INVESTIGATIONONS ON USE OF JAROSITE AS SET CONTROLLER IN CEMENT

INVESTIGATIONONS ON USE OF JAROSITE AS SET CONTROLLER IN CEMENT S K Agarwal, Puneet Sharma, Mithlesh Sharma and M M Ali National Council for Cement and Building Materials, Ballabgarh & B K Singh and Vikas Sharma Hindustan Zinc Limited, Udaipur ABSTRACT In the process of extraction of zinc metal through hydrometallurgical process, a by-product waste called Jarosite is generated which consists of soluble sulphates and small amount of ZnO and PbO. The chemico-mineralogical composition of jarosite was found to be compatible for use in the manufacture of cement. The present paper highlights the investigations carried out on replacement of mineral gypsum by varying proportions of jarosite in controlling the setting behaviour of Ordinary Portland Cement (OPC) and Portland Pozzolana Cement (PPC). Replacement of 20% mineral gypsum by jarosite in OPC was found optimum for controlling the setting time along with adequate compressive strength development. However, in case of PPC, replacement up to 40% mineral gypsum by jarosite was found suitable. This could be attributed to the combined effect of fly ash and jarosite on controlling cement hydration with regulated properties. As the jarosite sample was found to contain heavy elements, the hardened neat cement cubes were prepared and immersed in different solutions and leachates determined at different curing ages. Investigations up to 6 months have not indicated any leachability indicating fixation of heavy elements in the hardened cement matrix. The above investigations indicated potentials of utilizing jarosite as set controller for conservation of natural mineral gypsum. INTRODUCTION In the manufacture of ordinary Portland cement (OPC), the clinker is inter-ground with mineral gypsum for providing sulphate ions to regulate hydration of calcium aluminate phase resulting in retardation of cement setting. In the samples of jarosite, the presence of SO 3 as predominant oxide and presence of gypsum mineral phase showed resemblance with mineral gypsum and therefore, investigations was carried out at NCB on the technical suitability of jarosite partly replacing mineral gypsum as set controller along with compressive strength development. In addition, the strength development of resultant cement samples of OPC and PPC cured under different solutions and the leachability of heavy elements from hardened neat cement paste was also highlighted in the present paper. EXPERIMENTATION Chemical analysis of jarosite sample was carried out according to method specified in IS: 4032-1985 (Re 2005). The sample was analyzed for heavy elements using Inductive Coupled Plasma-Optical Emission Spectrometer (ICP -OES) [Model: VISTA-MPX, Varian Make]. Mineralogical analysis of jarosite was carried out by X-ray Diffractometer (XRD) [Rigaku International, Japan, D-MAX 2200V/PC, using CuK radiation (λ=1.5405å)]. The lime reactivity of the fly ash was determined as per the method laid down in Indian Standard IS: 1727-1967. 1

In order to investigate the role of jarosite as set controller in place of mineral gypsum in OPC and PPC, different cement samples were prepared by using clinkers CLK-1/CLK-2, fly ash (FA) and 5% gypsum which is made up of mixtures of mineral gypsum (GYP) and jarosite (JS -D) separately, the former decreasing from 4 to 1% and the later simultaneously increasing from 1 to 4% in the same proportion (Table 1). The above OPC and PPC samples were prepared by inter-grinding of all components in laboratory ball mill. Particle size distribution (PSD) of resultant cement samples was conducted on Microtrac S 3500. Performance evaluation of above OPC and PPC samples prepared