A Novel Method to Significantly Improve the Mechanical Properties of n-Type Bi_(1−x)Sb_x Thermoelectrics Due to Plastic Deformation

Round 1

Reviewer 1 Report

The manuscript by N. Sidorenko et. al. reports an interesting extrusion method that significantly enhance the mechanical properties in n-type Bi-Sb crystalline materials. The experiments were carefully done, and the results look reasonable, which may provide a novel strategy in strengthening the Bi-Sb single crystals or even other thermoelectric materials. Before publication, I have some minor comments below for your considerations.

• I suggest the authors to add more general introductions and related references on why the mechanical properties matter in thermoelectric applications. And what are the state-of-art methods that can enhance the mechanical properties in thermoelectric materials? If including these introductions, the readers could have a basic and general idea on the current progress in this area.

• How will the thermoelectric properties change after the plastic deformation? The enhancement of mechanical properties would be meaningless if the thermoelectric performance was significantly suppressed. I suggest the authors to add some discussion based on this question.

• Are the mechanical properties of Bi-Sb single crystals after extrusion better as compared to the textured polycrystalline Bi-Sb in literatures? Textured polycrystalline materials sometimes show high mechanical strength.

• As many thermoelectric materials in reports are polycrystals by sintering methods, will their mechanical performance be improved using the authors’ method?

Author Response

First  Reviewer

1. I suggest the authors to add more general introductions and related references on why the mechanical properties matter in thermoelectric applications. And what are the state-of-art methods that can enhance the mechanical properties in thermoelectric materials? If including these introductions, the readers could have a basic and general idea on the current progress in this area.

Our comment 1. (hereinafter, the line numbers of the edited copy of the manuscript are indicated.)

As noted in the manuscript (lines 36-40; 53-57), single crystals of n-type Bi–Sb solid solutions are the most effective thermoelectric (TE) materials at temperatures of T ≤ 180 K. However, Bi-Sb single crystals cannot withstand any bending along their trigonal axis, although it is in this direction that the figure of merit is the best.

The following is added to the text of the manuscript (lines 81 to 85):

 However, sintering methods offer the advantage of improving the mechanical strength of alloys, but unfortunately degrade the thermoelectric performance as a result of random orientation. Besides, impurities such as oxygen during power elaboration could be as the hindering factor for the improvement of the figure of merit [2].

2. How will the thermoelectric properties change after the plastic deformation? The enhancement of mechanical properties would be meaningless if the thermoelectric performance was significantly suppressed. I suggest the authors to add some discussion based on this question.

Our comment 2. In fact, TE properties after plastic deformation at extra low temperatures are only slightly lower than that of single crystals: the temperature-dependent dimensionless figure of merit ZT of extruded Bi_0.91Sb_0.09 crystals at 80 K is from 0.42 to 0.36 when Ke changes from 1.2 up to 10, while for single crystals it is 0.44. [9], (Sidorenko, 2020). However, the ultimate bending strength of polycrystalline samples obtained by extrusion of single crystals is 28-70 MPa versus 18 MPa for the starting material (Figs. 7, 8, see also lines 505-510).

3. Are the mechanical properties of Bi-Sb single crystals after extrusion better as compared to the textured polycrystalline Bi-Sb in literatures? Textured polycrystalline materials sometimes show high mechanical strength.

       Our comment 3. As noted above, sintering offers the advantage of significant improving the mechanical strength of TE materials, but unfortunately degrades the thermoelectric performance as a result of random orientation [2 - Lenoir B.; Scherrer H. 2001].

4. As many thermoelectric materials in reports are polycrystals by sintering methods, will their mechanical performance be improved using the authors’ method?

Our comment 4. As noted in our comment 1, sintering methods can include some content of impurities such as oxygen which cannot be removed by a new extrusion process.

Reviewer 2 Report

This paper has shown several interesting results regarding mechanical property improvement in n-type Bi-Sb thermoelectric alloy system. An effort to explain the improvement of mechanical properties in single- and poly-crystalline forms in terms of crystal orientation effect in this material. The subject of the present paper itself would be appropriate for this journal. Though experiments and the related theory development are quite well done, some evidences seem to be needed to assure this result. Following comments should be considered for correction.

1. 1^st paragraphs in “introduction” shall be supported by relevant references.
2. Many Greek letters seemed to be missed such as b (= σ_b?) throughout the manuscript, m (=μm?) in line 73.
3. The sentence in lines 52-55 shall be supported by relevant references.
4. In figure 4, standard XRD pattern of the alloy should have been provided, in order to understand texture development better.
5. How was the texture ratio calculated? The computational model used shall be provided in detail.
6. What makes the extrusion ratio develop differently? It shall be discussed in terms of slip and/or twinning mechanism (with dislocation movement), not by “cleavage”. Fracturing into the powders from a single crystal, the term “cleavage” would be appropriate to explain. However, for the modified hot-extrusion process used in this study, the processing mechanism shall be explained by plastic deformation (such as slip, diffusion, grain boundary sliding and dislocation movement) at high pressure and temperatures. Some cases, dynamic recrystallization during the process can also be associated with the texture development. Please provide the brief mechanism regarding the plastic deformation and texture development in this process.
7. In order to discuss thermoelectric properties and mechanical properties, densities are of important. Densities for specimens should be provided and discussed.
8. It is not clearly understood how the bend strengths are improved. Please provide reasonable explanation with relevant references.
9. It would be better, if TE properties, along with the extreme texture cases at least, were provided.

