Scripta Materialia 243 (2024) 115963

Contents lists available at ScienceDirect

Scripta Materialia
journal homepage: www.journals.elsevier.com/scripta-materialia

Characterization of hydrogen traps in a co-precipitation steel investigated

by atom probe experiments without cryogenic transfer
Severin Jakob a, *, Mohammad Sattari a, Birhan Sefer b, Steve Ooi c, d, Mattias Thuvander a, *
a
Department of Physics, Chalmers University of Technology, Göteborg SE-412 96, Sweden
b
Group of Corrosion and Hydrogen Embrittlement, Materials Development Department, Swerim AB, Box 7047, Kista 164 07, Sweden
c
Maxwell Centre, Ovako Group R&D, JJ Thompson Avenue, Cambridge CB3 0HE, United Kingdom
d
Ovako Corporate R&D, Building 202, Hofors SE-813 82, Sweden

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

Keywords: Hydrogen (H) embrittlement in high-strength steels can be mitigated by introducing H traps into the micro­
Atom probe tomography structure. The co-precipitation model steel in this work contains intermetallic ß-NiAl and secondary Cr-carbides,
Co-precipitation which provide abundant trapping sites. Needle-shaped specimens are prepared for atom probe tomography
Hydrogen trapping
(APT) and electro-chemically charged in a solution of 0.1 M NaOH in D2O to introduce deuterium (D). D is
located at the finely dispersed Cr-carbides even after specimen transfer at room temperature (RT), which shows
that nano-sized Cr-carbides are strong H traps. This is in contrast to previous studies of weak traps where
cryogenic transfer was needed to detect any D.

High-strength steels with improved mechanical properties are chamber and actual H in the material is the use of the heavier isotope
desired among other in lightweight design of structural components. deuterium. It was shown that D has similar solubility and only slightly
With higher yield strength, smaller amounts of material are used in the lower diffusivity compared to H, making it suitable for this kind of ex­
same application, leading to reduced resource consumption and overall periments [18,19]. D is either introduced into the specimen by gas
lower costs. However, high-strength steels are sensitive to hydrogen charging [12,13], implantation [20] or electro-chemically by cathodic
embrittlement (HE) due to the high stress concentration and the sup­ charging of the specimen in D2O solutions [15,21–23]. To prevent H or
pression of dislocation emission at the crack tip caused by accumulation D from diffusing and escaping the material, the use of cryogenic transfer
of H [1–3]. It has been shown that a critical content of diffusible H leads has been emphasized [24–26]. So far, D was hardly measured without
to material failure [4]. fast cooling the specimen and maintaining a continuous cold transfer to
The idea to prevent HE by trapping H in the microstructure is not the analysis chamber.
new [5,6]. Grain boundaries and dislocations have been shown to only A promising material design concept for high strength and resistance
weakly trap H [3]. Therefore, other nanostructures in form of carbides, to HE is known as dual-hardening steel [27–30]. The high mechanical
such as NbC, TiC, VC and Mo2C, or ε-Cu precipitates were investigated as strength, good fatigue resistance, and adequate deformability are ach­
H traps [7–11]. The amount of trapped H and the strength of binding are ieved by precipitating two types of particles with high number densities
usually measured by thermal desorption mass spectrometry (TDMS) [9, in a martensitic matrix [27]. As shown for Hybrid steel® from Ovako AB,
10]. More recently, investigations into the actual trapping sites were the improved strength is maintained at elevated temperatures due to the
performed with APT instruments [12–15]. APT offers nearly atomic reduced coarsening of particles compared to other steels [30]. Similar to
spatial resolution with high sensitivity for all elements [16]. However, the development of Hybrid steel® [31], the model alloy in this paper
quantitative characterization of H is still challenging. Since H2 is the was designed to have a lean alloying concept with balanced mechanical
main residual gas in the ultra-high vacuum (UHV) chamber, the pres­ properties as well as a high resistance to HE. The Hybrid model alloy has
ence of H+, or in addition H+
2 depending on the measurement mode, ions two different types of particles, namely intermetallic ß-NiAl as well as
in the mass spectrum is a common artefact during APT experiments secondary Cr-carbides. Samples are charged with D and moved to the
[17]. A practical way to distinguish between H from the vacuum APT instrument without cryogenic transfer. It is shown from the APT

* Corresponding authors.
E-mail addresses: severin.jakob@chalmers.se (S. Jakob), mattias.thuvander@chalmers.se (M. Thuvander).

