated that there is no evidence of hydrogen trapping at
reversibility [12]. The TDS measurements were performed at heating the ferrite/cementite interface. On the other hand, the center of the peak
rates of 3.3, 6.6, and 13.3 ◦ C.min-1. The heating rate variation and the and the obtained activation energy for Peak1 is similar to the peak
corresponding peak shifts and the Choo-Lee formulation were used to corresponding to the interstitial lattice sites reported for ferritic micro­
calculate the desorption energy (Ed) of different hydrogen trap sites [13, structure [14,22,23].
14]: For peak2, the average temperature of peak centers (Table 1) is

435
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

Fig. 1. (a) Pearlitic steel indicating colony boundaries and pearlitic lamellar structure, (b) TEM microghraph showing tangled dislocations in ferrite and cementite,
(c) the inverse pole figure (IPF), and (d) grain boundary map; with the misorientation angles greater than 15◦ , i.e. pearlite block boundaries and the misorientation
angles lower than 15◦ , i.e. low-angle grain boundaries.

comparable to the previous studies [19,24]. As described in our previous The hydrogen is located in the interstitial lattice and dislocations in the
work [25], the dislocation density in ferrite was risen with an increase in ferritic phase as relatively reversible hydrogen traps, and in the dislo­
cold deformation, which was accompanied by an increase in hydrogen cations and vacancies in the cementite, as relatively irreversible
content of the trap site corresponding to Peak2 with the degree of hydrogen traps.
deformation. Thus, Peak2 originated from H trapped at dislocations in
the ferritic phase. These dislocations were also directly observed by
TEM, as shown in Fig. 1. This observation is in good agreement with the 3.3. Hydrogen embrittlement mechanism vs hydrogen-enhanced entropy
other reports [19,24]. The activation energies of the Peak3 and Peak4 (HEENT)
were calculated as 52.2 and 67.4 kJ/mol, respectively.
According to the recent results [26], hydrogen trapping at the Fig. 3a shows the evolution of progressive entropy generation during
cementite phase was detected. The traps were reported as dislocations the cycling of the hydrogen-charged specimen at different hydrogen
and vacancies in cementite bulk [24,26]. Cairney et al. [26] showed that uptakes with different current densities of hydrogen charging and the
there is a significant amount of carbon vacancies that are strong total generated entropy values in different hydrogen uptake contents.
hydrogen traps, which is evident by considering the carbon deficiency in The results showed that the fatigue life is reduced by increasing the
the Fe3C stoichiometry and directly observing the trapped hydrogen. hydrogen uptake (by increasing the charging current density). Inter­
Fig. 1 also shows that there are some dislocations in the deformed estingly, we clearly see that the accumulated generated entropy reaches
cementite. In addition, the desorption activation energy for the single to about 3.27 MJ K− 1m− 3, regardless of the charging current density and
peak prior to deconvolution was reported 65 kJ/mol in another study the testing environment. It is obvious that the there is an entropy limit
[27], which is comparable to the current findings. In addition, some that is independent of the charging condition.
evidence of acting vacancy as a H trap site has already reported by other Fig. 4 shows the detailed fractography for the hydrogen-charged
researchers [28]. In a previous study [24], hydrogen trapped at specimen at 10 mA cm-2 current density. From Fig. 4, it is discernible
cementite dislocations and vacancies, vacancies are relatively irrevers­ that there are tearing topography surfaces accompanied by quasi-
ible trap sites compared with the dislocations [26,26,29,30]. Therefore, cleavage fracture features. It was reported that these tearing topog­
the trapping sites in the present alloy have been summarized in Table 1. raphy surfaces are related to having a degree of localized plasticity. It is
described as a fracture surface characterized by ductile micro-plastic

