Mullitization process of andalusite concentrates – Role of natural inclusions - 1-s2.0-S0272884213013023-main

(1)

CERAMICS

INTERNATIONAL

Ceramics International 40 (2014) 5129–5136

Mullitization process of andalusite concentrates – Role of natural inclusions

Teresa Wala

^a,n

, Bronisław Psiuk

^a

, Jerzy Kubacki

^b

, Katarzyna Stec

^a

, Jacek Podwórny

^a

aRefractory Materials Division, Institute of Ceramics and Building Materials, Toszecka 99, 44-100 Gliwice, Poland

bA. Chełkowski Institute of Physics, University of Silesia, Universytecka 4, 40-007 Katowice, Poland

Received 14 August 2013; received in revised form 10 October 2013; accepted 10 October 2013 Available online 9 November 2013

Abstract

In the article the results of investigations into the mullitization process of two andalusite concentrates are presented. XRD, XRF and ICP investigations revealed nearly identical phase and chemical composition of both concentrates. However, the andalusites considerably differ in the rate of the mullitization process. XPS and LM investigations revealed that the factors responsible for mullitization include not only the content of mineral inclusions and grain size distribution but also impurities dispersion and the content of naturally occurring carbon in andalusite grains.

Keywords:B. Inclusions; Andalusite; Mullitization; XPS study; Microscopic study

1. Introduction

Andalusite Al₂SiO₅ crystallizes in an orthorhombic system in Pnnm space group. It occurs in a form of euhedral, pseudotetragonal columnar crystals, characterized by good

ssility, parallel to plane (110) and slightly poorer along plane (100) [1,2]. Theoretically, its chemical composition contains 62.93% of Al₂O₃and 37.07% of SiO₂, but in nature it occurs with various impurities, for example carbon-rich andalusite.

When carbon is specically located in a form of a cross visible in the crystal along (001) plane, it is known as chiastolite[3].

The biggest deposits of mined andalusite are metamorpho- genic (primary) deposits, related to plutonism. Another exten- sively exploited type are supergenic (secondary), elluvial and alluvial deposits. The above mentioned types of deposits may differ in their age of mineralization. The term“mineralization”

is understood as a process of andalusite formation from rocks which are rich in aluminum and poor in silica, referred to as metapellits, in the conditions of contact metamorphism of igneous intrusions[4].

The interest in andalusite is related to the fact that it is a useful raw material to be applied in the ceramic industry, including high-temperature applications (refractory materials).

The usefulness of the mineral in question results mainly from

its properties such as low thermal conductivity (the lowest in the group of silimanite minerals) [5], a possibility of using andalusite in a raw state for ceramic mixes, both in ne and coarse fractions, as well as its relatively low price[6–8]. From the point of view of refractory applications, an extremely important phenomenon is thermally activated transformation of andalusite into mullite. Mullite Al₆Si₂O₁₃is a phase characterized by high melting point and good thermo-mechanical properties, resistance to chemical corrosion and thermal shocks [9]. As a formality, it should be added that the mullitization process is accompanied by a release of amorphous phase rich in SiO₂ [10]. It is assumed that the process of andalusite mullitization is a diffusion assisted phase transformation with silica dissolution [6,11]. The andalusite mullitization process was investigated by many authors and in various experiments a high number of results were obtained for the process temperature (1100–14801C) [12–16]. In general, on the basis of literature review, it can be stated that transition temperature and kinetics depend on the grain size distribution and the degree of raw andalusite material purication [15,17,18].

Undoubtedly, the type and content of impurities that accom- pany andalusite are important. Impurities can be related to the genesis of the deposit[19].

It is worth emphasizing that in the analysis of andalusite transition into mullite the attention was focused mainly on the total content of impurities related to mineral inclusions,

www.elsevier.com/locate/ceramint

http://dx.doi.org/10.1016/j.ceramint.2013.10.036

nCorresponding author.

E-mail address:[email protected] (T. Wala).

(2)

while the manner of their distribution was ignored. Neither discussed the role of carbon naturally occurring in andalusite grains as a factor which might inuence the mullitization process. In this article we would like to discuss both above mentioned issues. In order to do that, we analyzed 2 andalusite concentrates differing considerably in their susceptibility to mullitization but having a very similar chemical composition as well as grain size distribution.

