materials

Article

In Situ Formation of TiB2 in Fe-B System withTitanium Addition and Its Influence on PhaseComposition, Sintering Process andMechanical Properties

Mateusz Skałon 1,2, Marek Hebda 2 , Benedikt Schrode 3 , Roland Resel 3 , Jan Kazior 2 andChristof Sommitsch 1,*

1 IMAT Institute of Materials Science, Joining and Forming, Graz University of Technology, Kopernikusgasse24/1, 8010 Graz, Austria; mateusz.skalon@tugraz.at

2 Institute of Materials Engineering, Cracow University of Technology, 24 Warszawska ave, 31-155 Cracow,Poland; mhebda@pk.edu.pl (M.H.); kazior@mech.pk.edu.pl (J.K.)

3 Institute of Solid State Physics, Graz University of Technology, Petersgasse 16/II, 8010 Graz, Austria;b.schrode@tugraz.at (B.S.); Roland.resel@tugraz.at (R.R.)

* Correspondence: christof.sommitsch@tugraz.at; Tel.: +43-316-873-7180

Received: 20 July 2019; Accepted: 11 December 2019; Published: 13 December 2019�����������������

Abstract: Interaction of iron and boron at elevated temperatures results in the formation of an E(Fe + Fe2B) eutectic phase that plays a great role in enhancing mass transport phenomena duringthermal annealing and therefore in the densification of sintered compacts. When cooled down, thisphase solidifies as interconnected hard and brittle material consisting of a continuous network ofFe2B borides formed at the grain boundaries. To increase ductile behaviour, a change in precipitates’stoichiometry was investigated by partially replacing iron borides by titanium borides. The powderof elemental titanium was introduced to blend of iron and boron powders in order to induce TiB2

in situ formation. Titanium addition influence on microstructure, phase composition, density andmechanical properties was investigated. The observations were supported with thermodynamiccalculations. The change in phase composition was analysed by means of dilatometry and X-raydiffraction (XRD) coupled with thermodynamic calculations.

Keywords: titanium diboride; titanium; boron; iron; dilatometry; liquid phase; phase composition; sintering

1. Introduction

Due to a stable microstructure and mechanical properties under exposure to long-term thermalneutron irradiation, borated stainless steels find extensive applications in the nuclear industry [1,2].Potential applications of boron steels in nuclear industry are, e.g., control/shutoff rods in nuclearreactors, sensor for neutron counting, shapes for neutron shielding, dry transportation casks, spent fuelrod storage racks. The advantage of borated stainless steel produced by powder metallurgy technologyin comparison to the conventionally produced cast/wrought products is that they contain much smallerand more uniformly distributed boride particles [1,2]. Moreover, it is possible to obtain a materialwhose properties have lesser decrease in ductility and impact toughness as the boron content increasescompared with similar boron-containing cast/wrought borated stainless steels. The application ofboron in powder metallurgy as addition to ferrous alloys has been of interest to numerous researchersover the last years [3–12]. The main cause of this interest is the eutectic reaction between iron and boron,resulting in the appearance of a liquid phase which efficiently improves mass transport by intensifyinga diffusion process [7]. Even small amounts of boron (0.4–0.6 wt %) added to ferrous powder may result

Materials 2019, 12, 4188; doi:10.3390/ma12244188 www.mdpi.com/journal/materials

Materials 2019, 12, 4188 2 of 12

in great densification of sintered compacts reaching almost full relative density [7,9]. In accordancewith a series of experiments [9,13–16], the sintering process is activated through the presence of a liquidphase by acting in three stages: (i) rearrangement (liquid spreading), (ii) solution-reprecipitation ofa base material and (iii) microstructural coarsening [9]. Unfortunately, boron is almost non-soluble inan iron-based solid matrix, so it remains at the grain boundaries after the sintering process in the formof hard and brittle borides creating an almost continuous network surrounding the individual grains,therefore significantly influencing the material properties. Controlling the borides’ morphology and/orcrystallographic network could considerably broaden the application field of steel parts manufacturedby Powder Metallurgy (PM). It was already proved that the addition of titanium to iron-based castalloys may result in the formation of titanium diboride (TiB2) [17,18]. Such an interaction, however,was not yet investigated in sintered particular materials. The objective of the present paper wasto investigate the influence of titanium addition on phase changes and sintering behaviour in theFe-B system. This topic is of high importance due to the possibility of improving the ductility of theparticulate borated steels.

2. Materials and Methods

Water atomised iron powder of >99% purity delivered by Sigma Aldrich (St. Louis, MO, USA)was used as a base powder. Boron of >99.7% purity (grain size < 1 µm) delivered by Sigma Aldrichwas added to the blend in an amount of 0.8 wt % in order to induce a eutectic liquid appearanceduring the sintering process and consequently the appearance of Fe2B borides. The titanium of 99.9%purity delivered by VWR company (Radnor, PA, USA) was also added to the blend as a powder ina stepwise increasing amount in two different grain size classes as listed in Table 1. The varying grainsize distribution of titanium allowed for control of the initial reaction surface between titanium and theeutectic liquid. Blends were mixed using a Turbula-type mixer for 12 h in order to assure homogeneityof blends.

Table 1. Composition of utilised powder blends.

Sample Description Titanium Grain Size Class/µm Titanium/wt % Boron/wt % Iron/wt %

