(Ti,Cr)C-Based Cermets with Varied Nicr Binder Content via Elemental SHS for Perspective Cutting Tools

: The e ﬀ ects of granulation of reactive mixtures Ti-Cr-C and Ti-Cr-C-Ni on the combustion temperature and velocity, as well as phase composition and mechanical properties (crushing ability) of combustion products, were studied. Granulation was associated with a 1.5-fold increase in combustion velocity, caused by a nearly 10-fold increase in gas permeability. Secondary reactions between TiC, Cr 7 C 3 , and molten Ni led to the formation of (Ti,Cr)C FCC solid solution and Ni 2.88 Cr 1.12 intermetallics. After the combustion of Ti-Cr-C-Ni mixtures, prolonged ﬂuorescence was registered, suggesting the exothermic nature of secondary phase formation reactions. The introduction of the nickel binder decreased the content of Cr in the solid solution (Ti,Cr)C owing to the formation of the Ni 2.88 Cr 1.12 phase. To prevent the Cr depletion from the carbide solid solution, Ni-20%Cr binder was added to the granulated 80%(Ti + C) / 20%(3Cr + 2C) mixture. Combustion of granulated mixture yielded brittle porous sinter cake, which was easy to crush and mill, whereas the combustion products from the powdered mixtures were more ductile and harder to crush.

The composition and nomenclature of the mixtures used in this work are given in Table 1. Granulation of the powder mixtures was performed in the following fashion. The initial powder mixture was prepared for 4 h in a conventional ball mill (60 rotations per minute, 6 liter volume) with the rotational axis shifted for 45 • in the vertical plane. The ball to powder ratio was 6:1. Steel jars and steel milling medium (8 mm balls). The balls and powders filled 40% of the mill's volume. After the mixing, a 4% solution of polyvinyl butyral in alcohol was added to the mixture and the obtained paste was rubbed through a mesh with the 1.25 mm cell size and rounded on a rotating horizontal surface. The granules were then dried during 10 h and the target fraction of 0.63-1.6 mm was separated using a vibrating screen. The composition of the initial granulated mixture (M1) and binder (B1) was chosen after the works of Vlasov et al. [80], where the solid solution was synthesized by SHS and then subsequently coated by nickel. The rationale behind the other composition choice has arisen during the investigation.
The features of combustion were investigated using a custom device ( Figure 1). The device allows for synthesis in both gas flow and static atmosphere, measurement of gas expenditure and pressure during the combustion, and video recording. Combustion velocity was measured using a video camera with a dark filter. To preclude the geometry creep of the non-combusted part of a specimen during combustion and for overall stabilization of the results, the chamber was washed with Ar before every measurement.
In this work, syntheses were carried out in the argon atmosphere. The microstructure of the combustion products was investigated by scanning electron microscopy (SEM) using the Ultra Plus microscope (Carl Zeiss, Oberkochen, Germany). The phase composition of the combustion products was studied using the DRON-3 diffractometer with the monochromatic CuKα-radiation. Diffraction patterns were recorded in the range 2θ = 20-80 • with a step of 0.2 • and analyzed by «Crystallographica Search Match» software with Powder Diffraction File database (PDF-2, ICDD, USA, Release 2011). Adiabatic combustion temperature (T ad ) was calculated using THERMO software (http://www.ism.ac.ru/thermo/).
In this work, syntheses were carried out in the argon atmosphere. The microstructure of the combustion products was investigated by scanning electron microscopy (SEM) using the Ultra Plus microscope (Carl Zeiss, Oberkochen, Germany). The phase composition of the combustion products was studied using the DRON-3 diffractometer with the monochromatic CuKα-radiation. Diffraction patterns were recorded in the range 2θ = 20-80° with a step of 0.2° and analyzed by «Crystallographica Search Match» software with Powder Diffraction File database (PDF-2, ICDD, USA, Release 2011). Adiabatic combustion temperature (Tad) was calculated using THERMO software (http://www.ism.ac.ru/thermo/). Gas permeability of powdered and granulated green mixture was measured according to Equation (1). Measurements of the argon flow rate were performed at 1 and 2 atm difference between the inlet and outlet of installation ( Figure 1).
where the k is the gas permeability of the medium (Darcy units), Q is the volumetric flow rate of gas (cm 3 /s) at atmospheric pressure, L is the height of the charge (cm), S is the cross-sectional area of the quartz tube (cm 2 ), P1 is the gas pressure at the upper end of the sample (atm), P2 is the gas pressure at the lower end of the sample (atm), and μ is the viscosity of the gas (cP).

