Journal

Paper

▲459The Journal of The South African Institute of Mining and Metallurgy VOLUME 105 NON-REFEREED PAPER AUGUST 2005

Introduction

The use of zircon (ZrSiO4) as a glaze opacifierhas become widespread in the white wareindustry. The presence of zircon crystals thathave a particle size that approaches thewavelength of incoming light allows for lightscattering and, thus, opacification1–2. Thethree parameters, which control the degree ofopacity obtained, are the following3:

➤ relative refractive index between thecrystals and the glass matrix

➤ number of independent light scatters➤ particle size and shape of crystals.

Zircon is used in industry in a wide rangeof applications, as zircon sand directly fromthe mine; as a milled product in the form ofzircon flour (95 per cent micronized to < 45 µm, used in the manufacturing of frit)and opacifier grades (100 per cent micronizedto < 6 µm, used in ceramic glazes for tiles and

sanitary ware); chemical processing or fusionof zircon sand (or baddelyite) to producezirconium chemicals and synthetic zirconia.The most important market for zircon is as anopacifier for ceramic tiles and sanitary ware.Both zircon flour and the micronized gradescan be used as opacifier. However, opacifiergrades or zircon opacifiers tend to refer tomicronized zircon (100 per cent < 6 µm) thatcan be introduced directly into the glaze.Typically, zircon glaze uses 8–12 per cent ofmilled zircon depending on the degree ofopacity requirements.

The particle size of zircon grades is alsovery important in terms of application.Different zircon grades are usually classifiedaccording to their d50 value (the mesh sizethrough which 50 per cent of the materialpasses). There are a number of milled zirconproducts available commercially, which vary ingrain size and purity. The following are themost commonly used grades of opacifier: (a) 9micron (d50 = 1.5–2.0 µm)—widely use inceramic glazes; (b) 6 micron (d50 = 1.2–1.5µm) —widely use in ceramic glazes and alsoused as cheaper substitute for the 5 micronmaterial in sanitary ware and fully vitrifiedunglazed tiles: (c) 5 micron (d50 = 0.8–1.2µm)—highest use product in fully vitrifiedunglazed tiles and sanitary ware. This gradetends to be the benchmark because of its highvalue-added/volume ratio4-5.

Zircon is the most commonly occurringmineral of zirconium. It is an extremelyefficient refractory mineral with a high meltingpoint, a low and regular coefficient of thermalexpansion and high thermal conductivity. The

The upgrading of an inferior gradezircon to superior opacifier for sanitaryware and glazesby E. Snyders*, J.H. Potgieter†, and J.T. Nel‡

Synopsis

The upgrading process of an inferior grade of zircon to a superioropacifier for ceramic applications was investigated. A calciningprocess in the presence of mineralizers was used to produce theproduct in a conventional muffle furnace. A pre-milled inferiorgrade zircon concentrate, with a relatively high iron and aluminacontent generally not suitable for high quality ceramic applications,was used in this study. The influence of the upgrading process onthe opacifying properties of the zircon material was compared tothat of commercially available prime grade zircon opacifier productscurrently being produced in South Africa. The influence of acalcining step in the presence of mineralizers with regard to themilling characteristics of the treated zircon material was alsoevaluated in this study.

The morphology of the treated and untreated zircon opacifiersamples were investigated by SEM micrographs and the colours ofzircon samples were characterized on the grounds of theCommission Internationale de l`Eclairage (CIE) standard procedure(CIE L* a* b* measurements) after application on a bisque ceramictile. It was found that the treatment improved both the milling aswell as the opacifying properties of the treated inferior grade zirconconcentrate and renders it equivalent to or better than prime gradezircon opacacifiers currently produced.

Keywords: ZrSiO4; opacifier; colour; milling; calcining.

* Department of Chemical and MetallurgicalEngineering, Tshwane University of Technology,Pretoria, South Africa.

† School of Process and Materials Engineering,University of the Witwatersrand, Wits, South Africa.

‡ South African Nuclear Energy Corporation Limited,Pretoria, South Africa.

© The South African Institute of Mining andMetallurgy, 2005. SA ISSN 0038–223X/3.00 +0.00. Paper received Apr. 2005; revised paperreceived May 2005.

The upgrading of an inferior grade zircon to superior opacifier for sanitary ware

mineral has a high hardness and is thus resistant toabrasion. A major feature of the mineral is the strength ofthe ZrO2-SiO2 bonds. Significant amounts of energy areneeded to break these bonds and therefore zircon is highlyresistant to chemical attack. Unaltered zircon has a highrefractive index, which makes it suitable for use as opacifier.Highly crystalline zircon does not absorb or reflect anyparticular wavelength of visible light and therefore givesceramics a white appearance. This unique property allowszircon to be used as an opacifier in ceramic applications.

