Influence of different bonding and fluxing agents on the sintering behavior and dielectric properties of steatite ceramic materials

Ceramics International xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Ceramics International journal homepage: www.elsevier.com/locate/ceramint

Influence of different bonding and fluxing agents on the sintering behavior and dielectric properties of steatite ceramic materials ⁎

Anja Terzića, , Nina Obradovićb, Jovica Stojanovićc, Vladimir Pavlovićb, Ljubiša Andrićc, Dragan Olćand, Antonije Đorđevićd a

Institute for Testing of Materials IMS, Vojvode Mišića Bl. 43, 11000 Belgrade, Serbia Institute of Technical Sciences, Serbian Academy of Sciences and Arts, Knez Mihailova St. 35, 11000 Belgrade, Serbia c Institute for Technology of Nuclear and Other Mineral Raw Materials, Franchet d’Esperey 86, 11000 Belgrade, Serbia d School of Electrical Engineering, University of Belgrade, Kralja Aleksandra Bl. 73, 11000 Belgrade, Serbia b

A R T I C L E I N F O

A BS T RAC T

Keywords: A. Sintering B. Electron microscopy B. Microstructure-final B. X-ray methods C. Dielectric properties D. MgO E. Insulators

The focus of the study was on providing insights into interconnections between sintering and development of the crystalline microstructure, and consequently variations in dielectric behavior of four steatites fabricated from a low-cost raw material, i.e. talc. The changes, induced by the alternations of the binders (bentonite, kaolin clay) and fluxing agents (BaCO3, feldspar), were monitored in the temperature range 1000° to 1250 °C in which complete densification and re-crystallization of the investigated structures were accomplished. The critical points in the synthesis of steatite materials were assessed by instrumental analyses. Crystallinity changes and mineral phase transition during sintering were monitored by X-ray diffraction technique. Microstructural visualization of the samples and the spatial arrangements of individual chemical elements were achieved via scanning electron microscopy accompanied with EDS mapping. The thermal stability was observed on the green mixtures using differential thermal and thermo gravimetric analyses. Electrical measurements recorded variations of the dielectric constant (εr) and loss tangent (tan δ) as a function of the sintering temperature. The investigation highlighted critical design points, as well as the optimal combinations of the raw materials for production of the steatite ceramics for advanced electrical engineering applications.

1. Introduction

economical due the facilitated forming and lower temperature of sintering [1]. Such predispositions are classifying this type of material for applications in field of electrical and electronic engineering: electrical insulation, regulator bases, switches and plug parts, sockets bases for halogen bulbs, heating element holders, NH-fuses, element formers and casings for thermostats, interlocking insulating beads, etc. [7–11]. Apart from the use in electrotechnics, the steatite ceramic is employed in medicine as a material for dental implants and artificial bones due to high hardness and bending resistance, and good biocompatibility [12–14]. Mineral enstatite is natural equivalent of the inorganic material with chemical formula: Mg3(Si4O10)(OH)2) [15,16]. The mineral shows 3.21 g cm−3 theoretical density, a melting point at 1557 °C, and an orthorhombic crystal lattice with silica-tetrahedra chains connected by two symmetrical octahedral Mg2+ sites [6,15–18]. The steatite ceramics can be manufactured completely from economic natural raw materials, as its mineral base for the synthesis generally comprises the talc mineral mixture for the calcination procedure (80–90%), plastic

Depending on choice of raw materials, the ceramic is indexed as either “triaxial”, which, besides clayey base and flux, comprises quartz or alumina, or “dielectric” that primarily rests on magnesium silicate [1–3]. The materials from the first group generally prevail in the construction field as they are predominantly being used for manufacturing of the stoneware, while the second group includes hightemperature resistant products that are mainly employed in electrical engineering. The magnesium silicate materials are produced via standard ceramic processing schemes (e.g. dry pressing, extrusion, casting, semi-wet pressing) and fabricated or sintered into a variety of forms [4,5]. The steatite ceramics is, besides cordierite, the main represent of “dielectric” materials with trademarks such as high mechanical strength, good electrical properties, low dielectric loss, and excellent resistance to high temperatures [6,7]. Also, steatite is appraised as a cost-effective alumina replacement, because it satisfies the performance requirements of alumina, yet its production is more

⁎

Corresponding author. E-mail address: [email protected] (A. Terzić).

http://dx.doi.org/10.1016/j.ceramint.2017.07.024 Received 12 April 2017; Received in revised form 20 June 2017; Accepted 3 July 2017 0272-8842/ © 2017 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

Please cite this article as: Terzic, A., Ceramics International (2017), http://dx.doi.org/10.1016/j.ceramint.2017.07.024

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

clay as a bonding agent (5–10%) and a melting agent (5–10%). This mixture of raw materials is fired at temperatures up to 1400 °C depending on the selected sintering procedure [1,4–6]. The final structure of steatite is formed though crystallization of mineral phases, as well as their fusion and dissolution during the vitrification. By the time the sintering process reaches the end, the steatite comprises about 70% of crystalline MgSiO3 (i.e. protoenstatite) and about 30% of glassy phase [11]. The flux (e.g., BaO, BaCO3, ZnO, feldspar) enables and controls the formation of amorphous matter during the sintering, and influences mechanical and dielectric properties of the material [1,5,18– 20]. The microstructure development and the sintering behavior of the steatite, therefore, depend on the ratio of formed crystalline and liquid phases, as well as on the amount of present pores. A vitreous phase surrounds the crystals merging them together [11,19]. Two most commonly used fluxes in the industrial ceramics are feldspar and barium carbonate. Feldspar, due to the presence of alkaline ions, may lead to the ceramic with lower performances, including increased dielectric losses, while barium carbonate usually produces better dielectric properties [19–21]. The dielectric properties of a steatite product are also strongly dependent on the temperature range and/or cycling assigned for the given synthesis procedure, largely because enstatite undergoes a complex series of polymorphic transitions [10,17]. Namely, steatite occurs in four magnesium metasilicate polymorphic forms: enstatite, protoenstatite, clinoenstatite, and hightemperature clinoenstatite [4,11,17,22–25]. At normal ambient conditions, enstatite appears as a low temperature polymorph – clinoenstatite (i.e., low clinoenstatite) in monoclinic symmetry, or as orthoenstatite in orthorhombic symmetry. At elevated temperatures (up to 1000 °C) crystals undergo a transition into an intermediate structure: e.g., the low clinoenstatite inverts reversibly to the high clinoenstatite [21,22]. The thermodynamically stable crystals of protoenstatite initially appear above 985 °C. The well-ordered and fully stable protoenstatite is usually obtained upon heating up to 1300 °C [21,22]. At lower temperatures, the crystals within the ceramic body are stabilized by the vitreous phase. When crystals are not properly stabilized, protoenstatite transforms to the room temperature polymorph clinoenstatite. The volume changes that occur during this transition lead to formation of cracks and deterioration of dielectric properties of the material [21,22]. Even though there are numerous literature reports on the production of various types of dielectric ceramic materials [26–32], a very few works are dealing with low-cost fabrication of steatite [1,5,6,19,20]. This study provides insight into dielectric behavior of steatite that is prepared from natural raw materials (talc, BaCO3, feldspar, bentonite, kaolin) and sintered at four temperatures: 1000°, 1100°, 1200° and 1250 °C. The focus of the experiment was on altering of the bonding and fluxing agents in order to reduce the temperature of sintering as the most common route to obtain cost effective production of steatite ceramics. The interconnections between sintering procedure, microstructural and crystalline development, and dielectric properties of four steatite types were established. The aim of the study is to detect the critical points in the design of steatite, to make improvements in its synthesis in order to satisfy the increasing demands of the dielectric material performances, and finally to assess the materials suitability for advanced engineering applications.

