Comparative kinetic study of mechanical activation process of mica and talc for industrial application

Composites: Part B 59 (2014) 181–190

Contents lists available at ScienceDirect

Composites: Part B journal homepage: www.elsevier.com/locate/compositesb

Comparative kinetic study of mechanical activation process of mica and talc for industrial application Ljubiša Andric´ a, Anja Terzic´ b,⇑, Zagorka Ac´imovic´-Pavlovic´ c, Ljubica Pavlovic´ a, Milan Petrov a a

Institute for Technology of Nuclear and Other Mineral Raw Materials, Belgrade, Serbia Institute for Materials Testing, Belgrade, Serbia c University of Belgrade, Faculty of Technology and Metallurgy, Belgrade, Serbia b

a r t i c l e

i n f o

Article history: Received 17 July 2013 Received in revised form 18 October 2013 Accepted 7 December 2013 Available online 15 December 2013 Keywords: E. Powder processing D. Electron microscopy C. Analytical modeling B. Fragmentation Mechanical activation

a b s t r a c t Mica and talc have wide areas of application as a raw material in a number of industrial branches. Mechanically activated mica has specific applications such as: capacitors, insulators, and pearlescent pigments. Talc is widely used as either a basic raw material or as filler. This paper presents a comparative analysis of mechanically activated samples of mica and talc in ultra-centrifugal mechano-activator ‘‘Retsch ZM-1’’. The following mechano-activator parameters were variable: number of rotor revolutions (rpm); sieve mesh size (lm); current intensity (A). In addition, the following parameters were monitored: duration of mechanical activation, t (min); circumferential rotor speed, v (m/s); capacity of mechano-activator, Q (kg/h); and specific energy consumption, We (kW h/t). It was observed that effect of mechanical activation of mica and talc increased with an increase of the load and rotor revolution of ultra-centrifugal mechano-activator. Both mica and talc were successfully treated by mechanical activation procedure. In the processing of mica, mechanical activation is suggested to be applied as a post-treatment, and in the talc processing as a pre-treatment, as the high quality talc is obtained by means of hydrometallurgical concentration method. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Mechanically activated mica and talc in the ultra-fine powdery form have wide application spectra as raw materials used in the manufacturing of various industrial products such as capacitors, insulators, plastic fillers and pearlescent pigments. Mica is often used as filler in the synthesis of various materials with the special accent on production of the ‘smart’ materials [35,39,13,21], due to its color, density, particle shape, size and structure, and reflection index, adding the extra quality to synthesized products. For example, technology for production of the mica based glass–ceramics, whose microstructure contained preferentially aligned mica particles, was developed by Barlow and Manning [8] and Cheng et al. [11]. Talc, as a non-metallic raw material accompanied with exceptional physico-chemical characteristics, is irreplaceable in a number of industrial applications: production of paints, ceramics, cast products, rubber, cables, paper, pharmaceutical products, insecticides and herbicides, but also in civil engineering and military industry (Ulusoy [37], Nkoumbou et al. [24], Aoun et al. [4]). The standard way of talc and/or mica processing (comminuting, classification, flotation) can satisfy only a small number of users’

⇑ Corresponding author. Tel.: +381 112651842. E-mail address: [email protected] (A. Terzic´). 1359-8368/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesb.2013.12.003

requirements, thus the mechanical activation as pre-treatment method was introduced. The primary effect of the mechanical activation is the fragmentation of the mineral particles, which eventually results in the changes in a number of physico-mechanical properties of an investigated system, which was recently confirmed by Mucsi et al. [23]. Balaz and Achimovicová [7] proved that mechanical activation influences the crystal structure of a mineral usually making it disordered and generating crystal lattice defects or other meta-stable forms. During the last couple of decades, the mechanical activation procedure conducted by means of a different type of mechano-activators was extensively investigated: high energy mills, stirred media mills, attritional mills, jet mills, planetary mills, vibratory mills, and mortar mill. It has been reported by Baláz [5], Balaz [6], Erdemoglu et al. [15]; Kristóf-Makó et al. [20] and Tkácová [36] that the application of mechano-activators enables significant change in the structure and in the surface properties of solid phases. Sanchez-Soto et al. [29] and Suraj et al. [34], and later, Pérez-Maqueda et al. [25], Pérez-Maqueda et al. [26], Pérez-Maqueda et al. [27] proved that collection of the submicron and nanometric particles in different mechanoactivators is possible. Grinding, a common industrial procedure, and sonication treatments can change starting characteristics of treated material, however in this research special attention was paid to the

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

182

comparison and evaluation of the changes occurring during talc and mica processing in ultra-fine grinders (mechano-activators). The mechano-activation treatment might promote: the amorphization of treated material, noticeable change of the microstructure, size and shape of particles, etc. Furthermore, ultra-fine grinding kinetic investigation indicates the mechano-chemical reduction of the original particles of talc/mica which appears to have reached a limit at 30 min grinding time. However, longer grinding times might produce an apparent increase in particle size as was proved by Yang et al. [38], Mahadi and Palaniandy [22] and Cho et al. [12]. Even though ultra-fine grinding might appear as an expensive process due to low mill capacity and high energy consumption, there are many studies which prove the benefits of the mechanical activation process on the structural and physico-chemical properties of a material: Baudet et al. [9] investigated mechanical activation of kaolin clay by attrition-milling in a stirred bead mill; Sanchez-Soto et al. [28] examined the effects of dry grinding, using ball-milling, on the structure of kaolinite powders and explained why differently ground kaolinite samples show altered behavior during high-temperature transformations, while Franco et al. [17] studied the influence of the particle-size reduction on the dehydroxylation process of kalolinite. Dellisanti and Valdré [14] and later Hrachová et al. [18] investigated the structural changes of Camontmorillonite produced by mechanical activation by means of high-energy ball milling. Based on the study of the available literature, it can be concluded that few papers are dealing with the investigation of the ultra-centrifugal mill as apparatus for talc and mica mechano-activation. Therefore, the aim of the study presented in this paper is to investigate the applicability of ultracentrifugal mill for mechanical activation of mica and talc from a process engineering point of view.