The quality of the zircon determines the suitability for aparticular application. Quality specifications for zircon aretypically quoted in terms of the zirconia content (expressedas ZrO2 + HfO2), iron (Fe2O3), alumina (Al2O3) and titania(TiO2). An iron content of less than 0.07 per cent, designatedas a ‘premium’ grade zircon, is usually required for ceramicopacifier applications. However, refractory producers havelittle concern for iron levels, which could be as high as 0.3per cent. The iron, as well as the alumina content5, areimpurities present in zircon that negatively influence theopacity properties. The solubility of iron in the glaze causesdiscoloration of re-crystallized particles in the glaze. Inindustry, the iron content is primarily controlled throughgrain surface cleaning and efficient mineral separation. Highalumina contents are normally an indication of contami-nation by alumina-silicate minerals. The minerals are mosteffectively separated from zircon by using wet gravitationtechniques, such as spirals shaking tables and Kelsey jigs. Insome cases, final cleaning is done by dry gravity separationusing air tables6.

Zircon has to be milled or micronized to suitable particlesizes before it can be introduced into many of its end-useapplications. However, the energy costs for milling increaseexponentially with increasing fineness. Excluding the zirconprice, electric power is the highest cost element in theproduction of zircon as an opacifier, followed by the cost ofthe grinding media.

The aim of the present work was to study the effect that atreatment process in the presence of mineralizers has on theopacity properties of an inferior grade of zircon with arelatively high iron (Fe2O3) and alumina (Al2O3) content thatis generally not suitable for high quality ceramic applications,i.e. opacifier for glazes. The treated zircon material wascompared to three South African commercially available 5micron zircon opacifier products, produced from high qualityzircon concentrate (prime grade material), after applicationon a ceramic tile. The influence of the treatment process inthe presence of the mineralizers on the milling characteristicsof zircon was also investigated.

Materials and methods

A South African grade zircon, with relatively high Fe2O3 andAl2O3 contents, was used in this study as opacifier material.The chemical composition (determined by XRF spectroscopy)of the zircon grade used is given in Table I (column labelledas ZS). The zircon material, designated as a ‘standard grade’zircon product (ZS), was milled in a conventional ball mill toa zircon flour product (95 per cent < 45 micrometre). Themilled zircon flour product was mixed with selectedmineralizers in a Y-cone tumbler mixer in order to ensure athorough homogeneous blend and thereafter calcined in an

air furnace to produce a raw calcined product, referred to astreated zircon opacifier (designated ZT). After calcining, theproduct was washed with water to remove the excessmineralizers and impurities present in the calcined product.The resultant washed product was then wet milled in anMMS series RAPID mill to a average particle size (d50-value)of between 0.8–1.2 µm. Yttria-stabilized zirconia millingmedia was used in order to eliminate any contamination thatcould result from the milling media during the millingprocess. A 12 per cent opacifier/transparent glaze mixturewas prepared, mixed and applied to a bisque ceramic tile bymeans of a high-pressure spray gun and fired in a mufflefurnace at a temperature of 1080°C. The colour of theopacifier product was assessed on the grounds of L*, a* andb* parameters, calculated from the diffuse reflectance specra,through the method recommended by the CommissionInternationale de l`Eclairage (CIE). According to the methodthe parameter L* represents the brightness of a sample; apositive L* value indicates a light colour, while a negativeone corresponds to a dark colour; a* represents the green (-)→ red (+) colour axis and b* the blue (-) → yellow (+) colouraxis.

In order to evaluate the quality of the treated zirconopacifier, three commercially available South African 5-micron zircon opacifier products, supplied by EggerdingSouth Africa, were used for comparison purposes. Theindustrial zircon opacifier products were designated ZP1, ZP2and ZP3 and a typical chemical composition of such a SouthAfrican prime grade 5-micron zircon opacifier product isgiven in Table II.

The influence of the calcining process in the presence ofthe mineralizers on the milling characteristics of zircon wasalso investigated. A 1.5 kilogram batch of the standard gradezircon flour product was calcined in the presence of theselected mineralizers and the calcined product was washedwith water to remove the excess mineralizers and impuritiespresent in the product. A quantity of 1.5 kg of the calcinedand washed standard grade zircon product, as well as 1.5kilogram of standard grade zircon flour product, were milledsimultaneously in two separate roller jar mills for comparisonpurposes with yttria-stabilized zirconia as milling media for aperiod of 16 hours. Particle size measurements on the milledsamples were carried out at regular time intervals to comparethe milling times of both zircon samples used.

Particle size measurements were carried out on aSedigraph 5100 particle size analyzer supplied byMicromeritics.