REF 0 0.00

0.80 balance

A63 45–63 0.47B63 45–63 0.93C63 45–63 1.40

A140 100–140 0.47B140 100–140 0.93C140 100–140 1.40

The mixed powders were single-axially compacted under a pressure of 600 MPa informs of (a)4 mm × 4 mm × 20 mm cuboid (dilatometric samples); (b) a 5 mm × 20 mm cylinder (microstructureand density check) and (c) a cuboid 6 mm × 12 mm × 30 mm for Transverse Rupture Strength test(TRS). During the consolidation of the samples, no lubricant was used. Dilatometric tests were carriedout in a Netzsch DIL 402C dilatometer (NETZSCH-Gerätebau GmbH, Selb, Germany) under hydrogenatmosphere of 99.9999% purity and a flow rate of 100 mL/min. The heating rate was equal to 20 ◦C/min,the cooling rate was 10 ◦C/min while the isothermal temperature and time were 1240 ◦C and 30 min,respectively. Nabertherm P330 tubular (Nabertherm, Lilienthal, Germany) furnace was used forsintering both the TRS and the cylindrical samples under the same conditions as dilatometric samples.Material porosity was checked using the standard water displacement method. Scanning ElectronMicroscope (SEM, TESCAN Brno s.r.o., Brno, Czech Republic) TESCAN Mira3-SEM equipped withan Energy Dispersive X-Ray (EDX) spectrometer was used to evaluate the localisation of titanium inthe microstructure. Specular X-ray diffraction patterns were collected on a PANalytical Empyreandiffractometer (Malvern Panalytical, Herrenberg, Germany) using a wavelength of 0.154 nm. On theprimary side, the machine was equipped with a 1/8◦ receiving slit, a 10 mm beam mask and a multilayer

Materials 2019, 12, 4188 3 of 12

mirror for monochromatisation and formation of a parallel beam. On the secondary side, an 8 mmanti-scatter slit, 0.02 Rad Soller slits and a PANalytical PIXcel 3D detector (Malvern Panalytical,Herrenberg, Germany) in scanning line mode were used. All thermodynamic calculations wereperformed using Thermo-Calc software v.3.0 (Thermo-Calc Software, Solna, Sweden) with the TCFE6database appended. State of full equilibrium was assumed.

3. Results and Discussion

3.1. Thermodynamic Calculations

Figure 1a presents the calculated phase diagram in the Fe-B-Ti system along with the temperatureincrease for various amounts of titanium addition. The calculations indicate that an increasing amountof titanium results in disappearance of Fe2B boride which gets replaced by TiB2 (Figure 1c). According toliterature, the reason for this replacement is the relatively high difference in Gibbs free energy offormation between the two borides reaching 150 kJ/mol throughout the whole temperature rangebetween 400–1800 ◦C [18]. Such a replacement should be also accompanied by a decreasing amount ofa eutectic liquid at sintering temperature of 1240 ◦C as presented in Figure 1b.

Materials 2020, 13, x FOR PEER REVIEW 3 of 13

detector (Malvern Panalytical, Herrenberg, Germany) in scanning line mode were used. All 85 thermodynamic calculations were performed using Thermo-Calc software v.3.0 (Thermo-Calc 86 Software, Solna, Sweden) with the TCFE6 database appended. State of full equilibrium was assumed. 87

3. Results and Discussion 88

3.1. Thermodynamic Calculations 89

Figure 1a presents the calculated phase diagram in the Fe-B-Ti system along with the 90 temperature increase for various amounts of titanium addition. The calculations indicate that an 91 increasing amount of titanium results in disappearance of Fe2B boride which gets replaced by TiB2 92 (Figure 1c). According to literature, the reason for this replacement is the relatively high difference in 93 Gibbs free energy of formation between the two borides reaching 150 kJ/mol throughout the whole 94 temperature range between 400–1800 °C [18]. Such a replacement should be also accompanied by a 95 decreasing amount of a eutectic liquid at sintering temperature of 1240 °C as presented in Figure 1b. 96

97

Figure 1. (a) Calculated phase diagram of the system (Iron + 0.8B wt %) + Ti, where: α, Ferrite; γ, 98 Austenite; δ, Delta Ferrite; L, Liquid; (b) calculated phase fraction of a liquid phase at the sintering 99 temperature of 1240 °C; (c) calculated mole fraction of borides as a function of titanium addition at 100 the room temperature. 101

Figure 1b presents the mole fraction of a liquid phase for the constant sintering temperature of 102 1240 °C, showing a decreasing amount of liquid phase with an increasing titanium addition. This is 103 connected to changes in borides composition due to the replacement of Fe2B by TiB2 borides (Figure 104 1c). As a result of the difference in molar volume (Fe2B: 16.76 mol/cm3, TiB2: 15.37 mol/cm3) and the 105 difference in metal-boron stoichiometry (each mole of TiB2 contains two times more boron than a 106 mole of Fe2B), titanium diboride is capable of binding the same amount of boron in lesser volume as 107 compared to Fe2B (Figure 1c). Moreover, due to the high melting point of TiB2 it is solid in the 108 sintering temperature; therefore, calculations indicate the decreasing amount of a liquid phase in 109 Figure 1b. For the experimental studies, the compositions of blends (Table 1) were selected based on 110 the presented thermodynamic calculations to assure a considerable but linear decrease of the eutectic 111 liquid amount at the sintering temperature (Figure 1a,b). 112

3.2. Dilatometric Analysis and Density Changes 113

To study the influence of the particle size of the additional titanium on the dilatometric 114 behaviour, length changes of the investigated samples were recorded over a broad temperature range 115 (Figure 2a). Figure 2b,c shows the results for the different particle sizes. Quantification of the 116

Figure 1. (a) Calculated phase diagram of the system (Iron + 0.8B wt %) + Ti, where: α, Ferrite; γ,Austenite; δ, Delta Ferrite; L, Liquid; (b) calculated phase fraction of a liquid phase at the sinteringtemperature of 1240 ◦C; (c) calculated mole fraction of borides as a function of titanium addition at theroom temperature.

Figure 1b presents the mole fraction of a liquid phase for the constant sintering temperature of1240 ◦C, showing a decreasing amount of liquid phase with an increasing titanium addition. This isconnected to changes in borides composition due to the replacement of Fe2B by TiB2 borides (Figure 1c).As a result of the difference in molar volume (Fe2B: 16.76 mol/cm3, TiB2: 15.37 mol/cm3) and thedifference in metal-boron stoichiometry (each mole of TiB2 contains two times more boron than a moleof Fe2B), titanium diboride is capable of binding the same amount of boron in lesser volume ascompared to Fe2B (Figure 1c). Moreover, due to the high melting point of TiB2 it is solid in the sinteringtemperature; therefore, calculations indicate the decreasing amount of a liquid phase in Figure 1b.For the experimental studies, the compositions of blends (Table 1) were selected based on the presentedthermodynamic calculations to assure a considerable but linear decrease of the eutectic liquid amountat the sintering temperature (Figure 1a,b).

