Microstructural evaluation and wear behavior of B4C particle reinforced TiO2-SiO2-ZnO compound by hot-pressing method

In this study, different wt% of B4C material added to TiO2, ZnO, and SiO2 materials known with distinguished properties by hot-pressing method to produce new materials. Characterization of the materials was performed by using XRD, SEM and EDS. The micro-hardness of the materials was measured and the wear characteristics of B4C-added materials were determined by subjecting them to wear loads of 5 N, 10 N and 15 N. In XRD, it was observed that the intensity of the peak was decreased depending on the increase in B4C content. In SEM, it was determined that there was a two-phase structure with complex boundaries and indefinite grain boundary formation. Besides, it was determined that elements in the EDS and the compounds in the XRD supported each other. In wear analyses, it was seen that as the wear load increased, the depth and width of the wear track in all materials increased. Furthermore, as the weight ratio of B4C material in the composition was increased, the weight loss at different wear loads was decreased and the lowest weight loss was found in Ti45Si15Zn20B20 material. Moreover, it was observed that the wear track and its depth were inversely proportional to the increasing B4C content in the compound. Friction coefficients of the materials under 5 N wear load varied between 0.15–0.43, whereas under 10 N and 15 N wear load, it varied between 0.09–0.28 and 0.06–0.29, respectively. While the hardness value of B4C-free Ti60Si20Zn20 was 260 HV, the highest hardness value was seen in Ti45Si15Zn20B20 material as 1010 HV.


Introduction
Powder metallurgy is a material processing technique utilized to have desired solid metal, alloy or ceramic materials by using both metal and/or non-metals. Choosing the optimal processing route is highly dependent on the knowledge and insight of the operator [1]. It has a wide range of utilization from complex parts such as automotive components to simple parts such as household appliances. TiO 2 , ZnO, SiO 2 and B 4 C materials are being widely used in powder metallurgy applications because of their distinguished properties [2][3][4][5][6][7].
TiO 2 biomaterial is being widely used in composite production, by itself or together with alumina, as filler owing to its superior properties. The mechanical and viscoelastic properties of the composites according to different particle size and filler concentrations of TiO 2 were examined [8]. They established that fillers improved the properties of base composite material and the tensile strength, fracture toughness, glass transition temperature and storage modulus were increased more in the case of nano-sized TiO 2 [8]. Nayak et al [9] investigated nano-TiO 2 enhanced glass fiber reinforced polymer composites and revealed that with the addition of nano-TiO 2 , reduced water absorption and improved residual flexural strength by 19% and interlaminar shear strength by 18% compared to base composites. Chatterjee and Islam [10] investigated the effect of introducing nano-size TiO 2 and stated that thermal, mechanical and viscoelastic properties were improved but noted that it is highly dependent on the dispersion state of the TiO 2 . Kumar et al [11] found similar observations. Huang et al [12] found similar improvement in functional properties with TiO 2 incorporation and stated that any further addition of TiO 2 materials over 2 weight percent (wt%) might cause a decrement in the base material's inherent properties. Likewise, Kavimani et al [13] considered the synergistic effect of Graphene oxide and TiO 2 and observed an increase in mechanical and thermal properties. Bora et al [14] studied the effect of different wt% concentrations of TiO 2 -reinforced composites and concluded that the stiffness, toughness, maximum sustained strain, and resistance to crack propagation of the matrix could be simultaneously improved [15]. Shirkavand and Moslehifard stated that a significant increase in tensile strength occurred in 1 wt% TiO 2 nanoparticles but with further addition, found a decrease in the tensile strength [16].
ZnO biomaterial, today, is widely used in solar cells and other electronic nanodevices, this biomaterial is also employed in pharmaceutical and aesthetic applications due to its biocompatibility like Ti and its alloys. Owing to its physical properties, it is strongly demanded as a filler for different applications [17]. Promsawat et al [18] surveyed ZnO nanoparticulate addition on the properties of PMNT ceramics. Results revealed that hardness and fracture toughness were increased in PMNT ceramic with the addition of 0.5 to 1.0 wt% ZnO nanoparticulate. Ramezanzadeh et al [3] stated that the mechanical properties of nanocomposites and the curing behavior of epoxy coatings could be influenced effectively with the addition of the ZnO nanoparticles. Carrion et al [19] stated that with the presence of ZnO nano-filler, better tribological performance and more resistant nanocomposite were obtained. Bandyopadhyay et al [20] observed that for HAp and TCP materials, microhardness was improved, densifications of both ceramics occurred and cell spreading and cell to cell interactions were decreased with ZnO increasing.