by part/total replacement of mineral gypsum by jarosite was carried out according to Indian Standard methods described in IS: 4031.Leaching of heavy elements from above samples was studied by immersing the hardened cement cubes in above solutions for 3 and 6 months. Compressive strength developments of optimized OPC and PPC samples cured under different solutions such as normal water, sulphate (0.33N Na 2 SO 4 ), chloride ( 0.5N NaCl) and alkaline ( 0.3N NaOH) for 3 and 6 months was carried out by immersing the mortar cubes in above solutions. Table 1: Composition of OPC & PPC samples Cement code CLK-1 CLK-2 FA GYP JS-D OPC samples OPC-C 95.0 - - 5.0 0.0 OPC-1 95.0 - - 4.0 1.0 OPC-2 95.0 - - 3.0 2.0 OPC-3 95.0 - - 2.0 3.0 OPC-4 95.0 - - 1.0 4.0 OPC-5 95.0 - - 0.0 5.0 PPC samples PPC-C - 75.0 20.0 5.0 0.0 PPC-1-75.0 20.0 4.0 1.0 PPC-2-75.0 20.0 3.0 2.0 PPC-3-75.0 20.0 2.0 3.0 PPC-4-75.0 20.0 1.0 4.0 PPC-5-75.0 20.0 0.0 5.0 RESULTS AND DISCUSSIONS Characterization of jarosite, plant clinker and gypsum samples Chemical analysis of jarosite sample (JS -D) indicated presence of 27.13% Fe 2 O 3, 28.14% SO 3, 7.24%CaO, 12.02% SiO 2, 7.37% Al 2 O 3 and 0.33% MgO (Table 2) along with 2.07% moisture content and 7.60% insoluble residue (IR). Major heavy elements were determined to be 0.034% barium, 0.041% cadmium, 0.067% copper, 2.79% lead and 2.29% zinc. The solubility of the sulphate ions present in mineral gypsum and jarosite samples at 1 hour in normal water (ph~7) and at ph~10 are given in Table 3 and were found to be higher (61-63%) in case of jarosite as compared to solubility of SO 3 in mineral gypsum (47-48%). Mineral composition of jarosite sample showed the predominance of natrojarosite [NaFe 3 (SO 4 ) 2 (OH) 6 ] along with gypsum, quartz and anhydrite [CaSO 4 ]. Glass content in the sample was found to be 42% with average grain size of 36µ. The minimum and maximum sizes of glass particles were 2 and 65µ respectively. The Blaine s fineness of jarosite was determined to be 6000 cm 2 /gm. The particle size distribution showed the finer nature of the sample as 42.93% fraction was passed through 9.25µ and 20% particles were found to be below 5.50µ. The lime reactivity of the sample was determined to be 1.4 MPa indicating poor pozzolanic reactivity of the material. 2

Table 2: Chemical analysis of jarosite and other materials Oxide constituents JS-D FA-CH CLK-1 CLK-2 GYP LOI 33.88 2.29 0.40 0.35 - CaO 7.24 2.38 64.92 63.19 26.28 SiO 2 12.02 57.93 19.97 20.30 - Al 2 O 3 7.37 26.03 5.90 5.31 1.93 Fe 2 O 3 27.13 5.83 4.57 4.10 0.94 MgO 0.33 0.96 1.79 4.63 0.66 SO 3 28.14 0.39 1.08 0.92 32.77 Na 2 O 1.84 0.35 0.20 0.17 0.07 K 2 O 0.47 0.89 0.54 0.57 0.05 TiO 2 0.61 1.62 0.32 0.44 0.13 Mn 2 O 3 0.11 0.006 0.08 0.02 0.006 P 2 O 5-0.009 0.008 0.02 0.006 Chloride 0.019 0.009 0.10 0.029 0.006 Type of sample Mineral gypsum Jarosite (JS-D) SO 3 content (%) Table 3: Solubility study of SO 3 in jarosite samples SO 3 content in 10 gm Soluble SO 3 content in 10 gm Solubility at 1 hour Normal water % of Soluble SO 3 soluble SO 3 content in at ph~7 10 gm ph=10 30.77 3.08 1.46 47.40 1.47 47.70 28.14 2.81 1.72 61.21 1.76 62.63 % of soluble SO 3 at ph~10 Chemical analysis of plant clinker samples CLK-1 and CLK-2 showed presence of usual oxide constituents (Table 1). Free lime and insoluble residue in clinker samples were found to be 0.44 & 0.27 and 0.13 & 0.12% respectively. Semi-quantitative estimation of above clinkers by XRD showed presence of 61 & 59% alite, 19% belite, 3% C 3 A and 12 & 14% C 4 AF. Chemical analysis of gypsum sample showed 32.77% SO 3, 26.28% CaO and 20.23% SiO 2 +IR. The percentage of combined water was found to be 14.59%. The purity of gypsum based on SO 3 