In search of factors that are evidently different for both above quoted andalusites, several experimental techniques were applied: optical light microscopy LM, X-rayuorescence spectrometry XRF, X-ray diffraction XRD, plasma emission spectrometry ICP and X-ray photoelectron spectroscopy XPS.

The aim of this article is to enrich the discussion on factors which may play a signicant role in the process of andalusite phase transformation into mullite, which is of great practical importance.

2. Experimental

2.1. Investigated andalusite concentrates

Commercial concentrates of andalusite from two separate deposits, differing in their genesis and mineralization age, were used as a research material: the elluvial deposit from the Republic of South Africa, hereinafter referred to as andalusite A, and the autochthonous deposit from France, hereinafter referred to as andalusite B. Admittedly, andalusite A comes from a secondary, elluvial deposit but its primary mineralization is related to igneous granite intrusion of Bushveld complex, which occurs in Transwal in South Africa, the age of which is estimated as 1.95 billion years (1.95 Ga)[20]. The deposit of andalusite B is related to granite intrusion in the Central Armorican Massif in France, whose age is evaluated as 340 million years (340 Ma)[21].

2.2. Testing methods

Investigations into the raw materials were conducted by various research techniques. Phase composition was determined on the basis of XRD powder diffraction using an X'Pert PRO MPD(Cu Kα) diffractometer produced by PANalytical.

Chemical composition was determined by the XRF method, using a Magi'X spectrometer produced by PANalytical.

Investigations into the kinetics of andalusite mullitization were carried out by means of the same difractometer, using an Anton Parr HTK 2000 high-temperature chamber. The kinetic research methodology has been presented in detail in an earlier study [22]. Grain size distribution of the powdered concentrates was controlled by the low angle laser light scattering method (LALLS), using a Mastersizer S analyzer produced by Malvern Instruments Ltd. Trace impurities were identied by inductively coupled plasma/optical emission spectrometry (ICP/OES), using a JY36 spectrometer produced by Jobin Yvon. The electron structure of main elements in the samples was compared by X-ray photoelectron spectroscopy (XPS), using PHI 5700 (AlKα) equipment produced by Physical

Electronics. Measurements taken by this method were also used to determine the atomic concentration of samples.

Microstructural investigations by light microscopy (LM) were conducted using MeF2 microscopes produced by Reichert and NU microscopes produced by Karl Zeiss Jena.

3. Results and discussion

3.1. Characteristics of the raw materials

As mentioned in the introduction, andalusites A and B considerably differ in their susceptibility to mullitization. The results of degree of conversion α_B determined at 13501C, 14001C and 14501C are presented in a form of αB¼f(t) dependence inFig. 1. Transformation of andalusite B is evidently slower – lower values αB are reached for the same periods of transformation compared to andalusite A. Kinetic measurements of the mullitization process were carried out for powdered samples with grain size distribution presented inFig. 2.

Degree of conversion αB was expressed by a change in the number of moles of andalusiteΔnin relation to the initial number of andalusite molesΔn₀αB¼Δn=n₀. The kinetic research methodology has been presented in detail in an earlier manuscript[22].

Considerable differences in the mullitization process were observed despite the fact that the examined andalusite concentrates had a very similar chemical composition, mineral composition and grain size distribution (Tables 1–4 and Fig. 2). The results of investigations into the functional properties of both andalusites were presented earlier, in separate publications [19,22], therefore, they will not be discussed in detail at this point. Comparisons of raw material quality requirements obtained by XRF (chemical composition) and XRD (phase composition) as well as grain size distribution of powders for which measurements of the mullitization process kinetics were taken will be once more presented together for the sake of discussion clarity.

The results in Tables 1–4 give an “average picture” of both investigated andalusite concentrates, as they were obtained in measurements taken for averaged powder samples (ground grains).

Fig. 1. Time dependent degree of conversion of andalusite A and andalusite B concentrates obtained at 13501C, 14001C and 14501C.

