I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL 423–428 CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOLS KARAKTERIZACIJA ORODIJ IZDELANIH IZ RAZLI^NIH WC-Co KARBIDNIH TRDIN Iztok Nagli~ 1* , Adam Zaky 1 , Bla` Leskovar 1 , Miha ^ekada 2 , Bo{tjan Markoli 1 1 University of Ljubljana, Faculty of Natural Sciences and Engineering, Ljubljana, Slovenia 2 Jo`ef Stefan Institute, Department of Thin Films and Surfaces, Ljubljana, Slovenia Prejem rokopisa – received: 2022-04-21; sprejem za objavo – accepted for publication: 2022-07-05 doi:10.17222/mit.2022.478 This paper deals with the characterization of three different commercial, WC-Co cemented-carbide tools in the form of saw blades, one group of which exhibits more frequent cracking. Since the properties of these materials largely depend on the microstructure, a detailed characterization was carried out using scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS) and X-ray diffraction (XRD). The SEM image analysis included a determination of the binder content and the mean WC grain area. The average chemical composition of these materials was also determined using an X-ray fluorescence (XRF) analyser. The results show that despite the same content of binder-forming elements in all three WC-Co cemented-car- bide materials, the material that cracked more frequently contained a smaller amount of binder and a lower mean WC grain area, both of which are known to reduce the toughness of such a material. Keywords: microstructure, WC-Co cemented carbide, scanning electron microscopy, X-ray diffraction Prispevek obravnava karakterizacijo treh orodij v obliki `aginih listov izdelanih iz razli~nih komercialnih WC-Co karbidnih trdin, od katerih ena skupina pogostej{e poka. Ker so lastnosti teh materialov v veliki meri odvisne od mikrostrukture, je bila podrobna karakterizacija izvedena z uporabo vrsti~ne elektronske mikroskopije (SEM), energijsko disperzijske rentgenske spektroskopije (EDS) in rentgenske difrakcije (XRD). Analiza SEM slik je vklju~evala dolo~anje vsebnosti veziva in povpre~ne povr{ine WC zrn. Z rentgenskim fluorescen~nim (XRF) analizatorjem smo dolo~ili tudi povpre~no kemi~no sestavo teh materialov. Rezultati ka`ejo, da je kljub enaki vsebnosti elementov, ki tvorijo vezivo, v vseh treh materialih pogosteje pokala karbidna trdina, ki je vsebovala manj{i dele` veziva in manj{o povpre~no povr{ino WC zrn. Znano je, da oba parametra zmanj{ujeta `ilavost tak{nega materiala. Klju~ne besede: mikrostruktura, WC-Co karbidna trdina, vrsti~na elektronska mikroskopija, rentgenska difrakcija 1 INTRODUCTION WC-Co cemented carbides are an important tool ma- terial that has been known for nearly 100 years. These materials are widely used in manufacturing industry due to their excellent combination of hardness, toughness and thermal conductivity. 1,2 WC-Co cemented carbides consist of a hard hexagonal WC phase with high thermal conductivity and a face-centred cubic (fcc) cobalt binder phase. The binder phase is the minority phase in these materials. Cobalt is the most used binder metal in ce- mented carbides due to its excellent wettability of WC and its mechanical properties. In addition to cobalt, nickel, iron and their combinations are also used as bind- ers in various proportions in certain applications. 1,3–5 Pure cobalt exhibits a hexagonal structure at tempera- tures up to 417 °C, while the fcc structure is stable above that. After sintering is complete, the binder cools and the fcc structure is retained. This can be attributed to the dis- solution of tungsten in cobalt and the higher thermal ex- pansion of cobalt compared to WC. After cooling, the binder phase remains loaded with tensile stress. 1 Al- loying elements such as Ni, which has infinite solid solubility with Co, also stabilises the fcc structure. 6 The microstructure of cemented carbides is of great importance as it affects the properties of these materials. Therefore, by adjusting the binder content and the WC grain size, the mechanical properties such as toughness, hardness and thermal conductivity can be adjusted. 