content was 70.46%. Gypsum [Ca(SO 4 )(H 2 O) 2 ] along with small amount of calcite [CaCO 3 ] and -quartz [SiO 2 ] are the mineral constituents of gypsum sample. Chemical analysis of fly ash sample (FA) indicated the presence of 57.93% SiO 2, 2.38% CaO, 26.03% Al 2 O 3, 5.83% Fe 2 O 3, 0.39% SO 3 and 0.96% MgO. The insoluble residue and reactive silica in fly ash sample were found to be 90.47 and 28.98% respectively. The physical test results of above sample showed Blaine s fineness-322 m 2 /Kg, retention on 45µ-26.23%, lime reactivity-6.06 MPa, compressive strength at 28 days-84.5% of control and autoclave soundness-0.062%. The above values showed conformity of fly ash to IS: 3812 (Part -I)-2003 and found suitable to be used as component in the manufacture of PPC. The glass content in the sample was determined to be 63%. Mineral composition showed predominance of quartz [(SiO 2 ] with considerable amount of aluminium silicates in the form of mullite [Al 6 Si 2 O 13 ]. Microstructure of fly ash sample showed presence of rounded shaped glass grains in majority along with platty, lath, needle and prismatic shaped glass grains. Setting behaviour of cement samples (mineral gypsum vis-à-vis jarosite) In order to investigate the role of jarosite as set controller, different cement samples OPC-1A & 1B and OPC-2A & 2 B were prepared by taking 3 and 5% doses of mineral gypsum and jarosite as per the 3

designed compositions given in Table 4. SO 3 content in above cement samples were found to be in the range of 1.89-2.66%. The setting behavior of ground clinker CLK-1 (fineness: 326 m 2 /kg & without gypsum addition) sample i.e. cement without set retarder showed flash set whereas initial and final setting times of cement samples OPC-1B and OPC-2B (prepared using 3 and 5% jarosite) were found to be 150 & 220 and 180 & 235 minutes and found to be almost comparable to cement samples OPC-1A (180 & 255 minutes) and OPC-2A (215 & 280 minutes) prepared using equal doses of mineral gypsum. This could be attributed to the availability of higher concentration of sulphate ions from jarosite due to relatively more solubility as compared to solubility of SO 3 in mineral gypsum (Table 3). The above trend of setting behavior of cement samples showed the potential of using jarosite as set controller in place of conventional mineral gypsum. Table 4: Setting behavior of cement samples using clinker CLK-1 Sample code CLK-1 Retarder Dose (%) SO 3 (%) Fineness (m 2 /kg) Setting time (minutes) IST FST CLK-1 100.00 - - 1.08 326 Flash set OPC-1A 97.0 Gypsum 3.0 2.03 330 180 255 OPC-1B 97.0 Jarosite 3.0 1.89 331 150 220 OPC-2A 95.0 Gypsum 5.0 2.66 325 215 280 OPC-2B 95.0 Jarosite 5.0 2.43 334 180 235 Properties and performance evaluation of OPC samples prepared using jarosite as set controller Chemical requirements: Chemical parameters such as LOI, IR and SO 3 specified for OPC-43 grade cement are 5.0 (max), 3.0 (max) and 2.50% (max) respectively (IS: 8112-1989) and all the resultant cement samples analysis showed these values meeting the above requirements (Table 5). Table 5: Chemical parameters of OPC samples Cement sample Percentage LOI IR SO 3 OPC-C 1.62 1.29 2.66 OPC-1 1.54 1.07 2.61 OPC-2 1.50 0.92 2.57 OPC-3 1.43 1.13 2.53 OPC-4 1.31 1.26 2.48 OPC-5 1.30 0.94 2.43 IS:8112-1989 5.0, max 3.0, max 2.5, max Particle size distribution: The particle size distribution of cement samples OPC-1 to OPC-5 showed finer nature of samples as higher percentage of particle fractions passed through 9.25, 5.5, 2.75 and 2.31µ as compared to control cement. Physical characteristics: Blaine s finenesses of cement samples OPC-C and OPC-1 to OPC-5 were determined to be 325, 327, 325, 328, 330, 334 m 2 /kg respectively (Table 6) and indicating marginal increase in the fineness with the increasing doses of jarosite. Water requirement of resultant cements were found to be 26.0-27.0% and found to be comparable to control (26.0%). The initial and final setting times of cement samples OPC-1 to OPC-5 were found to be 210 & 265, 195 & 245, 175 & 235, 160 & 215 and 180 & 235 minutes respectively (tested as per IS: 4031(5) -1988) against the setting time of control cement OPC-C (215 and 280 minutes) indicating comparable setting trend of cement samples. However, 4

