ISSN 0020-1685, Inorganic Materials, 2016, Vol. 52, No. 7, pp. 669–676. © Pleiades Publishing, Ltd., 2016. Original Russian Text © E.I. Istomina, P.V. Istomin, A.V. Nadutkin, V.E. Grass, 2016, published in Neorganicheskie Materialy, 2016, Vol. 52, No. 7, pp. 726–733.
Effect of Chemical Modification Conditions on the Sintering Behavior of TiC Powders E. I. Istomina, P. V. Istomin, A. V. Nadutkin, and V. E. Grass Institute of Chemistry, Komi Scientific Center, Ural Branch, Russian Academy of Sciences, Pervomaiskaya ul. 48, Syktyvkar, 167982 Komi Republic, Russia e-mail:
[email protected] Received July 17, 2015; in final form, December 9, 2015
Abstract—Dense ceramics have been produced from a chemically modified titanium carbide powder. Chemical modification was carried out by siliciding titanium carbide powder in a gaseous SiO atmosphere at 1350°C. This treatment produced a Ti3SiC2 layer (up to 19 wt %) on the surface of the TiC particles. Hot pressing at a temperature of 1600°C and pressures from 10 to 20 MPa ensured effective densification of the modified powders. The density of the resultant material reaches 4.8 g/cm3, with a residual porosity under 2%. Its bending strength and fracture toughness are 330 ± 50 MPa and 6.2 ± 0.6 MPa m1/2, respectively. Keywords: titanium carbide, silicidation, hot pressing, ceramic materials DOI: 10.1134/S0020168516070074
INTRODUCTION Titanium carbide (TiC) is a refractory carbide. Its melting point exceeds 3000°C and it has very high hardness and wear resistance, which makes it promising for use under extremely high thermal and mechanical loads [1]. In particular, TiC-based materials are widely used in the manufacture of cutting tools. It is known that TiC powders are difficult to sinter to a pore-free state. Because of this, to produce dense TiCbased ceramics use is typically made of hot-pressing processes at temperatures higher than 1800°C and pressures above 30 MPa [2–4]. The use of sintering aids makes it possible to some extent to produce titanium carbide ceramics under milder conditions [5–7]. A potentially attractive approach for resolving this issue is to utilize chemically modified TiC powders after high-temperature silicidation in a gaseous SiO atmosphere. As shown earlier [8, 9], the initial stage of silicidation under such conditions is Ti3SiC2 growth on the surface of TiC particles according to the reaction scheme 3TiC + SiO(g) = Ti3SiC2 + CO(g). The Ti3SiC2 phase is known to have a nanolaminate structure [10] and, owing to this, it exhibits plasticity at temperatures above 1100°C. Moreover, it is structurally and chemically compatible with TiC. It is reasonable to expect that the ability to produce a Ti3SiC2 layer on the surface of each particle of TiC powder will make it possible to resolve two technolog-
ically important issues pertaining to the sintering behavior of TiC powders. First, this will ensure a uniform sintering aid distribution throughout the material. Second, this will significantly improve the thermomechanical performance of the TiC powder and will ensure effective densification and sintering of the powder in the course of hot pressing. In this context, the purpose of this work was to investigate how the chemical modification of TiC powders by silicidation in a gaseous SiO atmosphere influences the dynamics of their densification in the course of hot pressing. EXPERIMENTAL In our preparations, we used reagent-grade titanium carbide powder with an average particle size of 2–5 μm. A gaseous SiO atmosphere was generated using a granulated equimolar Si + SiO2 powder mixture as a reactive source. According to experimental data and thermodynamic calculations, this reaction mixture is an effective SiO vapor source, with negligible amounts of other species in the gas phase. The SiO partial pressure generated by this source depends on temperature and is about 400 Pa at 1350°C [8, 9, 11]. The TiC powder was silicided in a gaseous SiO atmosphere in an SShVE-1.2.5/25-I2 electric vacuum furnace at a temperature of 1350°C under dynamic vacuum in the chamber. To ensure effective silicidation and protect the furnace equipment from undesirable SiO vapor condensation at temperatures below
669
670
ISTOMINA et al.