436
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

Fig. 2. (a) Hydrogen desorption rate versus time for four different hydrogen charging current densities, (b) thermal desorption spectra with the measured data points
and the theoretically fitted hydrogen traps of the hydrogen charged pearlitic steel at 10 mA.cm-2 current density, (c) the desorption spectra in three different heating
rates: 200, 400, 800 ◦ C/h, (d) thermally ramped desorption data calculated using Eq. (2) and Figs. (c), indicating the desorption energy of each trap as the slopes of
the lines.

pearlitic steel can diffuse to the tip of the fatigue crack and lead to
Table 1 quasi-cleavage crack propagation. In fact, the fatigue crack growth is
Representative trapping data determined from thermal desorption spectroscopy
supported by the plasticity-induced transfer of material from the crack
(TDS).
tip [31,32].
peak Approx. peak Trap site Desorption energy, Ed [kJ Some of the current authors recently reported that [25,39–41] the
position [◦ C] mol-1]
crack is propagated transgranularly and {011} slip planes have a
1 75-90 Interstitial lattice 30.5 contribution to hydrogen-enhanced planar slip and crack propagation,
sites
resulting in a quasi-cleavage fracture. In recent works about other
2 120-150 Dislocations in 39.4
ferrite bcc/bct steels, we also suggested that [39,40] the {011} plane itself
3 310-330 Dislocations in 52.2 plays an important role in crack propagation of hydrogen-induced
cementite fractures. The quasi-cleavage fracture occurred along the {011} planes
4 370-390 Vacancies in 67.4 bcc/bct steels shows that fracture at {011} plane is an intrinsic char­
cementite
acteristic of hydrogen-assisted failure in quasi-cleavage mode.
H-enhanced planar slip is related to the shielding of the strain field
tearing on a very fine scale [31,32]. The tearing topography surfaces can around the dislocation by hydrogen and the hydrogen-enhanced local­
be formed while the fracture initiates close to the crack tip, and even­ ized plasticity (HELP) mechanism [42,43]. The fracture surface of
tually, the crack propagates transgranularly and can be seen as cyclically loaded hydrogen-charged pearlitic steel showed the domi­
quasi-cleavage fracture features [19,29,33]. In comparison with the nance of the quasi-cleavage fracture, indicating that the HELP mecha­
cyclically tested specimen in air [34–38], hydrogen-assisted fracture nism is dominant in the alloy.
changes the fracture modes from ductile striations to quasi-cleavage In this paper, we propose experimentally, for the first time, that the
fracture with serrated markings. Based on the trap site analysis, we entropy limit is an intrinsic value for the hydrogen-induced fatigue
have a portion of high-energy trap sites related to trapping hydrogen by fracture and can be used for the prediction of fatigue fracture life of the
cementite. The weakly trapped hydrogen atoms at interstitial lattice pearlitic steels. The model of Novak et al. [17,44] explained that there is
sites and dislocations in ferrite (Peak1 and Peak2 in Fig. 2 and Table 1, a dislocation-based plasticity model with a focus on dislocation gener­
respectively) can play a significant role during the cyclic loading of the ation, mobility, and the rate of annihilation in steels. The dissipation
H-charged pearlitic steel. The reversibly trapped hydrogen in the inequality equation is reported as

437
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

Fig. 3. (a) Evolution of accumulated entropy generation during cycling of the hydrogen-charged specimen at different hydrogen uptakes with different current
densities of hydrogen charging, the fracture surface observed by SEM-EBSD and HRTEM at the designated areas near the crack path for sample with (b) 1.2 ppm
hydrogen and (c) 8.2 ppm hydrogen.

Fig. 4. Detailed fractography for the hydrogen charged specimen at 10 mA.cm-2 current density, (a) general view of the fractured specimen, (b) the center of the
specimen, (c) the region near the surface of the sample.