Similar to TiO 2 and ZnO, silica(SiO 2 )-based materials are another biomaterial that is frequently used in biomedical applications due to their biocompatibility and advantageous biological effects. Silica, sphericalshaped, nanoparticles doped composites exhibit higher mechanical properties and corrosion resistance when used as filler material [21][22][23][24]. Corrosion resistance enhancement with SiO 2 addition in a composite system was reported [25]. Furthermore, hydrogen-bonding interaction between hydroxyl groups and the chain-like structure of aggregated SiO 2 nanoparticles were the reason for corrosion enhancement in composite [26]. Pourhashem et al [27] found that the SiO 2 -graphene oxide nanocomposite exhibits outstanding corrosion protection performance. Yu et al [28] reported that enhanced water adsorbing and hydrophilic properties of hybrid membrane doped with SiO 2 nanoparticles. Hasan et al [29] observed 45% and 26% improvement in tensile modulus and ultimate tensile strength respectively by using SiO 2 as a filler compared to the plain matrix.
Namini et al [30], reported that with the addition of B 4 C, macro/micro Vickers hardness was increased but a decrease in UTS and elongation was obtained in Ti-based composites at room temperature. Shirvanimoghaddam et al [31] investigated the effect of B 4 C, ZrSiO 4 and TiB 2 particles in Al-matrix composites. They reported an improvement in hardness and tensile strength. Ni et al [32] studied the B 4 C particle size effect in Ti-matrix composites and found that the size of B 4 C directly affects the mechanical properties and microstructure. Özler et al [33] studied B 4 C material reinforcement influence by not only gradient coating but also powder laying processes to analyze coating structure, micro-hardness and wear. Mohammad Sharifi et al [4] stated that the wear resistance of Metal-Matrix-Composites could be improved by using B 4 C particles. Ramkumar et al [34] found that the optimum wear resistance, coefficient of friction and hardness were obtained in that 10 wt%.
In this study, it was aimed to take advantage of excellent properties such as ductility of TiO 2 biomaterial, biocompatibility of ZnO, the strong surface energy of SiO 2 biomaterial and high hardness of B 4 C to produce new materials that combine the best properties of the starting materials. To do that, varying amounts of B 4 C in TiO 2 , ZnO, and SiO 2 biomaterials were used by the hot-pressing method. The effect of varying amounts of B 4 C, such as B 4 C-free for the control group, 5, 10, 20, and 30 wt% on wear behavior were investigated. The materials produced by the hot-pressing method were characterized by SEM, EDS and XRD analysis. Micro-hardness properties of materials were studied and tribological properties under different loads, such as 5 N, 10 N and 15 N were evaluated for every material. It was determined that with the addition of varying amounts of B 4 C to the starting materials, the wear losses generally decreased and the wear track and depth were found to be inversely proportional regarding the increased B 4 C content for the same wear load.

Preparation of materials and specimens
New materials were created by taking into account the weight percentages of several biomaterials in this study. In order to examine the impact of a doped material, one of the existing materials must be weighted down based on this circumstance. The ratios of ZnO and Silica biomaterials were attempted to be kept as constant as possible since silica increases mechanical qualities while ZnO material addition improves wear resistance as stated in the introduction section. Due to this, one substance-in this example, TiO 2 -was seen as a matrix structure, and its ratios were altered as B 4 C was added. Additionally, attention was taken when creating biomaterials to make sure that the weight percentages of the additives were smaller than the TiO 2 that was chosen as the matrix material.
The biomaterials, whose weight % ratios were given in table 1, were produced by the hot-pressing method. TiO 2 (99% pure, AppliChem), SiO 2 (99% pure, Sigma Aldrich), ZnO (99% pure, Fluka) and B 4 C compounds were used. Chemical compounds were mixed mechanically with a mixer for an hour. All materials were hot pressed to 2 mm in diameter and 2 mm in height under 45 MPa force at 950°C unidirectionally (figure 1). During hot-pressing, 10 −4 mbar vacuum pressure was applied to the materials and the temperature was increased to 950°C in 10 min. The pressure and temperature were kept under 45 MPa force and 950°C for 30 min respectively. Thereafter, the production of the materials was completed by cooling them to room temperature in 5 min. These parameters used for production were reached after dozens of experiments. When its mechanical properties and morphological structure were examined, it was seen that the best result was obtained at 950°C and 45 MPa. The wear behavior of produced materials based on different powders used in this study are very important for different applications, especially for biomedical application.

