Revealing dehydrogenation effect and resultant densification mechanism during pressureless sintering of TiH2 powder

Article abstract of DOI:10.1016/j.jallcom.2021.159792

The use of TiH₂ powder as a sintering precursor can produce nearly full-density titanium and titanium alloys with good mechanical properties. Unfortunately, there is a lack of research on the effect of lattice defects generated during the dehydrogenation of TiH₂ powder, and the underlying sintering diffusion mechanism and activation energy have yet to be determined. In this work, we report a two-step sintering strategy to reveal the dehydrogenation effect and resultant densification mechanism during the pressureless sintering of a TiH₂ powder precursor. The results show that, compared with hydrogenated-dehydrogenated (HDH) Ti powder, TiH₂ powder, an intermediate of HDH-Ti powder, exhibited a higher instantaneous densification rate, greater onset relative density, rapid grain growth, and thus a smaller grain size. It also showed a grain boundary diffusion mechanism below 91% relative density and half the sintering activation energy in the intermediate sintering stage. Fundamentally, this was attributed to lattice defects generated during the dehydrogenation of TiH₂ powder, which was confirmed by the greater relative density of a sintered TiH₂ compact due to the two-step sintering strategy designed herein. Interestingly, the sintered sample obtained from the TiH₂ powder precursor has a satisfying combination of strength and ductility that is far superior to other bulk Ti materials, especially sintered bulk Ti obtained from HDH-Ti powder. The results obtained in this paper provide theoretical guidance for using pressureless sintering to produce nearly full-density Ti and Ti alloys with good mechanical properties for structural applications.

Full text of DOI:10.1016/j.jallcom.2021.159792

Journal of Alloys and Compounds 873 (2021) 159792
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jalcom
Revealing dehydrogenation effect and resultant densification mechanism
during pressureless sintering of TiH2 powder
T. Chena, C. Yanga,⁎, Z. Liua, H.W. Maa, L.M. Kangc, Z. Wanga, W.W. Zhanga, D.D. Lib, N. Lib,
Y.Y. Lia,b
a National Engineering Research Center of Near-net-shape Forming for Metallic Materials, South China University of Technology, Guangzhou 510640, China
b State Key Laboratory of Materials Processing and Die & Mould Technology, Huazhong University of Science and Technology, Wuhan 430074, China
c Guangzhou Railway Polytechnic, Guangzhou 510430, China
a r t i c l e i n f o
a b s t r a c t
Article history:
The use of TiH2 powder as a sintering precursor can produce nearly full-density titanium and titanium alloys
with good mechanical properties. Unfortunately, there is a lack of research on the effect of lattice defects
generated during the dehydrogenation of TiH2 powder, and the underlying sintering diffusion mechanism and
activation energy have yet to be determined. In this work, we report a two-step sintering strategy to reveal the
dehydrogenation effect and resultant densification mechanism during the pressureless sintering of a TiH2
powder precursor. The results show that, compared with hydrogenated-dehydrogenated (HDH) Ti powder,
TiH2 powder, an intermediate of HDH-Ti powder, exhibited a higher instantaneous densification rate, greater
onset relative density, rapid grain growth, and thus a smaller grain size. It also showed a grain boundary
diffusion mechanism below 91% relative density and half the sintering activation energy in the intermediate
sintering stage. Fundamentally, this was attributed to lattice defects generated during the dehydrogenation of
TiH2 powder, which was confirmed by the greater relative density of a sintered TiH2 compact due to the two-
step sintering strategy designed herein. Interestingly, the sintered sample obtained from the TiH2 powder
precursor has a satisfying combination of strength and ductility that is far superior to other bulk Ti materials,
especially sintered bulk Ti obtained from HDH-Ti powder. The results obtained in this paper provide theoretical
guidance for using pressureless sintering to produce nearly full-density Ti and Ti alloys with good mechanical
properties for structural applications.
Received 13 January 2021
Received in revised form 13 March 2021
Accepted 27 March 2021
Available online 8 April 2021
Keywords:
Titanium hydride
Sintering mechanism
Microstructure
Mechanical properties
© 2021 Elsevier B.V. All rights reserved.
1. Introduction
separate from grain boundaries, remain in the grain interior, and
finally entrapped in the sintered products [7]. The relative density of
Due to their low density, high specific strength, good ductility,
and excellent corrosion resistance [1,2], titanium and its alloys are
wildly used as structural materials in the aerospace, automotive,
military, and chemical industries. As an extremely important ma-
terial forming approach, powder sintering allows the low-cost pro-
duction of Ti components while reducing material waste [3–6].
Generally, it is difficult to achieve nearly full-density Ti components
via the pressureless sintering of Ti powder, and they are commonly
produced by the hydrogenation-dehydrogenation (HDH) method.
The reason, the densification of HDH-Ti powder is insufficient to
eliminate the majority of pores before rapid grain growth, which
gives rise to pore-boundary separation, in which residual pores
pressureless-sintered components obtained using HDH-Ti powder is
typically lower than 98% [8,9].
In addition to pressure-assisted sintering methods such as spark
plasma sintering [10–14], hot isostatic pressing [15,16], vacuum hot
pressing [17,18], novel additive manufacturing [19–21] and sever
plastic deformation methods [22], an effective strategy for obtaining
nearly full-density bulk Ti and Ti alloys is to use TiH2 powder instead
of HDH-Ti powder as the starting material. This strategy has three
well-known advantages. First, HDH-Ti powder is produced by hy-
drogenating Ti raw material, milling titanium hydride into a powder,
and finally dehydrogenating milled titanium hydride powder;
therefore, TiH2 powder is an intermediate of HDH production and is
cheaper than HDH-Ti powder. Second, the hydrogen in TiH2 powder
can be employed as a temporary alloying element to control phase
transformation and induce grain refinement, which can improve the
mechanical properties [5,23,24]. For example, an ultrafine lamellar
⁎
Corresponding author.
E-mail address: cyang@scut.edu.cn (C. Yang).
https://doi.org/10.1016/j.jallcom.2021.159792
0925-8388/© 2021 Elsevier B.V. All rights reserved.