1,2,7 It has been reported that WC-Co cemented-carbide wear increases with increasing cobalt content and WC grain size. 8 The microstructures of these materials have been described by many authors with different quantities and relationships between them. These quantities are WC grain size, carbide contiguity, volume fraction of the binder, and binder mean free path, and many of them are more-or-less obviously related to each other. 1 Carbide contiguity is a quantity that measures the amount of WC /WC contacts. Various models relate it to the volume fraction of the binder and the WC grain size. In contrast, the binder mean free path is related to the average WC grain size and the volume fraction of the binder, which in turn is proportional to the cobalt content. Consequently, most of the properties of these materials can be de- scribed by simply relating them to the cobalt or binder content and the WC grain size. The correlations between Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428 423 UDK 549.2:669.018.25 ISSN 1580-2949 Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 56(4)423(2022) *Corresponding author's e-mail: iztok.naglic@ntf.uni-lj.si (Iztok Nagli~) these two quantities and the main properties of these ma- terials are well established and can be explained by sim- ple correlations. A higher thermal conductivity is ob- tained by a larger WC grain size and a low binder content, while a high hardness is obtained by a small WC grain size and a low binder content. On the other hand, high toughness can be achieved by a large WC grain size and a high binder content, while high edge toughness is achieved by a small WC grain size and a high binder content. 1,2,7 Several elements were found to inhibit the grain growth of WC grains during the sintering process. These elements are vanadium, titanium, chromium, tantalum, molybdenum and niobium. In addition to the grain size, these inhibitors also affect the shape of the grains. 9,10 It has been reported that the grain-growth inhibitors act by forming a thin cubic (M,W)C layer on the surface of the WC grains, which lowers the interfacial energy and acts as a kinetic barrier. 3,11–13 Chromium additions have also been found to increase the corrosion resistance of WC-Co cemented carbides. 14 Carbon content must be controlled during the manu- facturing process of cemented carbides. A low carbon content leads to the formation of -phase, which are ter- nary carbides composed of carbon, cobalt and tungsten. The M 6 C carbides can range from Co 3 W 3 CtoCo 2 W 4 C, while M 12 C has a fixed composition of Co 6 W 6 C. The presence of relatively large amounts of the brittle -phase leads to a degradation in the mechanical proper- ties of these carbides. 1,2,15 First-principles calculations of the elastic modulus and hardness of -phases revealed that they have lower elastic moduli than WC and are sig- nificantly softer than WC and W 2 C. 16 On the other hand, a high carbon content during the manufacturing process leads to the formation of graphite precipitates, which de- grade the mechanical properties. Three different commercial WC-Co cemented-car- bide tool materials were used as saw blades. Cracking of tools occurred more frequently in the saw blades made from one group of materials than in the other two. It is well known that the microstructure of these materials strongly affects the mechanical properties. Therefore, in this work we performed the characterization of the microstructure of these three different commercially available WC-Co carbide materials to find the most probable cause of this frequent cracking. 