Characterization
X-Ray Diffractometer (XRD), Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS) was used to analyze the microstructures and chemical properties of the materials. Firstly, a Rigaku-Miniflex600 diffractometer were used to examine the x-ray diffraction patterns of products. Materials were irradiated from Cu-Kα radiation at 40 kV voltage and 15 mA current. XRD analyses were conducted at a wavelength of 1.5406 (λ) between 10°and 80°with a step speed of 0.02°and a rate of scanning speed of 2°per min. Then, JEOL/JSM-7001F device was used for SEM analyses for examining surface characterization and wear zones of materials. Lastly, Oxford/INCA-EDS equipment was used for investigating the elemental analysis of the materials.

Pin-on-disc wear testing
Pin-on-disc wear tests were conducted under varying loads of 5 N, 10 N, and 15 N to examine the wear resistances of materials. Tribometer T10/20 equipment operating according to ASTM G99-17 was used for wear test. The materials were produced 2 mm in diameter and 2 mm in height. A 6 mm diameter stainless steel ball was used for the pin-on-disc wear test. Materials were weighed with a 10-4 accuracy weighing device before and after each wear test. All materials were subjected to 5 N, 10 N and 15 N varying loads to determine the wear resistances. The sliding distance, rotational speed and sliding speed were set to 300 m, 240 rpm, and 0.005 m s −1 , respectively.

Micro-hardness measurement
After wear tests, micro-hardness tests were applied by using a micrometer-controlled Qness-Q10M equipment to the back surfaces of the materials. In the preliminary hardness tests, it was observed that accurate hardness  measurements could not be obtained due to surface topography. Therefore, all materials were abraded using abrasive papers to reduce surface roughness. The rough surfaces of all materials were treated with 100, 400, 600, 800, 1000 and 1200 SiC abrasive papers. After the materials were polished, hardness tests were carried out thereafter with the relevant equipment. A load of 3000 gr (HV3) was applied to the materials, and 10 s of dwelling time conditions were considered. The micro-hardness tests of materials were repeated five times and the average value was considered. Each measurement point was chosen according to not being affected by the other measurements.

XRD
The main purpose of x-ray diffraction analysis is the determination of different phases in multi-phase materials. The concentration of a mixture is a determinant of the intensity of the diffraction pattern of a phase or phases [35]. Therefore, the crystallinities, phases and peaks of hot-pressed materials were examined by x-ray Diffraction technique. XRD analyses were conducted between 10°and 80°and the XRD results were given in figure 2. In order to compare the diffraction patterns of all materials in the same graph, all intensity values of XRD data of each material were normalized according to the largest intensity (cps-deg.) value and the results were plotted as per normalized-value. To see the peek value changes occurring depending on the B 4 C content, the intensity of peek values of each material was also shown on the normalized graph. Crystal structure analysis results were compared to Crystallography Open Database (COD) standard patterns for the identification of phases in the compounds as per DB card numbers. The peek values of each phase were indicated in the XRD graph by using different geometrical shapes for comparison (figure 2 and table 2). In all XRD graphs in figure 2, it was observed that composed compound peaks were compatible with the starting elements. Besides, it was seen that the composed materials had a crystalline structure evidently from the peaks formed in the graphics. When the XRD graph in the B 4 C-free was examined, it was observed that the most severe peak occurred around 2θ∼26. Moreover, it was observed that the relevant peak intensity firstly increased (5 wt% B 4 C doping), then it was decreased in general and even lower peak intensities than the Ti60Si20Zn20 material was acquired depending on B 4 C content. Furthermore, it was seen that all materials gave similar peaks in XRD results (figure 2). In addition, the most severe peaks occurred in the case of Ti45Si15Zn20B20 material which is a clear sign of more crystalline structure than in any other materials.
XRD results showed that Titanium-Oxide, Coesite phases were obtained for the first material (table 2) whereas Graphite, Zinc, Titanium-Oxide and Quartz phases were determined for the second material. The third material's XRD results showed Graphite, Zincite, Quartz, Heptaoxotetratitanate (O 7 Ti 4 ) phases and Graphite,

SEM-EDS analysis
SEM and EDS images of Ti60Si20Zn20 material were given in figures 3(a) and 3(b), respectively. When the SEM in figure 3(a) was examined, it was observed that two different phase structures, namely white and black phases, were formed. The general EDS analysis results of figure 3(a) containing two layered phases were given in figure 3(b). These analysis results showed that Ti60Si20Zn20 material had Ti, O, Zn, and Si materials.