SiO
4
SiO SiO
5 SiO
SiO SiO
SiO
6
SiO SiO
7
1 8
2 3
Fig. 1. Schematic of the reactor used for titanium carbide powder silicidation: (1) chamber of the vacuum furnace, (2) heaters, (3) gas-conducting channel, (4) gas-permeable graphite membrane, (5) BAU-A activated carbon absorber, (6) glassy carbon crucibles, (7) titanium carbide powder, (8) granulated mixture of silicon dioxide and silicon powders.
1200°C, we used a cascade silicidation process (Fig. 1). The process was run in a multisection reactor consisting of nested glassy carbon crucibles, which were connected to each other by gas-conducting channels (Fig. 1). The gaps between the crucibles were sealed with graphite foil. The reactive SiO vapor source was placed into the bottom section of the reactor, and the intermediate sections were filled with the TiC powder. The thickness of the TiC powder layer did not exceed 10 mm in order not to hinder the penetration of SiO into the bulk of the material to be silicided. Driven by the SiO partial pressure gradient between the sections, the silicidation agent moved from the
lower sections of the reactor to the upper sections through the gas-conducting channels. The unreacted SiO was captured via chemical binding with an absorber substance loaded into the top section of the reactor. The absorber substance used was BAU-A activated carbon ranging in particle size from 1.6 to 2.5 mm. The reactor was loaded with 6 g of the reactive SiO vapor source, 12 g of the TiC powder in each intermediate section, and 3 g of activated carbon. The heat-treatment conditions of the TiC powders are summarized in Table 1. Uniaxial hot pressing of the silicided and nonsilicided TiC powders was carried out in graphite press INORGANIC MATERIALS
Vol. 52
No. 7
2016
EFFECT OF CHEMICAL MODIFICATION
671
Table 1. Heat-treatment conditions of the TiC powders and the Si/Ti atomic ratio in the silicided samples Powder
Temperature, °С
TC 1 TC 2 TC 3
1500 1350 1350
Heat-treatment time, h 1 10 10
Additional information
MSi = Si/Ti
Vacuum heat treatment without SiO High SiO concentration Low SiO concentration
0.004 0.158 0.111
Table 2. Hot-pressing conditions of the TiC powders and the density of the resultant ceramics Sample HP 1 HP 2 HP 3 HP 4
Starting TiC powder Temperature, °С TC 1 TC 2 TC 3 TC 3
Pressure, MPa
Time, h
20 10 10 20
1 2 1 1
1600 1600 1600 1600
dies at a temperature of 1600°C and pressures from 10 to 20 MPa. The hot-pressing process comprised loading, heating at a rate of 2000°C/h to a predetermined temperature, isothermal holding for 1–2 h, cooling, and unloading. The hot-pressing conditions for each sample are summarized in Table 2. To monitor the linear shrinkage of the samples during the hot-pressing process, we measured the displacement of the movable die. The dynamic powder densification curves of the powders were determined as the difference between the movable die displacement curves obtained in experiments with and without the samples. The phase composition of the samples was determined by X-ray powder diffraction on a Shimadzu XRD 6000 diffractometer (Ni-filtered CuKα radiation). The quantitative content of crystalline phases in the samples was estimated by the Rietveld method using the PowderCell 2.4 program [12]. The elemental composition of the powders was determined by X-ray fluorescence analysis on a Horiba MESA 500 energy dispersive X-ray spectrometer. From the elemental analysis data, we evaluated the Si/Ti atomic ratio (MSi). The silicided powder samples and hot-pressed