(3)

XRD investigations (Table 1) revealed an identical qualitative and quantitative phase composition (within the uncertainty range). The tested materials had very similar grain curves (Fig. 2). Subtle differences in chemical composition determined by the XRF method concerned the content of calcium, potassium and ignition losses (Table 2). They cannot account for such a different course of the mullitization process, though.

In order tond differences, the powders were subjected to further tests by ICP spectroscopy. However, ICP investigations revealed that on the ppm level both andalusite concentrates had the same type of impurities (Table 3). It has to be added that the copper impurity might have been due to the manner of preparation. The ground grains of minerals were passed through a sieve with a copper mesh. Taking into consideration

the presence of the same type of impurities having a similar content in the matrix of both andalusites, an attempt was undertaken to nd other potential factors responsible for the different behavior of both concentrates. In order to do that, XPS and light microscopy investigations were conducted.

3.2. Photoemission and microscopic investigations into andalusites

XPS investigations into powdered andalusites conrmed the similarity of phase composition in andalusite A and andalusite B (Fig. 3andTable 5). Apart from the survey spectrum, which provided a basis for calculating the atomic concentration, the spectra of elements with the highest content in the sample were

Fig. 2. Grain size distribution of andalusite A and B.

Table 1

Qualitative and quantitative mineral composition of andalusite A and andalusite B from XRD.

Mineral composition Chemical formula Quantitative contribution [wt%]

Andalusite A Andalusite B

Andalusite Al2SiO5 95.370.3 95.470.4

Quartz SiO2 2.870.1 2.770.1

Muscovite 2M1 (K, Na)Al2(Si,Al)4O10(OH,F)2 1.570.3 1.870.3

Kaolinite 1Md Al2Si2O5OH4 0.370.1 0.170.1

Rutile TiO2 Trace amount Trace amount

Table 2

Chemical composition of andalusite A and andalusite B from XRF.

Andalusite Chemical composition based on oxides [wt%]

L.O.I. (10251C) SiO2 Al2O3 Fe2O3 CaO MgO K2O Na2O TiO2 P2O5 Cr2O3

A 1.01 38.83 58.61 0.78 0.16 0.07 0.33 0.06 0.13 0.01 0.02

B 0.54 39.35 58.86 0.70 0.07 0.07 0.21 0.03 0.16 0.01 0.02

Theoretical – 37.07 62.93 – – – – – – – –

(4)

measured in high resolution mode.Figs. 4and5present Al2p, Si2p, O1s and C1s spectra. The position of the spectra on the binding energy scale was calibrated on the basis of the position of peak Si2p. The Si2p spectrum was taken as a reference because for all the minerals in the silimanite group as well as for SiO₂ the binding energy is very similar, reaching ca 102.3 eV [23]. A comparative analysis of both examined materials on the basis of XPS lines of aluminum, silicon and oxygen did not show any signicant differences in the binding energy or the shape of the spectra (Fig. 4). On the other hand, an analysis of C1s spectrum revealed differences in both the position of line maximum and its shape (Fig. 5). In order to emphasize the differences, the C1s spectra were tted with component lines. It has to be remembered that C1s spectra

tting with component lines may be ambiguous. This results from the fact that in various XPS measurements performed for carbon-containing samples the position of C1s line is calibrated in different ways (frequently without quoting a reference point). For example, the positions of C1s peak ascribed to

graphite in various publications may be different by 1 eV.

In order to avoid this problem, the applied procedure of XPS spectra tting was identical for both samples. It is a well- known fact that irrespective of the calibration position choice, the components of C1s spectrum in a particular chemical environment are arranged according to the increasing binding energy in the following order: carbides, single particles of carbon, graphite, organic compounds, carbonates [25,24].

Therefore, selection of a calibration position does not have a great impact on a comparative analysis of XPS spectra. The

tting of spectra collected for both andalusite concentrates was based on components whose energy positions were as follows:

282 eV, 284 eV, 284.8 eV and, in the case of andalusite B, additionally 289.4 eV. The latter line can be ascribed to carbonates present in the sample [25], which indicates that their content in sample B is higher than in sample A, but their quantity is still slight in comparison with other forms of carbon represented by the remaining components. Line 248.8 eV is usually ascribed to adsorbates which are captured by the sample from the air [26]. The peak in this energy position can also represent amorphous carbon [27]. The component having a binding energy of 284 eV can represent graphite[28], whereas the component with the lowest binding energy probably originates from carbides [25]. Therefore, the C1s spectrum tting in Fig. 5 clearly showed a difference in the contents of the types of carbon occurring in both andalusites.