Author Response

Second reviewer

• 1^stparagraphs in “introduction” shall be supported by relevant references.

Our comment 5 (hereinafter, the line numbers of the edited copy of the manuscript are indicated). 

The 1^st and 2^nd paragraphs (lines 30-40) are supported by relevant references; the Ref. [4] by Goldsmid, 2007 was replaced by Ref. [3] by Yim and Amith, 1972.

• Many Greek letters seemed to be missed such as b (= σ_b?) throughout the manuscript, m (=μm?) in line 73.

Our comment 6. These typos have been corrected. They were not in the original manuscript.

• The sentence in lines 52-55 shall be supported by relevant references.

Our comment 7. This has been done, see line 56.

• In figure 4, standard XRD pattern of the alloy should have been provided, in order to understand texture development better.

Our comment 8.  Figure 4 shows XRD patterns for both the textureless and textured alloy after extrusion and annealing with Miller indices, which are given in the hexagonal setting for the [111] planes. As noted in the text (lines 262 to 265), according to the data presented in Table 2, the pole density P₀₀₃ with Miller indices in multiples of (001) significantly exceed P_hkl  of other observed diffraction lines. This indicates the predominant orientation of the trigonal axes of the crystallites along the extrusion axis or, in other words, the dominance of the <111> texture.

• How was the texture ratio calculated? The computational model used shall be provided in detail.

Our comment 9.The texture ratio (W_<111> content) was calculated by Eq. 3 using the results of magnetometric measurements, see lines 150-184 and  255-274 as well as Tables 1, 2.

• What makes the extrusion ratio develop differently? It shall be discussed in terms of slip and/or twinning mechanism (with dislocation movement), not by “cleavage”. Fracturing into the powders from a single crystal, the term “cleavage” would be appropriate to explain. However, for the modified hot-extrusion process used in this study, the processing mechanism shall be explained by plastic deformation (such as slip, diffusion, grain boundary sliding and dislocation movement) at high pressure and temperatures. Some cases, dynamic recrystallization during the process can also be associated with the texture development. Please provide the brief mechanism regarding the plastic deformation and texture development in this process.

Our comment 10. This remark is not entirely clear. Extrusion ratio, as the ratio of the length of the polycrystal (after extrusion) to the length of the original single crystal, increases with increasing work pressure. The extrusion process used in this study is not hot extrusion, but cold extrusion (300 K). As noted in the discussion (lines 386 to 392), plastic deformation of single crystals develops under all-round initial pressure in the range from 800 to 1100 MPa, which means that  s₁₁=  s₂₂=  s₃₃= −P at the initial state. The effect of the stress configuration is the most significant and characterized by the mean normal stress s_m = 1/3 ( s₁+  s₂+  s₃), which may not be low enough to avoid cracking or fracture during the deformation process of a brittle material.

The following is added to the text of the manuscript (lines 492 to 499):

Plastic deformation at compression by an all-round load can develop both due to sliding and due to twinning depending on the orientation of single crystals, temperature, and other factors. For example, for Bi single crystals deformed at a rate of ~10^-4 s^-1 Steegmuller and Daniel [27] found a slip mode at room temperature for the crystal orientation {11}; (111) and {100} and twinning mode for orientations (011), (101) and (110). In the process of bending of extruded samples, the increased density of dislocations and other structural defects, the development of grain boundaries (GBs) determine a significant resistance to twinning.

• In order to discuss thermoelectric properties and mechanical properties, densities are of important. Densities for specimens should be provided and discussed.

Our comment 11. No pores or cracks were identified on the surface of extruded samples. Their density is ~99% of the theoretical density (lines 291 to 293).

• It is not clearly understood how the bend strengths are improved. Please provide reasonable explanation with relevant references.

Our comment 12.  In the text (lines 193 to 199) is noted that for n-type branches from Bi - Sb crystals, the most dangerous deformation is bending perpendicular to the direction of maximum TE efficiency…….

   The significantly higher bending strength of Bi-Sb extruded alloys in comparison with single crystals is associated with the development of numerous grains with a high boundary surface, as well as structural defects such as dislocations that accumulate at grain boundaries. The boundary surface of extruded samples with an average grain size of less than 0.080 mm (lines 300, 301) is many times larger than that of single crystals, in which only blocks are present, predominantly elongated in the direction of crystal deformation with transverse dimensions from 0.1 to 2.0 mm. A similar substructure in the form of weakly misoriented single-crystal blocks was also observed earlier in Bi-Sb crystals grown by the Czochralski method [5, 9, 12-14](lines 235 to 237).

• It would be better, if TE properties, along with the extreme texture cases at least, were provided.

Our comment 13. In fact, TE properties after plastic deformation at extra low temperatures are only slightly lower than that of single crystals: the temperature-dependent dimensionless figure of merit ZT of extruded Bi_0.91Sb_0.09 crystals at 80 K is from 0.42 to 0.36 when Ke changes from 1.2 up to 10, while for single crystals it is 0.44. [9], (Sidorenko, 2020). However, the ultimate bending strength of polycrystalline samples obtained by extrusion of single crystals is 28-70 MPa versus 18 MPa for the starting material (Figs. 7, 8, see also lines 505-510). 

Round 2

Reviewer 2 Report

It seems to be duly addressed per each queries.