In figure 4(a), an SEM image of Ti60Si15Zn20B5 material was given. It was seen that two different regions were observed as so in the SEM image of the Ti60Si20Zn20 material. Therefore, it could be stated that there were two different phases in figure 4(a) and there was no homogeneous distribution between these two phases. In figure 4(b), an SEM image of Ti55Si15Zn20B10 material was given. This image had a similar structure to the materials given in figure 3(a) and figure 4(a) meaning there were two different phases, manifesting themselves in black and white, and a heterogeneous layered structure was formed. The differences of this image from the previous ones were that the black regions, which were considered as the matrix/main phase, were more dominant and the white clusters were lesser than Ti60Si20Zn20, Ti60Si15Zn20B5 material. SEM image of Ti45Si15Zn20B20 material was given in figure 4(c). Compared to the previous images of other materials, this material showed the same structure but white clusters dominated by Ti, O, and C elements evidently from the EDS results of ( * ) point in figure 4(d).
SEM image of Ti35Si15Zn20B30 material was given in figure 4(d). There were two different phase structures in two colors in common in figure 4(d), the grain boundaries were not clear and complex boundaries with a heterogeneous structure were formed as in all former materials [36,37]. General and specific ( * ) point EDS analysis results in figure 4(d) were given in figures 4(e) and 4(f), respectively. The general EDS analysis results in figure 4(e) showed that C, O and Ti elements were dominant, while specific ( * ) point EDS analysis results in figure 4(f) showed dominant Ti and O elements and very few B elements, respectively. Besides, SEM images of Ti60Si15Zn20B5 with 1.00 K magnitude and 2.00 K magnitude were given in order to support the discussion of structure formed in figures 4(g) and 4(h), respectively.

Wear test and weight losses
In this study, the wear characteristics of new materials, produced by adding 0, 5, 10, 20 and 30 weight percent of B 4 C material to TiO 2 , SiO 2 and ZnO compounds with the hot-pressing method, under 5 N, 10 N and 15 N applied loads were investigated. In the wear tests, a 300 m sliding distance and constant sliding velocity were chosen. Each material was weighed with a precision scale of a sensitivity of 10 −4 g before the wear test. Subsequently, every material was weighed again on the same scale and the weight losses after wear tests were measured. SEM image of Ti60Si20Zn20 material subjected to 5 N, 10 N and 15 N wear loads was given in figure 5(a). At 5 N wear load, it was observed that the wear marks were not deep and there were local flattening zone-by-zone. The depth and width of the wear track were increased as the wear load increased to 10 N. In addition, under this load, the wear marks were more evident contrary to the 5 N wear load. Besides, the white areas were ruptured generally in the wear track and pits were formed due to rupturing. It could be stated that the depth and width of the wear track were clearly higher in the case of 15 N wear load than not only 5 N but also 10 N wear loads. Likewise, both the amount and size of the ruptured areas broken off and the pitted structures associated with them were more than any other wear loads. Besides, grooved wear tracks were more obvious in the case of 15 N wear load.
SEM image of Ti60Si15Zn20B5 material after wear test of 5 N, 10 N and 15 N loads were given in figure 5(b). When the wear tracks were examined, it was observed that the wear tracks were evident for all loads and grooved tracks were formed due to the pieces broken-off in the wear trajectory. Besides, it was determined that the biggest wear track depth and width occurred under 15 N wear load as in the case of figure 5(a). In addition, it was observed that as the wear load increased, the amount and the size of the fragments broken-off increased and the depth of the pitted structures increased accordingly. Besides, Ti60Si15Zn20B5 material had less depth and width tracks under 10 N and 15 N wear loads than Ti60Si20Zn20 material. SEM image of Ti55Si15Zn20B10 material after the wear test was given in figure 5(c). It was seen that the wear tracks under 5 N wear load were not deep compared to other loads and there was some flattening due to friction. In addition, the wear track and depth at 10 N wear load were more evident and larger than the 5 N wear load. Besides, under 10 N wear load, the pieces broken off in the wear track and therefore the pitted structure occurrence were less than both Ti60Si20Zn20 and Ti60Si15Zn20B5 materials. In the case of the 15 N wear load, the wear track and depth were more profound than both 5 N and 10 N wear loads and pitted structures due to the pieces/fragments broken off under the 15 N wear load were more severe than any other wear load. Grooved wear tracks were more obvious in the case of 10 N and 15 N wear loads.