composed of fine particles, thus it was directly subjected to mechanical activation. On the other hand, the talc sample consisted of coarse particles; therefore the particle reduction was necessary. Comminuting of talc samples was performed by crushing and afterwards grinding to the particle size that could be used as input for mechanical activation procedure. Primary crushing of talc sample was carried out in a jaw crusher with 10 mm output opening working in a closed circle with a screen. After primary crushing, the talc sample was subjected to secondary crushing in a roll crusher with 5 mm output opening. The grinding of the secondary crushed sample was conducted in the ceramic-lined ball mill. The mill, in which ceramic balls were used as the grinding media, was working in the closed circle with an air classifier, which enabled the liberation of the minerals (Table 2.). Iron minerals were partly liberated, while other minerals (chlorite, quartz, and calcite) were kept. It was afterwards confirmed by microscopic analysis. The mica and talc samples particle size distribution used as an input for mechanical activation procedure is shown in Table 3. Ultra-centrifugal mechano-activator ‘‘Retsch ZM-1’’ was used in the investigation of the mechanical activation and improvement of the material reactivity. The original mica sample and the ground sample of talc were subjected to quantitative characterization. Physical characterizations of both samples were determined by

Table 3 Particle size distribution of mica and talc.

2. Experimental The mica (KAl2(Si3Al)O10(OH,F)2) from flotation concentration plant Samoljica deposit in Bujanovac, Serbia and the talc from ‘‘Bela Stena’’ mine, Serbia were used in this investigation. Chemical compositions of mica and talc samples are given in Table 1. The physico-chemical characteristics for both mica and talc samples were determined by means of the standard laboratory procedures. The following results were achieved: density (according to Standard SRPS EN 725-8:2010 [31]) was 3.00 g/cm3 for mica, 2.70 g/cm3 for talc; humidity before/after drying (according to Standard SRPS EN 13286-46:2012 [32]) for mica and talc were: 10.00/0.5% and 6.00/0.50%, respectively; bulk density (according to Standard SRPS EN 725-9:2010 [33]) for mica was 1.17 g/cm3 and for talc 1.50 g/cm3. The preparation of mineral samples usually implies the reduction of their particle size, as preparation for the subsequent mechanical activation procedure. However, the mica sample was

Class of coarseness (mm)

Mass portion (%)

D, undersize (%)

R, oversize (%)

Mica 0.833 + 0.589 0.589 + 0.417 0.417 + 0.295 0.295 + 0.208 0.208 + 0.147 0.147 + 0.104 _ 0.104 + 0.074 0.074 + 0.063 0.063 + 0.053 0.053 + 0.040 0.040 + 0.000

0.10 4.40 22.50 29.00 23.00 10.00 6.00 1.40 1.10 0.80 1.40

0.10 4.50 27.0 56.0 79.0 89.0 95.0 96.4 97.5 98.6 100.00

100.00 99.0 99.5 73.0 44.0 21.0 10.5 4.5 3.1 2.0 1.4

Talc 0.589 + 0.417 0.417 + 0.295 0.295 + 0.208 0.208 + 0.147 0.147 + 0.104 0.104 + 0.074 0.074 + 0.063 0.063 + 0.053 0.053 + 0.040 0.040 + 0.030 0.030 + 0.020 0.020 + 0.000

6.67 13.89 19.18 16.60 15.47 10.40 3.33 3.74 3.13 2.80 2.22 2.57

6.67 20.57 39.74 56.35 71.82 82.22 85.55 89.28 92.41 95.21 97.43 100.00

100.00 93.33 79.43 60.26 43.65 28.18 17.78 14.45 10.72 7.59 4.79 2.57

Table 1 Chemical composition of mica and talc. Oxide Mica Talc

(%) (%)

SiO2

Al2O3

CaO

MgO

Na2O

K2O

Fe2O3

MnO

TiO2

PS

FeO

LoI

57.60 49.18

25.50 0.68

0.30 3.87

0.60 28.00

1.90 0.02

11.6 0.01

1.50 2.60

0.03 0.42

0.17 0.01

0.40 –

– 3.49

0.40 11.72

Table 2 Mineralogical composition of ground talc sample. Mineral

Talc

Chlorite

Quartz

Magnetite

Hematite

Carbonate (Fe, Mg)

Calcite

Limonite

Other

Content (%)

54.1

8.30

2.20

3.40

0.70

29.20

1.00

0.50

0.60

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

183

The experiments concerning mechanical activation were based on the grinding kinetic model. The obtained results were described by Rosin-Ramler graph – exponential function and equation (Eq. (1)). The dependence of average particle size (d0 ) of the ultracentrifugal mechano-activator circumferential rotor speed (v) was used for the description of mechanical activation procedure of both samples, as in the work of Yvon et al. [41]. The layout of this chart is presented in Fig. 1 (as used in paper by Andric´ et al. [3]).

d ¼ dx þ dk ¼ dk þ d0 expkðv v k Þ 0

0

0

0

0

ð1Þ 0 dk

0

Fig. 1. Dependence of average grain size (d0 ) of circumferential rotor speed (v).

means of the coulter counter: Coulter Electronics-Coulter Multisizer. The differential thermal analysis (DTA) of samples was performed with a Shimadzu DTA-50 apparatus. Approximately 30 mg of a sample was used for a DTA testing along with a-Al2O3 powder as a reference sample. The sample was heated under a nitrogen atmosphere from 20 up to 1100 °C at heating rate of 10 °C/min. Powdery samples were analyzed by means of X-ray powder diffraction (XRD). The XRD patterns were obtained on a Philips PW-1710 automated diffractometer using a Cu tube operated at 40 kV and 30 mA. The microstructure of the samples was characterized by scanning electron microscopy method (SEM) using a JEOL JSM-6390 Lv microscope. The samples were covered with gold powder for better reflection to be obtained and measurements performed.