Materials 2019, 12, 4188 4 of 12

3.2. Dilatometric Analysis and Density Changes

To study the influence of the particle size of the additional titanium on the dilatometric behaviour,length changes of the investigated samples were recorded over a broad temperature range (Figure 2a).Figure 2b,c shows the results for the different particle sizes. Quantification of the dilatometric changeswas performed by calculating the length changes in specific regions (measurement scheme and detailedvalues in supplementary). Independently of the titanium particle size, the measurements showed rapidswelling of the sample at the temperature of 478 ± 2.3 ◦C and ends up at the temperature of 509 ± 1.9 ◦C.It is then followed by a rapid shrinkage starting at 646 ± 0.3 ◦C. The former reflects the temporaryformation of titanium hydride (TiHx) due to interaction of elemental titanium and hydrogen derived fromthe atmosphere. With an increase of titanium content, the swelling effect was more pronounced. On theother hand, the shrinkage observed at around 646 ± 0.3 ◦C reflects the intensive release of hydrogen dueto thermally induced dehydrogenation, i.e., decomposition of TiHx. This decomposition is finished beforethe temperature of 905 ◦C is reached because TiHx becomes unstable below this temperature [19,20].This also means that above 905 ◦C, titanium is available in elemental form. These observations stay ina good agreement with previous studies [19–21]. The swelling of titanium particles originated from itshydrogenation can introduce large tensile stress on neighbouring iron particles and therefore can bea potential source of an additional porosity in the sintered compact. After the dehydrogenation process,the transition of the base iron powder from the crystallographic form of BCC to FCC starts at 905 ± 4.4 ◦Cand ends at 934 ± 0.3 ◦C (Figure 2b,c).

Materials 2020, 13, x FOR PEER REVIEW 4 of 13

dilatometric changes was performed by calculating the length changes in specific regions 117 (measurement scheme and detailed values in supplementary). Independently of the titanium particle 118 size, the measurements showed rapid swelling of the sample at the temperature of 478 ± 2.3 °C and 119 ends up at the temperature of 509 ± 1.9 °C. It is then followed by a rapid shrinkage starting at 646 ± 120 0.3 °C. The former reflects the temporary formation of titanium hydride (TiHx) due to interaction of 121 elemental titanium and hydrogen derived from the atmosphere. With an increase of titanium content, 122 the swelling effect was more pronounced. On the other hand, the shrinkage observed at around 646 123 ± 0.3 °C reflects the intensive release of hydrogen due to thermally induced dehydrogenation, i.e., 124 decomposition of TiHx. This decomposition is finished before the temperature of 905 °C is reached 125 because TiHx becomes unstable below this temperature [19,20]. This also means that above 905 °C, 126 titanium is available in elemental form. These observations stay in a good agreement with previous 127 studies [19–21]. The swelling of titanium particles originated from its hydrogenation can introduce 128 large tensile stress on neighbouring iron particles and therefore can be a potential source of an 129 additional porosity in the sintered compact. After the dehydrogenation process, the transition of the 130 base iron powder from the crystallographic form of BCC to FCC starts at 905 ± 4.4 °C and ends at 934 131 ± 0.3 °C (Figure 2b,c). 132

133

Figure 2. (a) Temperature profile of dilatometric tests; (b) dilatometric curves of samples A63, B63 134 and C63; (c) dilatometric curves of samples A140, B140 and C140. 135

Figure 2. (a) Temperature profile of dilatometric tests; (b) dilatometric curves of samples A63, B63 andC63; (c) dilatometric curves of samples A140, B140 and C140.

Materials 2019, 12, 4188 5 of 12

Consequent heating results in the eutectic reaction between iron and boron starting at a temperaturearound 1172 ◦C, which is in good agreement with previous studies [8,10]. The presence of the eutecticliquid enhances mass transport causing densification of the material. As a consequence, shrinkingbecomes a more dominant effect than thermal expansion. This is observed as a length reduction of thesample in the dilatometric curves. This effect is presented in Figure 3a,b in detail. As can be easily seenin Figure 3a,b the more titanium was introduced the higher was swelling at temperatures of eutecticliquid occurrence. This shows that the more titanium was added the less effectively mass transportoccurred at a temperature above 1172 ◦C. Moreover, the total dimensional changes (Figure 3c) showthe same trend. The dilatometric results suggest that fine titanium particles reacted more effectivelywith the eutectic liquid rather than coarse ones.

Materials 2020, 13, x FOR PEER REVIEW 5 of 13

Consequent heating results in the eutectic reaction between iron and boron starting at a 136 temperature around 1172 °C, which is in good agreement with previous studies [8,10]. The presence 137 of the eutectic liquid enhances mass transport causing densification of the material. As a consequence, 138 shrinking becomes a more dominant effect than thermal expansion. This is observed as a length 139 reduction of the sample in the dilatometric curves. This effect is presented in Figure 3a,b in detail. As 140 can be easily seen in Figure 3a,b the more titanium was introduced the higher was swelling at 141 temperatures of eutectic liquid occurrence. This shows that the more titanium was added the less 142 effectively mass transport occurred at a temperature above 1172 °C. Moreover, the total dimensional 143 changes (Figure 3c) show the same trend. The dilatometric results suggest that fine titanium particles 144 reacted more effectively with the eutectic liquid rather than coarse ones. 145

146

Figure 3. Dilatometric changes: (a) in the region between the occurrence of the eutectic phase and 147 isothermal step; (b) at an isothermal step and (c) total changes. 148

Such an observation indicates a decreasing amount of the liquid phase along with the increasing 149 addition of the titanium. The smaller was the particle size of titanium the stronger was its influence 150 dilatometric behaviour. As presented in Figure 3c, use of high additions of titanium of fine particles 151 (C63) may lead even to swelling in respect to the compacted sample. This may affect negatively the 152 final relative density of the sintered compacts. The cooling part of dilatometric curves (Figure 2a–c) 153 shows no evidence of elemental titanium presence as it was observed in the heating stage. The only 154 visible effects are a solidification of borides at 1147 ± 2.1 °C and transformation of austenite back to 155 ferrite starting at 893 ± 2.4 °C. 156

Figure 3. Dilatometric changes: (a) in the region between the occurrence of the eutectic phase andisothermal step; (b) at an isothermal step and (c) total changes.