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Ti–6Al–4V alloy with an ultimate strength of 1431 MPa and 4.5%
elongation was obtained by the hot-extrusion of TiH2/60Al40V
blended powder [23]. Third and most importantly, compared with
HDH-Ti powder, the use of TiH2 powder as the starting sintering
material induces faster densification and produces sintered Ti
components with a higher relative density. Specifically, the relative
density can reach as high as 98.5–99.5% by pressureless vacuum
sintering [25,26] or hydrogen sintering and phase transformation
[27–29], which is equivalent to the corresponding pressure-assisted
sintering methods.
determine the sintering activation energy. The related sintering
parameters were selected to be: heating to 1250 °C and holding for
4 h at various heating rates of 2 °C/min, 5 °C/min, 10 °C/min, and
15 °C/min, respectively. Isothermal sintering was used to estimate
the diffusion mechanism, by heating to 950 °C, 1000 °C, 1050 °C,
1100 °C, and 1150 °C at 15 °C/min, with a dwell time of 2 h. To ex-
amine the grain size evolution under different sintering tempera-
tures and holding times, the two types of green compacts with
dimensions of 62.5 mm × 13.5 mm× 15 mm were heated to 1050 °C,
1100 °C, and 1150 °C at 10 °C/min without holding, or to 1250 °C at
10 °C/min and holding for 0, 1, 2, and 4 h, respectively. This was
followed by rapid cooling to ambient temperature in a flowing high-
purity Ar atmosphere. The grain size was measured optically and
determined using ImageJ software. A two-step sintering strategy
was designed and employed to illustrate the dehydrogenation ef-
fects. Specifically, the dehydrogenation behavior of a ϕ3 mm× 2 mm
TiH2 green compact was studied under a high-purity Ar atmosphere
in a NETZSCH STA 449F3 DSC by heating to 870 °C at 10 °C/min,
cooling to ambient temperature, then heating to 1250 °C at 10 °C/
min, and holding for 4 h. All samples for microstructure observations
were polished by abrasive paper, etched by Kroll solution (2 mL HF:
4 mL HNO3: 100 mL H2O), and finally characterized by scanning
electron microscopy (SEM, FEI Nova Nano SEM 430) and optical
microscopy. To evaluate the mechanical properties, 62.5 mm × 13.5
mm× 15 mm green compacts were sintered at 1100 °C and 1250 °C at
10 °C/min with a dwell time of 4 h under Ar protection. The tensile
samples with a gauge size of 16 mm, width of 2.5 mm, and thickness
of 2 mm, were machined according to ASTM E9-81 from as-sintered
bulk Ti and then tested using a universal testing machine under
quasi-static loading at a strain rate of 5 × 10−4 s−1. The oxygen con-
tents were measured by a TC600 Nitrogen/Oxygen analyzer (LECO
Co., USA). The density of the as-sintered specimens was analyzed by
Archimedes’ method.
Powder shrinkage and densification are macroscopic manifesta-
tions of pore elimination and atomic diffusion under the influence of
various factors, including the intrinsic properties of powders [7],
pressing and sintering parameters [30], defect concentration [31],
impurities [32,33], etc. In essence, densification mechanisms are
associated with specific physical quantities including the densifica-
tion rate ( ), time exponent (n), activation energy (Q), comprehen-
sive impact factor, etc. [31,34,35]. To date, some efforts have been
made to investigate the sintering behavior of TiH2 and HDH-Ti
powders, yet only the differences in shrinkage and final relative
density were compared [32,36]. Notably, no research has quantita-
tively compared the sintering densification mechanism of the two
types of powders from the perspective of the diffusion mechanism
and sintering activation energy, which are significant issues for
powder sintering. Two main effects account for the greater densifi-
cation of TiH2 powder. The first is the elimination of atomic diffusion
by producing a hydrogen-thinned oxide layer due to evolved hy-
drogen. The second is the acceleration of densification due to extra
defects generated by phase transformation during dehydrogenation
[37]. Revealing the related sintering mechanism and illustrating the
limitations of these two effects are crucial for achieving nearly full-
density bulk materials. Some investigations have verified a reduction
in the oxygen content by water generation during the dehy-
drogenation of TiH2 powder [36,38–41] and, thus, the promotion of
sintering densification [25,42]. Unfortunately, no work has clarified
the existence and impact of the aforementioned defects generation,
which might plays a critical role in the sintering of TiH2 powder.
The purpose of this work was to reveal the dehydrogenation ef-
fect and resultant densification mechanism during the pressureless
sintering of TiH2 powder to supplement the sintering mechanism of
TiH2 powder and help select appropriate sintering parameters to
fabricate nearly full-density Ti components with good mechanical
properties. To that effect, TiH2 and HDH-Ti compacts were sintered
through isothermal and non-isothermal methods, and their
shrinkage during sintering and mechanical properties were quanti-
tatively compared. In particular, grain growth, diffusion mechanism,
and activation energy were determined, and the effects of dehy-
drogenation on densification and their limitations were discussed in
detail.
3. Results
3.1. Shrinkage behavior
Fig. 2a presents the shrinkage-temperature curves of the TiH2
and HDH-Ti green compacts at different heating rates. It is found
that the HDH-Ti compacts began to shrink near 878 °C with a total
linear shrinkage of about 3.5% until 1250 °C. In contrast, the TiH2
compacts had a far lower initial shrinkage temperature around
320 °C and, thus, a far greater linear shrinkage of about 11%. The DSC
curve of the TiH2 compact (Fig. 2a inset) measured at 10 °C/min
shows a remarkable endothermic peak starting at 332 °C and ending
at 861 °C (lower than the phase transition temperature of HCP Ti→
BCC Ti). The corresponding TG curve presents a 3.09% total weight
loss in the same temperature range. These two events indicate that
the TiH2 compacts were dehydrogenated from 332 °C to 861 °C. Since
the phase transformation from FCC TiH2 to HCP Ti is accompanied by
volumetric shrinkage and weight loss induced by dehydrogenation,
the shrinkage of the TiH2 compacts before approximately 861 °C was
attributed to dehydrogenation. The exact dehydrogenation ending
temperatures of the TiH2 compacts were obtained by analyzing the
shrinkage curves with various heating rates of 2 °C/min, 5 °C/min,
10 °C/min, and 15 °C/min, and are 856 °C, 859 °C, 873 °C, and 892 °C,
respectively. To directly compare the sintering densification beha-
viors of the Ti and TiH compacts (transformed from the HDH-Ti and
TiH2 green compacts), the shrinkage curves of the TiH2 compacts
were corrected by removing the shrinkage of dehydrogenation ef-
fect, as shown in Fig. 2b. Both Ti and TiH compacts began to densify
near the HCP Ti→BCC Ti phase transition temperature due to the
enhanced self-diffusivity in the BCC Ti region [28,43]. Furthermore,
the shrinkage of both compacts increased upon decreasing the
heating rate at a given temperature, possibly because the lower
2. Materials and experimental methods
The as-received powders were − 325 mesh TiH2 (O: 0.24 wt%) and
HDH-Ti (O: 0.19 wt%) powders supplied by CW Nano Technology Co.,
Ltd., China, whose morphologies are shown in Fig. 1. The two pow-
ders have irregular shapes with the same particle size; therefore, the
powder shape and size should have a negligible influence on the
sintering densification in this work. The two types of powders were
uniaxially pressed at a pressure of 600 MPa and held for 30 s to
produce green compacts by a hydraulic machine. Subsequently, all
green compacts were subjected to pressureless sintering. A dilat-
ometer (NETZSCH DIL 402 PC) was applied to collect shrinkage data
of TiH2 and HDH-Ti compacts with dimensions of 5 mm × 5
mm× 20 mm. All compacts heated in the dilatometer were protected
by a high-purity (99.999%) Ar atmosphere. Non-isothermal sintering
was conducted to investigate powder shrinkage and further
2

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Fig. 1. SEM images of the as-received TiH2 (a and c) and HDH-Ti (b and d) powders.
3
heating rate provides a longer sintering time. Interestingly, the TiH
compacts have a distinctly higher linear shrinkage rate than the
corresponding Ti compacts at a given heating rate at temperatures
below approximately 1170 °C (Fig. 2b). Meanwhile, the TiH compacts
displayed a total linear shrinkage of about 6%, far greater than that
(3.5%) of the Ti compacts. Undoubtedly, these results indicate that
there is a unique densification mechanism due to dehydrogenation
during the pressureless sintering of TiH2 compacts.
1
1
Lf /L0
Lx/L0
=
x
f
(1)
where f (%) refers to the final relative density of the sintered com-
pact, L0 is the initial length of the green compact, Lf is the final
length change after sintering, and Lx is the instantaneous length
change at any temperature corresponding to the relative density
.
x
Table 1 lists the initial and final densities, peak densification rate,
and corresponding temperature of the TiH2 and Ti compacts at
various heating rates. The initial density of both compacts after
pressing was measured to be about 84% (Fig. 2b). In contrast, the
During pressureless sintering, the relative density of the sintered
compact x (%) at any temperature can be calculated from the linear
shrinkage by [44]:
Fig. 2. (a) Shrinkage-temperature curves of the TiH2 and HDH-Ti green compacts at different heating rates. a inset shows the TG and DSC curves of the TiH2 green compact. (b)
Corresponding shrinkage curves after removing shrinkage resulting from dehydrogenation.
3