0

Table 4 Parameters of ultra-centrifugal mechano-activator operation performed on mica sample. Parameters of mechano-activator Technological parameters of mechanoactivation Experim. sequence No. I

1 2 3 4 5 6 7 8

II 1 2 3 4 5 6 7 8 III 1 2 3 4 5 6 7 8 IV 1 2 3 4 5 6 7 8

NRR (rpm)

Mesh size Current intensity (A) (lm)

10000 80

20000 80

10000 120

20000 120

10000 200

20000 200

10000 500

20000 500

0

where d is the average grain size (lm); the constant part of d ; dx the variable part of d0 ; v the circumferential rotor speed (m/s); vk the constant circumferential rotor speed (m/s) which corresponds 0 to d0 ¼ 1; k is the constant of micronization rate which depends on experimental conditions and micronized material. In all of the series of mechanical activation experiments, the following operational parameters of mechano-activator were variable: number of rotor revolutions, (rpm); sieve mesh size, (lm); and current intensity, (A). During mechanical activation the following parameters were monitored: mechanical activation time, t (min); circumferential rotor speed, v (m/s); capacity of ultracentrifugal mechano-activator, Q (kg/h); and specific energy consumption, We (kW h/t). The mechanical activation of mica and talc were assessed through the following parameters: d1 and d2 – the mesh sizes of the used sieves, (lm); R1 and R2 – cumulative oversize, (%); d 0 – average grain size (parameter which depends on particle size distribution); d95 – sieve mesh size that is appropriate to 95% cumulative undersize of the micronized product, (lm); St – calculated-theoretical specific surface, (m2/kg) – (Eq. 2).; Sr – real

Energy consump. (kW h/t)

Mechano-activation product parameters d1 d2 (lm) (lm)

Speed (m/s)

Capacity (kg/h)

1.40 2.40 3.10 3.80 2.10 2.40 2.90 3.60

36.00 15.00 10.00 6.00 25.00 10.00 8.00 6.00

75.50 57.00 47.15 40.35 117.00 112.50 100.00 86.90

0.150 0.187 0.258 0.536 0.156 0.286 0.400 0.520

3234.40 3181.43 2772.00 1732.50 4370.26 2125.20 1720.90 1620.30

5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00

40.00 53.00 53.00 53.00 30.00 30.00 30.00 30.00

92.66 94.80 94.90 95.25 90.02 90.60 91.50 92.00

3.60 1.80 2.18 2.90 5.10 7.00 8.00 9.00

21.58 24.90 25.00 27.55 17.22 17.28 18.88 19.43

1.79 1.81 1.81 1.89 1.84 1.85 1.86 1.87

1.40 1.75 2.80 3.65 2.20 2.55 3.10 4.00

7.00 5.00 2.50 1.50 7.00 5.50 4.50 3.00

75.80 70.90 52.13 45.50 115.90 107.26 99.20 84.50

0.520 1.100 2.080 3.099 0.550 0.620 0.760 1.210

646.80 369.60 311.85 269.85 1016.40 962.49 924.00 769.99

5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00

53.00 63.00 63.00 74.00 40.00 53.00 63.00 63.00

95.70 96.20 97.60 97.80 94.00 95.70 96.00 96.50

4.00 1.50 6.00 4.30 5.00 2.00 1.90 2.80

28.84 29.10 38.09 41.47 23.08 25.80 29.18 32.78

1.40 2.00 3.00 3.95 2.20 2.40 3.00 3.90

5.00 3.50 2.50 1.50 6.50 4.00 2.00 1.50

75.50 63.12 50.58 42.60 117.70 107.37 98.10 88.63

0.700 1.020 1.600 3.300 0.600 0.900 2.000 6.100

484.86 425.00 415.67 292.64 1018.00 701.00 350.00 160.17

5.00 53.00 5.00 63.00 5.00 74.00 5.00 53.00 5.00 53.00 5.00 53.00 5.00 63.00 5.00 147.00

95.20 2.20 96.80 3.60 98.40 8.50 98.90 16.00 94.00 0.90 95.40 2.00 96.00 3.00 97.40 5.00

26.77 32.40 45.20 55.22 23.90 25.00 29.90 35.90

1.20 2.20 2.95 3.40 2.20 2.50 2.90 3.50

4.50 1.50 0.55 0.27 4.50 1.00 0.70 0.30

81.20 63.50 51.10 49.00 116.80 110.10 105.24 95.60

0.950 3.250 12.500 18.660 1.150 7.500 10.120 19.000

369.60 150.60 57.10 42.20 677.60 99.90 79.90 41.85

5.00 10.00 10.00 10.00 5.00 5.00 5.00 10.00

147.00 147.00 147.00 104.00 147.00 147.00 147.00 147.00

R1 (%)

99.43 98.50 98.70 98.75 99.36 99.50 99.60 98.30

R2 (%)

d0 (lm)

Duration (min)

n

d95 (lm)

St (m2/kg)

Sr (m2/kg)

39.06 45.78 46.72 49.56 30.38 32.26 33.22 34.09

190.44 158.78 156.98 149.20 229.09 225.28 208.97 201.86

571.34 476.34 470.95 447.61 687.29 669.89 626.91 605.58

1.80 1.83 1.84 1.86 1.84 1.89 1.90 1.92

51.71 53.64 67.44 74.11 40.87 47.82 55.45 58.76

141.66 132.85 104.72 97.92 172.77 144.66 134.61 124.65

424.91 398.56 314.18 293.76 518.33 434.00 403.83 373.19

1.82 1.84 1.86 1.89 1.82 1.84 1.85 1.87

47.69 60.10 81.50 93.00 41.90 45.90 48.90 61.20

150.78 118.80 87.20 72.90 171.03 163.90 140.20 110.00

452.34 358.90 250.08 218.48 513.10 505.20 419.10 350.20

4.00 81.10 1.86 146.25 11.00 93.40 1.86 176.80 12.00 98.62 1.90 176.90 14.00 103.30 1.99 179.90 13.00 70.00 1.85 126.10 5.50 79.12 1.87 149.80 7.00 81.40 1.92 155.30 8.00 92.10 1.97 160.10

47.47 39.10 37.20 36.80 52.30 50.90 42.50 42.10

142.41 115.90 109.30 105.90 152.90 138.20 121.90 120.20

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

184

Table 5 Parameters of ultra-centrifugal mechano-activator operation performed on talc sample. Experim. sequence No.