Such an observation indicates a decreasing amount of the liquid phase along with the increasingaddition of the titanium. The smaller was the particle size of titanium the stronger was its influencedilatometric behaviour. As presented in Figure 3c, use of high additions of titanium of fine particles(C63) may lead even to swelling in respect to the compacted sample. This may affect negatively thefinal relative density of the sintered compacts. The cooling part of dilatometric curves (Figure 2a–c)shows no evidence of elemental titanium presence as it was observed in the heating stage. The only

Materials 2019, 12, 4188 6 of 12

visible effects are a solidification of borides at 1147 ± 2.1 ◦C and transformation of austenite back toferrite starting at 893 ± 2.4 ◦C.

The results of dilatometric tests (Figure 3c) correspond well with the density of sintered samples(Figure 4) in respect to the direction of changes. There is, however, no agreement between these resultsin respect to the magnitude of changes. The shrinkage observed in dilatometric samples was heavilyimpaired by the evaporation of the boron from the sample surface [22]. Missing boron resulted in creationof smaller amount of eyectic phase. This effect is the more pronounced the smaller is the sample and thelarger is its surface. As can be easily seen, the higher the addition of titanium, the lower the density afterthe sintering process. This effect was more pronounced for the 45–63 µm titanium particle size comparedto the size of 100–140 µm. This shows that the reaction between the titanium and the eutectic phase wasmore effective when fine particles of high specific surface area were added.

Materials 2020, 13, x FOR PEER REVIEW 6 of 13

The results of dilatometric tests (Figure 3c) correspond well with the density of sintered samples 157 (Figure 4) in respect to the direction of changes. There is, however, no agreement between these 158 results in respect to the magnitude of changes. The shrinkage observed in dilatometric samples was 159 heavily impaired by the evaporation of the boron from the sample surface [22]. Missing boron 160 resulted in creation of smaller amount of eyectic phase. This effect is the more pronounced the smaller 161 is the sample and the larger is its surface. As can be easily seen, the higher the addition of titanium, 162 the lower the density after the sintering process. This effect was more pronounced for the 45–63 µm 163 titanium particle size compared to the size of 100–140 µm. This shows that the reaction between the 164 titanium and the eutectic phase was more effective when fine particles of high specific surface area 165 were added. 166

167

Figure 4. (a) Density changes and (b) porosity in function of the added titanium of Ø 20 × 5 mm 168 samples. 169

3.3. Phase Identification 170

Results of specular X-ray diffraction experiments are shown in Figure 5a together with peak 171 positions from literature in Figure 5b. As a measure for expected intensity, the absolute value of the 172 squared structure factor is used. For clarity, the experimental curves are shifted vertically. A major 173 Bragg peak can be observed at 2 Theta = 44.7°, corresponding to the 110 reflex of the cubic iron phase 174 (a = 2.8665 Å ) [23]. Additional peaks of this phase can be found at 65.1°, 82.4°, 99.0° and 116.5°, 175 originating from the (200), (211), (220) and (310) planes (Figure S1 in supplementary data). In the 176 reference sample as well as all samples with titanium addition, additional peaks due to the presence 177 of Fe2B can be observed (35.1° and 42.6°, corresponding to the 200 and 002 plane) [24]. At high 178 titanium concentrations, a simultaneous presence of these peaks and the peak of the 100 reflex of TiB2 179 [25] can be found. Such observations stay in a good agreement with results of [17,18] who observed 180 the formation of TiB2 at the expense of Fe2B borides in cast iron-based alloys during the solidification 181 process. Moreover, a weak presence of crystalline titanium (40.2°) was noticed, informing that the 182 reaction between titanium and boron was not complete [23]. 183

Figure 4. (a) Density changes and (b) porosity in function of the added titanium of Ø20× 5 mm samples.

3.3. Phase Identification

Results of specular X-ray diffraction experiments are shown in Figure 5a together with peakpositions from literature in Figure 5b. As a measure for expected intensity, the absolute value of thesquared structure factor is used. For clarity, the experimental curves are shifted vertically. A majorBragg peak can be observed at 2 Theta = 44.7◦, corresponding to the 110 reflex of the cubic iron phase(a = 2.8665 Å) [23]. Additional peaks of this phase can be found at 65.1◦, 82.4◦, 99.0◦ and 116.5◦,originating from the (200), (211), (220) and (310) planes (Figure S1 in supplementary data). In thereference sample as well as all samples with titanium addition, additional peaks due to the presence ofFe2B can be observed (35.1◦ and 42.6◦, corresponding to the 200 and 002 plane) [24]. At high titaniumconcentrations, a simultaneous presence of these peaks and the peak of the 100 reflex of TiB2 [25]can be found. Such observations stay in a good agreement with results of [17,18] who observed theformation of TiB2 at the expense of Fe2B borides in cast iron-based alloys during the solidificationprocess. Moreover, a weak presence of crystalline titanium (40.2◦) was noticed, informing that thereaction between titanium and boron was not complete [23].

Comparison of observed borides (REF and C63 samples) was presented in Figure 6a,b. Figure 6ashows borides in a network of rib-like structures observed in REF sample, whereas Figure 6b showsinterconnected globular structures observed in titanium modified sintered compacts. This change inmorphology was caused by crystallization of TiB2 as a result of the reaction of titanium and boron.The crystallisation process requires binding boron from a eutectic liquid resulting in precipitation ofiron atoms in surrounding space. As the TiB2 is solid at sintering temperature it consisted a barrierfor grain growth. This can be seen as uneven grain boundaries in samples with titanium addition(Figure 6b). Without the presence of titanium, Fe2B crystallizes at lower temperatures as the last phase(Figure 1a), therefore its morphology is determined by already existing spherical grains. Ex situ SEMEDX measurements (Figure 6c,d) after the sintering process showed the presence of small and nearlyspherical titanium based particles in all samples with titanium additions. Moreover, the addition oftitanium causes separation of previously (REF) connected borides, which enhances the connection amongneighbouring grains (figures available in supplementary data in the folder: Eutectics distribution).