Namely, andalusite A has a denitely higher content of carbon in compounds which do not contain oxygen than andalusite B.

It should be added at this point that the differences found in the occurrence of the type of carbon in the investigated andalusites correspond to the age of deposit mineralization. In meta- morphic processes, the carbonaceous substance in an older deposit (taking into account primary mineralization) was progressively graphitized to a larger extent[29].

A different content of“oxygen-free compounds”of carbon in the grains of andalusite from different deposits is a factor which undoubtedly might inuence the mullitization process.

This can be explained in two ways. Firstly, it has to be

Table 3

Composition of trace elements in andalusite A and andalusite B from ICP.

Andalusite Concentration of trace elements [mg/kg]

Zn Cd Co B Mn Cr Cu Sr Ba Ti

A 15 o1 2 12 10 9 101 17 28 70

B 8 o1 o1 15 17 3 129 7 21 110

Table 4

Atomic concentration of andalusite A and andalusite B calculated from XPS survey spectra.

Andalusite Atomic concentration [at%]

Al Si O C Fe Mg Ca N K Na Cu

A 13.3 11.7 53.3 20.0 0.4 0.3 0.2 0.4 0.2 0.1 0.2

B 14.0 11.6 55.1 17.5 0.4 0.4 0.2 0.5 0.1 0.2 0.1

Fig. 3. XPS survey spectra of andalusite A (top) and andalusite B (down) from powdering samples.

(5)

remembered that the mechanism of andalusite transformation into mullite, which is described in subject literature, is based on the process of crystallization (nucleation and crystal growth) and release of free SiO₂and that the phenomena of diffusion play an important role in this change [6,11,14].

Diffusion processes among others depend on the specic surface of a material [30]. It can be presumed that in the

process of heating in air atmosphere, carbon contained in the andalusite matrix, having direct contact with the air, is oxidized, which results in an increased specic surface of the grains. At this point attention should be paid to the photographs of microstructures presented in Fig. 6. As can be seen in the microscopic pictures, both andalusites contain charac- teristic crosses with carbonaceous substance, therefore, these

Fig. 4. XPS Al2p, Si2p, O1s spectra from powdering samples.

Fig. 5. XPS C1s spectra from powdering samples with additionaltting of both spectra.

Table 5

Average atomic concentration of fractured surfaces of andalusite A and andalusite B calculated from XPS survey spectra collected from 4 randomly selected grains of both andalusites.

Andalusite Atomic concentration, [at%]

Al Si O C Fe Mg Ca N K Na

A

Grain I 14.3 13.5 63.3 7.3 0.7 0.6 0.3 0.0 0.0 0.0

Grain II 14.4 13.9 62.4 6.9 1.1 0.6 0.2 0.4 0.1 0.0

Grain III 11.9 13.5 63.7 7.5 1.5 0.8 0.4 0.2 0.4 0.1

Grain IV 16.2 11.5 61.5 9.3 0.4 0.6 0.2 0.0 0.0 0.3

Average 14.2 13.1 62.8 7.7 0.9 0.6 0.3 0.2 0.1 0.1

Standard deviation 1.8 1.1 1.1 1.1 0.5 0.2 0.2 0.2 0.2 0.2

B

Grain I 12.2 13.7 59.8 9.2 2.4 0.9 0.4 0.2 0.0 2.4

Grain II 16.5 9.6 56.0 17.1 0.6 0.0 0.1 0.2 0.0 0.0

Grain III 18.5 10.1 53.3 17.2 0.0 0.0 0.1 0.2 0.0 0.6

Grain IV 11.1 13.9 62.8 8.6 0.4 0.8 0.5 0.3 0.1 1.7

Average 14.6 11.8 58.0 13.0 0.8 0.4 0.3 0.2 0.1 1.2

Standard deviation 3.5 2.3 4.2 4.8 1.1 0.5 0.3 0.1 0.1 0.1

(6)