Parameters of mechano-activator Technological parameters of mechano-activation Mechano-activation product parameters NRR (rpm)

Mesh size Current intensity (A) (lm)

Duration (min)

Speed (m/s)

Capacity (kg/h)

Energy consump. d1 d2 R1 (kW h/t) (lm) (lm) (%)

R2 (%)

d0 n (lm)

d95 St Sr (lm) (m2/kg) (m2/kg)

I

1 2 10000 80 3 4 5 6 20000 80 7 8

1.40 2.30 3.00 3.75 2.10 2.30 2.80 3.50

15.00 7.00 5.00 3.00 13.00 6.00 4.00 2.00

75.36 55.26 48.78 42.37 123.11 116.74 108.33 88.88

0.10 0.22 0.31 0.52 0.12 0.26 0.39 0.78

290.96 220.22 165.20 106.25 235.94 220.22 180.92 125.90

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

8.00 8.00 8.00 8.00 8.00 8.00 8.00 8.00

73.00 78.00 80.00 86.00 77.00 82.00 85.00 88.00

6.00 4.00 5.00 4.00 10.00 8.00 7.00 7.00

2.96 3.10 3.30 3.40 3.60 3.80 3.90 4.13

1.05 1.25 1.45 1.89 1.05 1.22 1.34 1.46

7.30 7.40 7.70 8.10 8.40 8.40 8.90 9.80

3839.45 2405.46 2170.52 1445.63 3261.76 2019.20 1595.13 1294.90

4607.34 2886.56 2604.62 1734.76 3914.12 2423.04 1914.16 1553.88

II

1 2 10000 120 3 4 5 6 20000 120 7 8

1.40 1.60 2.70 3.50 2.20 2.50 3.00 4.00

8.00 5.00 2.00 1.00 6.00 3.00 1.50 0.80

75.36 70.34 51.58 43.09 116.74 103.17 95.55 82.35

0.19 0.31 0.78 1.55 0.26 0.52 1.03 1.94

290.96 275.24 188.78 125.90 228.08 204.50 165.20 86.60

1.00 1.00 1.00 1.00 3.00 3.00 3.00 3.00

8.00 8.00 8.00 8.00 10.00 10.00 10.00 10.00

85.00 87.00 80.00 91.00 73.00 79.00 84.00 85.00

5.00 7.00 14.00 16.00 11.00 16.00 16.00 17.00

3.70 4.00 4.20 5.20 6.10 7.00 7.30 7.50

1.40 1.42 1.45 1.53 1.62 1.70 1.95 1.98

7.70 8.40 11.40 10.90 11.70 13.00 12.60 12.70

1556.81 1387.53 1084.05 1053.44 739.16 604.61 497.97 480.35

1868.17 1665.04 1364.86 1264.13 886.99 725.54 597.57 576.42

III

1 2 10000 200 3 4 5 6 20000 200 7 8

1.40 1.90 2.80 3.80 2.20 2.60 3.00 4.00

4.00 1.00 0.50 0.20 3.00 0.60 0.10 0.08

75.36 61.55 50.7 41.67 120.26 108.33 97.57 84.37

0.39 1.55 3.10 7.75 0.52 2.58 15.50 19.38

290.96 251.66 180.92 102.32 228.08 196.64 165.20 86.60

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

8.00 10.00 10.00 15.00 8.00 10.00 10.00 15.00

88.00 89.00 94.00 93.00 89.00 88.00 89.00 90.00

16.00 18.00 26.00 21.00 11.00 10.00 16.00 16.00

5.00 6.30 8.00 10.10 4.60 5.10 6.00 8.40

1.28 1.17 1.34 1.43 1.21 1.26 1.28 1.35

11.30 15.50 17.60 25.60 9.60 11.90 14.50 22.90

1368.19 1350.86 977.36 912.33 1224.65 1384.73 1326.33 1316.94

1641.83 1621.04 1332.83 1094.79 1669.58 1661.68 1591.59 1428.32

IV

1 2 10000 500 3 4 5 6 20000 500 7 8

1.20 2.00 2.80 3.20 2.20 2.40 2.80 3.20

3.00 0.50 0.10 0.06 2.00 0.70 0.06 0.02

75.36 61.55 51.58 48.78 116.74 108.33 103.17 95.55

0.52 3.10 15.50 25.83 0.78 2.21 25.83 77.50

306.68 243.80 180.92 149.48 228.08 212.36 180.92 149.48

1.00 2.00 2.00 1.00 2.00 2.00 5.00 5.00

10.00 10.00 12.00 15.00 13.00 15.00 17.00. 19.00

85.00 80.00 92.00 92.00 92.00 92.00 89.00 91.00

19.00 22.00 17.00 17.00 28.00 27.00 28.00 29.00

6.00 7.10 8.60 9.00 11.00 12.30 15.00 17.00

1.01 1.19 1.71 1.83 1.46 1.47 1.95 1.93

17.10 17.10 19.90 22.90 22.70 26.60 27.70 29.30

2196.60 1149.66 1142.89 1033.73 483.11 484.46 243.11 217.41

2635.92 1379.59 1321.46 1240.48 579.74 581.35 291.73 260.89

specific surface, (m2/kg) – (Sr = 3St); n – level of kinetics of micronization, parameter which depends on particle size distribution of the sample.

St ¼

6:39

q  d0

e

1:795 n2

ð2Þ

where q is density of the sample (kg/m3). In order to achieve high quality talc with low iron content (Fe2O3 = 1.00–1.50%), the talc sample was additionally submitted to hydrometallurgical treatment and its influence on the Fe2O3 recovery was investigated. The representative sample of mechanically activated talc sample was subjected to a leaching test by hydrochloric acid (15 wt.%). The leaching was conducted in a three-necked bottle at constant temperature (T = 80 °C) and within the time period of 3.5 h. 3. Results and discussion Mechanical activation by means of ultra-centrifugal mechanoactivator was performed on mica and talc samples in four series (mesh size being 80, 120, 200 and 500 lm for series I–IV, respectively). The detailed experimental results of kinetics of mica and talc mechanical activation, i.e. process parameters (number of rotor revolutions – NRR; mesh size; current intensity), technological parameters (duration; circumferential rotor speed; mechanical capacity; energy consumption), and parameters used to monitor and evaluate process of mechanical activation (mesh sizes of used sieves; coarseness; cumulative oversize, parameter of particle size distribution; and specific surface) are listed in Tables 4 and 5, respectively.