Materials 2019, 12, 4188 7 of 12

Materials 2020, 13, x FOR PEER REVIEW 7 of 13

184

Figure 5. (a) Specular X-ray diffraction patterns of the Fe-B system with titanium additions; (b) 185 expected peak positions from known crystal structures and their square of the absolute value of the 186 structure factor SF as a measure for expected intensity. 187

Comparison of observed borides (REF and C63 samples) was presented in Figure 6a,b. Figure 6a 188 shows borides in a network of rib-like structures observed in REF sample, whereas Figure 6b shows 189 interconnected globular structures observed in titanium modified sintered compacts. This change in 190 morphology was caused by crystallization of TiB2 as a result of the reaction of titanium and boron. 191 The crystallisation process requires binding boron from a eutectic liquid resulting in precipitation of 192 iron atoms in surrounding space. As the TiB2 is solid at sintering temperature it consisted a barrier 193 for grain growth. This can be seen as uneven grain boundaries in samples with titanium addition 194 (Figure 6b). Without the presence of titanium, Fe2B crystallizes at lower temperatures as the last phase 195 (Figure 1a), therefore its morphology is determined by already existing spherical grains. Ex situ SEM 196 EDX measurements (Figure 6c,d) after the sintering process showed the presence of small and nearly 197 spherical titanium based particles in all samples with titanium additions. Moreover, the addition of 198 titanium causes separation of previously (REF) connected borides, which enhances the connection 199 among neighbouring grains (figures available in supplementary data in the folder: Eutectics 200 distribution). 201

Figure 5. (a) Specular X-ray diffraction patterns of the Fe-B system with titanium additions; (b) expectedpeak positions from known crystal structures and their square of the absolute value of the structurefactor SF as a measure for expected intensity.

Materials 2020, 13, x FOR PEER REVIEW 8 of 13

202

Figure 6. SEM pictures of borides: (a) REF sample; (b) C63 sample; (c) TiB2 neighbouring the Fe2B; (d) 203 EDS line scan of sample C63 presented in Figure 6c. 204

Furthermore, the titanium addition influences also the microstructure of the sintered material as 205 presented in Figure 7a–e. The reference sample characterizes with a nearly continuous network of 206 borides located on grain boundaries and uniaxial grains. Contrary to the REF sample, the more 207 titanium was introduced, the more developed the grains’ shapes became. The presence of the TiB2 in 208 the matrix of the material may effectively hinder the grain growth, as was demonstrated in research 209 carried out, e.g., by Namini et al., Sobhani et al. and Gan et al. [26–28]. Moreover, the more titanium 210 was added, the more and the larger pores were observed. This is believed was a side effect of limiting 211 the amount of liquid phase during the sintering process. 212

213

Figure 7. Representative microstructure micrographs of the tested samples: (a) REF; (b) A63; (c) A140; 214 (d) C63; (e) C140 (detailed micrographs in supplementary data). 215

Figure 6. SEM pictures of borides: (a) REF sample; (b) C63 sample; (c) TiB2 neighbouring the Fe2B;(d) EDS line scan of sample C63 presented in Figure 6c.

Furthermore, the titanium addition influences also the microstructure of the sintered materialas presented in Figure 7a–e. The reference sample characterizes with a nearly continuous networkof borides located on grain boundaries and uniaxial grains. Contrary to the REF sample, the moretitanium was introduced, the more developed the grains’ shapes became. The presence of the TiB2 inthe matrix of the material may effectively hinder the grain growth, as was demonstrated in researchcarried out, e.g., by Namini et al., Sobhani et al. and Gan et al. [26–28]. Moreover, the more titanium

Materials 2019, 12, 4188 8 of 12

was added, the more and the larger pores were observed. This is believed was a side effect of limitingthe amount of liquid phase during the sintering process.

Materials 2020, 13, x FOR PEER REVIEW 8 of 13

202

Figure 6. SEM pictures of borides: (a) REF sample; (b) C63 sample; (c) TiB2 neighbouring the Fe2B; (d) 203 EDS line scan of sample C63 presented in Figure 6c. 204

Furthermore, the titanium addition influences also the microstructure of the sintered material as 205 presented in Figure 7a–e. The reference sample characterizes with a nearly continuous network of 206 borides located on grain boundaries and uniaxial grains. Contrary to the REF sample, the more 207 titanium was introduced, the more developed the grains’ shapes became. The presence of the TiB2 in 208 the matrix of the material may effectively hinder the grain growth, as was demonstrated in research 209 carried out, e.g., by Namini et al., Sobhani et al. and Gan et al. [26–28]. Moreover, the more titanium 210 was added, the more and the larger pores were observed. This is believed was a side effect of limiting 211 the amount of liquid phase during the sintering process. 212

213

Figure 7. Representative microstructure micrographs of the tested samples: (a) REF; (b) A63; (c) A140; 214 (d) C63; (e) C140 (detailed micrographs in supplementary data). 215

Figure 7. Representative microstructure micrographs of the tested samples: (a) REF; (b) A63; (c) A140;(d) C63; (e) C140 (detailed micrographs in supplementary data).

A comparison of a grain size presented in Figure 8 shows that the introduction of titanium ofparticle size of 63 µm (Figure 8a) results in grain refinement, which stays in good agreement withother researchers’ findings [26–28]. A similar trend was observed when titanium was introduced inthe form of bigger particles (140 µm). The grain size refinement is caused by the presence of a TiB2,which consists of an effective barrier for grain growth.

