The effects of hydrogen desorption on the properties of titanium hydride powder and nanocomposites containing corresponding metals and thermally expanded graphite were investigated. As can be seen from Fig. 1, which shows a typical SEM image of compacted nanocomposites based on hydrogenated titanium with 0.53 wt. % TEG before and after vacuum annealing at ~ 650°C for 1.5 hours, the particles of initially hydrogenated titanium have fewer microcracks and a finer structure compared to dehydrogenated particles. The thermal treatment also results in partial TEG destruction on the composite surface.
The study of the specific electrical conductivity σ of hydrogenated titanium with different grain sizes showed that for particles not separated (≤ 200 µm) and separated (200 − 100, and ≤ 100 µm) by size, the values of σ are equal to 2.98, 3.1, and 0.47 (Ohm‧cm)−1, respectively. To prepare a nanocomposite with TEG, the hydrogenated titanium with a particle size ≤ 200 µm was taken. The process of hydrogen desorption from TiHx shows that x = 2.74 wt.% (or 1.32 at. %) and leads to the appearance of dehydrogenated titanium powder (TidH), which was also used with TEG for the production of nanocomposites. A study of the electrophysical properties of both TiHx+TEG and TidH + TEG nanocomposites and their comparative analysis were carried out.
The hydrogen desorption process for the studied composite samples starts at a temperature close to 430°C, which is confirmed by the DTA curve in Fig. 2 (see Exp. 1, point 1). The peak of the endothermic effect of the dehydrogenation reaction is observed at 572°C (see Exp. 1, point 2) and a loss of 2.98% of the sample mass is observed. Some large mass losses in the composite containing 0.53 wt. % TEG compared to pure titanium hydride (2.74 wt. %) are explained by partial TEG destruction at high temperatures and different gas desorption processes in TEG. The latter is confirmed by the additional two small peaks at 321°C and 415°C on the DTA curve during the first heating of the composite sample. The second heating (see Exp. 2 in Fig. 2) of the sample does not show any reactions in the composite, i.e. the dehydrogenation process ends during the first heating (about 1.3 hours of heating up to 800°C).
The dependence of the specific electrical conductivity, σ, of the hydrogenated titanium before and after the hydrogen desorption on the powder density, ρ, for 4 cycles of the resistivity measurement is shown in Fig. 3. The initial powder of TiHx is substantially compacted after the first compression (Fig. 3a), and its electrical conductivity increases from the value of 2.44∙10− 3 (Ohm‧cm)−1 at ρ = 2.20 g/cm3 to the maximum of 2.98 (Ohm‧cm)−1 at ρ = 2.86 g/cm3. The diapason of ρ corresponding to stress relaxation during powder unloading is very narrow. The subsequent loading-unloading cycles leave the maximum density of the powder almost unchanged (close to 2.86 g/cm3), but lead to some reduction in conductivity (up to 0.79 (Ohm‧cm)−1) for these values of ρmax. For the first load-unload cycle of the TidH sample (Fig. 3b), the electrical conductivity has the initial value of 1.07 (Ohm‧cm)−1 at the density ρ0 = 1.94 g/cm3. The next compression of the sample is accompanied by an increase of σ, which reaches the maximum value σmax = 21.8 (Ohm‧cm)−1 at ρσmax = 2.47 g/cm3. The value of electrical conductivity at maximum possible (hardware limited) compression (2.85 g/cm3) is σρmax = 16.15 (Ohm‧cm)−1. During the unloading process, the inverse part of the dependence σ(ρ) is observed in a small range of the density, indicating the almost absent elastic component in the titanium powder. For subsequent compression cycles, the density of the transition to the conductive state increases by 12% and the maximum value of the electrical conductivity drops by 30% for the 4th loading-unloading cycle, which is also observed at the maximum compression of the studied samples. During the cyclic loading and unloading processes, the samples are ordered, i.e. the particles are packed and form a conductive system with the most effective contacts between them.
The maximum value of the electrical conductivity of titanium after hydrogen desorption exceeds 7.3 times the corresponding value for the initial powder of hydrogenated titanium (Figs. 3 and 4). This is due to the reduction of the oxide shells of the particles and changes in the structure of the metal particles during hydrogen desorption.
Analogously, we measured series of σ(ρ) dependences for different types of composites based on TiHx and TidH with the addition of different concentrations c of TEG and we analyzed their characteristic parameters (Table 1). The mixtures of metal and TEG powders have different electrical and mechanical properties compared to the original components and their averaged properties, i.e. nanocomposites are formed.