The kinetic experiments of mechanical activation were carried out in four series (I–IV), separately for each sample. The assessment of mechanical activation of mica and talc is based on the results given in Tables 4 and 5. For all sieve mesh sizes (80, 120, 200, 500 lm) at nominal number of rotor revolution transition from n0 = 100,00 (rpm) and n0 = 200,00 (rpm). The samples of each series are activated at 100,00 rpm, and 20,000 rpm, four experimental sequences for each nominal number of rotor revolutions. Current intensity, as one of the process parameters, was increasing during each experiment sequence (1–4 and 5–8). From the over-all comparison of the results given in Tables 4 and 5. it can be seen that, in both cases – mica and talc, by increasing the mechano-activator load and increasing the sieve mesh size from 80 to 500 lm, the current intensity was increased. The current intensity ranged from minimum of 1.20 to the maximum of 4.00 A in both cases – mica and talc samples. Minimum was obtained during activation of series IV at 500 lm mesh size and 10000 rpm both for mica and talc. Maximum of 4.00 A was obtained for mica sample in series II at 120 lm mesh size and 20,000 rpm, and for talc sample in series II at 120 lm mesh size and 20,000 rpm and in series III at 200 lm mesh size and 20,000 rpm. It was noticed that technological parameters (duration of mechano-activation procedure; circumferential rotor speed, capacity and energy consumption) were changing with the increasing of the mechano-activator load both for mica and talc (Tables 4 and 5). The decreasing of the duration of mechanical activation (t) directly influenced the increasing of mechano-activator capacity (Q). At the same time, the capacity was decreasing and the energy consumption was increasing. Maximal activation duration (t) was noticed during first experimental sequence for series I at 80 lm

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

185

Fig. 2. Average grain size (d0 ) of mechanically activated mica vs. circumferential rotor speed (v) for sieve mesh size: (a) 80 lm; (b) 120 lm; (c) 200 lm and (d) 500 lm.

mesh size and 10,000 rpm: 36 min for mica sample, and 15 min for talc sample. Minimal necessary time for mechanical activation procedure was 0.30 min and 0.20 min for mica and talc, respectively. Minimal values in both cases were noticed in eighth experimental sequence at 500 lm mesh size and 20,000 rpm. Achieved experimental results of mechanical activation were observed trough the following parameters: d1 and d2 – the mesh sizes of the used screens, (lm); R1 and R2 – cumulative oversize, (%); d0 – parameter which depends on particle size distribution of the sample. It characterizes sample coarseness at the same time represents the measure of the mesh size for oversize cumulative R = 36.79%; n – parameter which depends on particle size distribution of the sample; d95 – sieve mesh size that is appropriate to 95% cumulative undersize of the micronized product, (lm); St – calculated-theoretical specific surface, (m2/kg) (Eq. 2.); and Sr – real specific surface, (m2/kg). The increase of the values of parameter (d95), which defines a fineness of mechanical activation, for each of the experimental sequences for both talc and mica was noticed during the mechanical activation process. Also, increasing of the parameter (d95) values indicated on the increasing of coarseness (d0 ) of both mica and talc particles and on the reduction of the circumferential rotor speed (v). Following previously stated changes – increasing of the d95 parameter, decreasing of the real specific surface (Sr), theoretical

specific surface (St), and specific energy consumption were also noticed. Change in specific surface of the mica and talc indicates that certain changes in the texture and particle size occurred. Thus, it can be concluded that the application of ultra-centrifugal mechano-activator enabled changes both in the structure and in the surface properties of investigated mica and talc samples. By observing of the presented parameters of the mechanical activation, it could be concluded that by increasing the sieve mesh size, the rate of rotor revolution and mechano-activator load, the mechanical activation rate, as the main characteristic of the mechanical activation kinetics also increases. The experimental results of the mechanical activation of both mica and talc are explained by parameter (d0 ). By inserting the experimental values from Table 4. in (Eq. 1) the diagrams of the 0 average grain size vs. circumferential rotor speed d = f(v) for all sieves mesh sizes were obtained. The diagrams of dependency of average grain size (d0 ) on circumferential rotor speed for sieve mesh sizes 80, 120, 200 and 500 lm for mica, are presented in Fig. 2(a–d), respectively. After comparison of the diagrams given in Fig. 2 it can be concluded that the best results are obtained for 80 lm mesh size. Parameter (d0 ) as average grain size after activation ranges from 21.58–27.55 lm at 10,000 rpm and 17.22–19.43 lm at 20,000 rpm for 80 lm mesh size. In other cases, i.e. for mesh sizes

186

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

Fig. 3. Average grain size (d0 ) of mechanically activated talc vs. Circumferential rotor speed (v) for sieve mesh size: (a) 80 lm; (b) 120 lm; (c) 200 lm and (d) 500 lm.

Fig. 4. DTA curve of the mica sample recorded: (a) before mechanical activation and (b) after 30 min of mechanical activation.

120, 200 and 500 lm, bottom levels of average grain size (d0 ) for 10,000/20,000 rpm are 28.84/23.08 lm, 26.77/23.90 lm and 81.10/70.00, respectively. Following the same calculation, now using results from Table 5 and inserting them in (Eq. 1) the dependency of average grain size (d0 ) on circumferential rotary speed for various sieve mesh sizes for

talc are obtained. Diagrams for sieve mesh sizes 80, 120, 200 and 500 lm are given in Fig. 3(a–d), respectively. As in the case of mica samples, the best results of talc mechanoactivation are obtained for 80 lm mesh size. For 80 lm mesh size, the parameter (d0 ) ranges from 2.96 to 3.40 lm at 10000 rpm and 3.60–4.13 lm at 20,000 rpm. For mesh sizes 120, 200 and 500 lm,

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

187

Fig. 5. DTA curve of the talc sample recorded: (a) before mechanical activation and (b) after 30 min of mechanical activation.