samples were examined by scanning electron microscopy (SEM) on a TESCAN VEGA3 SBU in backscattered electron (BSE) imaging mode. The local elemental composition of the samples was determined using an X-ACT (EDS) energy-dispersive X-ray microanalyzer combined with the electron microscope. The fracture toughness (KIc) and bending strength (σ) of the hot-pressed materials were determined by four-point bending tests using 24 × 2 × 2 mm beam specimens with and without a notch, and an IR 5057-50 testing machine. INORGANIC MATERIALS
Vol. 52
No. 7
2016
Density, g/cm3 Relative density 3.84 4.79 4.29 4.66
0.78 1.0 0.87 0.95
RESULTS AND DISCUSSION In this study, the cascade silicidation process was utilized for the first time. Compared to the flow process used previously [8, 9], it enabled larger amounts of materials to be loaded into the reactor and allowed the silicidation rate to be controlled by varying the dimensions of the gas-conducting channels. Table 1 specifies the Si/Ti atomic ratio (MSi) in the TiC samples. These data demonstrate that heat treatment in a gaseous SiO atmosphere does indeed cause silicidation of the TiC powders. The chemical composition of the powders becomes appreciably enriched in silicon, so that the Si/Ti atomic ratio rises from a zero level to values in the range 0.11–0.16 mol/mol, depending on the gaseous atmosphere. X-ray diffraction data indicated that, when the process was run at a high SiO concentration (sample TC 2), a major silicidation product was Ti3SiC2 (Fig. 2b): its content was about 19 vol %. An electron micrograph of this sample shows many well-defined steps (slopes) and growth terraces of a new phase on the surface of the TiC particles (Fig. 3a). This suggests that the formation of Ti3SiC2 crystals follows the layer growth mechanism and may proceed to the point of the formation of individual platelike Ti3SiC2 grains having a nanolaminate structure. According to EDS analysis data, the silicon content of the growth steps of the new phase is 11–12 at %, which corresponds to the percentage of silicon in the Ti3SiC2 phase (Table 3, Fig. 3a, spectra 2, 3). In addition, there are areas darker in contrast on the surface of the grains in sample TC 2. According to EDS X-ray mapping results, the Si content in these areas exceeds 16 at % and there is a small amount of Ti (Fig. 3a; Table 3, spectrum 4). It seems likely that there is a SiC grain in the region of spectrum 4. In the X-ray diffraction pattern of this sample, no well-defined reflections from SiC are
672
ISTOMINA et al.
(a)
32
34
36
38 40 2θ, deg
42
(b)
32
34
36
38
40 2θ, deg
44
46
TiC Ti3SiC2
42
44
46
(c) β-SiC (cubic)
32
34
36
38 40 2θ, deg
42
44
46
Fig. 2. Portions of the X-ray diffraction patterns of the TiC powders before and after silicidation: powders (a) TC 1, (b) TC 2, and (c) TC 3.
detectable, but there is a weak reflection at 2θ = 35.7°, which can be indexed as the 102 peak of the SiC phase. In the X-ray diffraction pattern of sample TC 3, which was obtained by siliciding the TiC powder at low SiO concentration, no Ti3SiC2 phase was detected (Fig. 2c). At the same time, sample TC 3 had a rather large Si/Ti atomic ratio, 0.111 mol/mol, and its microstructure was very similar to that of sample TC 2 and