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Table 1
The initial and final density, peak densification rate, and corresponding temperature
for the TiH2 and Ti compacts sintered at various heating rates.
Compact
Heating
rate
(°C/min)
Initial
density
(%)
Final
density
(%)
Peak
Peak
densification
densification
temperature
(°C)
rate (×10−3%/s)
TiH2
Ti
2
5
10
15
2
5
10
15
84.3
84.6
84.5
84.2
84.3
84.6
84.7
84.2
97.5
96.5
95.8
95.0
94.6
94.1
92.4
91.8
2.3
5.5
10.4
14.8
1.4
3.1
6.0
10.1
1032
1058
1060
1082
1080
1115
1193
1194
density of the TiH2 compact after dehydrogenation decreased to
79–81% (Figs. 2 and 3a), decreasing upon increasing the heating rate
(Fig. 3a) due to the almost constant pore volume and the decreased
volume of the TiH2 sintered entity during dehydrogenation. A similar
phenomenon was also observed in the sintering of a TiH2 compact
[45]. Fig. 3b shows the instantaneous densification rate (d /dt) vs
temperature curves of the TiH and Ti compacts at various heating
rates. Evidently, the instantaneous densification rate increased upon
increasing the heating rate at a given temperature for each sintered
compact (Fig. 3b). Meanwhile, the TiH2 compacts displayed a far
greater peak densification rate and lower peak densification tem-
perature than the corresponding Ti compacts at a given heating rate
(Table 1). Taking 10 °C/min as an example, the TiH compact reached a
peak densification rate of 10.4 × 10−3%/s at 1060 °C, while the cor-
responding Ti compact was 6.0 × 10−3%/s at 1193 °C. The higher peak
densification rate and lower peak densification temperature make it
possible to produce nearly full-density bulk materials via low-tem-
perature sintering of the TiH2 compacts. Generally, powder compacts
pressed with a lower initial relative density usually densify faster
during sintering, but they achieve lower final densities after sin-
tering [7]. On the contrary, in this work, although the TiH compacts
after dehydrogenation obtained a lower relative density (79.3%) than
the Ti compact (84.7%) at 10 °C/min, it reached a higher relative
density of 95.8% than the Ti compact (92.4%) after sintering at
1250 °C (Fig. 3a). It can be concluded that there are significant causes
accounting for the higher peak densification rate and lower peak
densification temperature and higher final relative density of the TiH
compacts.
Fig. 4. Tensile stress-strain curves of TiH and Ti samples sintered by heating to 1100 °C
and 1250 °C at 10 °C/min and dwelling for 4 h. The insets show the corresponding
optical microstructures of the two types of bulk samples.
3.2. Mechanical properties
Fig. 4 displays the tensile stress-strain curves of the TiH and Ti
samples sintered by heating to 1100 °C and 1250 °C at 10 °C/min and
dwelling for 4 h. The corresponding tensile mechanical property
values are listed in Table 2. The relative density of the TiH samples
exceeded 98%, a threshold density that ensures ductility, while the
relative density of the Ti samples was lower than this. Sample TiH-
1100, when sintered at a far lower temperature (1250 °C) than ty-
pically used in sintering [25], has a density as high as 98.5%, com-
pared to only 96.2% of the sample Ti-1100. This high densification of
TiH-1100 was attributed to enhanced atomic diffusion in the TiH2
compact sintered at 1100 °C, coinciding with its peak densification
temperature of 1060 °C (Fig. 3b and Table 1). The TiH samples have a
satisfying combination of yield strength, ultimate strength, and
elongation, which are respectively 93 MPa, 133 MPa, and 3.3% higher
than the corresponding values of their Ti counterpart. The sintering
temperature also significantly affected the relative density and me-
chanical properties of the Ti samples sintered from HDH-Ti powder.
Sample Ti-1250 had a relatively high density of 97.2% as well as a
large elongation of 17.8%, yet the corresponding values were 96.2%
and 8.25% for sample Ti-1100. This means that 1100 °C is an
Fig. 3. Relative density (a) and instantaneous densification rate (b) vs temperature curves of the TiH and Ti compacts at various heating rates.
4

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Table 2
Relative density and tensile mechanical properties of the two types of bulk samples fabricated in Fig. 4. VS: Vacuum sintering; SPS: Spark plasma sintering; VHP: Vacuum hot
pressing; SLM: Selective laser melting; Ti-4: Wrought Ti grade 4.
Samples
Yield strength (MPa)
Ultimate strength (MPa)
Elongation (%)
Relative density (%)
Oxygen content (wt%)
TiH-1250
TiH-1100
Ti-1250
Ti-1100
Ti-VS [46]
Ti-VS [47]
Ti-SPS [14]
Ti-SPS [13]
Ti-VHP [18]
Ti-SLM [21]
Ti-SLM [20]
Ti-Forging [48]
Wrought Ti-4 [14]
533
529
409
436
458
407
595
286
404
533
502
490
483
635
659
502
516
549
515
720
364
659
714
600
619
550
21.1
20.8
17.8
8.5
6.4
18.4
18
39.4
4.84
17.8
34.3
29
99.0
98.5
97.2
96.2
97.5
94.5
99.0
>99.0
99.54
99.5
99.6
100
0.28
0.27
0.32
0.30
0.37
0.23
0.31
~
0.36
0.18
0.11
0.19
~
15
100
n
inappropriate sintering temperature for HDH-Ti powder. In addition
to the higher relative density (Table 2) and the finer grain size (Fig. 4
inset) of the TiH samples, the higher oxygen content in a certain
range also improves the yield strength. It is interesting to find that
the TiH samples circumvented the strength-ductility trade-off, and
their excellent combination of strength and ductility are far superior
to bulk Ti materials produced by vacuum sintering [46,47], and are
equivalent to those obtained by pressure-assisted sintering (SPS and
VHP) [13,14,18], forging [48], and selective laser melting [20,21], as
shown in Table 2. Therefore, applying TiH2 powder as a starting
material is an effective approach for fabricating nearly full-density
bulk Ti and Ti alloys, even via pressureless sintering in our case.
Additionally, considering the accelerated densification behavior and
reduced peak densification temperature of the TiH2 compacts, it is
possible to sinter Ti bulk materials at 1100 °C or even lower. If the
homogenization of alloying elements is taken into consideration,
this approach can also be used to optimize pressureless sintering
parameters of Ti alloys, such as Ti–6Al–4 V.
H (T)
Gm
L
L0
=
tn
(2)
where H (T) represents a function based on temperature and is re-
lated to geometry and property parameters of power particles: G
(μm) is the grain size, m and n are exponents associated with the
diffusion mechanism, and t (s) is the isothermal holding time. For
each dominant diffusion mechanism during solid sintering, Eq. (2)
varies primarily in the time exponent n, grain size exponent m, and
form ofH (T). Taking the logarithm of both sides, Eq. (2) can be
written as:
H (T)
Gm
L
L0
ln
= n ln t + n ln
(3)
During an isothermal hold, H (T) can be regarded as constant.
Assuming negligible grain growth (increase < 20%), n can be ob-
tained by fitting the slope of the function ln( L/L0)versus lnt.
Specifically, diffusion mechanisms can be classified as either grain
boundary diffusion or lattice diffusion when n = 0.25–0.35 or
n = 0.4–0.5 [35,51–53], respectively. Hence, the dominant diffusion
mechanism was determined herein by isothermal sintering at var-
ious temperatures inside the intermediate sintering stage.
Fig. 5 presents the relationship between the grain size and re-
lative density of the TiH and Ti samples, which were obtained by
measuring the two types of bulk samples prepared by heating the
4. Discussion
The results described above clearly demonstrate that sintered
TiH2 powder displays a higher densification rate, greater final re-
lative density, and higher mechanical properties than those of sin-
tered HDH-Ti powder. To quantitatively compare the shrinkage
behavior of the two types of powders, the diffusion mechanism and
sintering activation energy associated with densification were de-
termined, and the effects of dehydrogenation on the densification
mechanism were discussed as follows.
4.1. Diffusion mechanism
Powder shrinkage and densification during powder sintering are
usually divided into three overlapping stages according to the re-
lative density [49], i.e., the initial (< 66%), intermediate (66–92%),
and final stages (> 92%). The relative density change in the initial
stage was limited to 2–3% [7], and the most evident shrinkage and
densification occurred during the intermediate stage. The maximum
densification difference between the TiH and Ti samples in this work
arose during the intermediate stage (Fig. 3). As for the mass trans-
port responsible for the sintering densification mechanism in this
stage, there are two diffusion mechanisms to consider: lattice dif-
fusion and grain boundary diffusion.
In general, linear shrinkage during powder sintering follows a
power-law relationship versus the isothermal holding time t. The
relationship can be used to fit a large part of the sintering process,
including the initial and intermediate stages, when grain growth is
limited [50]. The corresponding equation can be expressed
as [31,51]:
Fig. 5. The relationship between the grain size and relative density of the TiH and Ti
samples after heating green compacts to 1050, 1100, 1150, and 1250 °C and holding for
0, 1, 2, and 4 h at 10 °C/min. The insets illustrate the microstructure evolution of the
TiH samples with relative densities of 86.8% and 92.4%.
5