all cement samples conformed to the setting requirements of Indian Standards for OPC i.e. IS: 8112-1989 (OPC-43) and IS: 12269-1987 (OPC -53). The compressive strength development of cement samples OPC-1 and OPC-2 were found to be comparable to control cement whereas in cement samples OPC-3 to OPC-5, the compressive strength was found to be marginally lowered at all ages as shown in Fig 1. However, all resultant cement samples conformed to the strength requirements of IS: 8112-1989 (OPC-43 grade). The results of linear and autoclave expansion, as determined by IS: 4031(3)-1988, showed high dimensional stability of the resultant cements. Table 6: Physical properties of OPC samples prepared using jarosite as set controller Properties OPC-C (5:0)* OPC-1 (4:1)* OPC-2 (3:2)* OPC-3 (2:3)* OPC-4 (1:4)* OPC-5 (0:5)* Blaine s fineness (m 2 /kg) 325 327 325 328 330 334 Consistency (%) 26.0 26.0 26.0 26.5 26.5 26.5 Setting time (minutes) Initial 215 210 195 175 160 180 Final 280 265 245 235 215 235 Compressive strength (MPa) 3 days 34.0 33.0 33.0 32.0 31.0 32.0 7 days 45.0 43.0 42.0 41.0 41.0 41.0 28 days 54.0 53.0 52.0 49.0 48.0 47.0 Soundness Le-chatelier (mm) 1.0 1.0 1.0 1.0 1.0 1.0 Autoclave (%) 0.03 0.03 0.03 0.04 0.03 0.04 *Mineral gypsum:jarosite 60 Comp. Strength (MPa) 50 40 30 20 10 0 OP C - C OP C - 1 OP C - 2 OP C - 3 OP C - 4 OP C - 5 Sample code 3D 7D 28D Linear (28D) Linear (7D) Linear (3D) Fig 1: Trend of compressive strength development of OPC samples prepared using jarosite as set controller Hydration studies of cement samples OPC-C and OPC-5 at 3 and 7 days showed formation of portlandite and ettringite phases. Unreacted calcium silicates were found be slightly higher at 3 and 7 days in control cement as compared to OPC prepared replacing mineral gypsum by jarosite i.e OPC-5 indicating higher rate of hydration. 5

Properties and performance evaluation of PPC samples prepared using jarosite as set controller Chemical requirements: The values of LOI, SO 3, IR and MgO showed that all resultant PPC samples prepared using different doses of jarosite as set controller fulfilled the chemical requirements of PPC cements as specified in IS:1489 (I)-1991 (Table 7). Table 7: Chemical parameters of PPC samples Cement Percentage sample LOI IR MgO SO 3 PPC-C 1.42 17.23 4.87 2.41 PPC-1 2.04 18.61 4.95 2.36 PPC-2 2.42 18.13 4.42 2.31 PPC-3 2.38 18.08 4.88 2.27 PPC-4 2.22 18.37 4.66 2.23 PPC-5 2.42 18.19 4.96 2.18 IS:1489-1991 5.0,max 23.2, max 6.0,max 3.0,max Particle size distribution: The particle size distribution of cement samples PPC-1 to PPC-5 showed wide distribution of particles (352.0-2.312µ) as compared to control cement (209.3-6.54µ) along with finer nature of cement samples as compared to control with percentage fraction passed through 9.25µ was found to be in the range of 16.67-18.79% against 12.33% in case of control PPC. Physical characteristics: The Blaine s finenesses of the cement samples PPC-1 to PPC-5 were found to be slightly increased with increasing doses of jarosite owing to the finer nature of jaroite sample (Table 8). Water requirements of all resultant PPC samples were found to be marginally increased