Results
To assess the influence of the granulation process on the combustion velocity and phase composition of the products, the aforementioned SHS installation (Figure 1) was employed to investigate the combustion of powdered and granulated mixtures. Gas permeability of powdered and granulated green mixture was measured according to Equation (1). Measurements of the argon flow rate were performed at 1 and 2 atm difference between the inlet and outlet of installation ( Figure 1).
where the k is the gas permeability of the medium (Darcy units), Q is the volumetric flow rate of gas (cm 3 /s) at atmospheric pressure, L is the height of the charge (cm), S is the cross-sectional area of the quartz tube (cm 2 ), P 1 is the gas pressure at the upper end of the sample (atm), P 2 is the gas pressure at the lower end of the sample (atm), and µ is the viscosity of the gas (cP).

Results
To assess the influence of the granulation process on the combustion velocity and phase composition of the products, the aforementioned SHS installation (Figure 1) was employed to investigate the combustion of powdered and granulated mixtures.
The granulated mixtures demonstrated 1.5-fold higher combustion velocity as compared with the powdered counterparts ( Table 2). For M1 composition, combustion velocity was 22 mm/s for granulated mixture and 14 mm/s for the powdered mixture. For M2 composition, velocities were 9 and 6 mm/s for the granulated and powdered mixtures, correspondingly. The combustion of powdered and granulated M5 and M6 mixtures differed considerably. Granulated mixtures combusted in a stable way, whereas powdered mixtures combusted in a pulsating mode. In the simplest case of steady propagation, all of the wave points move at a constant and identical velocity. When the steadiness is upset, the system may undergo planar autooscillations in the front velocity (pulsating combustion), localization of reaction in one or several hot spots that move along the spiral trajectory (spinning waves), and chaotic motion of numerous hot spots (chaotic solid flames) [82]. For purely gas-phase reactions, the reactants are completely intermixed and the measured intrinsic activation energy corresponds to the isolated binary collisions of molecules that lead to a chemical reaction. However, in solid−gas, solid-liquid, and solid−solid reactions, the structural properties have a pronounced effect on the measured kinetic parameters [82,83]. The combustion front is usually separated into the pre-heating zone, combustion zone, after-combustion zone, and secondary structure formation zone. In the pre-heating zone, the reactive mixture receives a flux of heat from the nearby combustion front; the heating rate depends mainly on the combustion temperature and thermal conductivity of the reactive mixture. Ignition is usually achieved when the less refractory component melts. In the case of Ti-C mixtures, a Ti-based melt is formed, and the dissolution of carbon in a Ti melt is the main driver of the combustion process [84][85][86].
The particle size of reactants and the heterogeneity of the system, in general, have a significant influence on the conversion of solid-phase reagents. In a solid reactive mixture, the larger particles create a clear temperature difference between the gas and the surface of the solid reactants, whereas the temperatures of the gas and surface of finer particles are almost identical. The consecutive stages of drying, devolatilization, and combustion of a reactive medium are overlapped for large particles, whereas they are sequential for fine mixtures. This allows for a considerable simplification in the analysis of the combustion of finer reactive mixtures. The propagation of planar combustion waves is usually described in terms of the model of gasless combustion, using the equations of thermal conductivity and chemical kinetics [87]. For example, the recently developed thermochemical model simplifies the combusting process by ignoring the processes occurring on the surface of the combusting particles [88]. In the framework of such models, one could expect that the granulation would decrease the combustion velocity in solid-state reactive mixtures owing to the increased porosity and heterogeneity of the medium. However, our measurements demonstrate a clear increase in the combustion velocity of granulated mixtures.