Materials 2020, 13, x FOR PEER REVIEW 9 of 13

A comparison of a grain size presented in Figure 8 shows that the introduction of titanium of 216 particle size of 63 μm (Figure 8a) results in grain refinement, which stays in good agreement with 217 other researchers’ findings [26–28]. A similar trend was observed when titanium was introduced in 218 the form of bigger particles (140 μm). The grain size refinement is caused by the presence of a TiB2, 219 which consists of an effective barrier for grain growth. 220

221

Figure 8. Histograms of grain size distribution for series (a) 63 and (b) 140. 222

The presence of TiB2 (induced by addition of Ti) changed also the grain shape, i.e., grains in REF 223 sample were of polygonal shape with rounded corners due to presence of large amount of borides. 224 Moreover, the more titanium was added, the more irregular was the shape of the grains and the more 225 dispersed were the borides. Furthermore, as appears from the analysis of grain size distribution a 226 large amount of titanium (1.40 wt % – C) promotes also the appearance of large grains (diameter 100–227 150 μm) in the case of series 140. The medium-sized (50–100 μm) grains are promoted when titanium 228 is added in the form of a particle of size 63 μm. 229

3.4. Mechanical Properties 230

Transverse Rupture Strength (TRS) tests were performed in order to estimate the influence of 231 the changed phase composition on the mechanical behaviour of the samples. The results were 232 presented in a way to highlight the influence of the increasing amount of titanium addition on 233 mechanical behaviour (Figure 9a–c). The results bring the conclusion that high addition of titanium 234 (C-series, 1.40 wt % Ti) causes an increase in the maximum observed deformation on the expense of 235 maximum force (Figure 9a,b) and hence, the TRS as well (Figure 9c). Other additions of titanium (A 236 and B) show intermediate states—samples with 0.47 wt % titanium addition (A) show a decrease of 237 both TRS and maximum deformation, whereas the B-series (0.93 wt % titanium addition) is 238 characterized by lower peak force and higher maximum deformation with respect to the reference 239 sample (Figure 9). The increase of ductility remains in the agreement of the other works [17,18]. The 240 drop of the mechanical properties, however, stays in opposition to the recent findings [17,18]. The 241 increase in ductility of sample (C series) can be attributed to joint effect of few phenomena: (i) the 242 grains are not anymore separated with a network of brittle borides; therefore, grains are connected 243 with each other; (ii) formation of TiB2 reinforces the matrix; (iii) grains are not uniaxial anymore and 244 their shape is irregular; therefore, mutual interlocking is possible. The irregular shape of grains is 245 attributed to the joint effect of grain growth in conditions of solid-state sintering (induced by eutectic 246 phase disappearance) and liquid phase sintering. These changes were possible mainly due to the 247 change of manufacturing process change—in previous works, the casting was used whereas is the 248 present work the liquid phase sintering process was applied. For the densification of the sample, 249 effective mass transport is required, and when the titanium is added, it is impeded as presented in 250 Figure 4, which leads to the drop of the density (in relation to the REF sample). 251

Figure 8. Histograms of grain size distribution for series (a) 63 and (b) 140.

The presence of TiB2 (induced by addition of Ti) changed also the grain shape, i.e., grains in REFsample were of polygonal shape with rounded corners due to presence of large amount of borides.Moreover, the more titanium was added, the more irregular was the shape of the grains and the moredispersed were the borides. Furthermore, as appears from the analysis of grain size distribution a largeamount of titanium (1.40 wt % – C) promotes also the appearance of large grains (diameter 100–150 µm)in the case of series 140. The medium-sized (50–100 µm) grains are promoted when titanium is addedin the form of a particle of size 63 µm.

3.4. Mechanical Properties

Transverse Rupture Strength (TRS) tests were performed in order to estimate the influence of thechanged phase composition on the mechanical behaviour of the samples. The results were presented in

Materials 2019, 12, 4188 9 of 12

a way to highlight the influence of the increasing amount of titanium addition on mechanical behaviour(Figure 9a–c). The results bring the conclusion that high addition of titanium (C-series, 1.40 wt % Ti)causes an increase in the maximum observed deformation on the expense of maximum force (Figure 9a,b)and hence, the TRS as well (Figure 9c). Other additions of titanium (A and B) show intermediatestates—samples with 0.47 wt % titanium addition (A) show a decrease of both TRS and maximumdeformation, whereas the B-series (0.93 wt % titanium addition) is characterized by lower peak force andhigher maximum deformation with respect to the reference sample (Figure 9). The increase of ductilityremains in the agreement of the other works [17,18]. The drop of the mechanical properties, however,stays in opposition to the recent findings [17,18]. The increase in ductility of sample (C series) can beattributed to joint effect of few phenomena: (i) the grains are not anymore separated with a network ofbrittle borides; therefore, grains are connected with each other; (ii) formation of TiB2 reinforces the matrix;(iii) grains are not uniaxial anymore and their shape is irregular; therefore, mutual interlocking is possible.The irregular shape of grains is attributed to the joint effect of grain growth in conditions of solid-statesintering (induced by eutectic phase disappearance) and liquid phase sintering. These changes werepossible mainly due to the change of manufacturing process change—in previous works, the casting wasused whereas is the present work the liquid phase sintering process was applied. For the densificationof the sample, effective mass transport is required, and when the titanium is added, it is impeded aspresented in Figure 4, which leads to the drop of the density (in relation to the REF sample).Materials 2020, 13, x FOR PEER REVIEW 10 of 13

252

Figure 9. Representative Transverse Rupture Strength (TRS) curve for samples from series (a) 63 and 253 (b) 140; (c) comparison of TRS values of tested samples. 254

Fracture surfaces investigated using SEM (Figure 10) reveal that the more titanium was added 255 the more ductile was the behaviour of the fracture surfaces, which can be seen by the development 256 of the fracture surface in Figure 10a–e. The more titanium was introduced the less flat intra-granular 257 fractures were observed and the more signs of ductility were present, i.e., “mountain-chain shapes 258 and valleys”. The extent of the changes was relatively small. This, however, suggests a lesser amount 259 of precipitated brittle borides at the grains boundaries compared to the material without titanium 260 addition (REF). 261

Figure 9. Representative Transverse Rupture Strength (TRS) curve for samples from series (a) 63 and(b) 140; (c) comparison of TRS values of tested samples.

Materials 2019, 12, 4188 10 of 12

Fracture surfaces investigated using SEM (Figure 10) reveal that the more titanium was added themore ductile was the behaviour of the fracture surfaces, which can be seen by the development of thefracture surface in Figure 10a–e. The more titanium was introduced the less flat intra-granular fractureswere observed and the more signs of ductility were present, i.e., “mountain-chain shapes and valleys”.The extent of the changes was relatively small. This, however, suggests a lesser amount of precipitatedbrittle borides at the grains boundaries compared to the material without titanium addition (REF).