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Table 3
Relative density and corresponding grain size of TiH and Ti compacts heated to 1050, 1100, 1150, and 1250 °C and held for 0, 1, 2, and 4 h at 10 °C/min, respectively.
Sample
TiH
Sintering parameter
1050 °C for 0 h
1100 °C for 0 h
1150 °C for 0 h
1250 °C for 0 h
1250 °C for 1 h
1250 °C for 2 h
1250 °C for 4 h
ρ (%)
86.8
29.2
86.9
28.1
89.9
31.4
88.0
29.2
92.4
35.4
89.6
34.4
95.8
47.6
93.0
39.6
98.5
87.0
96.7
96.5
98.8
93.6
97.0
99.2
189.1
97.4
G (μm)
ρ (%)
Ti
G (μm)
106.6
239.2
two types of green compacts to various temperatures of 1050, 1100,
1150, and 1250 °C and holding for 0, 1, 2, and 4 h at 10 °C/min. The
obtained values are listed in Table 3. As an example, the insets show
that the grain size increased and the pore volume decreased as the
relative density increased. In other words, the grain sizes of 29.2 µm
and 35.4 µm correspond to relative densities of 86.8% and 92.4% for
the TiH samples. Furthermore, the grain size increased quickly to
189 µm at 99.2% relative density. Similarly, the grain growth in the Ti
samples followed the same trend. Sample TiH has a 60 µm smaller
grain size than that of the Ti sample fabricated under the same
sintering parameters, i.e., heating to 1250 °C and holding for 4 h at
10 °C/min. The pores, especially those located along grain bound-
aries, reduced the contact area of adjacent grains and, thus, impeded
atom diffusion and exchange across grain boundaries, thereby
slowing grain growth in the sintered compacts. As a result, grain
growth is usually restrained in the low relative density stage, spe-
cifically in the intermediate stage of powder sintering [54]. Based on
this notion, the onset of grain growth was delayed at a relative
density of approximately 90%. Also, the experimental results in Fig. 5
demonstrate that grain size in both samples remained almost un-
changed until the relative density exceeded 91%. Rapid grain growth
in the TiH samples occurred when the relative density exceeded 96%,
3% higher than the Ti samples. This indicates a finer grain size
(189 µm) in the nearly full-density TiH sample compared with that
(239 µm) in the Ti sample (Table 3 and Fig. 5), which supports the
idea that applying the TiH2 powder as a starting material produces
bulk Ti and Ti alloys with finer-grained structures and better me-
chanical properties.
associated with the diffusion mechanisms were calculated and de-
creased from 0.33 at 950 °C to 0.25 at 1050 °C and then further de-
creased to 0.13 at 1150 °C for the TiH samples, and from 0.34 at 950 °C
to 0.27 at 1100 °C and then further to 0.16 at 1150 °C for the Ti
samples (Fig. 6), respectively. When n = 0.33–0.25, corresponding to
grain boundary diffusion [35,51], it can be deduced that powder
densification at a relative density below 91% is controlled by grain
boundary diffusion in the TiH samples from 950 to 1050 °C and in the
Ti samples from 950 to 1100 °C, respectively. Herein is the first report
of a specific sintering densification mechanism for TiH compacts
obtained from TiH2 powder. The deduced grain boundary diffusion
mechanism in this paper is similar to that when sintering 45 µm
Ti–6Al–4V powder up to 1100 °C and 90% relative density [34], sin-
tering 20 µm titanium powder up to 1150 °C [55], sintering 40 µm
copper powder [56] and 15 µm stainless steel powder below 90%
relative density [57], et al. In addition, the grain boundary diffusion
in the TiH samples occurs at a lower sintering temperature than the
Ti samples, making it possible to achieve nearly full-density at a
lower sintering temperature.
4.2. Sintering activation energy
Powder shrinkage densification in the intermediate sintering
stage can be theoretically expressed as [7]:
0.5
d
dt
Db Vm
1
P
v
b
= 854
RG4T
(4)
Fig. 6 displays the relationship between the logarithm of the
sintering time and relative shrinkage for the TiH and Ti samples.
Since grain growth is negligible at a relative density below 91%
(Fig. 5), the relative shrinkage data were selected from 2 h iso-
thermal sintering curves when heated to 950, 1000, 1050, 1100, and
1150 °C at 15 °C/min. The values did not exceed the critical value of
91% and were fitted according to Eq. (3). As such, the n values
where Db (m2/s) is the grain boundary diffusion coefficient, b (nm) is
the diffusion distance of the grain boundary, which is equal to the
grain boundary thickness, γ is the surface energy (J/m2), Vm(m3/mol)
is the molar volume of sintered compacts, R (8.314 J/K mol) is the gas
constant, T (K) is the absolute temperature, and
P
v
is the volume
fraction of pores.
Fig. 6. The relationship between the logarithm of the sintering time and relative shrinkage for the TiH (a) and Ti (b) samples by isothermal sintering for 2 h at 950, 1000, 1050,
1100, and 1150 °C (heating rate: 15 °C/min).
6

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
the determined values of Q are shown in Fig. 7. Q for the Ti samples
are 142.4, 138.2, and 133.5 kJ/mol at heating rates of 2, 5, and 10 °C/
min, respectively. These results accord with those (113–158 kJ/mol)
reported for β-Ti grain boundary diffusion [55,59]. Surprisingly, the Q
values for the TiH samples are lower than the corresponding values
of the Ti samples, which are 71.3, 67.6, 62.7, and 56.4 kJ/mol at dif-
ferent heating rates, respectively. This indicates that the Ti samples
have a slower atomic diffusion rate than the TiH samples in the in-
termediate sintering stage before rapid grain growth. For instance, at
a heating rate of 10 °C/min, the grain boundary diffusion coefficient
of TiH and Ti samples were calculated using Eq. (5) as and
DTi = D0 exp( 133.5/RT). It has been reported that if sufficient
densification is achieved before pore-boundary separation, sintered
compacts can have a relatively high final density and relatively fine
grains [7,49]. This is confirmed in Fig. 5, in which the TiH samples
reached 96% density before pore-boundary separation, thereby
producing a higher density (99.2%) and finer grain (189 µm) com-
pared with the Ti samples. In addition, the lower Q of the TiH sam-
ples explains why the TiH2 powder can also be sintered at a lower
temperature (1100 °C or even lower) while still fabricating nearly
full-density bulk materials (>98%) with comparable tensile proper-
ties (Fig. 4).
Fig. 7. Plots of ln(TP0.5(d /dt)) vs. 1/T of the TiH2 and HDH-Ti compacts at various
heating rates. The inset lists the sintering activation energies of grain boundary dif-
fusion for the TiH and Ti samples.
Simultaneously, the grain boundary diffusion coefficient Db is de-
pendent on the temperature and follows an Arrhenius relationship [58]:
4.3. Effect of dehydrogenation on the densification mechanism
Q
RT
Herein, the physical properties affecting powder shrinkage and
densification, such as the morphology and average particle size of
initial TiH2 and Ti powders (Fig. 1), and the fabrication parameters,
including the heating rate, were held constant. The only phase in
both Ti and TiH samples in the intermediate sintering stage (T > α↔β
transition temperature) was BCC β-Ti; therefore, the accelerated
densification behavior of the TiH samples at the intermediate sin-
tering stage was mainly attributed to dehydrogenation of the TiH2
precursor from 320 to 860 °C (T < α↔β transition temperature).
TiH2 has an FCC lattice structure with H atoms located in tetra-
hedral voids and H atoms that can be easily released from the Ti lattice
at elevated temperatures in a vacuum. There are two main effects
accounting for the accelerated densification of the TiH2 precursor
depending on the dehydrogenation mechanism (Fig. 8). One is a
self-purification effect through water (H2O) generation in which re-
leased H reacts with O on the surface of TiH2 powder, which reduces
Db = D0 exp
(5)
where D0(m2/s) is the pre-exponential factor related to the frequency of
atomic jumps, and Q (kJ/mol) is the activation energy for grain boundary
diffusion. Subsequently, applying and rearranging Eqs. (4) and (5) and
taking the logarithm of both sides, the powder shrinkage behavior can
be written as:
d
D0 Vm
Q
RT
b
ln TP0.5
= ln 854
dt
RG4
(6)
When grain growth is negligible, ln(854D
as a constant, and the activation energy Q is determined as the slope
of the linear fit of ln(TP0.5(d /dt)) versus 1/T. The selected data used
for linear fitting abide by two rules: located inside the intermediate
sintering stage and belonging to grain boundary diffusion. Finally,
) can be regarded
V
4
m
0
b
RG
Fig. 8. Schematic diagram of the two types of dehydrogenation effects in the TiH2 compact: (I) self-purification and (II) defects generation.
7