had characteristic signs of the growth of a new phase in the form of steps and terraces on the surface of the TiC particles (Fig. 3b). According to EDS analysis data, the growth steps contained some amount of silicon (Fig. 3b; Table 3, spectra 5, 6, 8). It seems likely that this corresponds to the formation of an X-ray amorphous silicide layer similar in composition to Ti3SiC2. According to X-ray diffraction data, sample TC 3 contained a small amount of SiC (Fig. 2c). In an electron micrograph of this sample, SiC grains had the form of micron-sized prismatic particles (Fig. 3b, region of spectrum 7) that appeared darker than the major phase. Thus, we are led to conclude that the cascade silicidation of TiC in a gaseous SiO atmosphere leads to the formation of a silicide layer on the surface of the TiC particles and that the layer is uniform in thickness throughout the sample. The presence of a small amount of SiC is due to the fact that, in our experiments, we used carbon materials, such as graphite foil and glassy carbon. In spite of their low reactivity, when in contact with SiO at high temperatures these materials reacted with it to form a thin layer consisting of SiC particles on their surface. Subsequently, as a result of mechanical forces acting on the reactor walls during unloading of the silicided material, the SiC particles were incorporated into the TiC powder. Table 2 summarizes the conditions used to produce ceramics from the silicided and nonsilicided TiC powders and indicates the density of the hot-pressed samples. It is seen that the chemical modification of the TiC powder by silicidation in a gaseous SiO atmosphere does indeed have a significant advantageous effect on its densification and sintering. The relative density of the ceramics obtained by hot-pressing the nonsilicided powders was as low as 0.78. The influence of temperature and pressure was in this case obviously insufficient for complete densification of the material. Figure 4 shows an electron micrograph of the ceramic produced from nonsilicided TiC, in which pores are clearly seen. The porosity of the material is at a level of 20–30%. By contrast, in the case of the ceramics produced from the presilicided TiC powders, the degree of densification was much higher. The relative density of the samples ranged from 0.87 to 1.0, and their porosity was within 13% (see Table 2). Figure 5 shows electron micrographs and EDS analysis data of the ceramics prepared from the silicided TiC powders (samples HP 2 and HP 4). Both samples had a relatively uniform microstructure. According to EDS analysis data, silicon is present essentially throughout the surface of sample HP 2 (Fig. 5a, Si X-ray map). In contrast, the surface of sample HP 4 has silicon-free areas (Fig. 5b, Si X-ray map; Table 3, spectrum 15). Note that, INORGANIC MATERIALS
Vol. 52
No. 7
2016
EFFECT OF CHEMICAL MODIFICATION
673
Table 3. Local elemental composition according to EDS analysis data Elemental composition, at % Spectrum Sample Figure С Si Ti 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
TC 2
3а
TC 3
3b
HP 2
5a
HP 4
5b
59.0 52.8 56.5 69.5 62.1 49.7 50.8 39.6 50.4 48.7 52.2 46.1 53.5 46.8 47.3 51.1
1.0 12.0 10.6 16.5 0.2 0.7 24.8 7.8 0.4 5.3 5.5 23.3 8.4 0.1 0.0 24.2
Si/Ti atomic ratio
40.0 35.2 32.9 14.0 37.7 49.6 24.4 52.6 49.2 46.0 42.3 30.6 38.1 53.1 52.7 24.7
0.03 0.34 0.32 1.18 0.01 0.01 1.02 0.15 0.01 0.12 0.13 0.76 0.22 0.00 0.00 0.98
(b)
(a) 4
7
5 7 8
3 6
2 4
1
5 μm Si Kα1
Ti Kα1
5 μm
5 μm
5 μm Si Kα1
Ti Kα1
5 μm
5 μm
Fig. 3. Surface microstructure (SEM in BSE imaging mode) and EDS element X-ray maps of powders (a) TC 2 and (b) TC 3. INORGANIC MATERIALS
Vol. 52
No. 7
2016
674
ISTOMINA et al.
5 μm Fig. 4. Microstructure (SEM in BSE imaging mode) of the ceramic sample HP 1.