SEM image of Ti45Si15Zn20B20 material subjected to 5 N, 10 N and 15 N wear loads was given in figure 5(d). It was observed that an obvious wear track was formed due to friction at 5 N wear load and there were particles (blue arrows) broken-off in the wear trajectory and pits were generated. In the case of the 10 N wear load, the wear tracks were more evident than the 5 N wear load, and the wear depth was greater as was the case for the other materials. Moreover, at 10 N wear load, it was determined that there were obvious pieces broken off in the wear trajectory like at the 5 N wear load. But the largest wear track and depth occurred in the case of 15 N wear load and the pitted structure due to the broken-off pieces were also seen at this wear load. Besides, grooved wear tracks were obvious for all wear loads but were the severest in the case of 15 N wear load. EDS analysis of the worn surfaces, rectangle in figure 6(a), of materials subjected to the wear test was examined. In addition to the general EDS analysis, the elemental contents of the materials produced by the hotpressing method were also examined. In this context, the EDS analysis and the elemental contents of EDS analysis results taken from the worn surface of the Ti60Si15Zn20B5 material were shown in figure 6(a) and figure 6(b) respectively. It could be seen from figure 6(b) that the element C most probably originating from the B 4 C compound was dominant on the worn surfaces and in terms of presence, this element was followed by Ti, Zn and Si elements, respectively.
Weight losses of materials were measured at each stage of the wear tests under different wear loads with a scale having a sensitivity of 10 −4 g. These weight losses of materials after the wear tests were given in figure 7. When this graph was examined, it was determined that the lowest weight loss under 5 N wear load was found for the Ti45Si15Zn20B20 material among the others. Besides, it was observed that the highest weight loss occurred in the Ti60Si15Zn20B5 material for the same wear load.
At 10 N wear load, it was determined that the lowest weight loss occurred in the Ti45Si15Zn20B20 material as was the case for the 5 N wear load. Moreover, it was observed that the highest weight loss occurred in Ti35Si15Zn20B30 material under the 10 N wear load. For the 15 N wear load, it was observed that the lowest weight loss occurred in the Ti45Si15Zn20B20 material as was the case for 5 N and 10 N wear loads. Furthermore,

Friction coefficients
Friction coefficient graphs of the materials were given in figures 8(a)-(c) respectively. Although the friction coefficients of the wear test in all graphs were initially started at zero, these situations were not visible due to the majority of the data being transferred to the chart. When figure 8(a) was examined, it was seen that the friction coefficients of the materials under 5 N wear load varied between 0.15 and 0.43. After a very short sliding distance at the beginning of the wear test, the friction coefficients of all materials reached a relative equilibrium state. All materials showed a similar trend in terms of coefficient of friction under 5 N wear load. That is, at the beginning of the wear test, it was observed that the materials passed the run-in-zone quickly and then showed a stable behavior in terms of friction coefficient. Moreover, it was observed that Ti35Si15Zn20B30 material showed the highest coefficient of friction.
When figure 8(b) was examined, it was seen that the friction coefficients of the materials under 10 N wear load varied between 0.09 and 0.28. All materials showed a generally similar friction coefficient trend for the relevant wear load. In other words, the run-in-zone was passed very quickly and then the friction coefficients showed a stable behavior likewise in figure 8(a). It was observed that, unlike the others, only the friction coefficient of Ti35Si15Zn20B30 material was increased to 0.21 value and became stable at an average friction coefficient value of 0.14. Under this wear load, in contrast to the 5 N wear load, Ti45Si15Zn20B20 material showed the highest friction coefficient.
When figure 8(c) was examined, it was seen that the friction coefficients of the materials under 15 N wear load varied between 0.06 and 0.29. All materials showed an overall similar trend in terms of friction coefficient under 15 N wear load as was the case in figures 8(a) and 8(b). However, it was observed that the friction coefficient of Ti35Si15Zn20B30 material after run-in-phase was decreased to an average of 0.05 and shortly after that it was increased to 0.18 and then relatively became stable following 50 m of sliding wear distance. The friction coefficient value of the B 4 C-free material reached a stable state of around 0.15 after a sudden increase of 0.22. Moreover, while Ti55Si15Zn20B10 and Ti45Si15Zn20B20 materials were initially stable at 0.22 friction coefficient, after 100 m of sliding distance, their friction coefficients showed a very small linear increase behavior, and they ended the test with a friction coefficient of 0.25. Under relevant wear load, contrary to the 5 N and 10 N wear load, Ti55Si15Zn20B10, Ti45Si15Zn20B20 and Ti55Si15Zn20B10 materials showed higher friction coefficients.