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
the oxide layer thickness [25,42]. Since O can act as an obstacle to
atomic diffusion, Ti powder with a purified surface is beneficial for
powder shrinkage and densification [33,60]. Another is defects gen-
eration, in which extra vacancies and dislocations are introduced by
large volumetric changes due to dehydrogenation during the trans-
formation from FCC-TiH2 to HCP-Ti. Therefore, the surface energy in-
creases, and atomic diffusion accelerates [36,37,45]. To date, the self-
purification effect has been theoretically supported by ΔG for the re-
action between H atoms and oxides (Ti2O3, TiO2, and TiO) [38], and
experimentally by the water generation and weight loss during dehy-
drogenation [32] and the oxygen content decrease of 0.05 wt% after
sintering a TiH2 compact [42]. However, few investigations have tes-
tified the defects generation on dehydrogenation, which is crucial for
fabricating nearly full-density Ti bulk materials.
The FCC-TiH2 to HCP-Ti phase transformation is multi-step process,
in which the TiH2 (a = 0.444 nm) phase transforms into metastable FCT-
TiH (a = 0.420 nm, c = 0.470 nm) and finally to HCP-Ti (a = 0.295 nm,
c = 0.468 nm) [61]. Fig. 8 illustrates the starting and ending states of the
lattice. The total volume change by determining the volume per atom
(V) is calculated as below. The volume per atom (V) was obtained by
dividing the volume of a single lattice cell by the nominal number of
atoms in the corresponding lattice (4 atoms for an FCC lattice and 6 for
ended at 387 °C, exhibiting an area of 580.2 J/mol during cooling
(Fig. 9b). It is well established that lattice defects raise the Gibbs free
energy of metallic materials, and stored energy accumulated by lattice
defects can be released via exothermic events at the corresponding
recovery temperature (0.1–0.5 Tm) by the relaxation and recovery of
defects [31]. This is corroborated by the fact that milling resulted in an
exothermic enthalpy of 0.82 0.15 kJ/mol and 0.50 kJ/mol for Fe [62]
and Ti [31] powders, respectively. Since the exothermic enthalpy of
580.2 J/mol (Fig. 9b) is almost equivalent to the values due to lattice
defects in milled Fe and Ti powders, the exothermic peaks in DSC
curves (Fig. 9b) likely correspond to the release of stored energy from
lattice defects. The initial stored defects, such as disorder, dislocations,
and vacancies from the powder preparation and pressing, recovered
during the first heating stage, and dropped to a relatively low con-
centration. Hence, the recovered defects during the cooling stage al-
most certainly resulted from dehydrogenation via the FCC-TiH2→HCP-
Ti transformation. This is consistent with the aforementioned conclu-
sion reached by analyzing the lattice volume strain (Fig. 8) and mac-
roscopic shrinkage (Fig. 2).
Furthermore, a two-step sintering experiment was designed and
applied to investigate the contribution and limitation of defects gen-
eration effects on powder densification. In step I, the TiH2 compact was
heated to 870 °C, which is exactly between the end temperature of
dehydrogenation and the onset temperature of powder densification. It
was then cooled to room temperature under a high vacuum. In this
case, the released hydrogen reacted with the oxide layer, which pur-
ified the powder surface (Fig. 8) because O contamination from the
surroundings was negligible due to high-vacuum sintering; thus, a self-
purification effect appeared, which promoted powder densification
during step I. Regarding the defects generation effect, the extra defects
introduced from dehydrogenation recovered to a normal concentration
when cooling from 870 °C to room temperature, as shown in Fig. 9,
which means that the defects generation effect was eliminated during
the sintering in step I.
HCP). The calculated values of FCC-TiH2 (Va) and HCP-Ti (Vb) are
2.19 × 10−29 m3 and 1.76 × 10−29 m3, respectively. Therefore, the volume
change for the FCC↔HCP transition calculated using the for-
mula(V
Vb)/V was 19.6% (Fig. 8). In addition, the total volume
a
a
change from FCC to HCP calculated by the deformation gradient tensors
of the lattice [61] was 19.4%, which is consistent with the above results.
The details of the calculations are shown in Appendix A. Considering
the porosity and hypothesizing that the defect concentration remained
unchanged during dehydrogenation, the macroscopic volume
shrinkage after dehydrogenation of the TiH2 compact was calculated by
ρ0 × 19.6% and was about 16.5%. Nevertheless, the shrinkage for the
TiH2 compacts measured by dilatometry (Fig. 2) was 14.2%, 13.9%, 13.6%,
and 13.5% at various heating rates of 2, 5, 10, and 15 °C/min, respec-
tively. The shrinkage difference of 2.3 – 3% might be attributed to an
increase in the number of vacancies or dislocation concentration during
dehydrogenation.
Fig. 9 shows the DSC curves of TiH2 powder heated to 870 °C,
subsequently cooled to room temperature, and reheated to 870 °C. The
endothermic area during heating was calculated to be 98.5 kJ/mol
(Fig. 9a), which is ascribed to the phase transformation from FCC-TiH2
to HCP-Ti induced by the dehydrogenation of the TiH2 powder. It can be
seen that an exothermic event with multiple peaks began at 791 °C and
In step II, the sample obtained in step I was re-heated to a sintering
temperature of 1250 °C, which produced a densified sintered compact
(named TS-Ti). The TS-Ti sample obtained by two-step sintering dis-
played only a self-purification effect, which accelerated powder den-
sification compared with the TiH compacts; thus, it is possible to
determine the contribution of the defects generation effects on powder
densification by comparing the density of the TS-Ti and TiH samples.
Specifically, if the density of the TS-Ti sample is lower than that of the
TiH sample and higher than that of the Ti sample, defects generation
significantly contributed to powder densification.
Based on the designed two-step sintering experiment, the TiH2
compact was heated to 870 °C at 10 °C/min, subsequently cooled to
50 °C in a sintering furnace under 10−4 Pa vacuum, and then reheated
to 1250 °C at 10 °C/min, with a dwell time of 4 h. The resulting
sample was called TS-Ti. For comparison, the TiH2 and HDH-Ti
compacts were heated to 1250 °C at 10 °C/min with a dwell time of
4 h under 10−4 Pa vacuum, which were respectively named TiH and Ti
(consistent with the contexts). Fig. 10 shows an SEM image of the
microstructure of the sintered Ti, TS-Ti, and TiH samples. The Ti and
TiH samples displayed the maximum and minimum proportions of
residual pores, respectively, which suggests that the densification
degree of the TS-Ti sample is superior to the Ti sample and inferior to
the TiH sample. Moreover, the densification degree of the three types
of samples was confirmed by their relative densities, which were
97.0%, 98.2%, and 99.0% for samples Ti, TS-Ti, and TiH, respectively.
Since only a self-purification effect was preserved to accelerate
densification in the TS-Ti sample, the defects generation effect in the
TiH sample promoted the powder densification process. Again, these
results prove that dehydrogenation introduced extra defects into the
sintered samples.
Unlike in pressure-assisted sintering, no extrinsic pressure is
applied in pressureless sintering. The extrinsic pressure shortens the
Fig. 9. DSC scans of TiH2 powder during heating to 870 °C (a), subsequent cooling to
room temperature (b), and reheating to 870 °C (c), respectively.
8