https://doi.org/10.1016/j.scriptamat.2023.115963
Received 5 July 2023; Received in revised form 21 November 2023; Accepted 27 December 2023
Available online 5 January 2024
1359-6462/© 2024 The Authors. Published by Elsevier Ltd on behalf of Acta Materialia Inc. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
S. Jakob et al. Scripta Materialia 243 (2024) 115963

measurements that nano-scale Cr-carbides are strong traps for H since D [34,35].
is still present in the measurements even after RT transfer. To investigate the de-trapping activation energy, TDMS was con­
The Hybrid model alloy in this work (0.42 C, 6.5 Cr, 2.5 Al, 3 Ni and ducted using a Bruker Galileo G8 analyzer in combination with a mass
0.25 Mo in wt.%) was solution annealed and quenched before aging for spectrometer from InProcess Instrument GmbH for H detection. Nitro­
20 h at 500 ◦ C. The microstructure was characterized by scanning gen was used as the carrier gas and pure hydrogen gas was used to
transmission electron microscopy (S/TEM), using an FEI Titan 80–300 calibrate the instrument. Samples with dimensions of 1 × 10 × 13 mm3
TEM operating at 300 kV as well as APT. Thin foils were prepared by were prepared with surface finish of P600 SiC sanding paper and
mechanically thinning a 3 mm disk to around 100 µm thickness, fol­ charged in 3 wt.% NaCl and 0.3 wt.% NH4SCN electrolyte at -1 mA/cm2
lowed by electrolytic twin-jet polishing in 10 % perchloric acid solution for 24 h. After degassing for 24 h at 50 ◦ C, one set of samples was
cooled to -30 ◦ C at 20 V. Samples for APT were cut to 0.3 × 0.3 × 15 measured with 0.3, 0.6 and 1.2 ◦ C/s heating rates from RT to 800 ◦ C. To
mm3 and needle-shaped specimens were prepared by a two-step electro- investigate the degassing behavior, one sample was measured directly
polishing process in 10 % and 2 % perchloric acid, respectively [16]. after cathodic charging and two more samples were degassed at 50 ◦ C
APT specimens were characterized without charging as well as for 72 h and 168 h before TDMS analysis using 0.6 ◦ C/s. Analysis of de-
cathodically charged in a solution of 0.1 M NaOH in D2O at -2.2 V trapping activation energies is based on the formulation of Kissinger
against a Pt counter-electrode for 30 s. After charging, the specimen was [36] and the simplified models thereof as shown by Lee and Lee [37] and
mounted and moved through the load-lock and buffer chamber of the Wei et al. [38].
atom probe instrument. The transfer-time at RT was about 15 min before Fig. 1(a) and 1(b) show TEM micrographs of the microstructure,
the APT needle was cooled to cryogenic temperature in the analysis showing martensite laths with particles of about 10 to 100 nm in size at
chamber. the boundaries. The composition of the particles was measured with
The specimens were measured on the latest generation of APT in­ energy dispersive X-ray spectroscopy (EDX) and revealed on one hand
struments, a LEAP 6000XR from CAMECA Instruments. This instrument Ni and Al in equal amounts and on the other hand Cr-carbides with C and
is equipped with a reflectron and has a detection efficiency of 52 %. APT Fe in different concentrations. The small size and hence contribution of
measurements were performed at 70 K in voltage mode with 20 % pulse the surrounding matrix to the EDX signal makes it difficult to measure
fraction, 0.5 - 1 % detection rate and pulse frequencies of 100, 200 and the exact composition of the particles.
333 kHz. The reconstructions were made in the commercial software AP High resolution TEM was performed within the laths, showing
Suite 6.3 (CAMECA) using the voltage method with an image smaller particles. The particle in Fig. 1(c) has about 4 nm width and
compression factor of 1.65 and a field factor of 4.0. The sizes of nano- shows an interplanar spacing of 2.06 Å. The closest match for this is
scale precipitates were characterized by applying the parameter free M23C6 (511) planes (ICCD card number 01–078–1499) with d = 2.05 Å.
method by Zhao et al. [32], which uses the radial distribution function. However, M7C3 with 2.10 Å plane spacing of (202) planes (ICCD card
The matrix contents were determined by the DIAM method that is based number 04-005-9650) cannot be excluded as a possible crystal structure.
on nearest neighbor distributions [33]. Particle sizes were calculated Another particle, shown in Fig. 1(d), with an interplanar spacing of 1.97
from the number of atoms, hence, avoiding local magnification effects Å can be best approximated by M7C3 (600) planes, having d = 1.99 Å

Fig. 1. (a–b) TEM micrographs showing the martensitic microstructure with particles at the lath boundaries (indicated by arrows). (c-d) show high resolution TEM of
precipitates within the laths and their interplanar spacing.