Materials 2020, 13, x FOR PEER REVIEW 11 of 13

262

Figure 10. Representative fracture surfaces of the tested samples: (a) REF; (b) A63; (c) A140; (d) C63; 263 (e) C140. 264

4. Conclusions 265

The influence of increasing titanium addition on the iron-boron system was under investigation. 266 Titanium was introduced in three various amounts of two different grain size distributions each. It 267 was shown that the addition of titanium partially binds the boron from the Fe2B + Fe eutectic liquid 268 phase and, therefore, reduces the shrinking rate by influencing the amount of the eutectic liquid 269 phase. It was also shown that the addition of titanium changes the phase composition of the final 270 material by creating crystalline TiB2 borides besides crystalline Fe2B borides. Changes in the 271 morphology from a network of rib-like structures for the pure Fe-B system to interconnected globular 272 structures in case of titanium additions were also observed due to TiB2 appearance. Both, increasing 273 amount of titanium addition and its fine grain size distribution, impede the shrinking of Fe-B system 274 by decreasing the amount of liquid phase. The addition of titanium enhances the plastic behaviour 275 of a sintered material at the expense of transverse rupture strength. This change was attributed to (i) 276 grain refinement; (b) irregular shape of grains, which allowed for mutual interlocking, and (iii) 277 decreasing amount of brittle borides on the grain boundaries, which enhanced the connection among 278 the neighbouring grains. 279

Contrary to the casting process, the sintering process was found to be not suitable for 280 strengthening iron-based sintered compacts via TiB2 in situ formation due to a negative influence on 281 the densification behaviour of the sintered material. 282

Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Figure S1: Specular X-283 ray diffraction patterns of the Fe-B system with titanium additions over the full measured range. Grey lines 284 denote expected peak positions of the cubic iron phase. 285

Author Contributions: Conceptualization, M.S., M.H. and J.K.; Data Curation, M.S., M.H., B.S. and R.R.; Formal 286 Analysis, M.H., R.R. and C.S.; Funding Acquisition, M.S., J.K. and C.S.; Investigation, M.S., M.H. and B.S.; 287 Methodology, M.S., M.H. and B.S.; Project Administration, M.S. and J.K.; Resources, M.S., R.R., J.K. and C.S.; 288 Supervision, R.R., J.K. and C.S.; Validation, M.H., R.R. and C.S.; Visualization, M.S. and B.S.; Writing—Original 289 Draft, M.S., M.H. and B.S.; Writing—Review & Editing, M.H., R.R. and C.S. 290

Funding: This research was supported by grant no. UMO-2014/13/N/ST8/00585, financed by the National 291 Science Centre, Poland. This research was also supported by internal funds of Graz University of Technology. 292 Open Access Funding by the Graz University of Technology. 293

Conflicts of Interest: The authors declare no conflicts of interest 294

Figure 10. Representative fracture surfaces of the tested samples: (a) REF; (b) A63; (c) A140; (d) C63;(e) C140.

4. Conclusions

The influence of increasing titanium addition on the iron-boron system was under investigation.Titanium was introduced in three various amounts of two different grain size distributions each. It wasshown that the addition of titanium partially binds the boron from the Fe2B + Fe eutectic liquid phaseand, therefore, reduces the shrinking rate by influencing the amount of the eutectic liquid phase. It wasalso shown that the addition of titanium changes the phase composition of the final material by creatingcrystalline TiB2 borides besides crystalline Fe2B borides. Changes in the morphology from a networkof rib-like structures for the pure Fe-B system to interconnected globular structures in case of titaniumadditions were also observed due to TiB2 appearance. Both, increasing amount of titanium addition andits fine grain size distribution, impede the shrinking of Fe-B system by decreasing the amount of liquidphase. The addition of titanium enhances the plastic behaviour of a sintered material at the expense oftransverse rupture strength. This change was attributed to (i) grain refinement; (b) irregular shape ofgrains, which allowed for mutual interlocking, and (iii) decreasing amount of brittle borides on the grainboundaries, which enhanced the connection among the neighbouring grains.

Contrary to the casting process, the sintering process was found to be not suitable for strengtheningiron-based sintered compacts via TiB2 in situ formation due to a negative influence on the densificationbehaviour of the sintered material.

Supplementary Materials: The following are available online at http://www.mdpi.com/1996-1944/12/24/4188/s1,Figure S1: Specular X-ray diffraction patterns of the Fe-B system with titanium additions over the full measuredrange. Grey lines denote expected peak positions of the cubic iron phase.

Author Contributions: Conceptualization, M.S., M.H. and J.K.; Data Curation, M.S., M.H., B.S. and R.R.; FormalAnalysis, M.H., R.R. and C.S.; Funding Acquisition, M.S., J.K. and C.S.; Investigation, M.S., M.H. and B.S.;Methodology, M.S., M.H. and B.S.; Project Administration, M.S. and J.K.; Resources, M.S., R.R., J.K. and C.S.;

Materials 2019, 12, 4188 11 of 12

Supervision, R.R., J.K. and C.S.; Validation, M.H., R.R. and C.S.; Visualization, M.S. and B.S.; Writing—OriginalDraft, M.S., M.H. and B.S.; Writing—Review & Editing, M.H., R.R. and C.S.

Funding: This research was supported by grant no. UMO-2014/13/N/ST8/00585, financed by the National ScienceCentre, Poland. This research was also supported by internal funds of Graz University of Technology. OpenAccess Funding by the Graz University of Technology.

Conflicts of Interest: The authors declare no conflicts of interest

References

1. Subramanian, C.; Suri, A.K.; Murthy, T.S.R.C. Development of Boron-based Materials for Nuclear Applications.Barc Newsl. 2010, 313, 14–22.

2. Lherbier, L. Solving Nuclear Storage Issues with Powder Alloys. 2019. Availableonline: https://www.carpentertechnology.com/en/alloy-techzone/technical-information/technical-articles/solving-nuclear-storage-issues (accessed on 20 July 2019).