Table 1 Chemical compositions of applied raw materials and steatite mixtures S1-4. Oxide, %

Feldspar

Talc

Bentonite

Kaolin

S1

S2

S3

S4

SiO2 K2O Na2O Fe2O3 CaO MgO Al2O3 TiO2 LoI

70.15 4.82 6.27 0.11 0.76 0.13 17.47 – 0.29

50.75 – – 0.42 0.56 25.3 0.03 – 22.94

68.48 0.42 0.23 2.42 2.10 2.28 17.20 – 6.87

70.83 0.07 0.64 0.11 0.72 – 20.12 0.23 7.46

58.6 0.5 0.65 0.65 0.57 25.8 5.2 – 8.03

69.7 0.3 0.40 0.65 0.57 25.2 6.2 0.02 6.96

55.8 0.1 0.65 0.65 0.57 25.1 4.7 – 12.43

56.9 0.01 0.05 0.62 0.78 25.1 6.3 0.02 10.22

and S2 mixtures and barium carbonate (BaCO3, 99+ %, synthetic, Acros Organics, India) in S3 and S4 mixtures. The chemical compositions of raw materials and starting steatite mixtures (S1–4) are given in Table 1. The chemical analysis was performed by means of the X-ray fluorescence method on an ED 2000 XRF spectrophotometer (Oxford Instruments, UK). Representative samples (100 g) were pulverized in a laboratory vibratory mill prior to the testing. The chemical composition of BaCO3 was provided by the manufacturer: Ba = 77%; and CO2 = 22.3%. The loss on ignition (LoI) was determined as a weight difference between 20° and 1000 °C. An ultra-centrifugal mill (ZM-1, Retsch, Germany) was used for the micronization of green steatite powders. The mill has a 300 ml working element which comprises a high-alloyed stainless steel rotor (∅100 mm, 600 W) and a variable mesh size ring sieves. The sieve with a 120 µm mesh size and trapezoid holes was used. The rotor velocity was adjusted at 20.000 rpm. The milling period was set at 30 min. Upon micronization, the binder-free S1–4 powders were pressed into cylindrical tablets using uniaxial double compression on a 40 mm diameter tool in a laboratory hydraulic press (4000 kg cm−2). Thus prepared samples were submitted to the sintering procedure. The same sintering regime was applied on all four steatite mixtures. The samples were fired at four different temperatures – 1000°, 1100°, 1200°, and 1250 °C, in a CWF 1300 laboratory furnace (Carbolite, UK). Sintering was conducted under a 10 °C/min heating rate. The samples were kept at the maximal projected sintering temperature for 60 min and extracted only when the furnace was completely cooled down. Mineralogical analyses were performed on the pulverized sintered steatite samples by means of the X-ray powder diffraction technique (XRD). The XRD patterns were obtained on a Philips PW-1710 automated diffractometer using a Cu tube. The instrument operates at 40 kV voltage and 30 mA current. It is equipped with a slanted graphite monochromator and a scintillation counter. The intensities of the diffracted CuKα X-ray emission (λ = 1.54178 Å) were measured at the ambient temperature in the intervals 0.02°2θ in the 2θ Bragg angle range from 4 do 65°2θ, counting for 0.5 s. The slits for the steering of the primary and diffracted beams were fixed at 1° and 0.1 mm. Differential thermal (DTA) and thermo-gravimetric (TGA) analyses were carried out in a static air flow with an automatic thermal analyzing system Setsys, SETARAM Instrumentation, Caluire, France. The steatite green powdery mixtures (approx. 30 mg) were loosely packed into an alumina holder and thermally treated at a constant heating rate of 10 °C/min in the temperature range between 25 °C and 1100 °C in air flow, in an alumina pan. The microstructure of the steatite samples S1–4 sintered at 1000 °C and 1250 °C was studied on a scanning electron microscope (JEOL, JSM-6390 LV). Parts of sintered tablets, whose surface was not polished prior to the imaging, were used in the analysis. The samples were covered with an Au film to improve the conductivity during recording. Energy-dispersive X-ray spectroscopic (EDS) analysis was conducted at the selected points on the S1–4 samples. The dielectric properties of the samples S1–4 were characterized by

2. Materials and methods Four steatite mixtures, marked as S1, S2, S3, and S4, were prepared for the experiment. All steatite materials were based on talc from Bela Stena deposit in Serbia. Talc comprised 80% of total mixture mass. Bentonite (Jelenkovac deposit, Vranjska Banja, Serbia) was applied as a bonding agent, in 10 wt% quantity, in the S1 and S3 mixtures. Kaolin clay (Crna Dolina deposit, Prijedor, Bosnia and Herzegovina) was used as binder (10 wt%) in the S2 and S4 mixtures. The fluxing agents, used in 10 wt% quantities, were: feldspar (Bujanovac deposit, Serbia) in S1 2

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

the relative permittivity (the relative dielectric constant, εr) and the loss tangent (tan δ), evaluated in the frequency range from 50 MHz to 500 MHz. The diameter of the sample tablets was around 8 mm. The height varied from sample to sample and it was in the range from 1.94 mm to 3.61 mm. The tablets were metallized on their top and bottom using a silver conductive paint (Electrolube), thus forming parallel-plate capacitors. Two test fixtures were used to measure the complex impedance of these capacitors [33]: a proprietary coaxial chamber and a simple open fixture fabricated using an SMA connector. The reflection coefficient of the fixture (with respect to a 50 Ω reference impedance) was measured using a vector network analyzer (Agilent E5061A). From the reflection coefficient, the complex relative permittivity of the samples was de-embedded using a numerical procedure, based on electrostatic models of the test fixtures and the tablets. The loss tangent was calculated from the real and imaginary parts of the complex permittivity.

(Py; JCPDS 15-0742), forsterite (F; JCPDS 34-0189), cristobalite (Cr; JCPDS 82-0512), and tridymite (T; JCPDS 42-1401). Enstatite (Mg2Si2O6), as the most abundant mineral phase registered in all four diffractograms, originated from talc. As it was shown in research studies by various authors, e.g., Jiao et al. [15], Smith [16], Reynard et al. [17], and Rigterink [18], enstatite mineral undergoes a series of polymorphic transitions when it is subjected to increasing and/or cycling of the temperature. The complex character of MgSiO3 polymorphism is controlled by a number of factors: temperature, pressure, dopants, internal stresses in grains, and grain sizes [10]. Thereby, magnesium metasilicate appears in four polymorphic forms: enstatite, protoenstatite, clinoenstatite, and high-temperature clinoenstatite [22]. At the room temperature, enstatite occurs in orthorhombic symmetry. Heating up to approximately 1000 °C results in polymorphs with intermediate structure and an inversion tendency. According to Smyth, protoenstatite appears to be stable form of MgSiO3 from about 1000 °C to 1300 °C [34]. In order to achieve a stable form of present minerals, the experimental S1–4 samples were analyzed upon thermal treatments at temperatures higher than or equal to 1000 °C. It was presumed that 30 min of mechanical activation and application of flux decreased the sintering temperature, as well as the temperature of the formation of stable enstatite. Quartz (SiO2) was the second most abundant phase. In the mineral phase composition of the S1 sample, quartz originated from bentonite or feldspar. The crystallinity of quartz was low. At 870 °C quartz ceases to be stable and, depending on the amount and type of flux, it converts into cristobalite and/or tridymite. The XRD analysis (Fig. 1a–d) registered the presence of all three quartz polymorphic transformations at temperatures from 1000 °C to 1250 °C. The reflections of both cristobalite and tridymite were feeble and superposed with other more significant peaks. Other phases, such as pyrope (Mg3Al2(SiO4)3 and forsterite (Mg2SiO4), were present in very small quantities.