(a)
throughout sample HP 4, its density is relatively high. Also present in samples HP 4 and HP 2 are isometric SiC particles ranging in size from 1 to 3 μm. Figure 6 shows typical densification curves obtained during hot pressing of the TiC powders. In the case of the silicided powder, densification began immediately after the temperature had been raised to above 800°C. In the case of the nonsilicided powder, the densification onset temperature was somewhat higher: about 1150°C. In both instances, the densification process continued throughout subsequent heating of the material, up to the isothermal holding temperature: 1600°C. Note that, in this stage of the process, the densification rate of the silicided powder exceeded that of the nonsilicided powder (Fig. 7). In the isothermal holding step, the densification rate of the silicided TiC powder decreased markedly and the final densification of the sample was reached during the first 30 min of isothermal holding. In the case of the nonsilicided powder, this took just 15 min. The overall linear shrinkage of the silicided powder was 50 to 60%
(b) 13 10
12 14 16
9 11 15
5 μm Si Kα1
Ti Kα1
5 μm
5 μm Si Kα1
5 μm
Ti Kα1
5 μm
5 μm
Fig. 5. Surface microstructure (SEM in BSE imaging mode) and EDS element X-ray maps of samples (a) HP 2 and (b) HP 4. INORGANIC MATERIALS
Vol. 52
No. 7
2016
EFFECT OF CHEMICAL MODIFICATION
h, mm 800°C 1200°C 1600°C 1600°C 8 1 7 6 5 2 4 3 2 Isothermal holding Heating 1 0
10
20
30
40
50
60
70
80 90 100 Time, min
Fig. 6. Densification curves obtained during hot pressing of the TiC powders: (1) silicided powder (sample TC 3), (2) nonsilicided powder (sample TC 1).
higher than that of the nonsilicided powder. Analysis of the above data indicates that, on the whole, the silicided and nonsilicided powders are similar in densification behavior. At the same time, all other factors being equal the presence of a Ti3SiC2 layer (powder TC 2) or an X-ray amorphous silicide layer (powder TC 3) on the surface of the TiC particles allows a substantially higher degree of densification to be reached. The best sintering and densification results among the ceramics produced from the modified TiC powders were obtained for sample HP 2: its residual porosity was under 2%. We determined strength characteristics of this sample. The bending strength and fracture toughness of sample HP 2 were found to be 330 ± 50 MPa and 6.2 ± 0.6 MPa m1/2, respectively. Accord-
ing to data in the literature [13–16], the bending strength of titanium carbide ceramics produced by hot pressing at temperatures from 1550 to 1700°C and pressures from 55 to 70 MPa is 450–560 MPa. The bending strength of sample HP 2 is somewhat lower, which seems to be due to the milder hot-pressing conditions. The fracture toughness of sample HP 2 exceeds that of the TiC ceramic produced without using sintering aids and is comparable to that of materials whose sinterability was improved by introducing nonmetallic additives. For example, in the case of materials prepared by spark plasma sintering (SPS) at 1600°C and 50 MPa, KIc evaluated from microindentation data is 3.7–5.0 MPa m1/2 for pure TiC [17] and 6.3 MPa m1/2 for TiC containing 3.5 wt % WC [5]. The high fracture toughness of sample HP 2 seems to be due to the presence of nanolaminate Ti3SiC2, which is helpful for damage localization at nanoscale structural elements and prevents macroscopic structure damage. CONCLUSIONS A new process for chemical modification of TiC powders has been proposed: silicidation in a gaseous SiO atmosphere. This approach makes it possible to significantly improve the thermomechanical performance of TiC powders and enables sintering to a porefree state at relatively low temperatures and pressures. As a result, dense ceramic materials with excellent strength characteristics can be obtained.
800°C 1200°C 1600°C
0.09 Densification rate, min−1
675
1600°C
0.08 0.07 0.06
1
0.05 0.04 0.03
2
0.02 0.01 0
Heating 10
Isothermal holding 20
30
40
50
60
70
80
90 100 Time, min
Fig. 7. Densification rate curves obtained during hot pressing of the TiC powders: (1) silicided powder (sample TC 3), (2) nonsilicided powder (sample TC 1).
INORGANIC MATERIALS
Vol. 52
No. 7
2016
676
ISTOMINA et al.