Hardness
Qness Q10M equipment was used for micro-hardness measurement and an example was given in figure 9 with the test parameters of HV3 and 10 s dwelling time. Five recurrences of hardness tests were applied to materials and average values were considered. When the hardness results were reviewed in figure 10, it was observed that Ti60Si20Zn20, material had a hardness value close to 260 HV. As the B 4 C amount was increased, the hardness was increased compared to Ti60Si20Zn20 material. The hardness value was increased to 940 HV approximately by introducing 5% of B 4 C material to compounds. Not only the highest hardness value but also the highest standard deviation was seen in Ti45Si15Zn20B20 material. It was also observed that after 20% B 4 C content, the hardness was decreased to 880 HV for Ti35Si15Zn20B30 material.

Discussion
In this study, varying amounts of B 4 C material were added to the TiO 2 , SiO 2 and ZnO compounds via the hotpressing method for having materials sustaining the distinguished properties of beginning materials. The characterization of the produced materials was carried out by XRD, SEM and EDS. In addition, the morphological and chemical wear characteristics of the materials were investigated under the different wear loads of 5 N, 10 N and 15 N and their micro-hardness measurements were studied.  In the XRD analysis, mostly sharp peaks were formed (figure 2) and these peaks show that most of the formed phases are in crystalline structure. On the other hand, the presence of broad peaks, which is a sign of less crystalline structure, indicates the presence of amorphous compound structures [38,39]. From figure 2, it was observed generally that as B 4 C was added to the compounds, the crystalline structure was decreased and the amorphousness structure was increased evidently from the decrease in peak intensities in general. In addition, the most severe peaks in terms of the normalized peak occurred in the case of Ti45Si15Zn20B20 material which is a clear sign of more crystalline structure than others. When the XRD results of the produced materials in figure 2 and table 2 were examined in terms of qualitative analysis results of compounds, it was found that Titanium Oxide, Coesite, Graphite, Zinc, Titanium Oxide, Quartz, Heptaoxotetratitanate (O 7 Ti 4 ), Ti 3 (BO 3 ) 4 , Cristobalite, Rutile, Zincite, SiO 2 and TiZn 2 O 7 phases were formed. Therefore, the crystalline phases formed in XRD analysis and the elements in the general EDS analysis for every material were found to be compatible.
It was observed in figure 3 and figure 4 that a heterogeneous structure, two phases with two different colors, complex grain boundaries, different height regions and white clusters in the black matrix structure were formed. In figure 3(a), (a) layered structure was formed between these white and black phases, but there was no clear grain structure. The absence of a clear grain boundary structure between the phases indicated that liquid phase sintering occurred during production. Furthermore, the presence of different layered phases was also an indication of the formation of a heterogeneous structure.
It was observed in figure 4(a) that the white were distributed within the black region, which could be called the matrix naturally. It could further be emphasized that layered regions in figure 4(a) were more but the white areas were more scattered than the Ti60Si20Zn20 material. But, the SEM image in figure 4(b) was different from the previous ones in terms of the black regions, which were considered as the matrix/main phase, were more dominant and the white clusters were lesser than Ti60Si20Zn20, Ti60Si15Zn20B5 material. On the other hand, the white clusters were occurred in figure 4(c) and these structures were scattered not only in the black matrix phase but at the layer boundaries. In addition, the layered structure was also formed in this image, but these regions were less involved as compared to the previous materials. In figure 4(d), it was observed that instead of a dominant black matrix structure as in the other images, the white and black regions were approximately covering half of the structure. Moreover, the layered structure was less than all other materials.
The absence of clear grain boundary structures in all of SEM images was thought to be caused by melting due to liquid phase sintering. The same formation was reported in the study [35], investigating the 1010 steel-based materials containing SiC, MgO, and H 3 BO 3 with varying amounts of B 4 C using the same production method. In addition, by further examining SEM images in figure 3 and figure 4, one could state that the layered structure was decreased with the increase of B 4 C amount material and the white agglomerated clusters in the structure were located more dispersedly. Besides, liquid phase sintering was evident in all images. It must be stressed that the decrease in peak intensities in the XRD results due to B 4 C addition and thus the decrease in crystallinity was not supported by the SEM results due to liquid phase sintering. A similar sintering effect due to the production method was reported on the Silicon Nitride films, implant materials of 316L and 1010 steel-based materials containing SiC, MgO, H 3 BO 3 and varying B 4 C amounts in [36,37,40] respectively.
General EDS analyses of each material were performed and it was observed that Ti, O, B, Zn, Si, and C elements were encountered in the results. So, elemental composition results obtained from EDS elemental analyses and crystalline phases (table 2) obtained from XRD analyses overlap not only with each other but also with starting materials. Therefore, it can be emphasized that the results of the two analyses support each other. According to the EDS analysis results taken from the SEM images, it was determined that Ti and O were dominant in the white regions, while C, Zn and Ti elements were dominant in the black regions (figures 4(e) and 4(f)). Furthermore, in the EDS analysis of the worn surfaces, it was observed that element C was more richly present than others ( figure 4(b)). This situation was thought to be due to the B 4 C compound.