3. Results and discussion The first stage in the experiment was to monitor the sintering induced changes within mineral phase compositions of the steatite mixes with respect to the employment of different bonding agents (i.e., bentonite and kaolin) and alternations in the used fluxing material (i.e., feldspar and BaCO3). All four steatites were submitted to the X-ray diffraction analysis subsequent to the sintering procedures conducted at four different temperatures: 1000 °, 1100 °, 1200 °, and 1250 °C. The acquired diffractograms are presented in Figs. 1–4. As illustrated by diffractograms in Fig. 1, the qualitative mineral phase composition of the S1 steatite remained unchanged by increasing of the sintering temperature from 1000 °C to 1250 °C. The S1 steatite comprised the following mineral phases: enstatite (E; JCPDS 86-0430), quartz (Q; JCPDS 46–1045), plagioclase (P; JCPDS 41-1480), pyrope

Fig. 1. The XRD diffractograms of the S1 steatite after thermal treatment at: a) 1000 °C; b) 1100 °C; c) 1200 °C, and d) 1250 °C.

3

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 2. The XRD diffractograms of the S2 steatite after thermal treatment at: a) 1000 °C; b) 1100 °C; c) 1200 °C, and d) 1250 °C.

enstatite, quartz, plagioclase, pyrope, and forsterite. Enstatite and quartz were the most abundant phases. Also, quartz was identified only in its primary modification, which means that the combination of feldspar as flux and kaolin as binder postponed the formation of other quartz polymorphs – cristobalite and tridymite. The crystallinity of the enstatite peaks located at 26° and 27° decreased from starting 175 a.u. to 125 a.u., i.e., 150 a.u. due to the increase in the sintering temperature. The reflection recorded at 32°, which also corresponds to enstatite, changed its size for about 60 a.u. between these two temperatures. Two additional crystalline phases appeared in the mineral-phase composition of the sample S2 after the thermal treatment conducted at 1200 °C (Fig. 2c): cristobalite and tridymite. These phases were located at 22° as a singular superposed reflection, referring to the small quantities of newly-formed quartz polymorphs. Nevertheless, the crystallinity of these new phases increased after sintering at 1250 °C (Fig. 2d). The thermal treatment at the given maximal temperature did not induce significant changes in the crystallinity of the S2 steatite. Namely, enstatite reflections located at 26° and 32° were constant, only the peak at 27° increased from the starting 225 a.u. to 375 a.u. at 1250 °C. The amount of glassy phase remained unchanged. The results of the XRD analysis of the S3 steatite mixtures are presented in Fig. 3a–d. Unlike the two previous steatite mixtures, the S3 steatite showed a wider variety of crystalline phases at the initial sintering temperatures. At 1000 °C, enstatite, pyrope, forsterite, cristobalite, tridymite, α (low) cristobalite (Crα; JCPDS 89-3606), α-BaAl2Si2O8 (Cα; JCPDS 881048), and sanbornite (S; JCPDS 72-0171) were identified, (Fig. 3a). New phases appeared after sintering at 1100 °C: celsian (C; JCPDS 381450), Ba2SiO4 (*; JCPDS 70-2113), and BaSiO3 (+; JCPDS 70-2112)) (Fig. 3b). These mineral phases, i.e. α-BaAl2Si2O8, Ba2SiO4, BaSiO3,

The most significant XRD reflections (Fig. 1a–d) are located between 20° and 40°. Two main peaks which correspond to enstatite were found at 27° and 32°, respectively. The intensities of these peaks at 1000 °C were approximately the same, counting up to 220 arbitrary units (a.u. in further text) (Fig. 1a). The recorded XRD diffractogram of the S1 sample, treated at 1100 °C (Fig. 1b), shows the same crystalline phases as the previous image (Fig. 1a). However, the crystallinity changed as a result of the increase in the temperature of the treatment. Namely, the enstatite peak located at 27° decreased to 150 a.u., while the corresponding peak at 32° remained unaffected. There were no significant changes in the crystallinity of the remaining less-prominent reflections. After the thermal treatment at 1200 °C, the crystallinity of the S1 steatite notably changed as the majority of reflections increased for at least 20% (Fig. 1c). The enstatite peak that was recorded at 27° increased up to 305 a.u., while the intensity of the reflection located at 32° remained consistent. Also, the increase in the temperature of sintering from 1100 °C to 1200 °C induced the elevation of the peak at 22°, which is a superposed quartz and enstatite reflection, from the starting 100 a.u. to final 160 a.u. Similarly, the peak located at 16°, which predominantly corresponds to quartz and tridymite, increased from 100 a.u. to 175 a.u.. The maximal crystallinity was achieved upon sintering conducted at 1250 °C (Fig. 1d). The intensities of the enstatite peaks located at 22°, 27°, and 32° were 200 a.u., 425 a.u., and 300 a.u., respectively. The base lines of the diffractograms were not elevated and they remained unchanged during the experiment, which refers to the presence of a relatively small amount of the amorphous phase. The diffractograms of the S2 steatite, recorded upon thermal treatments of the powdery mixtures in the interval 1000–1250 °C, are given in Fig. 2a–d. The sintering of the S2 steatite samples conducted at 1000 °C and 1100 °C resulted in identical mineralogical compositions (Fig. 2a, b):

4

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 3. The XRD diffractograms of the S3 steatite after thermal treatment at: a) 1000 °C; b) 1100 °C; c) 1200 °C, and d) 1250 °C.

main enstatite peaks reached values of 310 a.u. and 275 a.u.. At 1250 °C, the reflection at 32° decreased for approximately 40 a.u.. A similar occurrence took place on the reflection located at 22°. This can be a consequence of superposition of enstatite and quartz reflections on the diffractograms. Namely, due to the transformation of a quartz polymorph (cristobalite or tridymite), the corresponding peak shifts or changes which can cause a decrease in length of enstatite reflection. The comparison of the diffractograms in Figs. 1–4 highlighted that enstatite is the most abundant mineral phase because all steatite mixtures (S1-4) were primarily based on talc. Quartz, which originated from either bentonite or kaolin clay, was identified in moderate quantities. Primary quartz modification was detected in samples S1 and S2. Due to the application of BaCaO3 as a flux, the steatites S3 and S4 comprised only high-temperature quartz polymorphs – cristobalite and tridymite. Therefore, barium carbonate, unlike feldspar, influenced the changes in the mineralogical composition of the steatites. Other phases, such as pyrope and forsterite, which also originate from talc, were present in very small quantities in all steatites. Even though the crystalline structure improved with the increasing temperature, the alternations in the applied binders and flux resulted in differences in the degree of crystallinity. At 1250 °C, S1 and S2 steatites showed the higher crystallinity degree than S3 and S4 mixtures. All investigated steatite samples showed a fully crystallized structure at 1250 °C with only a small amount of amorphous phase. The majority of refractions on diffractograms at 1250 °C (Figs. 1d-4d) corresponded to the stable enstatite, which justifies the application of this sintering procedure. The assessment of the thermal behavior of the green steatite powder mixes with respect to the employment of different bonding agents and flux pertained to the second stage of the experiment. Specific thermal effects were recognized and indexed for each sample (S1–4) by means of the differential thermal analysis (DTA) and thermogravimetry (TG). The acquired diagrams are presented in Figs. 5 and 6.