Table 1
Values of electromechanical parameters for TiHx–TEG and TidH–TEG nanocomposites: σmax — the maximum value of the specific electrical conductivity; σρmax — the value of the specific electrical conductivity at the maximum possible compression; ρ0 — the density at which the electrical conductivity is fixed (transition to the conductive state); ε — the coefficient of relative mechanical deformation.
Samples
|
TiHx + с wt. % TEG
|
TidH + с wt. % TEG
|
TEG
|
c, wt. % TEG
|
0
|
0.53
|
5
|
20
|
50
|
85
|
0
|
0.53
|
5
|
19
|
50
|
85
|
100
|
σmax, (Ohm‧cm)−1
|
2.98
|
4.2
|
5.98
|
5.12
|
4.86
|
5.58
|
21.8
|
26.47
|
18.21
|
3.68
|
1.59
|
1.91
|
≅1.0
|
σρmax, (Ohm‧cm)−1
|
2.98
|
3.14
|
0.79
|
1.56
|
1.8
|
0.9
|
16.15
|
26.47
|
14.48
|
3.57
|
0.7
|
0.2
|
≅0.9
|
ρ0, g/cm3
|
2.2
|
0.82
|
0.925
|
0.08
|
0.06
|
0.07
|
1.91
|
1.81
|
0.67
|
0.06
|
0.05
|
0.03
|
0.19
|
ε, %
|
9.11
|
5.62
|
9.0
|
34.78
|
44
|
9.72
|
4.0
|
10
|
7.3
|
41.8
|
62.7
|
64.5
|
27.94
|
As can be seen from Fig. 4 (and partially from Table 1), the addition of TEG to the hydrogenated titanium powder leads to the increase of the maximum values of the specific electrical conductivity σmax for all studied TEG concentrations c in the nanocomposites: the maximum values of the electrical conductivity are increased ≅ 2 and ≅ 6 times with respect to pure TiHx and pure TEG, respectively. The effect of TEG content on conductivity has the character of a weakly modified function for all TEG concentrations, except for c < 5 wt. % TEG and c ≈ 100 wt. % TEG.
When nanocomposites contain titanium particles after the process of hydrogen desorption, the character of concentration dependence of electrical conductivity is significantly changed (Fig. 4). The highest value of electrical conductivity (σmax) is observed for the nanocomposite TidH + 0.53 wt. % TEG (note, this TEG concentration is preferable for thermionic converter), which is ≅ 1.2 times larger than the electrical conductivity of pure dehydrogenated Ti and ≅ 4.4 times larger than the highest value of electrical conductivity of original (before hydrogen desorption) powder composite.[8] In contrast to the original TiHx composites, which show an increased value of σmax for almost all TEG concentrations, the composites containing TidH show a decrease in electrical conductivity for c > 0.53 wt. % TEG. For carbon concentrations higher than 20 wt. %, the conductivity values of Ti-based composites tend to a limit of pure TEG (especially TidH-based composites).
The difference between Figs. 4a and 4b is that the first of them shows the maximum values of electrical conductivity (σmax), and the second shows the values of electrical conductivity (σρmax) for the maximum achievable compression (ρmax) of the powder material in this experiment (in both cases, data are taken from resistometric studies). At low concentrations of TEG, titanium powder acts as a matrix, and when c > 5 wt. % TEG, the role of the matrix is taken over by TEG (0.53 wt. % TEG correspond to 33 vol. % TEG). Therefore, it is precisely in the area of small c values the mechanisms that determine the nature of the concentration dependence of the electrical conductivity of a heterogeneous system change.
In the case of a composite based on TiHx for values of c ≤ 5 wt. % TEG can be seen (Fig. 4b) that at the maximum possible compression, when a conductive frame made of titanium particles is still formed, a decrease in the electrical conductivity of the composite is observed. This effect demonstrates the transfer of charge carriers from metal particles to TEG.[2, 8, 17–19]
In addition, a decrease in the metal component of the electron concentration (due to charge flow on the TEG) should be accompanied by a certain shift of the Fermi level in the energy spectrum of the electrons. Based on calculations of the density of electronic states (DOS) of titanium and its hydrides,[20] in the Ti and TiHx systems (x ≈ 2), the Fermi level falls on sharp and narrow peaks on the DOS, so even a small decrease in electron concentration n should lead to a noticeable decrease in the value of the density of the electronic states at the Fermi level, and hence the electrical conductivity of the metal component and the composite as a whole (we assume here that the DOS itself does not change significantly with small changes in n).