2 EXPERIMENTAL PART We characterized the microstructure of three different commercial WC-Co cemented-carbide tool materials, shaped as saw blades with a diameter of 35 mm and a thickness of 0.4 mm. These different materials were des- ignated as sample 1, sample 2, and sample 3. Cracking occurred more frequently in the tools made from the sample 3 material than in the tools made from samples 1 and 2. Smaller pieces of the tools were mounted, ground, and polished. Final polishing was performed with 0.25-μm diamond paste. The average chemical composition of the tool materi- als was determined using the Thermo Scientific Niton XL3t GOLDD+ X-ray fluorescence (XRF) analyser. Scanning electron microscopy (SEM) at (8, 15 and 20) kV was performed using ThermoFisher Scientific Quattro S equipped with Oxford Ultim Max EDS detec- tors for energy dispersive X-ray spectroscopy (EDS) and JEOL JSM 7600F equipped with Oxford Instruments INCA Microanalysis Suite and X-Max 20 SDD-EDS de- tector. X-ray diffraction (XRD) was used to characterise the phases present in these materials. X-ray diffraction patterns were recorded with a PANalytical X’Pert PRO diffractometer using non-monochromatic X-rays gener- ated by Empyrean Cu anode tube. The microstructural analysis also included the deter- mination of the volume fraction of a binder in the sam- ples. The content of the binder phase was determined us- I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL 424 Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428 Figure 1: Backscattered-electron image of the microstructure of sample 2 for the determination of the binder content and the same image with red-coloured detected binder phase ing backscattered-electron images taken with SEM at 8 kV, showing a large contrast between the light WC grains and the dark binder phase. The volume fraction was determined according to ASTM 1245-03 (2016) by image analysis of six to twelve backscattered-electron images using ImageJ 1.53f51 software. Figure 1 shows the determination of the binder content of sample 2. The detected binder phase is coloured red. To determine the size of WC grains, which also af- fects the mechanical properties of WC-Co cemented-car- bide materials, the mean grain area was measured on backscattered-electron images taken at 15 kV using SEM. The mean grain area was determined according to ASTM E1382-97 by the grain count (planimetric) method using AxioVision software on six images of the microstructure of each sample. The grain area of a region with many grains was outlined and measured. This grain area was then reduced by the volume fraction of the binder phase and divided by the number of grains. Fig- ure 2 presents an example of the measurement of the mean grain area in samples 2 and 3 using the grain count (planimetric) method. 3 RESULTS AND DISCUSSION The chemical composition, excluding the carbon con- tent, of all three samples was determined by XRF. These results are presented in Table 1. The results show that all three samples contain tungsten, cobalt and chromium, while sample 3 also contains some nickel. These results indicate that the binder phase of samples 1 and 2 is mainly cobalt with traces of chromium, while the binder of sample 3 contains some nickel. Chromium is normally added as Cr 3 C 2 . Metallic chromium on the other hand, is soluble in cobalt. The chromium content in the samples is rather low. Therefore, if it is present in solid solution, it cannot significantly affect the mechanical properties by changing the content of the binder phase in these three samples. Nickel exhibits infinite solid solubility in cobalt. 6 Since the content of binder significantly affects the mechanical properties, it is crucial to determine the amount of binder in the studied samples. A comparison of the cobalt content shows that it is the highest in sam- ple 2 and the lowest in sample 3, which was the most prone to cracking. Consequently, the content of binder-forming elements (cobalt and nickel) in sample 3 is 3.79 w/%, which is between samples 1 and 2. This is inconsistent with the fact that sample 3 was the one that cracked more frequently. We therefore performed an ad- ditional analysis to determine the volume fraction of the binder phase and the mean grain area via SEM images. BE images of the microstructure of all three carbide tool materials are shown in Figure 3. The light phase represents the WC phase, while the dark