[ ] 1 ∂NT tensor, and the hydrogen-induced strain tensor, flow stress, the lattice
σ : ε̇P + σ : ε̇h − kε̇SS − q.∇T − J.∇μ − (γ b − μh )αθT P ε̇P ≥ 0 (3)
irr T ∂ε strain relevant to statistically stored dislocations (SSD), temperature,
heat flux, the hydrogen transport flux, the chemical potential of the
where σ , ε̇h , ε̇P , k, ε̇SS , T, q, J, μh, are the stress tensor, the plastic strain hydrogen solute atoms, respectively. In addition, γb, α, θT, NT are an

438
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

external scalar body force, the number of lattice traps per trap, the oc­ interests or personal relationships that could have appeared to influence
cupancy rate of H atoms in traps, and the density of traps per unit vol­ the work reported in this paper.
ume. Eq. (3) shows that energy dissipation due to the HELP mechanism
is a dominant mechanism in the current study. The generation of the Acknowledgement
entropy due to the HELP mechanism can be increased by increasing the
H content according to Eq. (3). Fig. 3b and c shows the crack path and M. Moshtaghi gratefully acknowledges financial support from the
the high-resolution TEM (HRTEM) micrographs of the area near the Kamei Corporation, Japan, the MEXT-Monbu Kagakushu grant.
crack path for the cyclically leaded specimens with 1.2 and 8.21 ppm
hydrogen. The results showed that the dislocation substructure is well-
References
developed in the higher content of hydrogen (8.21 ppm) compared to
the dislocations in the specimen with 1.2 ppm hydrogen. The high [1] Dmytrakh IM, Leshchak RL, Syrotyuk AM, Barna RA. Effect of hydrogen
content of hydrogen increases the possibilities of the activation of the concentration on fatigue crack growth behaviour in pipeline steel. Int J Hydrogen
HELP mechanism and the H-enhanced planar slip and higher mobility of Energy 2017;42:6401–8. https://doi.org/10.1016/j.ijhydene.2016.11.193.
[2] Ogawa Y, Iwata K. Resistance of pearlite against hydrogen-assisted fatigue crack
dislocation due to the presence of hydrogen, leading to a well-developed growth. GCGW-21 - Glob Warm Reduct Hydrog 2022;47:31703–8. https://doi.org/
dislocation substructure in the higher hydrogen content. This is in good 10.1016/j.ijhydene.2022.07.074.
agreement with the results reported by Ogawa et al. [38]. The cyclic [3] Yamabe J, Yoshikawa M, Matsunaga H, Matsuoka S. Hydrogen trapping and
fatigue crack growth property of low-carbon steel in hydrogen-gas environment.
plasticity due to the activation of the HELP mechanism contributes to Int J Fatig 2017;102:202–13. https://doi.org/10.1016/j.ijfatigue.2017.04.010.
the total generated entropy. The contribution of the microplasticity of [4] Naderi M, Amiri M, Khonsari MM. On the thermodynamic entropy of fatigue
the HELP mechanism to the total entropy can compensate for the total fracture. Proc R Soc Math Phys Eng Sci 2010;466:423–38. https://doi.org/
10.1098/rspa.2009.0348.
generated entropy reduction due to reduced fatigue life. While the fa­ [5] Naderi M, Khonsari MM. An experimental approach to low-cycle fatigue damage
tigue life of the alloy reduced considerably at the hydrogen content of based on thermodynamic entropy. Int J Solid Struct 2010;47:875–80. https://doi.