and sanborite (phyllosilicate with the formula Ba(Si2O5)), are all related to BaCO3 which has been applied as the flux in the composition of this ceramic material. The mentioned phases were not present and/ or detectable above 1100 °C (Fig. 3c, d). Celsian (BaAl2Si2O8) disappeared from diffractograms above 1200 °C. The most abundant phase was enstatite. The enstatite corresponding main reflections located at 27° and 32° expanded for 90 a.u. and 50 a.u., respectively, due to the initial increase in the sintering temperature (Fig. 3a, b). The number of identified mineral phases progressively decreased with the increasing temperature (Fig. 3a–d). Quartz was solely present in one of the high-temperature polymorphic transformations: α-cristobalite, cristobalite, and/or tridymite. This means that BaCO3 enabled quicker and more efficient quartz transition. Analogous to the diffractograms of the samples S1 and S2, the crystallinity of the present mineral phases intensified with increasing of the sintering temperature. However, the difference in crystallinity of the samples treated at 1200 °C and at 1250 °C was less considerable. Fig. 4a–d illustrates the X-ray diffraction analysis of the S4 steatite mixtures. Equivalently to the S3 steatite mixture, the S4 steatite maintained the same mineral phase composition up to 1200 °C. The following phases were detected: enstatite, pyrope, forsterite, cristobalite, tridymite, α cristobalite, α-BaAl2Si2O8, sanbornite, celsian, Ba2SiO4, and BaSiO3. Barium carbonate related phases were not identified upon sintering at the projected highest temperature of the investigation, i.e., 1250 °C. Quartz was not present in the phase composition in its primary modification. The crystallinity of the S4 steatite continuously intensified with the increasing temperature of sintering (Fig. 4a–d). At 1100 °C (Fig. 4b), the enstatite reflection identified at 27° increased from 100 a.u to 220 a.u., while the peak located at 37° elevated from 175 a.u. to 235 a.u.. The sample showed an additional improvement in overall crystallinity upon thermal treatment conducted at 1200 °C. Two 5

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 4. The XRD diffractograms of the S4 steatite after thermal treatment at: a) 1000 °C; b) 1100 °C; c) 1200 °C, and d) 1250 °C.

Fig. 5. Differential thermal curves of steatite powders S1–4.

Fig. 6. Thermogravimetry curves of the steatite powders S1–4.

The DTA curve of the steatite S1 comprised three endothermic effects (Fig. 5). Regarding the starting composition, the investigated steatite is primarily based on talc; therefore the identified effects and registered thermal behavior are nominally related to the talc dehydration. Namely, talc is characterized by the presence of both adsorption water and hydroxyl groups that constitute the space lattice elements [35]. Water is normally driven off during three stages dehydration (115° to 200 °C; 350° to 500 °C, and 600° to 1050 °C), which is consequently followed by three endothermic peaks on the DTA curve [35]. The DTA analysis of the S1 sample (Fig. 5) registered the first dehydration stage as an endothermic valley in the interval from 20 °C to 200 °C. This endothermic effect was not characterized by presence of

a noticeable peak. The second stage was not recognized on the DTA curve, because the original mineral composition of talc and its genuine behavior have been altered by the addition of bonding agent and flux. During the initial two stages only 0.4% of water was liberated. Therefore, the corresponding peaks are hardly noticeable and/or small [35], which is in the agreement with the results illustrated in Fig. 5. Since 5.1% of the total non-hygroscopic water was liberated during the third stage [35], more notable effects were expected in this section of DTA curve. The first effect, located at 577 °C, was very small and hardly detectable. It belongs to the thermal range of the third stage of the dehydration process. This effect corresponds to the quartz transformation. Namely, ambient-temperature quartz modification (α-quartz) 6

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

reversibly transforms into β-quartz at 573 °C [35,36]. As it was shown by the XRD analysis, the quantity of quartz was moderate, which resulted in relatively unnoticeable changes in the crystalline structure of steatite and an equally small effect on the DTA curve. The second effect, with the peak located at approximately 668 °C, was slightly more notable. The third endothermic effect was recognized as the most prominent one, and it is situated at 950 °C. The second and the third peak can be related to the crystallization of silica (i.e., quartz – SiO2) as cristobalite, and the crystallization of magnesium metasilicate, respectively [10,35,37]. At 668 °C, talc decomposes into magnesium oxide and silica due to the liberation of constitution water. This reaction would take place at 800 °C in pure talc; however the addition of flux decreased the temperature of decomposition for more than 100 °C. At 1100 °C, the decomposition of pure talc would normally lead to the formation of MgSiO3, i.e., stable enstatite modification - protoenstatite and amorphous SiO2 [10,35,37]. The temperature of 950 °C can be accepted as the initial stage of stable enstatite formation. Further heating up to 1250 °C contributed to the creation of a well-ordered crystalline structure and establishment of the stable enstatite polymorph as the predominant mineral phase in the steatite composition, as it was shown and supported by the results of XRD analysis (Fig. 1a– b). The shape of DTA curve of steatite S2 is an approximate to the shape of S1 diagram with one additional exothermic peak at 987 °C (Fig. 5). The first endothermic effect at 566 °C is slightly larger than that of S1 steatite, probably due to the additional quantity of quartz that originated from kaolin clay. Nevertheless, the second peak (693 °C) diminished almost to the point of undetectability. In the sample S2, cristobalite did not appear in the mineralogical composition before sintering at 1200 °C (Fig. 2c), which justifies the absence of the second peak. An exothermic hump is located in the relatively wide thermal area from 700 °C to 900 °C. This hump gradually proceeds into a significant endothermic effect with its lowest point situated at 957 °C, which can be adopted as the starting temperature of the crystallization of stable enstatite polymorph. The exothermic peak at 987 °C also corresponds to the crystalline rearrangements and might represent the enstatite reversible transition from low-clinoenstatite to high-clinoenstatite [21,22]. The baseline of the S2 curve is elevated in comparison with the S2 differential thermal curve, which is mostly perceptible above 500 °C, and means that the addition of a combination of kaolin clay and feldspar induced the development of a supplemental amount of amorphous phase. The DTA curve of the S3 steatite comprises a huge exothermal hump in the thermal interval 500° to 900 °C. A thin endothermic peak was incorporated in the background hump at 809 °C. This peak belongs to the third stage of talc dehydration and it can be related to the release of the adsorption water or to the reactions that involve hydroxyl groups from the talc space lattice elements. The wide exothermal effect proceeds into another endothermic peak (at 895 °C), which is followed by the exothermic peak at 914 °C. The peaks located above 890 °C can be related to the recrystallization within talc structure and formation of the stable enstatite polymorph. Above 914 °C the diagram proceeds into the endothermic valley. The baselines of S2 and S3 curves are on the approximately same level, which means that the combinations kaolin clay - feldspar and bentonite - BaCO3 formed equal amounts of glassy material that surrounded enstatite crystals. The baseline of the S4 steatite diagram is set significantly higher than rest of DTA curves in Fig. 5. The towering exothermic point of the S4 curve is positioned twice higher than those of steatites S1, S2, and S3, referring to a high portion of amorphous phases formed by addition of a combination of BaCO3 as flux and kaolin clay as binder. The DTA curve of S4 is relatively smooth up to 786 °C, where the first significant endothermic peak is located. The additional peaks were registered – one endothermic at 884 °C and one exothermic at 987 °C, while the entire curve proceeds into an overall endothermic effect. As it was the case with previously discussed steatite variations (S1–3), the peaks on the DTA