area between the WC crystals is the cobalt-based binder phase. A com- parison of the microstructures shows that the size of the WC crystals is smaller and the binder phase content is lowest in sample 3. The mechanical properties of such materials are affected by these two parameters. These two observations were analysed in more detail by deter- mining the average volume fraction of the binder phase and the mean grain area of the WC phase using backscat- tered-electron images of the microstructure. The contents of the binder phase determined by the analysis of the backscattered-electron images are listed in Table 2. I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428 425 Table 1: Chemical composition of the tool materials, determined with XRF (w/%) W Co Cr Ni Co+Ni Sample 1 96.34 ± 0.44 3.58 ± 0.08 0.09 ± 0.09 – – Sample 2 96.00 ± 0.44 3.84 ± 0.09 0.16 ± 0.06 – – Sample 3 95.81 ± 0.46 2.90 ± 0.08 0.40 ± 0.07 0.89 ± 0.06 3.79 ± 0.14 Figure 2: Backscattered-electron image of the microstructure with ex- ample for the determination of the mean grain area by grain count (planimetric) method: a) sample 2, b) sample 3 The results in Table 2 show that the average volume fractions of the binder phase V V in samples 1 and 2 are 0.0616 and 0.0600, respectively, while in sample 3 it is only 0.0433. Considering the standard deviation (s), the 95 % confidence interval (95 % CI), and the relative ac- curacy (% RA), it can be observed that the interval of V V ± 95 % CI of sample 3 does not overlap with samples 1 and 2, which means that the average volume fraction of binder in sample 3 is lower than in samples 1 and 2. This finding contradicts the previous result of the XRF analy- sis, which indicates that the content of the binder-form- ing elements (cobalt and nickel) is similar to the cobalt content in samples 1 and 2. The lower content of binder in sample 3 ultimately leads to a lower toughness of this material and more frequent cracking. Table 3 shows the results of the evaluation of the mean grain area of WC crystals. The mean grain area re- sults confirm the observation of the microstructures in Figure 1, where the mean grain area is lowest for sample 3 with 0.139 μm 2 , while these values are 0.246 μm 2 and 0.262 μm 2 for samples 1 and 2, respectively. Considering s, 95 % CI and % RA, it can be observed that the interval I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL 426 Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428 Figure 3: Backscattered-electron images of the microstructure of: a and b) sample 1, c and d) sample 2, e and f) sample 3 Table 2: Analysis of the backscattered-electron images Sample 1 Sample 2 Sample 3 Average volume fraction of binder (VV) 0.0616 0.0600 0.0433 Standard deviation (s) 0.0067 0.0086 0.0061 95% confidence interval (95 % CI) 0.0062 0.0091 0.0039 Relative accuracy (% RA) 10.0 15.1 8.9 of A ± 95 % CI of sample 3 does not overlap with the in- tervals of samples 1 and 2, which means that the mean grain area of the WC crystals in sample 3 is smaller than in samples 1 and 2. This result is consistent with the re- sult of the XRF analysis that the chromium content in sample 3 is significantly higher. Chromium is known to inhibit the grain growth and thus affects the final size of the WC grains. Nevertheless, it was not clear what role the nickel in sample 3 played or whether it had incorpo- rated into the Co-based fcc binder phase and thus con- tributed to the total amount of binder. Table 3: Results of the evaluation of the mean grain area of WC crys- tals Sample 1 Sample 2 Sample 3 Mean grain area of WC A (μm 2 ) 0.246 0.262 0.139 Standard deviation s (μm 2 ) 0.036 0.034 0.013 95% confidence interval (95%CI(μm 2 )) 0.038 0.036 0.013 Relative accuracy (% RA) 15.6 13.6 9.7 Interval of mean grain area (A±95%CI(μm 2 )) 0.208 – 0.284 0.226 – 0.298 0.125 – 0.152 We therefore also performed EDS analyses of the binder phase. The analyses were performed on the larg- est area of binder in the microstructure of each sample. The EDS analyses of the binder phase