8.21 ppm, the total generated entropy reaches almost the same value org/10.1016/j.ijsolstr.2009.12.005.
[6] Amiri M, Khonsari MM. Rapid determination of fatigue failure based on
(FFE value). This shows hydrogen-enhanced entropy (HEENT) that
temperature evolution: fully reversed bending load. Int J Fatig 2010;32:382–9.
compensates for the total generated entropy reduction due to reduced https://doi.org/10.1016/j.ijfatigue.2009.07.015.
fatigue life. The dissipation from macroplastic work σ : ε̇P is the main [7] Amooie MA, Khonsari MM. On the effect of environmental temperature on fracture
fatigue entropy. Int J Fatig 2023;168:107411. https://doi.org/10.1016/j.
dissipation change when we increase the hydrogen content in the alloy. ijfatigue.2022.107411.
This increase is less pronounced or zero in other dissipation terms in Eq. [8] Jang JY, Khonsari MM. Experimentally validated thermodynamic theory of metal
(3). As H content increases, the proportion of weakly trapped hydrogen fatigue. Mech Mater 2021;160:103927. https://doi.org/10.1016/j.
mechmat.2021.103927.
at interstitial lattice sites and dislocations in ferrite increases, and the [9] Jirandehi AP, Khonsari MM. Microstructure-sensitive estimation of fatigue life
possibility of reaching the hydrogen into the slip planes decreases. The using cyclic thermodynamic entropy as an index for metals. Theor Appl Fract Mech
H-enhanced planar slip leads to a higher frequency of localized plasticity 2021;112:102854. https://doi.org/10.1016/j.tafmec.2020.102854.
[10] Teng Z, Wu H, Boller C, Starke P. Thermodynamic entropy as a marker of high-
in higher hydrogen content. This implies that higher hydrogen content
cycle fatigue damage accumulation: example for normalized SAE 1045 steel. Fatig
tends to increase the contribution of hydrogen-enhanced entropy Fract Eng Mater Struct 2020;43:2854–66.
(HEENT). At the highest extent of hydrogen damage, the contribution of [11] Young Subbarayan. Maximum entropy models for fatigue damage in metals with
H-induced localized plasticity is maximized via the HEENT effect while application to low-cycle fatigue of aluminum 2024-t351. Entropy 2019;21:967.
https://doi.org/10.3390/e21100967.
the contribution of the fatigue damage accumulation decreases signifi­ [12] Safyari M, Moshtaghi M, Kuramoto S. On the role of traps in the microstructural
cantly due to a lower fatigue life. control of environmental hydrogen embrittlement of a 7xxx series aluminum alloy.
J Alloys Compd 2021;855:157300. https://doi.org/10.1016/j.
jallcom.2020.157300.
4. Conclusions [13] Choo WY, Lee JY. Thermal analysis of trapped hydrogen in pure iron. Metall Trans
A 1982;13:135–40. https://doi.org/10.1007/BF02642424.
[14] Moshtaghi M, Loder B, Safyari M, Willidal T, Hojo T, Mori G. Hydrogen trapping
The following results can be concluded from the present work: and desorption affected by ferrite grain boundary types in shielded metal and flux-
cored arc weldments with Ni addition. Int J Hydrogen Energy 2022. https://doi.
1 The concept of hydrogen-enhanced entropy (HEENT) for hydrogen org/10.1016/j.ijhydene.2022.04.260. S036031992201881X.