curve of the S4 steatite above 780 °C can be related to the crystallization of cristobalite, which was followed by the crystallization of magnesium metasilicate. The results of the thermo-gravimetric analysis are introduced and illustrated in Fig. 6. As it is indicated in Fig. 6, in the interval from 20 °C to 180 °C, the TG diagrams of S1 and S3 steatite powders have small initial mass losses counting up to 0.5% and 0.8%, respectively. This interval corresponds to the first dehydration stage of talc. The diagrams of the S2 and S4 steatites remained unchanged during this thermal period. Afterwards, the S1 and S3 diagrams followed an infinitesimal decreasing trend up to 550 °C. This is also in agreement with the previously given statement about the secondary phase of talc dehydration taking place in the thermal range from 350 °C to 500 °C. Regarding the fact that steatite is not based on pure talc (instead, it comprises additional binder and flux), these intervals were either shifted or extended from the original thermal ranges that are characteristic for the talc dehydration. From 550 °C to 720 °C, the S1 steatite sample underwent a mass loss (0.7%), after which the diagram proceeded as a straight line up to 840 °C. These changes were related to the cristobalite crystallization from the initial silica, and followed by the magnesium metasilicate transformations. From 840 °C to 1030 °C, the S1 steatite sample showed a 4.4% mass loss. This thermal interval can be considered as a period when the transformation of enstatite into its stable polymorph was initiated. Unlike the S1 steatite, the diagram of the S3 sample proceeded into a steep decreasing slope above 550 °C. The total mass loss for this sample, in the interval 550–1100 °C, counted up to 6.6%. The S1 and S3 steatites have the same bonding agent – bentonite, but different fluxing additives. BaCO3 addition produced a continual mass decrease from 550 °C to 1100 °C, and also induced a higher mass loss (for 2.2%). The addition of feldspar induced a two-step mass decrease: the first one being in the 550–720 °C interval and the second during the interval 840–1030 °C. The total mass loss for the S3 steatite was higher (7.4%) in comparison to the S1 steatite (5.3%). The S2 steatite showed two major intervals regarding the mass loss. The first interval started at 460 °C and ended at 870 °C; while the second interval was counted from 870 °C to 1030 °C. The first interval is in agreement with the results of the DTA analysis (Fig. 5), i.e., the recorded exothermal hump at 500–900 °C, which might correspond to the quartz polymorphic transformations superposed with reversible changes in the mineral enstatite. The second interval is congruent with the period of the talc final dehydration and the initiation of the recrystallization of the stable high-temperature enstatite forms. The registered mass losses were 1.7% during the first interval, and 3.6% during the second interval. The diagrams of the S2 and S4 samples followed identical patterns up to 280 °C. The S4 steatite proceeded into a single step mass loss above 280 °C, which counted up to 7.5%. Analogously to the two previous steatite samples (i.e., S1 and S2), feldspar induced a two-step mass loss in the S2 and S4 steatites, while BaCO3 gave continual, but higher total mass losses: 5.3% and 7.5%. Regarding the total mass loss, the S1 and S2 steatites, i.e., the S3 and S4 steatites were comparable due to the application of the same fluxing agent. The application of feldspar initiated a two-step mass loss, while BaCO3 was responsible for the continual mass loss. The samples S1 and S3, i.e., S2 and S4 can be comparable due to the application of the same bonding agents. These pairs of samples had the same thermogravimetric behavior and approximately the same mass losses up to 700 °C. However, above 700 °C the given pairs of the steatite mixtures showed divergence in their thermal behavior. This resulted in a higher mass loss recorded in the case of the S3 steatite, i.e., the S4 steatite in the other pair. Since the bonding agent was the same, the higher mass loss was induced by the flux, which was BaCO3 in both cases. The SEM microstructural analysis was conducted on the parts of the crushed steatite samples S1–4, which were previously sintered at 1000 °C and 1250 °C. The SEM micro-photographs are presented in Fig. 7. 7

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 7. The SEM analysis of steatites: S1 sintered at a) 1000 °C and b) 1250 °C; S2 sintered at c) 1000 °C and d) 1250 °C; S3 sintered at e) 1000 °C and f) 1250 °C; and S4 sintered at g) 1000 °C and h) 1250 °C.

crystals, acquire the protoenstatite form of stabilization, and enable additional reinforcement for the newly crystallized stable enstatite particles by creating a glassy matrix. Obtaining such a microstructure guarantees a level of certainty that there will be no reversible phase transformations which would induce cracking within the sintered ceramic material. The S1 steatite sample, sintered at 1000 °C (Fig. 7a), comprised a mixture of differently shaped particles with diameters ranging in a wide interval from 2.08 to 22.47 µm. The particles correspond to different mineral species, i.e., enstatite, quartz, cristobalite, tridymite, and pyrope, as it was previously identified. The majority of the particles,

Since the base for the synthesis of steatite comprised 80 wt% of the talc mineral mixture, presumption is that all untreated (i.e., nonsintered) steatite samples predominantly adopted a shell silicate structure with triclinic symmetry characteristic for talc [10]. The increasing temperature normally induces the formation of micropores, that usually size up to 500 Å, and the appearance of free SiO2 (quartz) in the talc structure, which are both unfavorable occurrences [10]. Also, orthorhombic enstatite that originates from talc transforms during the thermal treatment into a number of polymorphs, some of which are reversible. Therefore, the sintering procedure sought to decrease the microporosity within talc minerals, reduce the quantity of quartz 8