Fig. 6. XRD diffractogram of: (a) initial mica sample; (b) mica sample after 6 min of mechanical activation and (c) mica sample after 30 min of mechanical activation.

bottom levels of average grain size (d0 ) for 10,000/20,000 rpm are higher than in case of 80 lm mesh size: 3.70/6.10 lm, 5.00/ 4.60 lm and 6.00/11.00, respectively. In order to explain the influence of mechano-activation on the structure and characteristics of mica and talc and to confirm results obtained during mechanical activation, specific analyses were performed before and after activation procedure: differential thermal analysis (DTA); X-ray powder diffraction analysis (XRD) and scanning electron microscopy (SEM). Thermal characteristics of the processes taking place during mica thermal treatment from 20 to 1100 °C were identified by means of DTA method. The DTA curves for un-ground and ground mica is given in Fig. 4(a and b). The adsorbed water content in both un-ground and ground mica samples are lost up to a temperature of approximately 300 °C. Above 300 °C (approximately around 450 °C) the loss of hydroxyl water content is achieved. In some cases a small exothermic reaction caused by the autooxidation process appears around 400 °C (Schomburg et al. [30]). Typical high temperature phases appearing in mica are mullite, spinel, corundum and leucite. Strong formation of glassy phases can be obtained in Fe-rich varieties, especially after the heating up to 900 °C (Schomburg and Zwahr [30]). However, melting of the non-activated and activated mica samples was not recorded up to the temperature 1100 °C which means that activation promoted formation of thermally stable phases. Results of DTA presented in Fig. 5(a and b) characterized the thermal behavior of natural un-ground and ground talc samples, respectively. The processes taking place during heating below 800 °C are endothermic (Sanchez-Soto et al. [29]). The peaks below 300 °C are corresponding to the loss of water adsorbed on the surface of the particles, and the peaks above correspond to the loss of

OH groups. From Fig. 5a., it followed that on further heating of the un-ground talc sample an DTA endothermal effect was observed in the range of 900 °C, due to the loss of structural water and accompanied by the formation of enstatite (MgSiO3) and silica (Aglietti [1]). The DTA results in Fig. 5b. confirmed that mechanical activation enhanced the formation of high temperature phases of the talc (Filio et al. [16]). An exothermal effect at 830 °C characterized the crystallization of non-crystalline phases, formed by grinding of the talc sample, into orthorhombic enstatite [20]. No exothermal (Aglietti and Porto Lopez [2]). These results are in agreement with the previously observed particle size diminution and surface area evolution of the ground sample and are associated with the weekly bound OH groups on the broken edges of the ground talc particles. Mineralogical phase changes as well as the variations in crystallinity occurring in mica and talc samples were tracked by means of XRD analysis. In Fig. 6., the diffractograms of mica samples before and after mechanical activation are given. It was noticed that changes in the crystal structure of mica appeared within 6 and 30 min of the mechanical activation process. Mechanical reduction of the original particles of both investigated minerals appears to have reached a limit at 30 min grinding time and longer grinding times might produce contra effect: an increase in particle size. Namely, in light of XRD analysis, contra effect would probably be the increase in crystalinity of investigated samples (Boldyrev et al. [10]). The results of XRD analyses of initial mica sample and changes caused by mechano-activation are shown in Fig. 6(a–c). From Fig. 6. it can be seen that the duration of the mechanoactivation influences the crystallinity of mica samples, i.e. the level of crystallinity is decreasing with the increasing of the duration of mechanical activation.

188

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

Fig. 7. XRD diffractogram of: (a) initial talc sample; (b) talc sample after 6 min of mechanical activation and (c) talc sample after 30 min of mechanical activation.

The results of XRD analyses of talc samples before and after mechanical activation are given in Fig. 7. The changes of the crystal structure appeared within 6 and 30 min of mechanical activation in ultra-centrifugal mechano-activator, as it was also in the case of mica samples. Comparison of the talc samples before mechanical activation and after mechanical activation implies that the length of mechano-activation influences the crystallinity of talc samples, i.e. the level of crystallinity is decreasing with the increasing of the time of mechanical activation. It can be concluded that mechanical activation influences the crystal structure of both minerals – talc and mica, i.e. level of crystallinity decreased with increasing mechano-activation time. Namely, mechanical activation makes the structure disordered and generates crystal lattice defects or other meta-stable forms, as it was also proved by Balaz and Achimovicová [7]. The mechan-

Fig. 8. Microstructure of mica sample after mechanical activation.

o-activation treatment might promote the amorphization of treated material, noticeable change of the microstructure, size and shape of particles, etc. In Fig. 8., the SEM microphotograph of mica sample after mechanical-activation is given. Typical mica (muscovite) nonactivated particles are thin and wide having a more plate-like form as it was shown in work by Holopainen et al. [19]. It can be seen that after mechano-activation mica particles gained a pseudo hexagonal crystal form. The particles maintained its layered structure, but the size of particles is reduced. The shape of original talc particles is normally rounded, but slightly elongated as it was shown by Yekeler et al. [40]. Dimensions of original talc particles varied in a significant range. The SEM microphotograph of talc sample after mechanical-activation, which is given in Fig. 9., shows that the talc particles gained rather angular shape. The talc particles possess a semi-layered structure and their size is reduced and rather uniform. The SEM analysis of mica and talc samples did not reveal any sign of particle size increase or particles agglomeration, thus it can be concluded that 30 min is optimal grinding time which did not produce any of the stated contra effects. After successfully conducted mechanical activation on the samples of both minerals – mica and talc it was concluded that in the processing of mica mechanical activation should be applied as a post-treatment, while in the talc processing mechanical activation should be conducted as a pre-treatment. Namely, the high quality talc can be obtained only by means of hydrometallurgical concentration method. This, the mechanically activated talc sample was additionally subjected to leaching by hydrochloric acid. The talc sample had the following characteristics: d0 = 3.60 lm; t = 13 min; sieve mesh size 80 lm and n = 20000 rpm. During the process, calcite (green color) was separated at the bottom as a layer. At the same time, talc concentrate (white) was separated in the upper layer. CO2 was produced during the leaching process, carrying the fine talc particles concentrate, lighter than the chlorite particles which remained on the bottom. The products of the leaching were: talc concentrate, chlorite, intermediate product (talc-chlorite) and final tailing. In addition, talc concentrate in warm condition was rinsed by water, then filtered and dried. The same process was applied to

Table 6 Chemical composition of talc (concentrate, intermediate product and tailing). Product

Talc concentrate Intermediate product Tailing Total Fig. 9. Microstructure of talc sample after mechanical activation.