[15] Moshtaghi M, Safyari M. Temperature mitigates the hydrogen embrittlement
embrittlement prediction was firstly introduced and discussed. The
sensitivity of martensitic steels in slow strain rates. Vacuum 2022:111187.
contribution of the microplasticity of the HELP mechanism to the [16] Caballero FG, García de Andrés C, Capdevila C. Characterization and
total entropy, i.e. HEENT effect compensates for the total entropy morphological analysis of pearlite in a eutectoid steel. Mater Char 2000;45:111–6.
reduction generated due to reduced fatigue life. https://doi.org/10.1016/S1044-5803(00)00056-5.
[17] Novak P, Yuan R, Somerday BP, Sofronis P, Ritchie RO. A statistical, physical-
2 The accumulated generated entropy reaches about a constant value; based, micro-mechanical model of hydrogen-induced intergranular fracture in
i.e. FFE, regardless of the charging current density, and hydrogen steel. J Mech Phys Solid 2010;58:206–26. https://doi.org/10.1016/j.
content. FFE can be proposed as a property of the material. jmps.2009.10.005.
[18] Moshtaghi M, Safyari M, Kuramoto S, Hojo T. Unraveling the effect of dislocations
3 Hydrogen resides at interstitial lattice sites and dislocations in ferrite and deformation-induced boundaries on environmental hydrogen embrittlement
and dislocations and vacancies in cementite. behavior of a cold-rolled Al–Zn–Mg–Cu alloy. Int J Hydrogen Energy 2021;46:
4 The crack is propagated transgranularly and hydrogen-enhanced 8285–99. https://doi.org/10.1016/j.ijhydene.2020.12.028.
[19] Yu S-H, Lee S-M, Lee S, Nam J-H, Lee J-S, Bae C-M, et al. Effects of lamellar
planar slip results in the propagation of the crack along a certain structure on tensile properties and resistance to hydrogen embrittlement of
slip plane in ferrite. pearlitic steel. Acta Mater 2019;172:92–101.
5 The fracture surface of cyclically loaded H-charged pearlitic steel [20] Choo WY, Lee JY. Hydrogen trapping phenomena in carbon steel. J Mater Sci 1982;
17:1930–8.
shows predominantly quasi-cleavage fractures, indicating that the [21] Chen Y-S (Eason), Niu R, Liu P-Y, Burr P, Cairney J. Direct observation of hydrogen
HELP mechanism is active in the alloy. distribution in pearlite. Microsc Microanal 2022;28:1618–20. https://doi.org/
6 A direct evidence of HELP mechanisms with a well-developed 10.1017/S1431927622006468.
[22] Park C, Kang N, Liu S. Effect of grain size on the resistance to hydrogen
dislocation substructure is firstly reported in the pearlitic steels
embrittlement of API 2W Grade 60 steels using in situ slow-strain-rate testing.
with HRTEM observation near the H-assisted crack path. Corrosion Sci 2017;128:33–41. https://doi.org/10.1016/j.corsci.2017.08.032.
[23] Chen L, Xiong X, Tao X, Su Y, Qiao L. Effect of dislocation cell walls on hydrogen
adsorption, hydrogen trapping and hydrogen embrittlement resistance. Corrosion
Declaration of competing interest Sci 2020;166:108428. https://doi.org/10.1016/j.corsci.2020.108428.
[24] Takai K, Nozue A, editors. Hydrogen permeation and trapping characteristics of
cold-drawn pure iron and eutectoid steel evaluated by thermal desorption
The authors declare that they have no known competing financial spectrometry. Corrosion; 2000.