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

induced an increase in the average particle diameter. Namely, the diameters of the smaller particles were in the range 0.94–2.41 µm, while most of the monolithic blocks sized from 8.5 to 15.80 µm. The fully sintered S2 sample had extremely well defined enstatite crystals, and a very low porosity. The sample S3 showed a very high open porosity upon sintering at 1000 °C (Fig. 7e). It is possible that gaseous carbon dioxide from BaCO3 was released during high-temperature reactions and thereby contributed to the formation of the additional share of pores. The present pores were irregularly shaped, sizing 1– 5 µm. The increased porosity at 1000 °C also can mean that the mass transportation did not take place and that densification is yet to occur at higher temperatures. At this temperature, the detected crystallinity and the amount of glassy phase were on the same levels as in the S2 sample (Figs. 3a, 6). The dimensions of the particles varied from very small ones (0.96 µm) to extremely big elongated grains (32.08 µm). Small particles correspond to a quartz high temperature modification (cristobalite or tridymite) or one of the Ba related minerals. The huge particles of enstatite were mostly sheet-like assembled and elongately shaped. After sintering at 1250 °C (Fig. 7f), a change in the shape of pores took place (i.e., coalescence) as they merged and became even bigger and rounder (5–15 µm). Also, the densification within the solid material occurred, producing a specific structure that comprised a high amount of amorphous phase in which enstatite crystals are submerged. Therefore, the application of BaCO3 as flux contributed to the formation of an increased amount of amorphous matter and reduced the sintering temperature below 1250 °C; however it created a porous structure within the observed ceramic material. After sintering at 1000 °C, a multiphase system was created within the sample S4 (Fig. 7g). This composite system comprised several types of particles: smaller polygonally shaped enstatite crystals with average diameter of approximately 5 µm, big enstatite grains sizing from 10 µm to 20 µm, and small oval-shaped particles of approximately 0.5 µm with strong tendency towards agglomeration (quartz polymorphs or Ba-related minerals). Unlike the sample S3, the steatite S4 did not have increased porosity at 1000 °C. This means that BaCO3 and kaolin clay did not induce an intense high temperature reaction that resulted in the CO2 release as it was the case with BaCO3 and bentonite. At 1250 °C (Fig. 7h), coarser and micro particles present in steatite S4 merged to form a less layered structure. However, the S4 sample sintered at the maximal temperature showed increased porosity, probably due to the earlier formation of glassy phase which was induced by addition of BaCO3 as a flux (Fig. 6). This resulted in a lack of the amorphous material for the reinforcement of the newly crystallized stable enstatite and most likely is to lead to the degradation in dielectric properties. The EDS analysis of the characteristic points on the SEM microphotographs is given in Fig. 8. The conducted EDS analyses support the results of the chemical (Table 1) and mineralogical (Fig. 1–4) investigations of the steatite samples S1–4. The most abundant chemical elements detected in all investigated samples were silicon ( > 22%) and magnesium ( > 12%). Si and Mg are chemically merged in magnesium silicate, i.e., one of the mineralogical modifications of talc (mineral enstatite). Aluminum is present in a small quantity 1–3%, probably as Al2O3. The aluminum originates from the applied binder – bentonite or kaolin clay. The percentage of elemental oxygen is around 40–50%. The oxygen is present in the form of various oxides as it was previously shown in the chemical analysis (Table 1). Potassium is present in very small amounts, i.e., below 1%. The EDS elemental mapping on the SEM images gave an insight into the spatial arrangements of individual chemical elements in the recorded steatite samples upon sintering at 1000 °C and 1250 °C. Thereby, it can be seen that the majority of the element-related points on the microphotographs belong to Mg, Si, and O which originate from talc. These elements are constituents of the mineral enstatite. Mg, Si, and O are homogenously distributed on all images. Al dots are separately banked as they originate from bentonite or kaolin clay,

present at 1000 °C, are related to the mineral enstatite, which is endorsed by the results of the XRD analysis (Fig. 1a). The stable enstatite polymorph normally comprises massive prismatic crystals, which is in agreement with the measured diameters of the particles. Namely, the average diameter of a particle in the S1 steatite sample is set at approximately 9.5 µm, which relates to the average dimension (≤ 10 µm) of the stable enstatite polymorph – protoenstatite [16,17]. According to the literature [10], enstatite crystals are characterized by a lamellar or fibrous structure. A cluster of massive, lamellarly constituted particles was registered in the upper left corner of the SEM microphotograph presented in Fig. 7a. As it was indexed via the X-ray diffractometry (Fig. 1a), the sample S1 sintered at 1000 °C had low abundances of quartz and its high-temperature modifications – cristobalite and tridymite. Microcrystalline particles that correspond to quartz, cristobalite, and tridymite were noticed on the surface of large enstatite particles and agglomerated in various cavities. Smaller spherical to rounded particles (i.e., spherulites) visible on the SEM image correspond to the cristobalite, which normally appears in a tetragonal–trapezohedral crystal system [38]. Orthorhombic tridymite was present in platy to sheet forms or as pseudo-hexagonal crystals [38], which were set within cavities of the observed steatite sample. Quartz, which is usually prismatically or pyramidally shaped and finegrained [38], was also present within the bulk of agglomerations created from the smallest particles. The grain boundaries between large particles appear to be feeble at 1000 °C, as it was the lowest sintering temperature. Therefore, at this point, the connections between particles are still weak, which results in low physico-mechanical properties as well as unsatisfactory dielectric characteristics of the synthesized material. Also, the microstructural analysis of the S1 sample at 1000 °C pointed out to a relatively increased porosity. The pores were relatively wide and set between grains; thereby the observed voids do not belong to the category of micropores and by such they are not the result of the talc decomposition, but more feasibly the product of an incomplete sintering procedure. After sintering of the S1 steatite sample conducted at the 1250 °C (Fig. 7b), the particles retained approximately the same dimensions: 3.46–21.66 µm. Also, the shape of the present particles did not vary with the increasing temperature. Therefore, the mineral phase composition of S1 did not alter in its quality, only the quantity of certain phases was either increased (e.g., enstatite) or decreased (e.g., quartz). Small particles, with the diameters lesser than 1 µm, were mutually merged and additionally consolidated with “coarser” grains during the sintering procedure. Also, at elevated temperatures, a certain amount of amorphous matter was created, as it was detected by the DTA analysis given in Fig. 6. The glassy phase filled out the open pores present in the material and stabilized newly formed enstatite crystals, and thereby contributed to the fabrication of a very compact ceramics with no signs of structural cracking or similar defects. The microstructural analysis of the visible surfaces of the samples S2, S3, and S4, after sintering at two prescribed temperatures, showed small variations from the sample S1 with very few identified differences. The SEM microphotograph of the S2 steatite, after sintering at 1000 °C (Fig. 7c), gave a very good insight into the size differences between particles that belong to different mineral phases. The particles that correspond to quartz and its high-temperature modifications were rather small, with diameters sizing from 0.8 μm to 1 µm. Several times bigger particles, or so-called “monolithic blocks”, can be related to enstatite crystals. While the micro-particles from the first group were rounded or prismatically shaped, the blocks comprised levels of lamellar structures and sized around 4.5–5 µm. Also, there were even bigger blocky agglomerations of enstatite crystals sizing up to 16.5 µm. The distribution of the particles within the S2 steatite was more homogenous than in the previous sample (S1). The registered porosity was insignificant. After sintering at 1250 °C (Fig. 7d), the initial shapes of the particles remained the same; however the density of the S2 structure improved. The particles merged due to the sintering, which 9

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 8. EDS mapping (left), and elemental analyses (right) of the steatite samples from Fig. 7a–h.