We suggest that the release of admixture gases during the combustion is a major factor in the combustion front propagation, and the difference of gas permeability between powdered and granulated reactive mixtures is responsible for the difference in combustion velocity. Released gases, filtering through the granulated mixture, convectively heat up and ignite the surface of the reactive granules; the combustion front within individual granules propagates from the surface towards the center. In the powdered mixtures, the convective heating of green mixture by exuding hot gasses is hindered by the formation of viscous melt in the combustion front, which traps the hot gases and prevents the flux of hot gases from the combustion front into the preheating zone. This effect is also responsible for a significant elongation of the samples during combustion [87]. In the granulated mixtures, the melting occurs only within the bounds of individual granules and a network of interconnected open pores is always available, thus allowing for efficient convective heating of reactive granules in the preheating zone. The sources of outgassing are both the residual gaseous impurities in the green powders and the products of decomposition of polyvinyl butyral [89][90][91][92][93].
Crystals 2020, 10, 412 6 of 16 A similar problem was previously addressed in the area of the combustion of solid fuels. Some of the results indeed indicate that the larger fuel particles attain a higher propagation rate. The increase of fuel density was not associated with lower specific reactant conversion rates or significant changes in the structure of the combustion front [94]. To support this hypothesis, gas permeability was measured for the powdered and granulated mixtures of all compositions. The values can be found in Table 2 along with the calculated adiabatic combustion temperatures and measured combustion temperatures and velocities. The granulation was associated with a nearly 10-fold increase in gas permeability (from 2-3 to 25-28 Darcy).
For the composition without the metallic binder (M1, M5), X-Ray diffraction (XRD) of the combustion products did not reveal any differences for the powder and granulated mixtures. The calculated phases at combustion temperature and experimentally defined phase composition are provided in Table 3. The difference in the phase composition of combustion products at T ad calculated by Thermo and investigated by XRD at room temperature (Table 3)   The fluorescence is presumably related to the exothermic secondary phase formation in the aftercombustion zone, which is corroborated by the lack of a prolonged fluorescence and faster cooling of a specimen prepared from the granulated Ti + C mixture, although the combustion temperature for this mixture is higher than for Ti-Cr-C. One should note that the specimens without the metallic binder (M1, M5) featured the presence of TiC in addition to (Ti, Cr)C (Figure 3a), but TiC was absent in the combustion products produced in semi-industrial SHS reactor from M1 mixture (Figure 4a). The fluorescence is presumably related to the exothermic secondary phase formation in the after-combustion zone, which is corroborated by the lack of a prolonged fluorescence and faster cooling of a specimen prepared from the granulated Ti + C mixture, although the combustion temperature for this mixture is higher than for Ti-Cr-C. One should note that the specimens without the metallic binder (M1, M5) featured the presence of TiC in addition to (Ti, Cr)C (Figure 3a), but TiC was absent in the combustion products produced in semi-industrial SHS reactor from M1 mixture (Figure 4a).
The difference in the phase composition of products obtained under different conditions can be explained if we take into account the sequence of the chemical conversion process and the difference in the product cooling rate. The mass of the initial mixture is almost 100 times higher in the semi-industrial SHS reactor as compared with the lab reactor. Therefore, the cooling rate is degrees of magnitude lower, which provides sufficient time for diffusion processes to be completed. In the framework of this mechanism, the presence of TiC in the small-volume combustion products is inevitable, as it is the most refractory compound (Tmelt = 3200 K) among the intermediate combustion products. These samples cool down faster than the formation of the (Ti,Cr)C phase is concluded.
Let us discuss the influence of metallic binder content on the composition of synthesis products produced from the granulated mixtures. XRD of combustion products of the M1 mixture revealed the formation of the (Ti,Cr)C solid solution with a minor amount of TiC (Figure 3a). An addition of 20 w.% nickel binder (mixture M2) alters the phase composition ( Figure 3b) and leads to the formation of chromium carbide and nickel-chromium intermetallic. Zhang et al. [77] showed that peaks of the main phase (Ti,Cr)C shift towards the lower angles, signifying the decrease of chromium content in (Ti,Cr)C. This shift of (Ti,Cr)C peak is demonstrated in Figure 3 for compositions with the various content. XRD has also revealed that part of the Cr has bonded with the Ni as the nichrome phase has emerged in the pattern. combustion zone, which is corroborated by the lack of a prolonged fluorescence and faster cooling of a specimen prepared from the granulated Ti + C mixture, although the combustion temperature for this mixture is higher than for Ti-Cr-C. One should note that the specimens without the metallic binder (M1, M5) featured the presence of TiC in addition to (Ti, Cr)C (Figure 3a), but TiC was absent in the combustion products produced in semi-industrial SHS reactor from M1 mixture (Figure 4a).  Table 1); e-etalon TiC lines. Circle highlights the shift of one of the (Ti,Cr)C peaks.   Table 1); e-etalon TiC lines. Circle highlights the shift of one of the (Ti,Cr)C peaks.  Table 1). (c) shows the etalon TiC lines.