2
S. Jakob et al. Scripta Materialia 243 (2024) 115963

(ICCD card number 04–005–9650). whereas Mo is enriched. The depiction of a sharp interface is limited by
Fig. 2(a) shows a 5 nm thick sections of an APT reconstruction of an the lateral resolution of the APT measurement as well as local magni­
uncharged specimen (overall 5.7 million ions collected). The elements, fication effects between the different phases [34]. Hence, a significant
which are relevant for precipitation, are depicted. Iron, as the main proportion of Fe is recorded (not shown in Fig. 3) within the precipitates
element in the alloy, as well as some minor elements are not shown for and the C content will not reach a stochiometric composition for car­
visibility. The two different particle types can be clearly distinguished. bides of this size. During characterization of uncharged specimens of this
On one hand, ß-NiAl with about a 1:1 ratio of Ni and Al atoms and about material, two populations of nano-scale carbides were apparent, similar
20 at.% Fe forms spherical precipitates with an average diameter of 1.60 to the slightly larger particles from the TEM investigation. Proxigrams of
± 0.11 nm. It has been shown that ß-NiAl incorporates Fe, especially the different nano-scale precipitates in this material are shown in the
during the early stages of precipitation [39]. On the other hand, supplementary material. Most carbides have a measured C content of
Cr-carbide precipitates nucleate heterogeneously on ß-NiAl as well as on about 16 at.%. Considering C loss from multiples, these carbides
dislocations and grain/lath boundaries. Consequently, the particles are correspond to M23C6 [40]. The other carbide population has higher C
not spherical. The mean size, determined as average diameter, was and corresponds to M7C3. Fig. 3(c) displays the D, C and Cr contents. D is
calculated to be 2.15 ± 0.08 nm. in very low concentration at distances far from the Cr-carbide pre­
Fig. 2(b) shows the mass-spectrum of an uncharged specimen. The cipitates. However, it is enriched in the region where C and Cr are
distinct peaks in this mass-to-charge range correspond to H+ and C2+ starting to increase. Deuterium remains at an elevated concentration
ions. In comparison to that, the same mass-to-charge range of the mass- within the carbides. The error bars correspond to the counting statistics
spectrum for the charged specimen is presented in Fig. 2(d). Besides the as computed by AP Suite and are higher deeper inside the carbides since
expected H+ and C2+ peaks, a prominent peak at 2 Da corresponding to fewer ions are sampled.
D+ is present. The charged specimen, shown in Fig. 2(c) was measured with three
Fig. 2(c) shows a 5 nm thick section of the reconstruction of the different voltage pulse frequencies to employ the evaporation rate
charged specimen. Again, ß-NiAl as well as Cr-carbide precipitates are method as demonstrated by Meier et al. [41]. The H content from the
visible. In addition, the measured D ions are depicted as red spheres. The vacuum chamber is dependent on the time between evaporation events,
distribution of these ions seems to be closely connected to the Cr- as was previously shown by Sundell et al. [17]. By varying frequency
carbides. Interestingly, ß-NiAl does not seem to attract and trap D. and/or detection rate, the H content could be measured despite the
To visualize and analyze particles, iso-concentration surfaces are presence of a contaminant H background signal. 1.1 million ions were
widely used. Fig. 3(a) depicts the reconstruction of the charged spec­ collected with 200 kHz pulse frequency and 300 000 ions with 100 and
imen with 1 nm voxel size and 3 nm delocalization. The threshold for 333 kHz each. The concentrations of H, D, and C are displayed in Fig. 4.
iso-concentration surfaces was set to 7 at.% Cr and the proximity his­ The concentration of H clearly depends on the frequency and thus
togram, also known as proxigram, is plotted in Fig. 3(b). The proxigram confirming the expected supply mechanism from the UHV chamber. In
shows the composition as a function of distance to the iso-concentration contrast, the D concentration does not correlate with the frequency.
surface. Within the Cr-carbide precipitates, Ni and Al are depleted, Rather, higher D content is seen in sections of the run with higher C

Fig. 2. Distribution of precipitation forming elements as well as D within a 5 nm thick section of APT reconstructions and selected mass-spectrum ranges of un­
charged (a-b) and D charged (c-d) specimens. Solid arrows indicate exemplary carbides and open arrows show intermetallic particles.