Component and content (in%) M%, Mass portion

MgO

Fe2O3

LOI

Mineral

75.90 15.77 8.33 100.00

28.30 12.94 14.57 24.22

1.15 5.83 10.00 2.80

5.00 6.90 9.75 5.84

Pure talc Talc chlorite Chlorite –

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

chlorite, intermediate product (talc-chlorite) and final tailing. The results are shown in Table 6. 4. Conclusion The comparative testing and detailed investigation of the kinetics of mechanical activation procedure by means of ultracentrifugal mechano-activator applied to the mica and talc samples, as well as the analyses of the characteristics of non-activated and activated samples of both talc and mica led to the following conclusions:  In accordance with the Rosin–Ramler diagram and the mathematical equation for the average grain size (d0 ) dependency of the circumferential rotor speed (v), the kinetic model for mechano-activation is proposed and confirmed. This model can be used for the optimization and automation of mechanical activation in ultra-centrifugal mechano-activator with a peripheral comminuting path.  The mechanical activation rate, as the main characteristic of mechano-activation kinetics, increased with increasing sieve mesh size, rate of revolutions and load of the ultracentrifugal mechano-activator. The best results were obtained with a nominal mechano-activator load and for 80 lm mesh size, for both talc and mica samples.  Reduction of the original particles of both talc and mica conducted by mechanical activation procedure appeared to have reached a limit at approximately 30 min grinding time and longer activation times might produce an increase in particle size or agglomeration of the particles.  The increase of the parameter (d95) which defines a fineness of mechanical activation was followed by decreasing of the specific surface which indicates that certain changes in the texture and particle size occurred. Thus, it can be concluded that the application of ultra-centrifugal mechano-activator enabled changes both in the structure and in the surface properties of investigated mica and talc samples.  After analyzing DTA curves of both talc and mica samples (before and after activation) it was concluded that signs of melting (apparition of glassy phases) of neither of investigated minerals was not recorded up to the temperature 1100 °C which contributes to good refractory characteristics of mica and talc. Mechanical activation promoted formation of thermally stable phases.  X-ray analyses of non-activated and activated mica and talc samples confirmed that mechanical activation contributes to the decreasing of crystallinity. The mechano-activation treatment promotes amorphization of mineral, microstructural changes, and changes in size and shape of minerals particles.  Activated mica particles are characterized by a pseudo hexagonal crystal form and layered structure as it was proved by scanning electron microscopy. Due to mica starting specific mineralogical and mechanical properties, the particles are easily mechanically activated by ultra-centrifugal mechano-activator, which makes them applicable in a wide variety of industrial advanced materials. Mechanically activated mica is most commonly used as filler in industrial application.  Originally rounded and elongated talc particles of various sizes were also easily activated by ultra-centrifugal mechano-activator, giving then angular shape, semi-layered structure and relatively uniform size distribution. As such the talc particles were well-prepared for subsequent leaching treatment applied to ensure the high quality of the

189

material. Mmechano-activation of talc represented an initial step for hydrometallurgical treatment, which together represents a new approach for obtaining high-grade talc concentrate for application as a raw material for the production of various industrial products.  The SEM analysis of mica and talc samples did not reveal any sign of particle size increase or particles agglomeration which confirms that 30 min is optimal grinding time which does not produce any of the stated contra effects.  Mica and talc were successfully treated by mechanical activation procedure. In the processing of mica minerals, mechanical activation was applied as a post-treatment method, while in the case of talc mineral processing, mechanical activation process was applied as pretreatment. Even though mechanical activation procedure might appear as an expensive process due to low mill capacity and high energy consumption, activated products have advanced characteristics which open entirely new spectra of their industrial applications. Also, activated minerals significantly improve the performances of standard products in which they are used as raw material. Acknowledgements This investigation was supported and funded by the Ministry of Education and the Science and Technological Development of the Republic of Serbia and it was conducted under the Projects: 33007, 34006, 172057 and 45008.