are listed in Ta- ble 4. A low oxygen content is present only in sample 1, likely due to contamination. Carbon is present in all three samples, due in part to impurities and detection of the main phase in these samples, the WC phase. The tungsten content was the highest of all the elements de- tected in all three samples. Chromium, known to be an inhibitor of grain growth, is also present in all three sam- ples, although its content is significantly higher in sam- ple 3. This result is consistent with the XRF analysis pre- sented in Table 1 and the finding that the WC crystals are the smallest in this sample. Consequently, the binder phase in samples 1 and 2 consists mainly of cobalt, while sample 3 also contains a significant amount of nickel, as XRF analysis also shows. Assuming that only cobalt and nickel are present in the binder phase, the nickel content would be 22.6 w/%. To resolve our doubts, we also performed detailed XRD analyses to reveal the actual phases present in the microstructure of the three WC-Co cemented-carbide tools. The XRD patterns of all three samples are shown in Figure 4. The XRD patterns show that all three sam- ples contain more-or-less the same phases, among which the hexagonal WC dominates (COD 9007456). The binder phase Co (COD 9012949) with an fcc structure is also present. It is worth noting that the intensity of the two peaks for the binder phase Co in sample 3 is signifi- cantly lower than in the samples 1 and 2. This result in- dicates that sample 3 contains less binder phase than the other two samples and is consistent with the result ob- tained by determining the volume fraction of binder phase using backscattered-electron image analysis. In addition to the two previously mentioned phases, Co 6 W 6 C (ICCD 00-022-0597), Co 3 W 3 C (ICDD 00-006-0639), Cr 7 C 3 (COD 1009019) and Cr 3 C 2 (COD 9009906) are also present in all three samples. The content of cobalt in samples 1 and 2 and of co- balt and nickel in sample 3 is similar, as shown in Ta- ble 1. On the other hand, the content of the binder, based on image analysis and XRD results, is clearly lower in I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428 427 Figure 4: XRD patterns of samples 1, 2 and 3: a) unmagnified, b) magnified Table 4: EDS compositional analysis of a binder phase at 20 kV (w/%) C O Cr Co Ni W Sample 1 9.2 ± 0.5 2.0 ± 0.2 0.8± 0.1 42.4 ± 0.4 – 45.6 ± 0.5 Sample 2 20.0 ± 0.7 – 0.7 ± 0.1 31.3 ± 0.5 – 48.0 ± 0.6 Sample 3 8.9 ± 0.6 – 1.6 ± 0.1 18.5 ± 0.3 5.4 ± 0.2 65.6 ± 0.5 sample 3. Consequently, an additional phase containing binder elements should have formed in sample 3. Com- paring the XRD patterns of all three samples, it is notice- able that a small peak is present only in sample 3 with a 2 angle between 41° and 42°. This peak corresponds to the position of the most intense peak of the -phases Ni 2 W 4 C, Co 2 W 4 C (ICCD 98-009-0811) or their mixed carbide (Ni,Co) 2 W 4 C or simply M 2 W 4 C. 17 The formation of this additional -phase could explain the lower con- tent of the fcc Co binder phase and consequently, the lower toughness of such a material. One of the possible reasons for the formation of the additional -phase could be related to the lower carbon content in the production process of WC-Co cemented-carbide tools. 4 CONCLUSIONS Three different commercial WC-Co cemented-car- bide tool materials, used as saw blades were character- ized in this work, as one group exhibited a more frequent cracking. The results of our work show that the binder phase in sample 3 contains a significant amount of nickel in addi- tion to cobalt, while the other two samples do not. The results also show that despite having the same content of binder-forming elements in all three different commer- cial WC-Co carbide tools, the content of the binder phase in sample 3 is significantly lower. In addition, the size of the WC grains is also smaller in this sample. Both the binder content and the size of the WC grains lead to a decrease in the toughness of such a material, which re- sults in the more frequent cracking of such tools. The re- sults of this work also suggest that the cause of the lower content of the binder phase in sample 3 could be the for- mation of an additional -phase of type M 2 W 4 C, which reduces the content of the fcc binder phase and thus the toughness. Acknowledgements The authors would like to acknowledge the financial support of the Slovenian Research Agency through the research programme P1-0195. 