The difference in the phase composition of products obtained under different conditions can be explained if we take into account the sequence of the chemical conversion process and the difference in the product cooling rate. The mass of the initial mixture is almost 100 times higher in the semiindustrial SHS reactor as compared with the lab reactor. Therefore, the cooling rate is degrees of magnitude lower, which provides sufficient time for diffusion processes to be completed. In the framework of this mechanism, the presence of TiC in the small-volume combustion products is inevitable, as it is the most refractory compound (Tmelt = 3200 К) among the intermediate combustion products. These samples cool down faster than the formation of the (Ti,Cr)C phase is concluded.
Let us discuss the influence of metallic binder content on the composition of synthesis products produced from the granulated mixtures. XRD of combustion products of the M1 mixture revealed the formation of the (Ti,Cr)C solid solution with a minor amount of TiC (Figure 3a). An addition of 20 w.% nickel binder (mixture M2) alters the phase composition ( Figure 3b) and leads to the formation  Table 1). (c) shows the etalon TiC lines.
To investigate the influence of scale factor on the phase composition of the combustion products, we compared the XRD data from our specimens (Figure 3) with the specimens produced in a semi-industrial SHS reactor (Figure 4). In the SHS reactor, Vlasov et al. [79] synthesized M1 and M2 mixtures, herewith, the reactor load was 12 kg. As compared with the specimens produced using the lab reactor (Figure 3a), the specimens from the semi-industrial reactor (Figure 4a) do not contain the TiC phase. The addition of 20 wt.% of nickel into the mixture altered the composition of the products obtained from the semi-industrial SHS reactor (Figure 4b) and caused the formation of chromium carbide and nichrome, the same as in Figure 3b. Therefore, the change of phase composition upon the introduction of nickel Crystals 2020, 10, 412 9 of 16 binder does not pertain to the sample's geometry, but rather to the changes of the combustion and phase formation mechanisms.
Zhang et al. [78] suggested that the decreased content of Cr in solid solution (Ti,Cr)C is related to the decreased combustion temperature of reactive mixturex. Indeed, addition of 20 wt.% Ni decreases the adiabatic combustion temperature T ad from 2700 K (M1 composition) to 2350 K (M2 composition). The decrease of Tad lowers the combustion velocity from 22 mm/s to 9 mm/s. To increase the Cr content in (Ti,Cr)C and eliminate the chromium carbide phases, we have to increase the combustion temperature. This could be attained by increasing the content of Ti + C fraction in the mixture, for example, to M5 composition. For this composition, T ad is equal to 2950 K. An addition of 20 wt.% Ni binder (composition M6) leads to a decrease of T ad down to 2600 K, which is close to T ad for M1 composition. The combustion velocities of granulated compositions M5 and M6 increased to 35 and 16.5 mm/s, correspondingly. According to the XRD data, M5 combustion products contain (Ti,Cr)C with a small amount of TiC. Upon the addition of nickel (M6 composition), the content of Cr in TiC decreases drastically, in the same manner as in the composition M2. The peak of solid solution shifts to the left and is close to the pure TiC peak (Figure 5a,b). No direct dependencies between combustion temperature and Cr content in solid solutions were found in mixtures with a nickel binder.