ACKNOWLEDGMENTS This work was supported by the Russian Foundation for Basic Research, grant no. 15-08-08472. REFERENCES 1. Kiparisov, S.S., Levinskii, Yu.V., and Petrov, A.P., Karbid titana: poluchenie, svoistva, primenenie (Preparation, Properties, and Applications of Titanium Carbide), Moscow: Metallurgiya, 1987. 2. Stasyuk, L.F. and Neshpor, V.S., Effect of the particle size of titanium carbide on hot compaction under high pressure, Poroshk. Metall. (Kiev), 1987, no. 8, pp. 31–35. 3. Ono, T., Endo, H., and Uedi, M., Hot-pressing of TiC–graphite composite materials, J. Mater. Eng. Perform., 1993, vol. 2, pp. 659–664. 4. Zhoua, M., Rodrigoa, P.D.D., Wanga, X., Hub, J., Dongb, Sh., and Chenga, Y.-B., A novel approach for preparation of dense TiC–SiC nanocomposites by sol– gel infiltration and spark plasma sintering, J. Eur. Ceram. Soc., 2014, vol. 34, pp. 1949–1954. 5. Cheng, L., Xie, Z., and Liu, G., Spark plasma sintering of TiC ceramic with tungsten carbide as a sintering additive, J. Eur. Ceram. Soc., 2013, vol. 33, pp. 2971– 2977. 6. Cho, K.S., Kim, Y.W., Choi, H.J., and Lee, J.G., SiC– TiC and SiC–TiB2 composites densified by liquidphase sintering, J. Mater. Sci., 1996, vol. 35, no. 23, pp. 6223–6228. 7. Ahmoye, D. and Krstic, V.D., Reaction sintering of SiC composites with in situ converted TiO2 to TiC, J. Mater. Sci., 2015, vol. 50, no. 7, pp. 2806–2812. 8. Istomina, E.I., Istomin, P.V., and Nadutkin, A.V., Siliciding of titanium carbides with gaseous SiO, Russ. J. Inorg. Chem., 2012, vol. 57, no. 8, pp. 1058–1063.
9. Istomina, E.I., Istomin, P.V., Nadutkin, A.V., and Grass, V.E., Modification of titanium carbide powders by silicidation with gaseous SiO, Ceram. Trans., 2014, vol. 248, pp. 509–514. 10. Barsoum, M.W., The Mn + 1AXn phases: a new class of solids; thermodynamically stable nanolaminates, Prog. Solid State. Chem., 2000, vol. 28, nos. 1–4, pp. 201– 281. 11. Kazenas, E.K. and Tsvetkov, Yu.V., Isparenie oksidov (Vaporization of Oxides), Moscow: Nauka, 1997. 12. Kraus, W. and Nolze, G., Powder Cell—a program for the representation and manipulation of crystal structures and calculation of the X ray powder patterns, J. Appl. Crystallogr., 1996, vol. 29, pp. 301–303. 13. Andrievskii, A.R. and Spivak, I.I., Prochnost’ tugoplavkikh soedinenii i materialov na ikh osnove: Spravochnoe izdanie (Strength of Refractory Compounds and Related Materials: A Handbook), Chelyabinsk: Metallurgiya, 1989. 14. Svoistva, poluchenie i primenenie tugoplavkikh soedinenii. Spravochnoe izdanie (Properties, Preparation, and Applications of Refractory Compounds: A Handbook), Kosolapova, T.Ya., Ed., Moscow: Metallurgiya, 1986. 15. Miracle, D.B. and Lipsitt, H.A., Mechanical properties of fine-grained substoichiometric titanium carbide, J. Am. Ceram. Soc., 1983, vol. 66, no. 7, pp. 592–597. 16. Das, G., Mazdiyasni, K.S., and Lipsitt, H.A., Mechanical properties of polycrystalline titanium carbide, J. Am. Ceram. Soc., 1982, vol. 65, no. 2, pp. 104–110. 17. Cheng, L., Xie, Z., Liu, G., Liu, W., and Xue, W., Densification and mechanical properties of TiC by SPSeffects of holding time, sintering temperature and pressure condition, J. Eur. Ceram. Soc., 2012, vol. 32, pp. 3399–3406.
Translated by O. Tsarev
INORGANIC MATERIALS
Vol. 52
No. 7
2016