as compared to control PPC due to the higher surface area of cements. The setting behaviour of the PPC samples showed slightly delayed setting trend as compared to control PPC. Compressive strength of the above samples showed marginal improvement in later age strength with increasing replacement level of jarosite with mineral gypsum. The trend of compressive strength development at 3 & 7 days were found to be comparable whereas at 28-days showed marginal improvement in strength with increasing dose of jarosite (Fig 2). The results of soundness of all cement samples showed higher dimensional stability. Drying shrinkage of cement samples PPC-1 to PPC-5 were found to be in the range of 0.03-0.04% as compared to control PPC (0.03%) and conformed to the IS requirement ( 0.15%). Table 8: Physical properties of PPC samples prepared using jarosite as set controller Properties PPC-C (5:0)* PPC-1 (4:1)* PPC-2 (3:2)* PPC-3 (2:3)* PPC-4 (4:1)* PPC-5 (0:5)* Blaine s fineness (m 2 /kg) 352 349 351 377 385 398 Consistency (%) 28.0 29.0 29.0 29.0 29.0 29.5 Setting time (minutes) Initial 150 175 170 200 215 210 Final 315 345 340 325 330 280 Compressive strength (MPa) 3 days 31.0 29.0 29.0 31.0 31.0 30.0 7 days 40.0 39.0 41.0 42.0 42.0 39.0 28 days 48.0 48.0 51.0 52.0 52.0 50.0 Soundness Le-chatelier (mm) 1.0 1.0 1.0 1.0 1.0 1.0 Autoclave (%) 0.05 0.06 0.04 0.05 0.06 0.07 Drying shrinkage (%) 0.03 0.03 0.04 0.03 0.04 0.04 * Mineral gypsum:jarosite 6

6 0 Comp.Strength (MPa) 50 4 0 3 0 2 0 10 0 PPC - C PPC - 1 PPC - 2 PPC - 3 PPC - 4 PPC - 5 Sample Code 3D 7D 28D Linear (3D) Linear (7D) Linear (28D) Fig 2: Trend of compressive strength development of PPC samples prepared using jarosite as set controller Long term compressive strength development of OPC and PPC samples Based on above discussions, cement samples OPC-1 and PPC-2 prepared replacing 20 and 40% mineral gypsum with by-product jarosite in OPC and PPC respectively was optimized. Long term compressive strength (up to 6 months) of above cement samples cured under different solutions showed comparable strength development to control cement as shown in Fig 3a for OPC and 3b for PPC. Investigations up to 6 months have not indicated any leachability indicating fixation of heavy elements in the hardened cement matrix. 80 Compressive strength (MPa) 70 60 50 40 30 20 10 0 OPC-C OPC-1 OPC-C OPC-1 3M 6M Cenment sample (months) Water Sulphate Alkaline Chloride Fig 3a: Compressive strength development of OPC samples at 3 and 6 months cured under different solutions 7

Compressive strength (MPa) 80 70 60 50 40 30 20 10 0 PPC-C PPC-2 PPC-C PPC-2 3M Cenment sample (months) 6M Water Sulphate Alkaline Chloride Fig 3b : Compressive strength development of PPC samples at 3 and 6 months cured under different solutions CONCLUSIONS 1. Ordinary Portland cement sample OPC-1 prepared using 4 part mineral gypsum and 1 part jarosite was found to be suitable in regulating the setting of cement and therefore 20% mineral gypsum could be replaced by jarosite in OPC. 2. In PPC, 40% replacement level of mineral gypsum by jarosite i.e. 3 part mineral gypsum and 2 part jarosite was considered suitable for controlling setting of PPC and strength development. 3. The use of jarosite could result in gypsum saving. The potential of conserving mineral gypsum was found more in case of blended cement which contributes to the major share of cement production in India. ACKNOWLEDGEMENT The work reported in this paper forms a part of study sponsored by M/s Hindustan Zinc Limited, Udaipur, India. The authors are grateful to Director General, National Council for Cement and Building Materials, Ballabgarh, Haryana, India for allowing the publication of the paper. 8