Crystals 2020, 10, x FOR PEER REVIEW 9 of 16 decreases the adiabatic combustion temperature Tad from 2700 К (M1 composition) to 2350 К (M2 composition). The decrease of Tad lowers the combustion velocity from 22 mm/s to 9 mm/s. To increase the Cr content in (Ti,Cr)C and eliminate the chromium carbide phases, we have to increase the combustion temperature. This could be attained by increasing the content of Ti + C fraction in the mixture, for example, to M5 composition. For this composition, Tad is equal to 2950 К. An addition of 20 wt.% Ni binder (composition M6) leads to a decrease of Tad down to 2600 K, which is close to Tad for M1 composition. The combustion velocities of granulated compositions M5 and M6 increased to 35 and 16.5 mm/s, correspondingly. According to the XRD data, M5 combustion products contain (Ti,Cr)C with a small amount of TiC. Upon the addition of nickel (M6 composition), the content of Cr in TiC decreases drastically, in the same manner as in the composition M2. The peak of solid solution shifts to the left and is close to the pure TiC peak (Figure 5a,b). No direct dependencies between combustion temperature and Cr content in solid solutions were found in mixtures with a nickel binder.
The addition of 20 wt.% Ni as the binder leads to the formation of the Ni2.88Cr1.12 phase. Therefore, chromium must be (partly) forming a solid solution with nickel, as Cr and Ni are well-soluble in each other [81]. Apparently, Ni, which has a lower melting temperature, melts first and spreads on Cr particles with the formation of nichrome, which immobilizes Cr and prevents the formation of (Ti, Cr)C with the calculated composition. To prevent Cr depletion owing to the formation of nichrome, the following experiments were conducted with Ni replaced by the Ni-20%Cr and Ni-40%Cr alloys (binders B2, B3 in Table 1). XRD data for combustion products of compositions M3, M4, M7, and M8 are provided in Figure 3c Figure 3 demonstrates that the substitution of nickel with nichrome did not lead to the increase of Cr content in (Ti, Cr)C, that is, to the shift of the solid solution peak towards the larger angles (see The addition of 20 wt.% Ni as the binder leads to the formation of the Ni 2.88 Cr 1.12 phase. Therefore, chromium must be (partly) forming a solid solution with nickel, as Cr and Ni are well-soluble in each other [81]. Apparently, Ni, which has a lower melting temperature, melts first and spreads on Cr particles with the formation of nichrome, which immobilizes Cr and prevents the formation of (Ti, Cr)C with the calculated composition. To prevent Cr depletion owing to the formation of nichrome, the following experiments were conducted with Ni replaced by the Ni-20%Cr and Ni-40%Cr alloys (binders B2, B3 in Table 1). XRD data for combustion products of compositions M3, M4, M7, and M8 are provided in Figure 3c,d and Figure 5c,d. Figure 3 demonstrates that the substitution of nickel with nichrome did not lead to the increase of Cr content in (Ti, Cr)C, that is, to the shift of the solid solution peak towards the larger angles (see the (220) peak in the magnified fragment of XRD pattern in Figure 3). When the nichrome (Ni-40%Cr) alloy was used, chromium carbides and Cr 7 Ni 3 solid solution were formed, signifying the excess of Cr in the initial mixture. However, for composition M7, the substitution of Ni to nichrome (20%Cr) leads to the increased content of Cr in (Ti, Cr)C and little to no unwanted by-product phases (Figure 5c). For M8 composition (binder with 40% Cr), unwanted Cr-based solid solutions and chromium carbides are formed (Figure 5d).
It is noteworthy that synthesis products produced from all of the investigated granulated mixtures, both with and without the metallic binder, were in a form of brittle sinter cake comprised of granules with the same size as in the initial mixture. Figure 6 demonstrated the porous, easy-to-grind granules of combustion products.