5 REFERENCES 1 J. García, V. Collado Ciprés, A. Blomqvist, B. Kaplan, Cemented carbide microstructures: a review, International Journal of Refractory Metals & Hard Materials, 80 (2019), 40–68, doi:10.1016/j.ijrmhm. 2018.12.004 2 A. T. Santhanam, Metallography of Cemented Carbides, Metallo- graphy and Microstructures, Vol. 9, ASM Handbook, ASM Interna- tional, 2004, 1067–1078 3 S. Norgren, J. García, A. Blomqvist, L. Yin, Trends in the P/M hard metal industry, International Journal of Refractory Metals and Hard Materials, 48 (2015), 31–45, doi:10.1016/j.ijrmhm.2014.07.007 4 G. Gille, J. Bredthauer, B. Gries, B. Mende, W. Heinrich, Advanced and new grades of WC and binder powder – their properties and ap- plication, International Journal of Refractory Metals and Hard Mate- rials, 18 (2000), 87–102, doi:10.1016/S0263-4368(00)00002-0 5 B. Wittmann, W.-D. Schubert, B. Lux, WC grain growth and grain growth inhibition in nickel and iron binder hardmetals, International Journal of Refractory Metals and Hard Materials 20 (2002), 51–60, doi:10.1016/S0263-4368(01)00070-1 6 H. Okamoto, Desk Handbook: Phase Diagrams for Binary Alloys, ASM International, 2000, 253 7 J. M. Densley, J. P. Hirth, Fracture toughness of a nanoscale WC-Co tool steel, Scripta Materialia, 38 (1998), 239–244, doi:10.1016/ S1359-6462(97)00435-1 8 H. Saito, A. Iwabuchi, T. Shimizu, Effects of Co content and WC grain size on wear of WC cemented carbide, Wear, 261 (2006), 126–132, doi:10.1016/j.wear.2005.09.034 9 A. V. Shatov, S. A. Firstov, I. V. Shatova, The shape of WC crystals in cemented carbides, Materials Science and Engineering A, 242 (1998), 7–14, doi:10.1016/S0921-5093(97)00509-1 10 A. Delano¸, M. Bacia, E. Pauty, S. Lay, C. H. Allibert, Cr-rich layer at the WC/Co interface in Cr-doped WC–Co cermets: segregation or metastable carbide?, Journal of Crystal Growth, 270 (2004), 219–227, doi:10.1016/j.jcrysgro.2004.05.101 11 S. Lay, J. Thibault, S. Hamar-Thibault, Structure and role of the in- terfacial layers in VC-rich WC-Co cermets, Philosophical Magazine, 83 (2003), 1175–1190, doi:10.1080/1478643031000075759 12 S. A. E. Johansson, G. Wahnstr?m, A computational study of thin cubic carbide films in WC/Co interfaces, Acta Materialia, 59 (2011), 171–181, doi:10.1016/j.actamat.2010.09.021 13 J. Weidow, H.-O. Andrén, Grain and phase boundary segregation in WC–Co with small V, Cr or Mn additions, Acta Materialia, 58 (2010), 3888–3894, doi:10.1016/j.actamat.2010.03.038 14 G. Li, Y. Peng, L. Yan, T. Xu, J. Long, F. Luo, Effects of Cr concen- tration on the microstructure and properties of WC-Ni cemented car- bides, Journal of Materials Research and Technology, 9 (2020), 902–907, doi:10.1016/j.jmrt.2019.11.030 15 H. L. de Villiers Lovelock, Powder/processing/structure relationships in WC-Co thermal spray coatings: A review of the published litera- ture, Journal of Thermal Spray Technology, 7 (1998), 357–373, doi:10.1361/105996398770350846 16 J. Kim, Y. Jae Suh, I. Kang, First-principles calculations of the phase stability and the elastic and mechanical properties of -phases in the WC-Co system, Journal of Alloys and Compounds, 656 (2016), 213–217, doi:10.1016/j.jallcom.2015.09.214 17 V. B. Voitovich, V. V. Sverdel, R. F. Voitovich, E. I. Golovko, Oxida- tion of WC-Co, WC-Ni and WC-Co-Ni hard metals in the tempera- ture range 500–800 °C, International Journal of Refractory Metals and Hard Materials, 14 (1996), 289–295, doi:10.1016/0263- 4368(96)00009-1 I. NAGLI^ et al.: CHARACTERIZATION OF DIFFERENT WC-Co CEMENTED-CARBIDE TOOL 428 Materiali in tehnologije / Materials and technology 56 (2022) 4, 423–428