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
Fig. 10. SEM images of the microstructures of the sintered Ti (a), TS-Ti (b), and TiH (c) samples.
distance and increases the contact area between powder particles,
causing powder particles or sintered bulk materials to yield and
creep at higher sintering temperatures, which is crucial for elim-
inating pores [30]. Usually, pressureless-sintered bulk Ti can not
achieve nearly full-density, whose relative density is lower than 98%.
Fortunately, applying TiH2 as a starting material overcomes this
problem. During industrial production, HDH-Ti powder is prepared
by the dehydrogenation of TiH2 powder. It seems that dehy-
drogenation may introduce a high concentration of crystal defects;
however, the need for an unavoidable cooling stage can lead to the
recovery of crystal defects (Fig. 9b); therefore, nearly no extra de-
fects are persevered after cooling to ambient temperature. This was
verified by the lack of an exothermic peak in the reheating curve in
Fig. 9c. Despite the reduction in the O content due to self-purifica-
tion after dehydrogenation, the HDH-Ti powder again absorbed
considerable O contaminants during storage and pressing, resulting
in the formation of a thick oxide layer on the powder surface. The
above two aspects outline the limitation of the dehydrogenation
effects. Consequently, applying TiH2 powder as the sintering pre-
cursor can effectively circumvent limitations due to dehydrogena-
tion effects, and is beneficial for accelerating powder densification
and thus fabricating nearly full-density bulk Ti and Ti alloys with
high mechanical properties.
materials obtained by various other processing approaches, especially
sintered bulk Ti from HDH-Ti powder. Essentially, the excellent me-
chanical properties reported herein originated from accelerated densi-
fication and a decreased grain size, which were induced by 3% extra
defects generated via dehydrogenation during sintering of the TiH2
compact. This work reports the first quantification of the dehy-
drogenation effect and resultant densification mechanism during the
pressureless sintering of TiH2 powder, which can help produce Ti and Ti
alloys with better mechanical properties for use as structural materials.
CRediT authorship contribution statement
T. Chen: Methodology, Investigation, Visualization, Writing. C.
Yang: Writing-review & editing, Supervision, Funding acquisition,
Project administration. Z. Liu: Investigation, Formal analysis. H.W.
Ma: Formal analysis. L.M. Kang: Discussion. Z. Wang: Language
polish. W.W. Zhang: Discussion. D.D. Li: Formal analysis. N. Li:
Discussion. Y.Y. Li: Supervision, Discussion.
Declaration of Competing Interest
The authors declare that they have no known competing fi-
nancial interests or personal relationships that could have appeared
to influence the work reported in this paper.
5. Conclusions
The dehydrogenation and resultant densification mechanism were
revealed by a designed two-step sintering strategy that included the
pressureless sintering of a TiH2 powder precursor. Compared with the
HDH-Ti compact, the TiH2 compact has the same sintering diffusion
mechanism of grain boundary diffusion, a 53% lower sintering activation
energy, and a 3% greater onset relative density of rapid grain growth.
This led to a 2.3% higher relative density and 60 µm smaller grain size in
the TiH2 compact than the HDH-Ti compact. Interestingly, the bulk Ti
sintered from the TiH2 compact exhibited an ultimate strength of
659 MPa and 20.8% elongation, which were far superior to bulk Ti
Acknowledgments
This work was supported financially by the National Key Research
and Development Program of China (Nos. 2020YFB2008300 and
2020YFB2008304), the National Natural Science Foundation of
China (No. U19A2085), the Key Basic and Applied Research Program of
Guangdong Province (No. 2019B030302010), the Key-Area
Research and Development Program of Guangdong Province (No.
2020B090923001), and the University Scientific Research Project of
Guangzhou Education Bureau (No. 202032783).
Appendix A
The deformation gradient tensors of FCC-TiH2 to HCP-Ti are calculated by method of Woo [61]. The two most common orientation re-
¯
¯
¯ ¯
¯
lationships of FCC TiH2→HCP Ti are expressed as OR1: {111} //{0001} <110> // <1210> and OR2: {001} //{0001} <110> // <1210>
.
As for OR1, the phase transformation is accomplished by the passage of a/6 < 112>partial dislocations on the {111} plane [63,64]. The
coordinate systems for this case is shown as follows:
¯
¯ ¯
[2110]
x1
x2
x3
:
:
:
[110]
2 /2a
6 /2a
2/ 3 a
a
.
¯ ¯
[112]
[111]
¯
[0110]
3 a
.
.
[0001]
c
In dehydrogenation process, The FCC TiH2 phase is first transformed to metastable FCT γ-hydride (TiH, a = 0.420 nm, c = 0.470 nm) followed
¯ ¯
<112>
by dilating to the FCC lattice with a = 0.417 nm, and then shearing 2 /4 along
to HCP Ti generation.
direction to change of stacking sequence ABC/AB, leading
The first step in calculation is to dilate the FCC lattice to FCT, the matrix for the step, referring to coordinate system index is:
M1 = A 1T A
(A1)
1
9

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
1
1
6
1
6
1
3
1
3
1
3
2
1
2
where A =
(2).
2
0
6
0
0
0
1
0
and T =
(3).
1
1
0
0
2
with ξ1 = 0.420/0.444, ξ2 = 0.470/0.444.
The FCT lattice is further dilated to FCC lattice with a = 0.417 nm, where the deformation matrix is:
M2 = A 1T2A
(A4)
r
0
0
0
r
1
0
0
r2
1
where T2 =
(5).
0
with r1 = 0.417/0.420, r2 = 0.417/0.470.
¯ ¯
The FCC to HCP transformation was achieved by the passage of a/6 < 112> partial dislocations on the {111} plane, which is shearing 2 /4
¯ ¯
<112>
along
direction followed by dilating the (c/a) ratio for Ti from 1.633 (the ideal value for HCP) to 1.588. The matrix for this step is:
1 0
0
1 0
0
0
M3 =
0
0
1
0
0
1
0
2 /4
1
0 1.588/1.633
(A6)
(A7)
Combining these three steps, the deformation gradient tensor of OR1 FCC TiH2→HCP Ti transformation is:
B = M1M2M3
Which is calculated as:
0.9392
0
0
B =
0
0
0.9392 0.3320
0.9133
0
(A8)
(A9)
The change of the transformation from FCC to HCP is obtained by the determinant of the strain matrix:
VHCP
VFCC
= B = 0.806
Therefore, the volume change from FCC to HCP transition with OR1 is calculated by the formula (VFCC VHCP)/VFCC was about 19.4%.
As for OR2, the phase transformation is accomplished via pure-shuffle mechanism. The stacking sequence for FCC and HCP lattice
along <001> and <0001> direction is AB/AB, Thus, only atomic shuffle are considered in calculation. The coordinate systems selected in the
current work are shown as follows:
¯
¯ ¯
[2110]
x1
x2
:
:
[110]
2 /2a
2 a
a
.
3 a
.
¯
[110]
[0110]
x3 : [001]
[0001]
a
c .
The first step is FCC lattice transformed to metastable FCT lattice with atomic shuffle, in which the matrix can be expressed as:
s1
0
0
0
s1
0
0
0
s2
P1 =
(A10)
with s1 = 0.420/0.444, and s2 = 0.470/0.444.
The lattice matrix transformed from FCT to HCP in the second step is:
t1
0 t1
0
0
0
2
0
0
P2 =
0
6 /2 0
0
0 t2
0
0
1
(A11)
(A12)
with t1 = 0.295/0.420, and t2 = 0.468/0.470.
Combining these two steps, the deformation gradient tensor of OR2 FCC-TiH2→HCP-Ti transformation is:
C = P1P2
which is calculated as:
0.9396
0
0.8137
0
0
0
C =
0
0
1.0541
(A13)
10