Crystals 2020, 10, x FOR PEER REVIEW 10 of 16 the (220) peak in the magnified fragment of XRD pattern in Figure 3). When the nichrome (Ni-40%Cr) alloy was used, chromium carbides and Cr7Ni3 solid solution were formed, signifying the excess of Cr in the initial mixture. However, for composition M7, the substitution of Ni to nichrome (20%Cr) leads to the increased content of Cr in (Ti, Cr)C and little to no unwanted by-product phases ( Figure  5c). For M8 composition (binder with 40% Cr), unwanted Cr-based solid solutions and chromium carbides are formed (Figure 5d). It is noteworthy that synthesis products produced from all of the investigated granulated mixtures, both with and without the metallic binder, were in a form of brittle sinter cake comprised of granules with the same size as in the initial mixture. Figure 6 demonstrated the porous, easy-togrind granules of combustion products. Porosity arises from the emission of adsorbed gases and products of decomposition of polyvinyl butyral, which arrests the coagulation of droplets formed during combustion. The fractures of individual granules of combustion products are provided in Figure 7. Noticeably, all SEM images demonstrate that the grains of combustion products are 2-5 μm, which is an order of magnitude below the size of Cr and Ti particles owing to the self-dispersion of starting reagents during the combustion. Porosity arises from the emission of adsorbed gases and products of decomposition of polyvinyl butyral, which arrests the coagulation of droplets formed during combustion. The fractures of individual granules of combustion products are provided in Figure 7. Noticeably, all SEM images demonstrate that the grains of combustion products are 2-5 µm, which is an order of magnitude below the size of Cr and Ti particles owing to the self-dispersion of starting reagents during the combustion. Combustion products of granulated specimens with the metallic binder could be crushed into <1 μm powder both in a mortar and using the hydraulic press at approximately 2.45 MPa (Figure 8b). Combustion products from non-granulated mixtures could be ground only if they contained no metallic binder. Binder-containing non-granulated samples could be crushed to the 5-10 mm pieces using the hydraulic press at 2.45 MPa pressure ( Figure 8d) and then to 1-2 mm pieces using 12.25 MPa pressure (Figure 8e). Thus, in this work, we used SHS of granulated 80%(Ti + C)/20%(3Cr + 2C) mixture with a nichrome (Ni-20%Cr) binder to produce in one technologic operation the (Ti,Cr)C solid solution with Combustion products of granulated specimens with the metallic binder could be crushed into <1 µm powder both in a mortar and using the hydraulic press at approximately 2.45 MPa (Figure 8b). Combustion products from non-granulated mixtures could be ground only if they contained no metallic binder. Binder-containing non-granulated samples could be crushed to the 5-10 mm pieces using the hydraulic press at 2.45 MPa pressure ( Figure 8d) and then to 1-2 mm pieces using 12.25 MPa pressure (Figure 8e). Combustion products of granulated specimens with the metallic binder could be crushed into <1 μm powder both in a mortar and using the hydraulic press at approximately 2.45 MPa (Figure 8b). Combustion products from non-granulated mixtures could be ground only if they contained no metallic binder. Binder-containing non-granulated samples could be crushed to the 5-10 mm pieces using the hydraulic press at 2.45 MPa pressure ( Figure 8d) and then to 1-2 mm pieces using 12.25 MPa pressure (Figure 8e). Thus, in this work, we used SHS of granulated 80%(Ti + C)/20%(3Cr + 2C) mixture with a nichrome (Ni-20%Cr) binder to produce in one technologic operation the (Ti,Cr)C solid solution with Thus, in this work, we used SHS of granulated 80%(Ti + C)/20%(3Cr + 2C) mixture with a nichrome (Ni-20%Cr) binder to produce in one technologic operation the (Ti,Cr)C solid solution with nickel-containing binder and a minor amount of impurities. Products of combustion of agglomerated mixtures could be crushed and ground easily. This suggests a perspective of the application of granulation of reactive mixtures for the SHS of composite powders with an optimal combination of melting point, hardness, corrosion, and ablation resistance for various applications, including the flame spraying [95][96][97][98].

Conclusions
1. The granulation of reactive mixtures Ti + C, 70%(Ti + C)/30%(3Cr + 2C), and 80%(Ti + C)/20%(3Cr + 2C) was demonstrated to increase the combustion velocity by 1.5-fold; the granules in the combustion products retained their size and did not sinter.; 2. The compositions with metallic binder featured a prolonged fluorescence after the passing of the combustion front, presumably owing to the occurrence of exothermic secondary phase formation processes; 3. The addition of nickel decreases the Cr content in (Ti,Cr)C solid solution owing to the interaction of nickel melt with chromium and formation of NiCr intermetallic; 4. For the granulated 80%(Ti + C)/20%(3Cr + 2C) mixture, the substitution of Ni to Ni-20%Cr and Ni-30%Cr alloys resulted in the increase of Cr content in solid solution and the formation of (Ti, Cr)C with NiCr binder; 5. Combustion products procured from granulated mixtures could be more easily processed into powders owing to the decreased ductility of the combustion products.