3
S. Jakob et al. Scripta Materialia 243 (2024) 115963

Fig. 3. (a) APT reconstruction of the charged specimen; Cr and D atoms as well as the 7 at.% Cr iso-concentration surfaces are shown. (b) Proxigram of 7 at.% Cr iso-
concentration surfaces displaying atomic concentrations near and within the Cr-carbides; (c) Depiction of Cr, C and D concentration as a function of distance to the
iso-concentration surface.

intermetallic ß-NiAl as well as carbides. The ß-NiAl particles are

coherent and have a small misfit [42]. Therefore, it is not expected that
they provide significant H trapping sites. The alloy has only Cr (except
for a limited amount of Mo) as carbide forming element. In this way, the
trapping capability of Cr-carbides is investigated without interference
from other carbide types.
Since the size of the carbides in this material is in the same range as
the lateral resolution of concentration gradients, the interface of pre­
cipitates cannot be unambiguously distinguished from the interior of the
particles. However, the fact that D is already present at the beginning of
the C and Cr gradient in the vicinity of the precipitates suggests that the
interface is the more prominent trapping site. This is supported by
microautoradiography [43] as well as time-of-flight secondary ion mass
spectrometry [44]. TDMS of lab-cast Fe-C-Cr alloy suggests the interface
as trapping site by comparing the measured H content to the available
interface area of differently sized Cr23C6 precipitates [45]. Looking at
different carbides, it was shown that D resides within VC precipitates by
Fig. 4. Concentrations of H, D as well as C as a function of the time between
evaporation events (calculated from the pulse frequency and the detection rate) APT with fast cooling after charging [13,14]. The fact that trapping sites
during APT measurement. are within the carbides is supported by neutron scattering [46]. It is
known that VC has a high tolerance for C vacancies, thus offering
possible trapping sites for H [47]. In contrast, NbC seems to trap H at the
content, i.e., in regions where more carbides are present.
interfaces as documented by the same methods [15,48].
During APT measurements of steel in voltage mode, as opposed to
TDMS is a commonly used method to measure the de-trapping acti­
laser mode, the evaporation of H+ 2 ions is inhibited. The variation in
vation energy for H of microstructural features in a material. However,
pulse frequency, together with the absence of a 2 Da peak for the un-
the desorption peak shape and temperature is dependent on the charging
charged specimen, validates the assumption that the peak at 2 Da in
method as well as the initial H distribution [49,50]. We performed 24 h
the mass spectrum is indeed D+ and not H+ 2.
charging on the TDMS samples to ensure saturation and a homogeneous
The Hybrid model alloy in this study is precipitation strengthened by

4
S. Jakob et al. Scripta Materialia 243 (2024) 115963

distribution of H through the samples thickness. This was further could not be observed with certainty. They moved on to a gas charging
enhanced by degassing for 24 h at 50 ◦ C. Considering the temperatures cell with fast cooling capability. TiC has been reported to have a high
of the H desorption peaks for the three heating rates, a trapping energy saddle energy around the trapping sites [37]. It could be that electro­
of 15.6 kJ/mol was measured, as illustrated in Fig. 5(a). This is in good chemical charging was not suitable to overcome this saddle energy.
agreement with trapping energies of various carbides in martensitic steel Breen et al. investigated different workflows to measure H/D in pearlitic
investigated in a similar manner by Lee et al. [51]. wire [22]. They found a weak D signal at cementite interfacial regions in
The peak shape in our study as well as in [51] suggests that there is some of their RT samples after 1 h electrolytic charging at -1.2 V,
remaining H in the samples. We therefore performed further degassing however, no D was measured after 5 min of charging. The duration of
for 72 h and 168 h (see Fig. 5(b)). The peak shape in the TDMS curve RT-transfer was not reported, and the highest D concentration was
after 168 h degassing suggests that there is more than one trap. Hence, measured after cryogenic transfer.
deconvolution was performed and trap energies of 45.5 and 56 kJ/mol The findings in this work therefore suggest that nano-scale Cr-car­
were measured, as shown in Fig. 5(c). This is comparable with trapping bides act as strong H traps that can retain H/D for an extended period of
energies for interfaces of Cr23C6 precipitates, reported to be 49 to 55 time. To optimize the H trapping capability, the positions of trapping
kJ/mol for 50 to 100 nm sized precipitates [45] as well as 44.4 kJ/mol sites are important. Trapping inside the carbides would require a large
for Cr23C6 with sizes of about 1 µm [44]. These values are higher than volume fraction in the microstructure. However, since the D concen­
21.5 kJ/mol, reported for coherent Mo2C carbides [10]. Values reported tration is elevated at the interfaces, a large number density of nm-sized
for VC range from 33–35 kJ/mol [9] to 52–67 kJ/mol [52]. In studies carbides offers abundant trapping sites. This leads to a significant
regarding TiC the character of the interface is discussed extensively [38, trapping capacity for H, as can be seen by the amount of desorbed H in
53]. Semi-coherent TiC interfaces are prominent trapping sites with Fig. 5(d), and therefore most likely also resilience to HE. The investi­
de-trapping energies of 46 to 59 kJ/mol, whereas the incoherent TiC gated microstructure in this study is a promising step forward to produce
interfaces are considered irreversible traps for H with de-trapping en­ HE resistant steel with lean alloying design.
ergies larger than 68 kJ/mol [38]. In this work, we showed that D can be trapped in the vicinity of nano-
So far, the use of cryogenic transfer was thought to be necessary. scale Cr-carbides in a co-precipitation model alloy. The measurement in
Takahashi et al. [12] report on unsuccessful cathodic charging of voltage mode prevents the formation of H+ 2 ions overlapping with D .
+

nano-sized TiC containing steel. After a transfer time of 20 min at RT, D Furthermore, the evaporation rate method demonstrated by Meier et al.