References [1] Aglietti E. The effect of dry grinding on the structure of talc. Appl Clay Sci 1994;9:139–47. [2] Aglietti E, Porto Lopez J. Physicochemical and thermal properties of mechanochemically activated talc. Mater Res Bull 1992;27:1205–16. [3] Andric Lj, Miloševic V, Petrov M, Mihajlovic S. Operating technique of mills in the process of micronization milling of alumina. J Min Metall 2005;41A. 27-4. [4] Aoun S, Chebli C, Baboulène M. Noncovalent immobilization of chloroperoxidase onto talc: Catalytic properties of a new biocatalyst. Enz Microbial Technol 1998;23:380–5. [5] Baláz P. Extractive Metallurgy of Activated Minerals. Amsterdam: Elsevier; 2000. p. 278. [6] Balaz P. Mechanical activation in hydrometallurgy. Int J Min Process 2003;72:341–54. [7] Balaz P, Achimovicová M. Mechano-chemical leaching in hydrometallurgy of complex sulphides. Hydrometallurgy 2006;84:60–8. [8] Barlow SG, Manning AC. Influence of time and temperature on reactions and transformations of muscovite mica. Br Ceram Trans 1999;3:122–6. [9] Baudet G, Perrotel V, Seron A, Stellatelli M. Two dimensions comminution of kaolinite clay particles. Powder Technol 1999;105:125–34. [10] Boldyrev V, Pavlov S, Goldberg E. Interrelation between fine grinding and mechanical activation. Comminution 1996;1994:181–5. [11] Cheng K, Wan J, Liang K. Enhanced mechanical properties of oriented mica glass-ceramics. Mater Lett 1999;39:350–3. [12] Cho H, Lee H, Lee Y. Some breakage characteristics of ultra-fine wet grinding with a centrifugal mill. Int J Miner Process 2006;78:250–61. [13] Chou Y, Stevenson J, Hardy J, Singh P. Material degradation during isothermal ageing and thermal cycling of hybrid mica seals under solid oxide fuel cell exposure conditions. J Power Sources 2006;157:260–70. [14] Dellisanti F, Valdré G. Study of structural properties of ion treated and mechanically deformed commercial bentonite. Appl Clay Sci 2005;28:233–44. [15] Erdemoglu M, Ding Y, Ghadiri M. Intensive milling of siderite, in: XII. Internationa Mineral Processing Symposium, Cappadocia-Nevsehir, Turkey; 2010. p. 149–57. [16] Filio J, Sugiyama K, Saito F, Waseda Y. A study on talc ground by tumbling and planetary ball mills. Powder Technol 1994;78:121–7. [17] Franco F, Pérez-Maqueda LA, Pérez-Rodríguez JL. Influence of the particle-size reduction by ultrasound treatment on the dehydroxylation process of kaolinites. J Therm Anal Calorim 2004;78:1043–55. [18] Hrachová J, Madejová J, Billik P, Komadel P, Fajnor VŠ. Dry grinding of Ca and octadecyltrimethylammonium montmorillonite. J Colloid Interface Sci 2007;316:589–95. [19] Holopainen M, Hirvonen M, Komulainen H, Klockars M. Mica dust-induced TNFa production in mouse macrophages in vitro. Ann Occup Hyg 2002;46:300–3.

190

L. Andric´ et al. / Composites: Part B 59 (2014) 181–190

[20] Kristóf-Makó E, Juhász AZ. Effect of mechanical treatment on the crystal structure and thermal decomposition of dolomite. Thermochim Acta 1999;342:105–14. [21] Luna M, Colchero J, Gil A, Gómez-Herrero J, Baró AM. Application of noncontact scanning force microscopy to the study of water adsorption on graphite, gold and mica. Appl Surf Sci 2000;157:393–7. [22] Mahadi MI, Palaniandy S. Mechanochemical effect of dolomitic talc during fine grinding process in mortar grinder. Int J Miner Process 2010;94:172–9. [23] Mucsi G, Rácz A, Mádai V. Mechanical activation of cement in stirred media mill. Powder Technol 2013;235:163–72. [24] Nkoumbou C, Villieras F, Njopwouo D, Ngoune C, Barres O, Pelletier M, et al. Physicochemical properties of talc ore from three deposits of Lamal Pougue area (Yaounde Pan-African Belt, Cameroon) in relation to industrial use. Appl Clay Sci 2008;41:113–32. [25] Pérez-Maqueda LA, Duran A, Pérez-Rodríguez JL. Preparation of submicron talc particles by sonication. Appl Clay Sci 2005;28:245–55. [26] Pérez-Maqueda LA, Caneo OB, Poyato J, Pérez-Rodríguez JL. Preparation and characterization of micron and submicron-sized vermiculite. Phys Chem Miner 2001;28:61–6. [27] Pérez-Maqueda LA, Franco F, Avilés MA, Poyato J, Pérez-Rodríguez JL. Effect of sonication on particle-size distribution in natural muscovite and biotite. Clays Clay Miner 2003;51:701–8. [28] Sanchez-Soto PJ, de Haro MDJ, Pérez-Maqueda LA, Varona I, Pérez-Rodríguez JL. Effects of dry grinding on the structural changes of kaolinite powders. J Am Ceram Soc 2000;83:1649–57. [29] Sanchez-Soto PJ, Wiewiora A, Aviles MA, Justo A, Pérez-Maqueda LA, PérezRodríguez JL, et al. Talc from Puebla de Lillo, Spain. II. Effect of dry grinding on particle size and shape. Appl Clay Sci 1997;12:297–312.

[30] Schomburg J, Zwahr H. Thermal differential diagnosis of mica mineral group. J Therm Anal 1997;48:135–9. [31] Serbian standard: SRPS EN 725-8:2010 Advanced technical ceramics Methods of test for powders - Part 8: Determination of density. [32] Serbian standard: SRPS EN 13286-46:2012 Unbound and hydraulically bound mixtures – Part 46: Test method for the determination of the moisture condition value. [33] Serbian standard: SRPS EN 725-9:2010 Advanced technical ceramics Methods of test for powders – Part 9: Determination of bulk density. [34] Suraj G, Iyer CSP, Rugmini S, Lalithambika M. The effect of micronization on kaolinites and their sorption behaviour. Appl Clay Sci 1997;12:111–30. [35] Tenório Cavalcante P, Dondi M, Guarini G, Barros F, Benvindo da Luz A. Ceramic application of mica titania pearlescent pigments. Dyes Pigm 2007;74:1–8. [36] Tkácová T. Mechanical activation of minerals. Amsterdam: Elsevier; 1989. 293. [37] Ulusoy U. Application of ANOVA to image analysis results of talc particles produced by different milling. Powder Technol. 2008;188(2008):133–8. [38] Yang H, Du C, Hu Y, Jin S, Yang W, Tang A, et al. Preparation of porous material from talc by mechanochemical treatment and subsequent leaching. Appl Clay Sci 2006;31:290–7. [39] Yavuz F. Evaluating micas in petrologic and metallogenic aspect: Part II— Applications using the computer program Mica. Comput Geosci 2003;29:1215–28. [40] Yekeler M, Ulusoy U, Hicyılmaz C. Effect of particle shape and roughness of talc mineral ground by different mills on the wettability and floatability. Powder Technol 2004;140:68–78. [41] Yvon J, Villieras F, Michot L. Effect of Different Dry Grinding Procedures on the Immersion Heat of Talc and Chlorite. J Min Metall 2005;41 A:1–9.