Under the condition of incomplete compaction and small values of c, an increase in the electrical conductivity of the composite is observed (Fig. 4a), due both to the increase in the electron concentration in its carbon component and to the presence of a significant number of TEG interlayers in contact between the metal particles. The latter provides an increased contribution of TEG to the conductivity of the system due to its higher volume concentration ϑ1 compared to the state of maximum compression (Eq. (1)). At maximum compaction and small values of c (as long as the matrix material is TiHx), a decrease in the conductivity of the composite is observed (Fig. 4b), which actually indicates a decrease in the value of the density of electronic states at the Fermi level due to the shift of the latter from the maximum at the DOS of TiHx particles due to a decrease in n. A further increase in the content of TEG in the Ti hydride composite leads to an increase in the conductivity values at saturation due to the action of two opposing factors: i) a decrease in electrical conductivity due to a decrease in the concentration of free charge carriers in TEG with an increase in its volume fraction in the composite (a charge that redistributes the flow of metal particles over a larger volume of TEG); ii) compaction of TEG, which leads to an increase in the number of contacts between graphene flakes of TEG and an increase in electrical conductivity.
The second factor acts against the background of another effect. So, from Fig. 4b, it can be seen that when the system undergoes maximum compression at high concentrations of TEG (c > 50 wt. % TEG), then the compaction of the carbon component leads to a certain compensation for the increase in the number of contacts between TEG scales due to the increase in their deformation, which causes a decrease in the electrical conductivity of TEG. At the same time, the reduction factor n begins to prevail when c is directed to 100 wt. %, and the electrical conductivity of both the TEG and the composite as a whole decrease slightly. The deformation of the scales of TEG can also be explained by the mesh values of electrical conductivity in Fig. 4b compared to Fig. 4, and although the factor of growth in the number of contacts continues to operate.
The dehydrogenated titanium (TdH) based composite behaves similarly to the TiHx based composite at high TEG concentrations (c > 20 wt. % TEG). The main differences in the nature of the concentration dependence of these two types of composites are observed at low concentrations of TEG. As already mentioned, after hydrogen desorption, the electrical conductivity of pure titanium is significantly higher than that of titanium hydride with a thick oxide shell on the particles. At the same time, the addition of a relatively small amount of TEG (0.53 wt. % TEG), as in the case of ТіНх, leads to a significant increase in the maximum value of the electrical conductivity (Fig. 4a), due to the flow of part of the free charge carriers from the metal to the TEG. However, due to the somewhat larger width of the peak at the DOS of pure (dehydrogenated) Ti,[20] the Fermi level does not immediately fall on the sharp slope of the DOS, and therefore the values of the density of electronic states at the Fermi level and the electrical conductivity of Ti at the smallest values of c (small n) change little. This results in an increase in the electrical conductivity of the composite under maximum compression (Fig. 4b) at the initial stage of adding TEG to the metal matrix TidH, mainly due to a significant change in the conductivity of the carbon component due to an increase in the number of contacts and the concentration of free charge carriers.
As the amount of TEG in the composite increases, more and more metal particles end up as inclusions isolated in the carbon matrix. When they are all wrapped in TEG (Fig. 6), then the further flow of free charge carriers between the metal and carbon components of the composite stops. At the same time, the maximum value of the electron concentration in the TEG and the maximum shift of the Fermi level in the electronic energy spectrum of the metal is achieved due to a decrease in n in it. Under these conditions, the concentration dependences of electrical conductivity (Fig. 4) of the TidH-based composite are observed at c > 0.53 wt. % TEG a sharp decrease in electrical conductivity to values close to those for pure TEG. Therefore, the increase in electron concentration in TEG dominates only at the smallest values of c, and under the condition 5 < c < 20 wt. % TEG begins to be dominated by a decrease in electrical conductivity Ti due to a decrease in the density of electronic states at the Fermi level. At an even higher content of TEG (due to the dominant contribution of TEG), the two-factor mechanism of the formation of the saturation shelf on the dependence begins to operate, which for the TidH composite is the same as that described above for the TiHx composite.