3. Hye-Jin, K.; Soon-Hyeok, J.; Won-Seog, Y.; Byung-Gil, Y.; Yoo-Dong, C.; Heon-Young, H.; Hyun-Young, C.Effects of titanium content on hydrogen embrittlement susceptibility of hot-stamped boron steels. J. AlloysCompd. 2018, 735, 2067–2080.

4. Fedorova, I.; Liu, F.; Grumsen, F.B.; Cao, Y.; Mishin, O.V.; Hald, J. Fine (Cr,Fe)2B borides on grain boundariesin a 10Cr-0.01B martensitic steel. Scr. Mater. 2018, 156, 124–128. [CrossRef]

5. Yang, X.; Guo, S. Fe-Mo-B Enhanced Sintering of P/M 316 L Stainless Steel. J. Iron Steel Res. Int. 2008, 15,10–14. [CrossRef]

6. Uzunsoy, D. Investigation of dry sliding wear properties of boron doped powder metallurgy 316 L stainlesssteel. Mater. Des. 2010, 31, 3896–3900. [CrossRef]

7. Wu, M.W.; Cai, W.Z.; Lin, Z.J.; Chang, S.H. Liquid phase sintering mechanism and densification behavior ofboron-alloyed Fe-Ni-Mo-C-B powder metallurgy steel. Mater. Des. 2017, 133, 536–548. [CrossRef]

8. Sulikowska, K.; Skałon, M.; Kozub, B.; Kazior, J. Influence of mechanical alloying on surface layer of P/M AISI316 L powder with addition of FeNiMnSiB master alloy. Surf. Coat. Technol. 2016, 302, 142–149. [CrossRef]

9. Molinari, A.; Kazior, J.; Straffelini, G. Investigation of liquid-phase sintering by image analysis. Mater. Charact.1995, 34, 271–276. [CrossRef]

10. Luitjohan, K.E.; Krane, M.J.M.; Ortalan, V.; Johnson, D.R. Investigation of the metatectic reaction in iron-boronbinary alloys. J. Alloys Compd. 2018, 732, 498–505. [CrossRef]

11. Skałon, M.; Buzolin, R.; Kazior, J.; Sommitsch, C.; Hebda, M. Improving the dimensional stability andmechanical properties of AISI 316L + B sinters by Si3N4 addition. Materials 2019, 12, 1798. [CrossRef]

12. Skałon, M.; Hebda, M.; Sulikowska, K.; Kazior, J. Influence of FeNiMnSiB master alloy on the structure andmechanical properties of P/M AISI 316L. Mater. Des. 2016, 108, 462–469. [CrossRef]

13. Kwon, O.J.; Yoon, D.N. Closure of isolated pores in liquid phase sintering of W-Ni. Int. J. Powder Metall.1981, 17, 127–133.

14. Kang, S.J.L.; Kaysser, W.A.; Petzow, G.; Yoon, D.N. Elimination of pores during liquid phase sintering ofMo-Ni. Powder Metall. 1984, 27, 97–100. [CrossRef]

15. Lee, D.D.; Kang, S.G.L.; Yoon, D.N. A direct observation of the grain shape accommodation during liquidphase sintering of Mo-Ni alloy. Scr. Metall. Mater. 1990, 24, 927–930. [CrossRef]

16. Kang, S.J.L.; Azou, P. Trapping of pores and liquid pockets during liquid phase sintering. Powder Metall.1985, 28, 90–92. [CrossRef]

17. Lin, H.; Ying, L.; Jun, L.; Binghong, L. Effects of hot rolling and titanium content on the microstructure andmechanical properties of high boron Fe-B alloys. Mater. Des. 2012, 36, 88–93.

18. Liu, Y.; Li, B.; Li, J.; He, L.; Gao, S.; Nieh, T.G. Effect of titanium on the ductilization of Fe-B alloys with highboron content. Mater. Lett. 2010, 64, 1299–1301. [CrossRef]

19. Ma, M.; Wang, L.; Wang, Y.; Xiang, W.; Lyu, P.; Tang, B.; Tan, X. Effect of hydrogen content on hydrogendesorption kinetics of titanium hydride. J. Alloys Compd. 2017, 709, 445–452. [CrossRef]

20. Ma, M.; Liang, L.; Wang, L.; Wang, Y.; Cheng, Y.; Tang, B.; Tan, X. Phase transformations of titanium hydridein thermal desorption process with different heating rates. Int. J. Hydrogen Energy 2015, 40, 8926–8934.[CrossRef]

Materials 2019, 12, 4188 12 of 12

21. Baymakov, Y.V.; Lebedev, O.A. Metallurgy of Nonferrous Metals Transactions. Leningrad Polytechnic. Inst.1963, 223, 25–34.

22. Karwan-Baczewska, J. The properties of Fe-Ni-Mo-Cu-B materials produced via liquid phase sintering.Arch. Metall. Mater. 2011, 56, 789–796. [CrossRef]

23. Wyckoff, R.W.G. Crystal Structures, Second edition; Interscience Publishers: New York, NY, USA, 1963;pp. 7–83.

24. Bjurstroem, T. Arkiv for Kemi, Mineralogi Och Geologi; Almqvist & Wiksell: Stockholm, Sweden, 1933; pp. 1–12.25. Moehr, S.; Mueller-Buschbaum, H.K.; Grin, Y.; Schnering, H.G. Zeitschrift für anorganische und allgemeine

Chemie 5/2003. Z. Fur Anorg. Allg. Chem. 1996, 622, 1035–1037.26. Namini, A.S.; Azadbeh, M.; Asl, M.S. Effect of TiB2 content on the characteristics of spark plasma sintered

Ti–TiBw composites. Adv. Powder Technol. 2017, 28, 1564–1572. [CrossRef]27. Sobhani, M.; Arabi, H.; Mirhabibi, A.; Brydson, R.M.D. Microstructural evolution of copper-titanium alloy

during in-situ formation of TiB2 particles. Trans. Nonferr. Met. Soc. 2013, 23, 2994–3001. [CrossRef]28. Gan, G.; Yang, B.; Zhang, B.; Jiang, X.; Shi, Y.; Wu, Y. Refining Mechanism of 7075 Al Alloy by In-Situ TiB2

Particles. Materials 2017, 10, 132. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC BY) license (http://creativecommons.org/licenses/by/4.0/).