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
The change of the transformation from FCC to HCP is obtained by the determinant of the strain matrix:
VHCP
= C = 0.806
VFCC
Therefore, the volume change from FCC to HCP transition with OR2 is calculated by the formula (VFCC VHCP)/VFCC was about 19.4%.
(A14)
[26] H.T. Wang, M. Lefler, Z.Z. Fang, T. Lei, S.M. Fang, J.M. Zhang, Q. Zhao, Titanium and
titanium alloy via sintering of TiH2, Key Eng. Mater. 436 (2010) 157–163.
[27] P. Kumar, K.S.R. Chandran, Enhancement of fatigue resistance using the ac-
celerated diffusion/sintering phenomenon near beta transus temperature in
Ti–6Al–4V powder metallurgy alloy, Scr. Mater. 165 (2019) 1–5.
[28] F. Cao, K.S.R. Chandran, P. Kumar, New approach to achieve high strength
powder metallurgy Ti–6Al–4V alloy through accelerated sintering at β-transus
temperature and hydrogenation-dehydrogenation treatment, Scr. Mater. 130
(2017) 22–26.
[29] P. Sun, Z.Z. Fang, M. Koopman, Y. Xia, J. Paramore, K.S. Ravi Chandran, Y. Ren, J. Lu,
Phase transformations and formation of ultra-fine microstructure during hy-
drogen sintering and phase transformation (HSPT) processing of Ti–6Al–4V,
Metall. Mater. Trans. A 46 (2015) 5546–5560.
[30] X.X. Li, C. Yang, H.Z. Lu, X. Luo, Y.Y. Li, O.M. Ivasishin, Correlation between atomic
diffusivity and densification mechanism during spark plasma sintering of tita-
nium alloy powders, J. Alloys Comp. 787 (2019) 112–122.
[31] C. Yang, M.D. Zhu, X. Luo, L.H. Liu, W.W. Zhang, Y. Long, Z.Y. Xiao, Z.Q. Fu,
L.C. Zhang, E.J. Lavernia, Influence of powder properties on densification me-
chanism during spark plasma sintering, Scr. Mater. 139 (2017) 96–99.
[32] L.B. Mei, C.M. Wang, Y.H. Wei, S.F. Xiao, Y.G. Chen, Effects of hydrogen content on
powder metallurgy characteristic of titanium hydrides, Int. J. Hydrog. Energy 43
(2018) 7102–7107.
[33] D.Y. Li, H. He, J. Lou, Y.M. Li, Z.Y. He, Y.Z. Chen, F.H. Luo, Effect of oxygen contents
on predominant sintering mechanism during initial stage of pure titanium
powder, Powder Technol. 361 (2020) 617–623.
[34] X.Y. Xu, P. Nash, Sintering mechanisms of Armstrong prealloyed Ti–6Al–4V
powders, Mater. Sci. Eng. A 607 (2014) 409–416.
[35] T.S. Zhang, L.B. Kong, X.C. Song, Z.H. Du, W.Q. Xu, S. Li, Densification behaviour
and sintering mechanisms of Cu- or Co-doped SnO2: a comparative study, Acta
Mater. 62 (2014) 81–88.
[36] D.H. Savvakin, M.M. Humenyak, M.V. Matviichuk, O.H. Molyar, Role of hydrogen
in the process of sintering of titanium powders, Mater. Sci. 47 (2012) 651–661.
[37] V. Bhosle, E.G. Baburaj, M. Miranova, K. Salama, Dehydrogenation of nanocrys-
talline TiH2 and consequent consolidation to form dense Ti, Metall. Mater. Trans.
A 34 (2003) 2793–2799.
[38] C.M. Wang, L. Pan, Y.N. Zhang, S.F. Xiao, Y.G. Chen, Deoxidization mechanism of
hydrogen in TiH2 dehydrogenation process, Int. J. Hydrog. Energy 41 (2016)
14836–14841.
[39] Y.G. Zhang, C.M. Wang, Y. Liu, S.P. Liu, S.F. Xiao, Y.G. Chen, Surface character-
izations of TiH2 powders before and after dehydrogenation, Appl. Surf. Sci. 410
(2017) 177–185.
[40] V. Duz, M. Matviychuk, A. Klevtsov, V. Moxson, Industrial application of titanium
hydride powder, Met. Powder Rep. 72 (2017) 30–38.
[41] C.M. Wang, Y.A. Zhang, S.F. Xiao, Y.G. Chen, Sintering densification of titanium
hydride powders, Mater. Manuf. Process. 32 (2017) 517–522.
[42] O.M. Ivasishin, D.G. Savvakin, M.M. Gumenyak, O.B. Bondarchuk, Role of surface
contamination in titanium PM, Key Eng. Mater. 520 (2012) 121–132.
[43] B. Sarma, K.S. Ravi Chandran, Accelerated kinetics of surface hardening by dif-
fusion near phase transition temperature: mechanism of growth of boride layers
on titanium, Acta Mater. 59 (2011) 4216–4228.
[44] K. Matsui, Sintering kinetics at constant rates of heating: mechanism of silica-
enhanced sintering of fine zirconia powder, J. Am. Ceram. Soc. 91 (2008)
2534–2539.
[45] G. Chen, K.D. Liss, G. Auchterlonie, H.P. Tang, P. Cao, Dehydrogenation and sin-
tering of TiH2: an in situ study, Metall. Mater. Trans. A 48 (2017) 2949–2959.
[46] X. Lu, Y. Pan, W. Li, M.D. Hayat, F. Yang, H. Singh, W. Song, X. Qu, Y. Xu, P. Cao,
High-performance Ti composites reinforced with in-situ TiC derived from pyr-
olysis of polycarbosilane, Mater. Sci. Eng. A 795 (2020) 139924.
[47] S.D. Luo, B. Liu, J. Tian, M. Qian, Sintering of titanium in argon and vacuum: pore
evolution and mechanical properties, Int. J. Refract. Met. Hard Mater. 90 (2020)
105226.
References
[1] E. Alabort, D. Putman, R.C. Reed, Superplasticity in Ti–6Al–4V: characterisation,
modelling and applications, Acta Mater. 95 (2015) 428–442.
[2] A. Devaraj, V.V. Joshi, A. Srivastava, S. Manandhar, V. Moxson, V.A. Duz,
C. Lavender, A low-cost hierarchical nanostructured beta-titanium alloy with
high strength, Nat. Commun. 7 (2016) 1–8.
[3] Z.Z. Fang, J.D. Paramore, P. Sun, K.S.R. Chandran, Y. Zhang, Y. Xia, F. Cao,
M. Koopman, M. Free, Powder metallurgy of titanium–past, present, and future,
Int. Mater. Rev. 63 (2018) 407–459.
[4] J.D. Paramore, Z.Z. Fang, P. Sun, M. Koopman, K.S.R. Chandran, M. Dunstan, A
powder metallurgy method for manufacturing Ti–6Al–4V with wrought-like
microstructures and mechanical properties via hydrogen sintering and phase
transformation (HSPT), Scr. Mater. 107 (2015) 103–106.
[5] H.R. Zhang, H.Z. Niu, M.C. Zang, J.K. Yue, D.L. Zhang, Microstructures and me-
chanical behavior of a near α titanium alloy prepared by TiH2-based powder
metallurgy, Mater. Sci. Eng. A 770 (2020) 138570.
[6] L.M. Kang, C. Yang, A review on high-strength titanium alloys: microstructure,
strengthening, and properties, Adv. Eng. Mater. 21 (2019) 1801359.
[7] S.-J.L. Kang, Sintering: Densification, Grain Growth, and Microstructure, Elsevier
Butterworth-Heinemann, Oxford, 2005.
[8] Y. Yamamoto, J.O. Kiggans, M.B. Clark, S.D. Nunn, A.S. Sabau, W.H. Peter,
Consolidation process in near net shape manufacturing of Armstrong CP-Ti/Ti-
6Al-4V powders, Key Eng. Mater. 436 (2010) 103–111.
[9] S. Raynova, M.A. Imam, F. Yang, L. Bolzoni, Hybrid microwave sintering of
blended elemental Ti alloys, J. Manuf. Process 39 (2019) 52–57.
[10] L.M. Kang, Y.J. Cai, X.C. Luo, Z.J. Li, X.B. Liu, Z. Wang, Y.Y. Li, C. Yang, Bimorphic
microstructure in Ti–6Al–4V alloy manipulated by spark plasma sintering and
in-situ press forging, Scr. Mater. 193 (2021) 43–48.
[11] C. Yang, L.M. Kang, X.X. Li, W.W. Zhang, D.T. Zhang, Z.Q. Fu, Y.Y. Li, L.C. Zhang,
E.J. Lavernia, Bimodal titanium alloys with ultrafine lamellar eutectic structure
fabricated by semi-solid sintering, Acta Mater. 132 (2017) 491–502.
[12] J. Kozlík, H. Becker, J. Stráský, P. Harcuba, M. Janeček, Manufacturing of fine-
grained titanium by cryogenic milling and spark plasma sintering, Mater. Sci.
Eng. A 772 (2020) 138783.
[13] F.M. Zhang, T.F. Liu, Nanodiamonds reinforced titanium matrix nanocomposites
with network architecture, Compos. B Eng. 165 (2019) 143–154.
[14] Y.F. Yang, M. Qian, Titanium Powder Metallurgy!Spark plasma sintering and
hot pressing of titanium and titanium alloys, Titan. Powder Metall. (2015)
219–235.
[15] C. Cai, B. Song, P. Xue, Q.S. Wei, C. Yan, Y.S. Shi, A novel near α-Ti alloy prepared
by hot isostatic pressing: microstructure evolution mechanism and high tem-
perature tensile properties, Mater. Des. 106 (2016) 371–379.
[16] R. Guo, L. Xu, Z.Y. Chen, Q.J. Wang, B.Y. Zong, R. Yang, Effect of powder surface
state on microstructure and tensile properties of a novel near α-Ti alloy using hot
isostatic pressing, Mater. Sci. Eng. A 706 (2017) 57–63.
[17] J.B. Jia, K.F. Zhang, S.S. Jiang, Microstructure and mechanical properties of
Ti–22Al–25Nb alloy fabricated by vacuum hot pressing sintering, Mater. Sci. Eng.
A 616 (2014) 93–98.
[18] J.H. Shon, J.M. Park, K.S. Cho, J.K. Hong, N.K. Park, M.H. Oh, Effects of various
sintering methods on microstructure and mechanical properties of CP-Ti powder
consolidations, Trans. Nonferrous Met. Soc. China 24 (2014) s59–s67.
[19] C. Yang, Y.J. Zhao, L.M. Kang, D.D. Li, W.W. Zhang, L.C. Zhang, High-strength si-
licon brass manufactured by selective laser melting, Mater. Lett. 210 (2018)
169–172.
[20] J.Q. Zhang, Y.G. Liu, M. Bayat, Q. Tan, Y. Yin, Z.Q. Fan, S.Y. Liu, J.H. Hattel,
M. Dargusch, M.X. Zhang, Achieving high ductility in a selectively laser melted
commercial pure-titanium via in-situ grain refinement, Scr. Mater. 191 (2021)
155–160.
[21] D.W. Wang, Y.H. Zhou, J. Shen, Y. Liu, D.F. Li, Q. Zhou, G. Sha, P. Xu, T. Ebel, M. Yan,
Selective laser melting under the reactive atmosphere: a convenient and effi-
cient approach to fabricate ultrahigh strength commercially pure titanium
without sacrificing ductility, Mater. Sci. Eng. A 762 (2019) 138078.
[22] R. Kulagin, A. Mazilkin, Y. Beygelzimer, D. Savvakin, I. Zverkova, D. Oryshych,
H. Hahn, Influence of High Pressure Torsion on structure and properties of
Zr–Ti–Nb alloy synthesized from TiH2, ZrH2 and Nb powders, Mater. Lett. 233
(2018) 31–34.
[48] O. Ertorer, T. Topping, Y. Li, W. Moss, E.J. Lavernia, Enhanced tensile strength and
high ductility in cryomilled commercially pure titanium, Scr. Mater. 60 (2009)
586–589.
[49] R. German, Sintering: From Empirical Observations to Scientific Principles,
Elsevier Butterworth-Heinemann, Oxford, 2014.
[50] M.N. Rahaman, Sintering of Ceramics, CRC Press, Boca Raton, 2007.
[51] S. Majumdar, S. Raveendra, I. Samajdar, P. Bhargava, I.G. Sharma, Densification
and grain growth during isothermal sintering of Mo and mechanically alloyed
Mo-TZM, Acta Mater. 57 (2009) 4158–4168.
[52] R.L. Coble, Initial sintering of alumina and hematite, J. Am. Ceram. Soc. 41 (1958)
55–62.
[53] D.L. Johnson, T.M. Glarke, Grain boundary and volume diffusion in the sintering
of silver, Acta Metall. 12 (1964) 1173–1179.
[54] J. Wang, R. Raj, Estimate of the activation energies for boundary diffusion from
rate–controlled sintering of pure Alumina, and Alumina doped with Zirconia or
Titania, J. Am. Ceram. Soc. 73 (1990) 1172–1175.
[23] Y.F. Luo, Y.H. Xie, L. Cao, J.M. Liang, Y. Liu, D.L. Zhang, Harmonically structured
Ti–6Al–4V alloy with ultrahigh strength fabricated by thermomechanical con-
solidation of mechanically milled TiH2/Al60V40 powder blend, Mater. Sci. Eng. A
787 (2020) 139493.
[24] Y.F. Zheng, X. Yao, Y.J. Su, D.L. Zhang, High strength titanium with a bimodal
microstructure fabricated by thermomechanical consolidation of a nanocrys-
talline TiH2 powder, Mater. Sci. Eng. A 686 (2017) 11–18.
[25] O. Ivasishin, V. Moxson, Titanium Powder Metallurgy!Low-cost titanium hydride
powder metallurgy, Titan. Powder Metall. (2015) 117–148.
11