coaxial chamber was around 0.01. The loss tangent measured using the open test fixture included losses due to the radiation, which increased with increasing frequency. Thus, the accuracy of these results deteriorated at frequencies above approximately 200 MHz. Hence, it can be concluded from the diagrams illustrated in Fig. 9 that the dielectric losses in all samples are small, i.e., below the measurement threshold, which benefits and approves the utilization of the experimentally synthesized steatites S1–4 as dielectric materials in electrical engineering. Table 2 summarizes the relative permittivity (εr) of the four steatite mixtures, for all sintering temperatures (1000 °, 1100 °, 1200 ° and 1250 °C), measured at 200 MHz using the coaxial chamber. The relative permittivity of the samples sintered at 1250 °C was higher than the relative permittivity of the samples thermally treated at 1000 °C, which can be considered as the effect of the sintering. Namely, it has been widely found and accepted that higher densities result in the enhancement of the dielectric constant. Also, as it is well known, the sintering causes microstructural changes and rearrangements regarding the density and the amount of open and closed pores. This, in return, performs a significant influence on the permittivity as a final output for an insulation ceramic material [39]. Therefore, the higher relative permittivity can be attributed to the higher crystallinity of the samples fired at 1250 °C. The measured relative permittivities of the steatite samples S1, S2, and S3 were similar, i.e. all vales were above

and in the steatite mixtures they represent the binder. Accumulations of Al dots in the images might indicate the presence of the smallest particles (usually quartz) that are mutually connected in agglomerations by binder. With increasing temperature Al dots became more homogenously scattered, especially in the samples S2 and S3. Also, the sintering of the powder mixtures influenced the increase in the quantity of Mg, Si, and O, which can be related to the increase in crystallinity of enstatite and formation of new stable enstatite polymorphs during sintering at 1250 °C. The dielectric properties were evaluated for the samples sintered at various temperatures from 1000 °C to 1250 °C, for all four steatite mixtures. The reflection coefficient was measured in the frequency range between 1 MHz and 1 GHz, and the corresponding relative permittivity (εr) and the loss tangent (tan δ) of the tablets were computed. The results are shown in Fig. 9. The measurement error of the results for low frequencies is increased due to the very high impedances of the samples; the reflection coefficient for such impedances cannot be measured accurately. The accuracy of the results for high frequencies was deteriorated because the range of validity of the electrostatic models was exceeded. Hence, reliable results were obtained for the frequency range from 50 MHz to 500 MHz. In this range, the error of the results for the real part of the relative complex permittivity was below 5%. Furthermore, the minimal loss tangent that can be reliably measured using the

10

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

Fig. 9. The relative permittivity and the loss tangent, as a function of frequency, for the mixtures S1–4 fired at 1000 °C and 1250 °C: (a) S1 at 1000 °C, (b) S1 at 1250 °C, (c) S2 at 1000 °C, (d) S2 at 1250 °C, (e) S3 at 1000 °C, (f) S3 at 1250 °C, (g) S4 at 1000 °C, and (h) S4 at 1250 °C.

temperature were noticed in microphotographs of the majority of the sintered samples (i.e., S1–3). The sample S4 exhibited less visible microstructural changes, yet its permittivity increased between sintering at 1000 °C and 1250 °C. Therefore, in all cases the microstructural rearrangements were followed by the increase in the relative permittivity. Also, the samples S1–S3 had higher level of crystallinity than sample S4, as it was detected by the XRD analysis (Figs. 1–4). In the sample S4, amorphous phase was created too early as a result of the application of BaCO3 as flux. The glassy material failed to connect the newly crystallized enstatite and formed a porous structure, which gave a low relative permittivity. By observing the crystallinity and microstructural development induced by sintering, as well as the dielectric behavior, it can be said that talc – feldspar - kaolin clay (S2) and talc bentonite - BaCO3 (S3) are the optimal combinations of the raw materials for production of the steatite ceramics with an acceptable dielectric properties for practical applications.

Table 2 The relative permittivity (εr) of the steatite mixtures S1–4. Firing temperature

1000 °C 1100 °C 1200 °C 1250 °C

Steatite mixture S1

S2

S3

S4

4.61 4.75 5.07 5.16

4.55 4.62 5.12 5.19

4.55 4.71 5.20 5.25

4.32 4.49 4.83 4.95

5.15. However, the permittivities of the S1–3 steatites were higher than the relative permittivity of the sample S4 (4.95). The obtained results are in agreement with the recorded microstructures, presented in Fig. 7. Namely, in most cases the porosity was higher after sintering at the lower temperature (i.e. 1000 °C). The decrease in the porosity and an increase in density induced by the elevation of the sintering 11

Ceramics International xxx (xxxx) xxx–xxx

A. Terzić et al.

4. Conclusion [11]

This study provided an insight to the interconnections of sintering parameters, microstructural and crystalline development, and the changes in the dielectric behavior of four steatite types fabricated from low-cost raw materials. The steatite mixtures exhibited a tendency for establishing of the enstatite mono-phase system with the increase in temperature, as highlighted by the X-ray diffraction analysis. At 1250 °C, a stable polymorph of enstatite (i.e., protoenstatite) was detected as the predominant mineral phase, accompanied only by small abundances of SiO2 and/or Ba-related phases. The minor phases did not cause the deterioration in the steatite dielectric behavior. All steatite mixtures showed the increase in relative permittivity, from min. 4.32 to max. 5.25, registered during the sintering interval from 1000 °C to 1250 °C. The dielectric properties improved due to microstructural changes and rearrangements regarding density and the amount of open and closed pores. The increase in relative permittivity is also attributed to the improvement in crystallinity. The differential thermal analysis identified partially modified talc three-stage dehydration, with effects located above 890 °C that were related to the recrystallization within the talc structure and formation of the stable enstatite polymorph. An amorphous phase that appeared between 500 °C and 900 °C was responsible for the stabilization of newly formed protoenstatite crystals, the elimination of reversible reaction, as well as the steatite structure reinforcement which contributed to the prevention of cracking and deterioration. By observing the crystallinity and microstructural development induced by sintering, as well as the dielectric behavior, it can be concluded that talc – feldspar - kaolin clay (S2) and talc - bentonite - BaCO3 (S3) are the optimal combinations of the raw materials for production of the steatite ceramics with appropriate dielectric properties for practical electro-technical applications.

[12] [13]

[14]

[15] [16] [17] [18] [19]

[20]

[21] [22] [23] [24]

[25]

[26]

[27]

Acknowledgements [28]

This investigation was supported by the Serbian Ministry of Education, Science and Technological Development and it was conducted under the following projects: ON 172057 and III 45008. The authors would like to express their gratitude to Dr. Smilja Marković, Institute of Technical Sciences of SASA, Belgrade, Serbia, for providing help with the thermal analysis.