439
M. Moshtaghi et al. International Journal of Hydrogen Energy 53 (2024) 434–440

[25] Truschner M, Pengg J, Loder B, Köberl H, Gruber P, Moshtaghi M, et al. Hydrogen [36] Shinko T, Hénaff G, Halm D, Benoit G, Bilotta G, Arzaghi M. Hydrogen-affected
resistance and trapping behaviour of a cold-drawn ferritic–pearlitic steel wire. Int J fatigue crack propagation at various loading frequencies and gaseous hydrogen
Mater Res 2023;114:439–52. https://doi.org/10.1515/ijmr-2022-0376. pressures in commercially pure iron. Int J Fatig 2019;121:197–207. https://doi.
[26] Cairney J, Niu R, Li H, Liu P-Y, Burr P, Feng Y, et al. The effect of hydrogen on the org/10.1016/j.ijfatigue.2018.12.009.
deformation of pearlite in steels. Physical Sciences 2022. https://doi.org/10.21203 [37] Suresh S, Ritchie RO. Mechanistic dissimilarities between environmentally
/rs.3.rs-1325909/v1. influenced fatigue-crack propagation at near-threshold and higher growth rates in
[27] Takai K, Watanuki R. Hydrogen in trapping states innocuous to environmental lower strength steels. Met Sci 1982;16:529–38. https://doi.org/10.1179/
degradation of high-strength steels. ISIJ Int 2003;43:520–6. https://doi.org/ msc.1982.16.11.529.
10.2355/isijinternational.43.520. [38] Ogawa Y, Iwata K. Fatigue crack propagation in pearlitic steel under pressurized
[28] Doshida T, Takai K. Dependence of hydrogen-induced lattice defects and hydrogen gaseous hydrogen: influences of microstructure size and strength level. ISIJ Int
embrittlement of cold-drawn pearlitic steels on hydrogen trap state, temperature, 2023;63:1251–62. https://doi.org/10.2355/isijinternational.ISIJINT-2023-011.
strain rate and hydrogen content. Acta Mater 2014;79:93–107. [39] Pichler S, Bendo A, Mori G, Safyari M, Moshtaghi M. Inhibition of grain growth by
[29] Robertson IM, Sofronis P, Nagao A, Martin ML, Wang S, Gross DW, et al. Hydrogen pearlite improves hydrogen embrittlement susceptibility of the ultra-low carbon
embrittlement understood. Metall Mater Trans A 2015;46:2323–41. ferritic steel: the influence of H-assisted crack initiation and propagation
[30] Safyari M, Moshtaghi M, Kuramoto S, Hojo T. Influence of microstructure-driven mechanisms. J Mater Sci 2023;58:13460–75. https://doi.org/10.1007/s10853-
hydrogen distribution on environmental hydrogen embrittlement of an Al–Cu–Mg 023-08856-y.
alloy. Int J Hydrogen Energy 2021;46:37502–8. https://doi.org/10.1016/j. [40] Moshtaghi M, Maawad E, Bendo A, Krause A, Todt J, Keckes J, et al. Design of
ijhydene.2021.09.013. high-strength martensitic steels by novel mixed-metal nanoprecipitates for high
[31] Toribio J. HELP versus HEDE in progressively cold-drawn pearlitic steels: between toughness and suppressed hydrogen embrittlement. Mater Des 2023:112323.
Donatello and Michelangelo. Eng Fail Anal 2018;94:157–64. https://doi.org/ https://doi.org/10.1016/j.matdes.2023.112323.
10.1016/j.engfailanal.2018.07.026. [41] Ogawa Y, Nishida H, Nakamura M, Olden V, Vinogradov A, Matsunaga H. Dual
[32] Toribio J. Hydrogen-assisted cracking paths in oriented pearlitic microstructures: roles of pearlite microstructure to interfere/facilitate gaseous hydrogen-assisted
resembling donatello wooden sculpture texture (DWST) & mantegna’s dead christ fatigue crack growth in plain carbon steels. Int J Fatig 2022;154:106561. https://
perspective (MDCP). Procedia Struct Integr 2020;26:368–75. doi.org/10.1016/j.ijfatigue.2021.106561.
[33] Toribio J, Vergara D, Lorenzo M. Influence of loading rate on the hydrogen-assisted [42] Martin ML, Fenske JA, Liu GS, Sofronis P, Robertson IM. On the formation and
micro-damage in bluntly notched samples of pearlitic steel. Metals 2016;6:11. nature of quasi-cleavage fracture surfaces in hydrogen embrittled steels. Acta
[34] Birenis D, Ogawa Y, Matsunaga H, Takakuwa O, Yamabe J, Prytz Ø, et al. Mater 2011;59:1601–6. https://doi.org/10.1016/j.actamat.2010.11.024.
Interpretation of hydrogen-assisted fatigue crack propagation in BCC iron based on [43] Robertson IM. The effect of hydrogen on dislocation dynamics. Eng Fract Mech
dislocation structure evolution around the crack wake. Acta Mater 2018;156: 1999;64:649–73. https://doi.org/10.1016/S0013-7944(99)00094-6.
245–53. https://doi.org/10.1016/j.actamat.2018.06.041. [44] Novak PM. A dislocation-based constitutive model for hydrogen-deformation
[35] Sun Z, Moriconi C, Benoit G, Halm D, Henaff G. Fatigue crack growth under high interactions and a study of hydrogen-induced intergranular fracture. University of
pressure of gaseous hydrogen in a 15-5PH martensitic stainless steel: influence of Illinois at Urbana-Champaign; 2009.
pressure and loading frequency. Metall Mater Trans A 2013;44:1320–30. https://
doi.org/10.1007/s11661-012-1133-5.

440