Fig. 5. (a) TDMS measurements with different heating rates after 24 h of degassing at 50 ◦ C. The inset shows the activation energy (Ea) calculated from peak
temperatures (TP) and heating rates (HR). (b) Desorption curves with the same heating rate after different degassing times. (c) Deconvoluted TDMS spectrum after
168 h of degassing to two H peaks with different de-trapping activation energies. (d) Total amount of H in wppm evaluated under the desorption curves as a function
of degassing time. Labels indicate the values for measurements with 0.6 ◦ C/s heating rate.

5
S. Jakob et al. Scripta Materialia 243 (2024) 115963

[41] was employed and could confirm that the presence of D is not [16] W. Lefebvre-Ulrikson, F. Vurpillot, X. Sauvage (Eds.), Atom Probe Tomography:
Put Theory Into Practice, Academic Press, London, 2016.
related to measurement parameters. The amount of D is rather corre­
[17] G. Sundell, M. Thuvander, H.O. Andrén, Hydrogen analysis in APT: methods to
lated to C. We show that D is trapped already in the vicinity of control adsorption and dissociation of H2, Ultramicroscopy 132 (2013) 285–289,
Cr-carbides suggesting the interface as the major trapping site. No D is https://doi.org/10.1016/j.ultramic.2013.01.007.
located in connection to the intermetallic ß-NiAl. The remaining traps [18] Y. Hayashi, H. Hagi, A. Tahara, Diffusion coefficients of hydrogen and deuterium in
iron determined by permeation with gas, ion and electrochemical charging,
after 168 h of degassing at 50 ◦ C have trapping energies of 45.5 and 56 Zeitschrift Fur Phys. Chem. 164 (1989) 815–820, https://doi.org/10.1524/
kJ/mol measured by TDMS. Moreover, this work demonstrates that zpch.1989.164.Part_1.0815.
strong H trapping sites can be measured in APT without cryogenic [19] W. Raczynski, Permeability, Diffusivity, and Solubility of Hydrogen and Deuterium
in Pure Iron at 10 to 60 C, Phys. Status Solid. 48 (1978) K27–K30, https://doi.org/
transfer. 10.1002/pssa.2210480143.
[20] H. Takamizawa, K. Hoshi, Y. Shimizu, F. Yano, K. Inoue, S. Nagata, T. Shikama,
Y. Nagai, Three-dimensional characterization of deuterium implanted in silicon
Declaration of Competing Interest using atom probe tomography, Appl. Phys. Express. 6 (2013), https://doi.org/
10.7567/APEX.6.066602.
[21] D. Haley, S.V. Merzlikin, P. Choi, D. Raabe, Atom probe tomography observation of
The authors declare that they have no conflict of interest. hydrogen in high-Mn steel and silver charged via an electrolytic route, Int. J.
Hydrogen Energy 39 (2014) 12221–12229, https://doi.org/10.1016/j.
ijhydene.2014.05.169.
Acknowledgements [22] A.J. Breen, L.T. Stephenson, B. Sun, Y. Li, O. Kasian, D. Raabe, M. Herbig, B. Gault,
Solute hydrogen and deuterium observed at the near atomic scale in high-strength
The model alloy is produced under the Vinnova sponsored project steel, Acta Mater. 188 (2020) 108–120, https://doi.org/10.1016/j.
actamat.2020.02.004.
HY-Toolbox (2020-03115). The research is funded by the Swedish [23] I.E. McCarroll, Y.C. Lin, A. Rosenthal, H.W. Yen, J.M. Cairney, Hydrogen trapping
Research Council (2021-05072). The APT and TEM measurements were at dislocations, carbides, copper precipitates and grain boundaries in a dual
performed at Chalmers Materials Analysis Laboratory (CMAL). The precipitating low-carbon martensitic steel, Scr. Mater. 221 (2022) 114934, https://
doi.org/10.1016/j.scriptamat.2022.114934.
TDMS measurements were performed at Swerim AB. [24] L.T. Stephenson, A. Szczepaniak, I. Mouton, K.A.K. Rusitzka, A.J. Breen, U. Tezins,
A. Sturm, D. Vogel, Y. Chang, P. Kontis, A. Rosenthal, J.D. Shepard, U. Maier, T.F.