It was observed in the EDS analysis that as the B 4 C addition increased, the white regions rich in Ti and O elements ratios were also increased. Likewise, a decrease in crystalline structure and an increase in amorphousness were observed in XRD results due to an increase in B 4 C amount. Similar observations were reported in [38,39].
It was seen in the wear test of Ti60Si20Zn20 material that the wear track and depth increased as the wear load increased ( figure 5(a)). It was also observed that small wear tracks were formed under 5 N wear load and the wear trajectory was not clear compared to the bigger wear loads. In addition, it was determined that as the wear load increased, the pieces broken off from the surface were increased and pits formation occurred accordingly. The pitted structure and grooves on the worn trajectory in the wear test are an obvious indication of the abrasive and adhesive wear mechanism occurrence for the material [41][42][43].
When the SEM images of the wear test of the Ti60Si15Zn20B5 material in figure 5(b) were examined, it was observed that the largest wear track and depth were formed at 15 N load, as in the Ti60Si20Zn20 material. Besides, in the wear test of Ti60Si15Zn20B5 material, it was determined that the pieces broken off were increased regarding the increasing wear load and pits formation occurred accordingly. Same observations were reported in [44]. However, it was seen that the wear track and depth of the Ti60Si15Zn20B5 material were less than the Ti60Si20Zn20 material. This observation was in agreement with the results of [45]. The pitted structure and grooves on the worn trajectory were also formed.
When the wear images of Ti55Si15Zn20B10 material in figure 5(c) were examined, it was observed that the wear track and depth increased as the wear load increased, as for the other materials. In addition, for the same wear load, it was determined that the pieces broken off in the wear trajectory, and the pits formed accordingly, were less than both Ti60Si20Zn20 and Ti60Si15Zn20B5 materials. The pitted structure and grooves on the worn were formed too.
When the wear images of the Ti45Si15Zn20B20 material in figure 5(d) were examined, it was determined that there were also broken-off pieces in the wear trajectory. Moreover, the wear track and depth of the Ti45Si15Zn20B20 material increased as the wear load increased as was the case for other materials. The pitted structure and grooves on the worn trajectory were also formed for Ti45Si15Zn20B20. The plastic deformation encountered in all wear tests was a conclusion of the possible oxide compounds formed on the wear surfaces causing plastic deformation during wear and resulting in the removal of materials from the worn surfaces due to adhesion and cohesive failure [41,43]. Moreover, it was observed that a grooved structure formation in the wear track was due to the broken-off pieces remaining in the wear trajectory, which causes plastic deformation zones [43].
In figure 7, the weight losses of the materials under different wear loads were given. It was obtained that the lowest weight losses occurred in Ti45Si15Zn20B20 material for all wear loads. In addition, it was determined that as B 4 C addition was increased, the wear losses were generally decreased, but after 20 wt% B 4 C, which seems to be the optimal ratio in terms of weight loss, in the compounds, the weight losses were increased. Therefore, it could be stated that there was a decrease in the wear track and depth with respect to B 4 C addition in general, and the weight losses were decreased with the increase of the B 4 C content. The wear characteristics of B 4 C/Al6061 composites were investigated in [43] and it was reported that the wear width and depth decreased due to the increase in the B 4 C content in the composites. They stated that the wear loss amount and hardness properties were inversely proportional. Moreover, aluminum nanocomposites reinforced with B 4 C were investigated in [4]. Mechanical and tribological properties of nano-composites were evaluated and a decrease in ductility but an increase in wear resistance, hardness and compression strength was determined. The wear behavior of Al-2219 matrix reinforced with B 4 C, Molybdenum Di-sulphide (MoS 2 ), and graphite composites was studied and it was stated that the primary factor that had the most influence on the wear of composites was the load parameter. Moreover, it was reported that B 4 C and coconut shell ash utilization as a filler in the AA7075 had a 66% higher tensile strength and a 33% higher hardness effect compared to the AA7075 aluminum alloy [46]. Thus, it can be said that the weight losses resulting from the materials and their wear behavior are in line with the literature.