As the amount of TEG in the composite increases, more and more metal particles become isolated as inclusions in the carbon matrix. When they are all encapsulated in TEG (Fig. 6), the further flow of free charge carriers between the metal and carbon components of the composite stops. At the same time, the maximum value of the electron concentration in the TEG and the maximum shift of the Fermi level in the electronic energy spectrum of the metal are reached due to a decrease of n in the TEG. Under these conditions, the concentration dependence of the electrical conductivity (Fig. 4) of the TidH-based composite at c > 0.53 wt. % TEG shows a sharp decrease in electrical conductivity to values close to those of pure TEG. Therefore, the increase in electron concentration in TEG dominates only at the smallest values of c, and at the condition 5 < c < 20 wt. % TEG, the decrease in electrical conductivity Ti begins to be dominated by a decrease in the density of electronic states at the Fermi level. At an even higher content of TEG (due to the dominant contribution of TEG), the two-factor mechanism of the formation of the saturation shelf on the dependence σ(c) begins to operate, which for the TidH composite is the same as that described above for the TiHx composite.
All the effects mentioned above are related to the heterogeneous nature of the studied systems, in particular, the effect of the formation of saturation shelves on the concentration dependence of the electrical conductivity of composites based on TiHx and TidH at c > 20 wt. % TEG, allow not only the above qualitative explanations, but also a quantitative description based on the expression for the specific electrical conductivity of the matrix heterostructure:[21]
$$\sigma ={\sigma _0}\left( {1+\frac{{{\vartheta _1}}}{{{\vartheta _0}/3+{\sigma _0}/({\sigma _1} - {\sigma _0})}}} \right)$$
1
where ϑ0 = 1 − ϑ1 — volume fraction of the matrix, ϑ1 — volume concentration of inclusions, σ — specific electrical conductivity of the system, the index 0 denotes the properties of the matrix material, the index 1 — material properties of inclusions.
Expression (1) can be used to explain the setting of the shelf height values on the concentration dependence of the electrical conductivity at the level of the TEG electrical conductivity values, especially in the case of maximum densification of the composites (Fig. 4b). Note that for c ≤ 20 wt%, the TEG charge flow to the carbon component of the TidH-based composite leads to a less rapid decrease in the electrical conductivity of the dehydrated Ti particles compared to the initially hydrogenated particles, due to differences in the DOS of these metals around the Fermi level. However, already at c > 20 wt. % TEG, the electrical conductivity values of both hydrogenated and dehydrogenated Ti with reduced values of n become close to the corresponding value for pure TEG with additional free carriers (i.e., the values of σ0 and σ1 become close). This leads to the fulfillment of the condition σ0/(σ1 – σ0) > > 1 in expression (1) and the second term in the brackets goes to 0. In the case of TiHx, this condition is fulfilled at smaller values of c, when the matrix material is still TiHx and the value σ0 characterizes its conductivity, which, according to Fig. 4b quickly becomes comparable to the conductivity of the densified TEG with increasing c. At large values of c, when the matrix material becomes TEG, the electrical conductivity of composites based on both TiHx and TidH, at their maximum densification, approaches the value σ0 which characterizes the conductivity of the carbon matrix.
The difference in the height of the saturation shelves on the concentration dependence of the electrical conductivity of TiHx and TidH based composites (Fig. 4) can be explained by the different size of the titanium particles before and after hydrogen removal from them (Fig. 5). Since the average size (128.8 µm) of the titanium hydride particles is smaller than the corresponding size (157.1 µm) of the titanium particles after desorption (due to the sintering of small particles with large ones during the desorption heating), a slightly higher degree of compression of the carbon component of the composite is possible in the case of a composite based on TiHx. This leads to the dominance of the contribution of the compaction effect of the TEG scales to the electrical conductivity, both at smaller and at maximum deformations, which is reflected in higher values of the electrical conductivity in the region of the saturation shelf for the TiHx-based composite compared to the TidH-based composite.
A comparison of additional electromechanical parameters of the nanocomposites before and after hydrogen desorption is shown in Table 1. It can be seen that the density ρ0, at which the transition to the conductive state is fixed, decreases with increasing TEG content for all investigated nanocomposites. On the contrary, the coefficient of relative mechanical deformation ε increases under the same conditions. Both results are due to the growth of the TEG concentration in the composites and the peculiarity of the TEG structural changes, as well as the changes in the state of the metal-carbon interfaces during the compression of the composites. For both types of nanocomposites under compression, the elastic processes and many electrical conductivity effects are mainly determined by the ability of TEG particles to significantly compress. In fact, for TEG concentrations higher than 5 wt. %, the powder samples become solids (Fig. 6) even after small compression.