Kelly, D. Raabe, B. Gault, The Laplace project: an integrated suite for correlative
Supplementary materials
atom probe tomography and electron microscopy under cryogenic and UHV
conditions, (2018) 1–13. http://arxiv.org/abs/1805.10836.
Supplementary material associated with this article can be found, in [25] J.M. Cairney, I. McCarroll, Y.S. Chen, K. Eder, T. Sato, Z. Liu, A. Rosenthal,
the online version, at doi:10.1016/j.scriptamat.2023.115963. R. Wepf, Correlative UHV-Cryo Transfer Suite, Connecting atom probe, SEM-FIB,
Transmission electron microscopy via an environmentally-controlled glovebox,
Microsc. Microanal. 25 (2019) 2494–2495, https://doi.org/10.1017/
References s1431927619013205.
[26] I.E. McCarroll, P.A.J. Bagot, A. Devaraj, D.E. Perea, J.M. Cairney, New frontiers in
atom probe tomography: a review of research enabled by cryo and/or vacuum
[1] S. Lynch, Hydrogen embrittlement phenomena and mechanisms, Corros. Rev. 30
transfer systems, Mater. Today Adv. 7 (2020) 100090, https://doi.org/10.1016/j.
(2012) 105–123, https://doi.org/10.1515/corrrev-2012-0502.
mtadv.2020.100090.
[2] Z. Sun, C.H. Liebscher, S. Huang, Z. Teng, G. Song, G. Wang, M. Asta, M. Rawlings,
[27] Z.B. Jiao, J.H. Luan, M.K. Miller, Y.W. Chung, C.T. Liu, Co-precipitation of
M.E. Fine, P.K. Liaw, New design aspects of creep-resistant NiAl-strengthened
nanoscale particles in steels with ultra-high strength for a new era, Mater. Today
ferritic alloys, Scr. Mater. 68 (2013) 384–388, https://doi.org/10.1016/J.
20 (2017) 142–154, https://doi.org/10.1016/J.MATTOD.2016.07.002.
SCRIPTAMAT.2012.10.040.
[28] H.J. Kong, C.T. Liu, A Review On Nano-Scale Precipitation In Steels, Technologies
[3] J. Venezuela, Q. Zhou, Q. Liu, H. Li, M. Zhang, M.S. Dargusch, A. Atrens, The
6 (2018) 36, https://doi.org/10.3390/TECHNOLOGIES6010036.
influence of microstructure on the hydrogen embrittlement susceptibility of
[29] Z. Xiong, I. Timokhina, E. Pereloma, Clustering, nano-scale precipitation and
martensitic advanced high strength steels, Mater. Today Commun. 17 (2018) 1–14,
strengthening of steels, Prog. Mater. Sci. 118 (2021) 100764, https://doi.org/
https://doi.org/10.1016/J.MTCOMM.2018.07.011.
10.1016/J.PMATSCI.2020.100764.
[4] W.H. Johnson, On some remarkable changes produced in iron and steel by the
[30] J.E. Andersson, F. Lindberg, P. Ölund, Hybrid steel enhances component
action of hydrogen and acids, R. Soc. 23 (1875) 168–179.
performance at reduced weight, Light. Des. Worldw. 12 (2019) 34–39, https://doi.
[5] T.I. Ramjaun, S.W. Ooi, R. Morana, H.K.D.H. Bhadeshia, Designing steel to resist
org/10.1007/s41777-018-0063-3.
hydrogen embrittlement: part 1–trapping capacity, Mater. Sci. Technol. 34 (2018)
[31] J.E. Andersson, F. Lindberg, S. Ooi, Hybrid steel and its potential for bearing
1737–1746, https://doi.org/10.1080/02670836.2018.1475919.
applications. ASTM Special Technical Publication, ASTM International, 2020,
[6] H.K.D.H. Bhadeshia, Prevention of hydrogen embrittlement in steels, ISIJ Int. 56
pp. 436–454, https://doi.org/10.1520/STP162320190163.
(2016) 24–36, https://doi.org/10.2355/isijinternational.ISIJINT-2015-430.
[32] H. Zhao, B. Gault, D. Ponge, D. Raabe, F. De Geuser, Parameter free quantitative
[7] H.G. Lee, J.Y. Lee, Hydrogen trapping by TiC particles in iron, Acta Metall. 32
analysis of atom probe data by correlation functions: application to the
(1984) 131–136, https://doi.org/10.1016/0001-6160(84)90210-4.
precipitation in Al-Zn-Mg-Cu, Scr. Mater. 154 (2018) 106–110, https://doi.org/
[8] C. Zhang, Y. Liu, C. Jiang, J. Xiao, Effects of niobium and vanadium on hydrogen-
10.1016/j.scriptamat.2018.05.024.