are andalusites of chiastolite type. In andalusite A the crosses have a relic character and the carbonaceous substance is more scattered than in andalusite B. This fact, correlated with an analysis of XPS measurements (Table 4andFig. 5), showing a higher content of carbonaceous substance in andalusite A than in andalusite B, leaves no doubt that carbon burning would result in a more intense increase of the specic surface of andalusite A. Assuming the carbon burning model, the increased specic surface of andalusite A causes easier diffusion in the mullitization process than in the case of andalusite B. The second explanation applies to the case when carbon is trapped inside the grain and has no contact with the air in the heating process. In this case, it is worth quoting the results of the mullitization process conducted for kaolinite with an addition of graphite in neutral gas atmosphere[31]. As it has been demonstrated, an addition of graphite facilitated the formation of mullite at a certain temperature, when the heating process was conducted in an oxygen-free atmosphere, e.g. argon. The authors related this increased content of mullite in reaction products with a reductive effect of carbon (in an oxygen-free atmosphere) on the siliceous component. It seems that the same mechanism can facilitate the processes of andalusite mullitization inside the grains which contain carbon and do not have direct contact with the air in the heating process. For this model also light microscope observations can be important (Fig. 6). Andalusite A contains more visible microcracks around which there are mineral impurities, including carbon. Therefore, the area of contact between carbonaceous substance and„proper grain”is larger there. Moreover, in andalusite grains there are areas where graphite co-exists with quartz (Fig. 7).

The surfaces of contact between carbon and silicon minerals, including quartz, can become areas of facilitated mullitization, when there is no air access due to carbon's reductive effect on silica. Now, if the contact surface is larger in andalusite A than in andalusite B, also in this case the model of mullitization in an

“oxygen-free”atmosphere can explain the easier change in the case of therst andalusite. From the above analysis it can be concluded that the presence of carbon in andalusite grains facilitates the mullitization process.

Another explanation of differences in the investigated andalus- ties' susceptibility to mullitization is based on microstructural observations. The mineral inclusions (isotropic aggregates) occurring around microstructural defects, i.e. microcracks and cracks, are more scattered in the grains of andalusite A. It is not only visible in

microstructural images (Fig. 6), but has been directly conrmed by photoelectron spectroscopy measurements. XPS experiments were conducted for (randomly selected) fractures of andalusite grains by taking survey spectra of particular fracture areas.Table 5contains the results of atomic concentration calculated on the basis of these measurements for small areas (corresponding to andalusite A and andalusite B). These results also indicate a more uniform distribution of elements on the fracture surface for andalusite A grains. A kind of simplied numerical indicator of this regularity could be the values of standard deviation of the examined minerals' main components presented inTable 5, which in the case of andalusite A are lower. On the other hand, in andalusite B it is easier tond areas for which relative atomic concentration Al:Si is close to the theoretical value of pure andalusite (Al:Si¼2:1), but there are also areas where this ratio departs considerably from the theoretical one to the advantage of silicon. In general, the results obtained in XPS investigations (Table 5) conrmed earlier ndings based on microstructural photographs (Fig. 6), namely, that inclusions contained in andalusite A, while similar in their amount, are more scattered than in andalusite B. Therefore, greater susceptibility to mullitization in andalusite A grains should be explained by the fact that they contain more crystal nuclei than andalusite B. A greater dispersion of inclusions in the andalusite grains matrix should also be considered as a factor that can facilitate mullitization.

Fig. 7. Microstructure of mineral inclusions of quartz and graphite present in the matrix of andalusite A, reected light, non-polarized.

Fig. 6. Microstructure of andalusite A and andalusite B, in cross- shaped traces as“carbonaceous material”, thin section┴(001), transmitted light, non-polarized.

(7)

Summing up, it can be stated that in the process of mullitization an important role is played not only by the type of inclusions occurring in andalusite, but also by their distribution. Another factor inuencing this process might be the presence of carbon inside andalusite grains.

4. Summary

In the article the results concerning two types of andalusite– andalusite A, originating from the secondary (elluvial) deposit, and andalusite B, from the primary (autochthonous) deposit – have been presented. XRD, XRF and ICP investigations revealed a very similar phase and chemical composition of both samples.