T. Chen, C. Yang, Z. Liu et al.
Journal of Alloys and Compounds 873 (2021) 159792
[55] R.M. German, Titanium sintering science: a review of atomic events during
densification, Int. J. Refract. Met. Hard Mater. 89 (2020) 105214.
[56] F.B. Swinkels, M.F. Ashby, A second report on sintering diagrams, Acta Metall. 29
(1981) 259–281.
[60] M.X. Wu, Y. Yang, G. Yang, K.L. Huang, D.Q. Yin, Direct evidence for surface cleaning
mechanism during field-activated sintering, J. Alloys Comp. 784 (2019) 975–979.
[61] O.T. Woo, G.C. Weatherly, C.E. Coleman, R.W. Gilbert, The precipitation of γ-
deuterides (hydrides) in titanium, Acta Metall. 33 (1985) 1897–1906.
[57] S.J.L. Kang, Y. Il Jung, Sintering kinetics at final stage sintering: model calculation
and map construction, Acta Mater. 52 (2004) 4573–4578.
[58] X.X. Li, C. Yang, T. Chen, Z.Q. Fu, Y.Y. Li, O.M. Ivasishin, E.J. Lavernia,
Determination of atomic diffusion coefficient via isochronal spark plasma sin-
tering, Scr. Mater. 151 (2018) 47–52.
[62] Y.H. Zhao, H.W. Sheng, K. Lu, Microstructure evolution and thermal properties in
nanocrystalline Fe during mechanical attrition, Acta Mater. 49 (2001) 365–375.
[63] Y.N. Liu, H. Yang, Y. Liu, B.H. Jiang, J. Ding, R. Woodward, Thermally induced fcc ↔
hcp martensitic transformation in Co–Ni, Acta Mater. 53 (2005) 3625–3634.
[64] H.C. Wu, A. Kumar, J. Wang, X.F. Bi, C.N. Tomé, Z. Zhang, S.X. Mao, Rolling-in-
duced face centered cubic titanium in hexagonal close packed titanium at room
temperature, Sci. Rep. 6 (2016) 24370.
[59] B.B. Panigrahi, M.M. Godkhindi, K. Das, P.G. Mukunda, P. Ramakrishnan, Sintering
kinetics of micrometric titanium powder, Mater. Sci. Eng. A 396 (2005) 255–262.
12