induced delayed fracture in high strength spring steel, J. Iron Steel Res. Int. 18
[33] F. De Geuser, W. Lefebvre, F. De Geuser, W. Lefebvre, Determination of matrix
(2011) 49–53, https://doi.org/10.1016/S1006-706X(11)60077-0.
composition based on solute-solute nearest-neighbor distances in atom probe
[9] H. Asahi, D. Hirakami, S. Yamasaki, Hydrogen trapping behavior in vanadium-
tomography, Microsc. Res. Tech. 74 (2011) 257–263, https://doi.org/10.1002/
added steel, ISIJ Int. 43 (2003) 527–533, https://doi.org/10.2355/
jemt.20899.
isijinternational.43.527.
[34] F. Vurpillot, A. Bostel, D. Blavette, Trajectory overlaps and local magnification in
[10] D. Li, R.P. Gangloff, J.R. Scully, Hydrogen Trap states in ultrahigh-strength
three-dimensional atom probe, Appl. Phys. Lett. 76 (2000) 3127–3129, https://doi.
AERMET 100 steel, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 35 A (2004)
org/10.1063/1.126545.
849–864, https://doi.org/10.1007/s11661-004-0011-1.
[35] D.J. Larson, B. Gault, B.P. Geiser, F. De Geuser, F. Vurpillot, Atom probe
[11] Y.C. Lin, I.E. McCarroll, Y.T. Lin, W.C. Chung, J.M. Cairney, H.W. Yen, Hydrogen
tomography spatial reconstruction: status and directions, Curr. Opin. Solid State
trapping and desorption of dual precipitates in tempered low-carbon martensitic
Mater. Sci. 17 (2013) 236–247, https://doi.org/10.1016/j.cossms.2013.09.002.
steel, Acta Mater. 196 (2020) 516–527, https://doi.org/10.1016/j.
[36] H.E. Kissinger, Reaction kinetics in differential thermal analysis, Anal. Chem 29
actamat.2020.06.046.
(1957) 1702–1706, https://doi.org/10.1021/ac60131a045.
[12] J. Takahashi, K. Kawakami, Y. Kobayashi, T. Tarui, The first direct observation of
[37] J.Y. Lee, S.M. Lee, Hydrogen trapping phenomena in metals with B.C.C. and F.C.C.
hydrogen trapping sites in TiC precipitation-hardening steel through atom probe
crystals structures by the desorption thermal analysis technique, Surf. Coatings
tomography, Scr. Mater. 63 (2010) 261–264, https://doi.org/10.1016/j.
Technol. 28 (1986) 301–314, https://doi.org/10.1016/0257-8972(86)90087-3.
scriptamat.2010.03.012.
[38] F.G. Wei, T. Hara, K. Tsuzaki, Precise determination of the activation energy for
[13] J. Takahashi, K. Kawakami, T. Tarui, Direct observation of hydrogen-trapping sites
desorption of hydrogen in two Ti-added steels by a single thermal-desorption
in vanadium carbide precipitation steel by atom probe tomography, Scr. Mater. 67
spectrum, Metall. Mater. Trans. B Process Metall. Mater. Process. Sci. 35 (2004)
(2012) 213–216, https://doi.org/10.1016/j.scriptamat.2012.04.022.
587–597, https://doi.org/10.1007/s11663-004-0057-x.
[14] Y.S. Chen, D. Haley, S.S.A. Gerstl, A.J. London, F. Sweeney, R.A. Wepf, W.
[39] Z. Guo, W. Sha, D. Vaumousse, Microstructural evolution in a PH13-8 stainless
M. Rainforth, P.A.J. Bagot, M.P. Moody, Direct observation of individual hydrogen
steel after ageing, Acta Mater. 51 (2003) 101–116, https://doi.org/10.1016/
atoms at trapping sites in a ferritic steel, Science 355 (80) (2017) 1196–1199,
S1359-6454(02)00353-1.
https://doi.org/10.1126/science.aal2418.
[40] M. Thuvander, J. Weidow, J. Angseryd, L.K.L. Falk, F. Liu, M. Sonestedt, K. Stiller,
[15] Y.S. Chen, H. Lu, J. Liang, A. Rosenthal, H. Liu, G. Sneddon, I. McCarroll, Z. Zhao,
H.O. Andrén, Quantitative atom probe analysis of carbides, Ultramicroscopy 111
W. Li, A. Guo, J.M. Cairney, Observation of hydrogen trapping at dislocations,
(2011) 604–608, https://doi.org/10.1016/J.ULTRAMIC.2010.12.024.
grain boundaries, and precipitates, Science 367 (80) (2020) 171–175, https://doi.
org/10.1126/science.aaz0122.