However, the andalusites considerably differed in the rate of the mullitization process. A possible explanation of this problem was provided by investigations conducted by XPS and LM.

On the basis of the obtained results, it has been found that if the mineral composition is identical and the content of inclusions in the grains very similar, the arrangement of these mineral impurities is very important for the mullitization process. A greater dispersion of inclusions in the mineral matrix results in a larger number of crystal nuclei in the process of transformation into mullite.

An important factor for this process might also be the presence of carbonaceous substance (as well as the manner of its distribution) in the andalusite matrix and possibly the co-occurrence of graphite and poikilitic inclusions of quartz or other silicon compounds. In typical technological processes in air atmosphere, carbon naturally occurring in andalusite grains is burned, thus increasing the specic surface of the grains, which can facilitate diffusion processes that initiate mullitization. Inside the grain in which carbon co-exists with silicon oxides and has no contact with air atmosphere during the heating process, a carbothermal reduc- tion of silicon compounds can take place, which makes the mullitization process easier.

Therefore, the following should also be considered as factors facilitating andalusite transformation into mullite: (a) a greater dispersion of impurities, which are related to inclusions around microstructural defects, i.e. microcracks and cracks, (b) a higher content of carbon naturally occurring in andalusite grains.

References

[1] A. Manecki, M. Muszyński i inni, Przewodnik po petrograi (Guide to the petrography), Uczelniane Wyd. Naukowo-Dydaktyczne, Kraków, 2008, pp. 179–180, 446 (in Polish).

[2] M. Handke, Krystalochemia krzemianów (Crystallochem. Silicates), Uczelniane Wyd. Naukowo-Dydaktyczne, Kraków 2005 (in Polish).

[3]M. Mason, K.W. Burton, Yanming Yuan, Zhenbing She, Chiastolite, Gondawana Res. 18 (2010) 222–229.

[4]M. Ihlen Peter, Utilisation of sillimanite minerals, their geology, and potential occurrences in Norway-an overview, Norges Geologiske Under- sokelse Bull. 436 (2000) 113–128.

[5]J.K. Winter, S. Ghose, Thermal expansion and high-temperature crystal chemistry of Al2O3polymorphs, Am. Miner. 64 (1979) 573–586.

[6]P. Dubreil, V.M. Sobolev, Andalusite: a promising material for manu- facturing high–quality refractory, Refract. Indust. Ceram. v. 40 (no 3-4) (1999) 152–158.

[7] J. Szczerba, B. Studencka, Beton specjalny odporny na działanie soli alkaliów i żużli z przemysłu metali nieżelaznych (Special Concrete resistant to alkali salt and ferrous metals industry slags), Rudy i Metale, R 51 (no 9) (2006) 524–530 (in Polish).

[8] J. Poirier, M.L. Bouchetou, L. Colombel, P. Hubert, Andalusite: an attractive raw material for its excellent thermal shock resistance, Refract.

Worldforum 1 (1) (2009) 93–102.

[9] H. Schneider, J. Schreuer, B. Hildmann, Structure and properties of mullite –A review, J. Eur. Ceram. Soc. Volume 28 (Issue 2) (2008) 329–344.

[10] W. Pannnhorst, H. Schneider, The high temperature transformation of andalusite (Al2SiO5) into mullite (3Al2O3*2SiO2) and silica (SiO2), Miner. Mag. 42 (1978) 197–198.

[11] A. Hülsmans, M. Schmücker, W. Mader, H. Schneider, The transformation of andalusite to mullite and silica: part I. Transformation mechanism in [001]Adirection, Am. Min., 6 85 (7–8) (2000) 980–986.

[12] H.H. Wilson, Mullite formation from the sillimanite group of minerals, Am. Ceram. Soc. Bull 48 (8) (1962) 796–797.

[13] W. Krönert, G. Rehfeld, Untersuchungen im System Al₂O₃-SiO₂ das Umwandlungs verhalten von Kyanit, Andalusit und Sillimanit, Westdendscher Verlag Köln und Opladen 2133 (1971) 25–28.