When the weight losses respective to wear loads in figure 7 were further examined, it was seen that the highest and the lowest weight losses for all materials were observed under 15 N and 10 N wear loads, respectively. In addition, it was seen that there was no linear relationship between wear loads and weight losses. One of the probable reason for this situation was considered to be the heterogeneous structure of the materials originating from production. The other probable reason for this situation is that under 10 N wear load, unlike 5 N and 15 N, the parts broken off from the wear surface did not separate from the wear surface and caused a lubricating effect in the wear trajectory. As can be seen from the SEM images under 15 N wear load, it can be seen that the wear width, the number of parts broken off and its gaps increase compared to the wear behavior under 10 N wear load. For this reason, the weight loss was smaller at 10 N but larger at 15 N wear load [37,47].
When figures 4(a)-(d) and figure 7 were examined together, it was observed that liquid phase sintering occurred in all images and as the amount of B 4 C increased, the structure formation in the form of layers on the surface increased. On the other hand, it was observed that the amount and size of different phase structures increased regarding to B 4 C increasing. Weight loss under 5 N to 15 N wear loads generally decreased as the B 4 C increased in the composition. On the other hand, it was observed that under 10 N wear load, due to the remaining of broken parts in the wear trajectory cause a lubricating effect and the weight loss after wear was found lesser than other wear loads [48][49][50].
As indicated in the SEM images, the materials showed different wear marks at different B 4 C contents and under different wear loads. In some materials, such as Ti45Si15Zn20B20, the weight loss was reduced due to the lubricating effect of the parts broken off from Ti45Si15Zn20B20 material adhered to the worn zone, while the opposite wear behavior was found for the Ti35Si15Zn20B30 material. In addition, the width of the wear marks and the number of pits formed on the wear surface also differed for those materials. These appearance differences on the surface as a result of wear caused the weight loss data ( figure 7, figure 8) to be variable [37,[43][44][45].
When figure 8(a)-(c) were examined, it was observed that although the friction coefficients for materials fluctuate in general, they showed stable behavior after the first 50 m sliding distance. It could be emphasized that the B 4 C content was directly proportional to the friction coefficient in general for 5 N wear load, but a different situation occurred for 10 N and 15 N wear loads. At 10 N wear load, it was observed that the friction coefficients generally decreased after the first run-in fluctuations regarding the B 4 C addition. For 15 N wear load, it was found that the friction coefficient values were decreased with the addition of B 4 C in general. The main reason for this at the relevant wear load was thought to be the formation of lubricating oxide compounds during the wear test according to the B 4 C addition. Same observations were reported in the literature that there was a decrease in the friction coefficients generally depending on the B 4 C content [37,47]. Because it was reported that the B 2 O 3 compound was formed as a result of the contact of the B 4 C with air during wear tests. Moreover, it was stressed that such oxide compounds have a direct lubricating effect on the wear behavior and thus reduce the friction coefficient [37,[47][48][49][50][51]. Therefore, the friction coefficient results of this study are generally consistent with the relevant studies.
When the hardness results in figure 10 were examined, it was observed that an increase in the microhardness was experienced with the addition of B 4 C material. The lowest micro-hardness was encountered in the case of Ti60Si20Zn20 material and the highest hardness was observed for Ti45Si15Zn20B20 material. The same material had the biggest standard deviation as evidenced by the error bars in figure 10. After 20% B 4 C amount, the hardness and standard deviations decreased. The results showed that the material having the highest hardness value, generally showed the minimum weight loss. Therefore, it can be stated that the hardness results and weight losses of the wear test are inversely related as stated in [45]. Moreover, the fact that the Ti45Si15Zn20B20 material, which had the most crystalline structure according to the XRD results, also had the highest micro-hardness and therefore the lowest weight loss, showing that all analyses in this study were supporting each other.

Conclusion
In this study, different wt% of B 4 C material was added to TiO 2 , SiO 2 and ZnO compounds and the wear characteristics of produced materials were investigated under different loads. The conclusions drawn from this study could be summarized as; • In XRD results, it was observed that a crystalline structure was formed in all materials and the crystallinity decreased with the addition of B 4 C.
• In SEM results, it was observed that two complex phases were formed with unclear grain boundaries. The reduction of crystallinity in XRD results was not observed in SEM images due to liquid phase sintering.
• Ti, O, B, Zn, Si, and C elements were observed in the general EDS results. It was determined that the elements in the EDS analysis and the formed phases in the XRD supported each other.
• It was observed that as the wear load increased, the wear track and depth increased for all materials.
• It was found that the weight loss was decreased generally as B 4 C material was added to the compounds.
• The lowest weight loss and the highest micro-hardness were found for the Ti45Si15Zn20B20 material. Besides, XRD results, weight loss of the materials and micro-hardness test were found to be compatible with each other.