6
S. Jakob et al. Scripta Materialia 243 (2024) 115963

[41] M.S. Meier, M.E. Jones, P.J. Felfer, M.P. Moody, D. Haley, Extending estimating [47] J. Takahashi, K. Kawakami, Y. Kobayashi, Origin of hydrogen trapping site in
hydrogen content in atom probe tomography experiments where H 2 molecule vanadium carbide precipitation strengthening steel, Acta Mater. 153 (2018)
formation occurs, Microsc. Microanal. 28 (2022) 1231–1244, https://doi.org/ 193–204, https://doi.org/10.1016/J.ACTAMAT.2018.05.003.
10.1017/S1431927621012332. [48] M. Ohnuma, J. ichi Suzuki, F.G. Wei, K. Tsuzaki, Direct observation of hydrogen
[42] L.M. Pike, I.M. Anderson, C.T. Liu, Y.A. Chang, Site occupancies, point defect trapped by NbC in steel using small-angle neutron scattering, Scr. Mater. 58 (2008)
concentrations, and solid solution hardening in B2 (Ni,Fe)Al, Acta Mater 50 (2002) 142–145, https://doi.org/10.1016/j.scriptamat.2007.09.026.
3859–3879, https://doi.org/10.1016/S1359-6454(02)00192-1. [49] M. Enomoto, L. Cheng, H. Mizuno, Y. Watanabe, T. Omura, J. Sakai, K. Yokoyama,
[43] C. Paes de Oliveira, M. Aucouturier, P. Lacombe, Hydrogen trapping in BCC Fe-Cr H. Suzuki, R. Okuma, Hydrogen Absorption into austenitic stainless steels under
Alloys (7 ~ 9. 4 Wt% Cr) as studied by microautoradiography-contribution of high-pressure gaseous hydrogen and cathodic charge in aqueous solution, Metall.
carbon-hydrogen interaction-consequences on hydrogen cracking, Corrosion 36 Mater. Trans. E. 1 (2014) 331–340, https://doi.org/10.1007/s40553-014-0034-5.
(1980) 53–59, https://doi.org/10.5006/0010-9312-36.2.53. [50] E. Wallaert, T. Depover, M. Arafin, K. Verbeken, Thermal desorption spectroscopy
[44] T. Wang, X. Fang, H. Zhang, W. Lv, J. Ma, W. Liang, L. Zheng, Observation of evaluation of the hydrogen-trapping capacity of NbC and NbN precipitates, Metall.
hydrogen trap and hydrogen embrittlement of 430 ferritic stainless steel, Mater. Mater. Trans. A Phys. Metall. Mater. Sci. 45 (2014) 2412–2420, https://doi.org/
Lett. 313 (2022) 131728, https://doi.org/10.1016/j.matlet.2022.131728. 10.1007/s11661-013-2181-1.
[45] T. Depover, K. Verbeken, Hydrogen trapping and hydrogen induced mechanical [51] J. Lee, T. Lee, D.J. Mun, C.M. Bae, C.S. Lee, Comparative study on the effects of Cr,
degradation in lab cast Fe-C-Cr alloys, Mater. Sci. Eng. A. 669 (2016) 134–149, V, and Mo carbides for hydrogen-embrittlement resistance of tempered martensitic
https://doi.org/10.1016/j.msea.2016.05.018. steel, Sci. Rep. 9 (2019) 1–9, https://doi.org/10.1038/s41598-019-41436-2.
[46] B. Malard, B. Remy, C. Scott, A. Deschamps, J. Chêne, T. Dieudonné, M.H. Mathon, [52] T. Depover, K. Verbeken, Evaluation of the effect of V4C3 precipitates on the
Hydrogen trapping by VC precipitates and structural defects in a high strength Fe- hydrogen induced mechanical degradation in Fe-C-V alloys, Mater. Sci. Eng. A. 675
Mn-C steel studied by small-angle neutron scattering, Mater. Sci. Eng. A. 536 (2016) 299–313, https://doi.org/10.1016/j.msea.2016.08.053.
(2012) 110–116, https://doi.org/10.1016/j.msea.2011.12.080. [53] F.G. Wei, K. Tsuzaki, Quantitative analysis on hydrogen trapping of TiC particles in
steel, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 37 (2006) 331–353, https://
doi.org/10.1007/S11661-006-0004-3/METRICS.