[14] H. Schneider, A. Majdic´, Kinetics and mechanism of the solid-state high- temperature transformation of andalusite (Al₂SiO₅) into 3/2 mullite (3 Al₂O₃SiO₂) and silica (SiO₂), Ceram. Int. Vol. 5 (no 1) (1979) 31–36.

[15] M.L. Bouchetou, J.P. Ildefonse, J. Poirier, P. Daniellou, Mullite grown from red andalusite grains: role of impurities and of the high temperature liquid phase on the kinetics of mullitization and conse- quences on thermal shocks resistance, Ceram. Int. 31 (2005) 999–1005.

[16] Guang Yulong An, Xiaoqing Liu, Jianmin Zhao, Huidi Chen, Zhou, Guoliang Hou, Preparation and microstructure characterization of mullite coatings made of mullitized natural andalusite powders, J. Therm. Spray Technol. v. 20 (3) (2011) 479–485 (Match).

[17] M.L. Bouchetou, J.P. Ildefonse, J. Poirier, Production of a high mullite content material from natural andalusite: inuence of grain size, Refract.

Appl. News v. 10 (2005) 21.

[18] T. Wala, J. Podwórny, The role of grain size and the participation of impurities in the transformation process the andalusite in mullite, in:

Proceedings of the VIII International Conference on Metallurgy, Refrac- tories: Production, Test methods and Use in Metallurgy, Ustroń25–28 May, 1999, pp. 135–142 (in Polish).

[19] T. Wala, J. Podwórny, R. Suwak, K. Stec, Effect of mineral admixtures on the selected properties of Andalusite, Materiały Ceramiczne/Ceram.

Mater. 61 (4) (2009) 288–294 (in Polish).

[20] P.W. Owerbeek, Andalusite in South Africa, J. S. Afr. Inst. Min. Metall.

vol. 89 (no. 6) (1989) 157–171 (June).

[21] B. Schulz, C. Audren, C. Triboulet, Regional vs. contact metamorphism of garnet metapelites in the vicinity of Late Variscan granites (Central Armorican Domain, Brittany, France), Geol. Rundach 87 (1998) 78–93.

[22] J. Podwórny, T. Wala, High temperature XRD investigations of andalusite raw materials, in: Proceedings of the International Colloquium on Refractories, Eurogress Aachen.

[23] F.S. Ohuchi, S. Ghose, M.H. Engelhard, D.R. Baer, Chemical bonding and electronic structures of the Al2SiO5polymorphs, andalusite, sillimanite, and kyanite: X-ray photoelectron- and electron energy loss spectroscopy studies, Am. Mineral Volume 91 (2006) 740–746.

[24] N.M. Rodriguez, P.E. Anderson, A. Wootsch, U. Wild, R. Schlögl, Z. Paál, XPS, EM, and catalytic studies of the accumulation of carbon on Pt black, J. Catal. 197 (2001) 365–377.

[25] Handbook of X-Ray Photoelectron Spectroscopy, Physical Electronics Inc. 1992.

[26] G.N. Salaita, P.H. Tate, Spectroscopic and microscopic characterization of Portland cement based unleached and leached solidied waste, Appl.

Surf. Sci. 133 (1998) 33–46.

[27] P. Reinke, G. Francz, P. Oelhafen, Structural phase transitions of carbon observed with photoelectron spectroscopy, Thin Solid Films 290-291 (1996) 148–152.

[28] J. Luthin, Ch. Linsmeier, Carbon lms and carbide formation on tungsten, Surf. Sci. 454–456 (2000) 78–82.

(8)

[29] O. Beyssac, B. Goffe, C. Chopin, J.,N. Rouzaud, Raman spectra of carbonaceous material in metasedimentes a new geothermometer, J. Metamorph. Geol. 20 (2002) 859–871.

[30] K. Pigoń, Z. Róziewicz, Chemiazyczna (Phys. Chem.), Wydawnictwo Naukowe PWN, Warszawa, 2005 (in Polish).

[31]K.J.D. MacKenzie, R.H. Reinhold, W.M. Brown, G.V. White, The formation of mullite from kaolinite under various reaction atmospheres, J. Europ. Ceram. Soc. 16 (1996) 115–119.