[29]

[30]

References

[31]

[1] E. Vela, M. Peiteado, F. Garcia, A. Caballero, J. Fernandez, Sintering behaviour of steatite materials with barium carbonate flux, Ceram. Int. 33 (2007) 1325–1329. [2] W. Carty, U. Senapati, Porcelain: raw materials, phase evolution and mechanical behaviour, J. Am. Ceram. Soc. 81 (1) (1998) 3–20. [3] J. Liebermann, Microstructure, properties and product quality of strength stressed high voltage insulators, Am. Ceram. Soc. Bull. 82 (2) (2003) 39–46. [4] P. Rohana. Neufuss, J. Matejícek, J. Dubský, L. Prchlık, C. Holzgartner, Thermal and mechanical properties of cordierite, mullite and steatite produced by plasma spraying, Ceram. Int. 30 (2004) 597–603. [5] H. Soykan, Low-temperature fabrication of steatite ceramics with boron oxide addition, Ceram. Int. 33 (2007) 911–914. [6] M. Valaškov, J. Zdrálková, J. Tokarský, G. Martynková, M. Ritz, S. Študentová, Structural characteristics of cordierite/steatite ceramics sintered from mixtures containing pore-forming organovermiculite, Ceram. Int. 40 (2014) 15717–15725. [7] A. Terzić, Lj Andrić, J. Stojanović, N. Obradović, M. Kostović, Mechanical activation as sintering pre-treatment of talc for steatite ceramics, Sci. Sinter. 46 (2) (2014) 247–258. [8] L. Munoz, S. Cava, C. Paskocimas, Modelling of the vitrification process of ceramic bodies for whiteware, Ceramica 48 (308) (2002) 217–222. [9] P. Ctibor, J. Sedlacek, K. Neufuss, J. Dubsky, P. Chraska, Dielectric properties of plasma-sprayed silicates, Ceram. Int. 31 (2) (2005) 315–321. [10] W. Mielcarek, D. Nowak-Woźny, K. Prociów, Correlation between MgSiO3 phases

[32]

[33]

[34] [35] [36]

[37]

[38] [39]

12

and mechanical durability of steatite ceramics, J. Eur. Ceram. Soc. 24 (2004) 3817–3821. W. Lee, A. Heuer, On the polymorphism of enstatite, J. Am. Ceram. Soc. 70 (1987) 349–360. A. Shortall, X. Hu, P. Marquis, Potential countersample materials for in vitro simulation wear testing, Dent. Mater. 18 (3) (2002) 246–254. M. Ghazal, B. Yang, K. Ludwig, M. Kern, Two-body wear of resin and ceramic denture teeth in comparison to human enamel, Dent. Mater. 24 (4) (2008) 502–507. H. Zhawi, M. Kaizer, A. Chughtai, R. Moraes, Y. Zhang, Polymer infiltrated ceramic network structures for resistance to fatigue fracture and wear, Dent. Mater. 32 (11) (2016) 1352–1361. J. Jiao, X. Liu, W. Gao, C. Wang, H. Feng, X. Zhao, L. Chen, Two-step synthesis of witherite and tuning of morphology, Mater. Res. Bull. 45 (2010) 181–185. J.V. Smith, The crystal structure of protoenstatite, MgSiO3, Acta Crystallogr. 12 (1959) 515–519. B. Reynard, J.D. Bass, J.M. Jackson, Rapid identification of steatite–enstatite polymorphs at various temperatures, J. Eur. Ceram. Soc. 28 (2008) 2459–2462. M.D. Rigterink, Microscopic and X-ray investigation of some steatite bodies, J. Am. Ceram. Soc. 30 (7) (1947) 214–218. H. Gökce, D. Agaogulları, M. Lütfiovecoglu, I. Dumana, T. Boyraz, Characterization of microstructural and thermal properties of steatite/cordierite ceramics prepared by using natural raw materials, J. Eur. Ceram. Soc. 31 (2011) 2741–2747. P. Ptáceka, K. Lang, F. Soukal, T. Opravil, E. Bartonícková, L. Tvrdík, Preparation and properties of enstatite ceramic foam from talc, J. Eur. Ceram. Soc. (2014) 515–522. H. Thurnauer, A.R. Rodriguez, Notes on the constitution of steatite, J. Am. Ceram. Soc. 25 (15) (1942) 443–445. W.L. Brown, N. Moromoto, J.W. Smith, A structural explanation of the polymorphism and transition of MgSiO3, J. Geol. 69 (1961) 609–616. G.A. Rankin, H.E. Merwin, The ternary system MgO–Al2O3–SiO2, Am. J. Sci. 45 (1918) 301–325 (4th series). M. Tribaudino, A transmission electron microscope investigation of the C2/c-P21/c phase transition in clino pyroxenes along the diopside– enstatite (CaMgSi2O6– Mg2Si2O6) join, Am. Mineral. 85 (2000) 707–715. A. Goel, D. Tulyaganov, S. Agathopoulos, M. Ribeiro, J. Fereirra, Synthesis and characterization of MgSiO3-containing glass-ceramics, Ceram. Int. 33 (2007) 1481–1487. E. Sentürk, S. Duman, S. Bagcıa, H.S. Soykan, Z. Aslanogluca Sakarya, Humidity sensing properties of steatite ceramic containing B2O3, Sens. Actuators A 240 (2016) 80–84. L. Zhang, S. Olhero, J. Ferreira, Thermo-mechanical and high-temperature dielectric properties of cordierite-mullite-aluminaceramics, Ceram. Int. 42 (2016) 16897–16905. Y. Liu, F. Luo, J. Su, W. Zhou, D. Zhu, Z. Li, Enhanced mechanical, dielectric and microwave absorption properties of cordierite based ceramics by adding Ti3SiC2 powders, J. Alloy. Compd. 619 (2015) 854–860. N. Obradović, S. Filipović, N. Đorđević, D. Kosanović, S. Marković, V. Pavlović, D. Olćan, A. Đordjević, M. Kachlik, K. Maca, Effects of mechanical activation and two-step sintering on the structure and electrical properties of cordierite-based ceramics, Ceram. Int. 42 (12) (2016) 13909–13918. H. Li, Z. Huang, L. Cheng, S. Kong, S. Liu, Structure and dielectric properties of novel low temperature co-fired Bi2O3-RE2O3-MoO3 (RE = Pr, Nd, Sm, and Yb) based microwave ceramics, Ceram. Int. 43 (5) (2017) 4570–4575. Y. Wang, K. Miao, W. Wang, Y. Qin, Fabrication of lanthanum doped BaTiO3 finegrained ceramics with a high dielectric constant and temperature-stable dielectric properties using hydro-phase method at atmospheric pressure, J. Eur. Ceram. Soc. 37 (6) (2017) 2385–2390. Z. Xu, H. Qiang, Y. Chen, Z. Chen, Microstructure and enhanced dielectric properties of yttrium and zirconium co-doped CaCu3Ti4O12 ceramics, Mater. Chem. Phys. 191 (2017) 1–5. A. Đorđević, J. Dinkić, M. Stevanović, D. Olćan, S. Filipović, N. Obradović, Measurement of permittivity of solid and liquid dielectrics in coaxial chambers, Microw. Rev. 22 (2) (2016) 3–9. J.R. Smyth, Experimental study on the polymorphism of enstatite, Am. Miner. 59 (1974) 345–352. M. Wesolowski, Thermal decomposition of talc: review, Thermochim. Acta 78 (1984) 395–421. A. Terzić, L. Pezo, Lj Andrić, M. Arsenović, Effects of mechanical activation on the parameters of talc quality for ceramics production – chemometric approach, Compos. B Eng. 79 (2015) 660–666. A. Wiewiora, S. Sanchez-Soto, M. Avilos, A. Justo, L. Perez-Maqueda, J. PerezRodriguez, P. Bylina, Talc from Puebla de Lillo Spain I - XRD study, Appl. Clay Sci. 12 (1997) 233–245. Web Site: 〈http://webmineral.com〉. X. Kuang, X. Jing, Z. Tang, Dielectric loss spectrum of ceramic MgTiO3 investigated by AC impedance and microwave resonator measurements, J. Am. Ceram. Soc. 89 (2006) 241–246.