▲

460 AUGUST 2005 VOLUME 105 NON-REFEREED PAPER The Journal of The South African Institute of Mining and Metallurgy

Table I

Chemical analyses of zircon flour material

Composition Unit ZS ZT

ZrO2 (+HfO2) % 66.6 66.6SiO2 % 32.7 32.7TiO2 % 0.24 0.25Fe2O3 % 0.20 0.04Al2O3 % 0.53 0.12Cr2O3 % < 0.01 < 0.01MgO % 0.01 0.01CaO % 0.12 0.02P2O5 % 0.12 0.12

The samples were demarcated as follows:➤ a first letter that is relative to the product (Z: zircon)➤ a second letter that is relative to the description of the

product (S: standard grade, P: prime grade, T: treated)➤ a number that is relative to the different commercially

available zircon opacifier products used.

Results and discussions

Table I reports the chemical analyses of the ‘standard grade’zircon flour product (ZS) used in this study compared to thecalcined standard grade zircon flour product (ZT). Theinfluence of the calcining step in the presence of themineralizers, particularly on the Fe2O3, Al2O3 and CaOimpurity levels, is evident. A substantial decrease in theFe2O3 content of the ZT sample is evident. The Fe2O3 contentdecreases from 0.20 per cent for the ZS sample to 0.04 percent for the ZT sample. As suggested by Poynton7, a zircongrade with an iron impurity level of 0.06 to 0.07 per cent,must be treated with concentrated hydrochloric acid orsulphuric acid for long periods either at ambient or elevatedtemperatures to reduce the iron level to a minimum of 0.035per cent. The treatment process, as suggested in this study, istherefore a very effective method to reduce the iron impuritylevel present in the zircon material as discreet grains.

A large decrease in the alumina content of the calcined‘standard grade’ zircon flour product compared to theuntreated zircon flour product also occurs. The alumina(Al2O3) content decreased from 0.53 per cent for the ZSsample to 0.12 per cent for the ZT product. A typical chemicalanalysis of the unmilled standard grade zircon product, asstated in the specification sheet of the supplier, shows analumina content of 0.25 per cent. The treatment processtherefore not only reduces the alumina constituents presentin the zircon grains as alumina-silicate minerals, but alsoalumina present due to contamination of the alumina millingmedia used to mill the sample to a zircon flour product. Alsoevident from the chemical analyses in Table I is the decreasein the CaO content in the ZS product from 0.12 to 0.02 percent for the ZT sample. The removal of the iron, aluminiumand calcium content in the ZS sample during the calcinationprocess therefore suggest a possible mechanism ofvolatilization of these constituents to form variousintermediate volatile iron, aluminium and calcium species inthe presence of the mineralizers.

Table III reports the CIE L* a* b* parameters of the ZTsample compared to that of three commercially availableprime grade zircon opacifier products (5 µm product)available in South Africa. In this method, the L* parameterrepresents the brightness (whiteness) of a sample; a morepositive L* value indicates a light colour (white) while a morenegative one corresponds to a darker colour (black); a*represents the green (-) → red (+) colour axis and b* the blue(-) → yellow (+) colour axis. The largest L* value of theindustrial zircon opacifier product corresponds to sample ZP1with a value of 88.62, while samples ZP2 and ZP3 havevalues of 88.20 and 88.10 respectively. A substantialincrease in the L* value to 90.11 is observed for sample ZT,giving it a much whiter appearance compared to the commer-cially available 5 micrometre zircon opacifier samples. The ZTsample is also characterized by lower values of a* and b*,indicating a tendency to achromatism. The a* values for thecommercial opacifier samples, samples ZP1, ZP2 and ZP3,varies from 2.08 to 2.26 compared to 1.57 for the ZT sample.The same tendency is seen for the b* values. The latter variesfrom 5.65 to 6.14 for the commercial opacifier samples,compared to 3.44 for the ZT sample.

Figure 1 shows the comparison of the L* value of the ZTsample compared to the L* values of the commerciallyavailable zircon opacifier samples with regard to the effect ofthe wt per cent opacifier applied on a ceramic tile. A range ofbetween 8–12 wt per cent opacifier was selected to cover theamounts most commonly used in industry and alsorepresenting a reasonable change in the L* values. Tofacilitate the preparation of the test tiles, a fixed weight ofglaze/opacifier mixture was applied per unit area by means ofa high-pressure spray gun. Uniformity of application wasmonitored by first weighing the test tiles, and then sprayingthe mixture to a fixed weight gain. It is evident from the testresults that the L* values for the ZT sample are much higherthan the results obtained for the commercially availablezircon opacifier products with each of the 8, 10 and 12 wt percent opacifier additions. The results in Figure 1 also indicatethat the L* value of 88.39 obtained for the ZT sample with an8 wt per cent opacifier addition is even higher than thevalues obtained of 88.10 and 88.20 for the ZP2 and ZP3samples respectively with a 12 wt per cent opacifier addition.Only sample ZP1 with an L* value of 88.62 with a 12 percent opacifier addition is higher than the L* value obtainedfor sample ZT with a 8 wt per cent opacifier addition. The L*value of 89.21 for the ZT sample with a 10 wt per centopacifier addition surpasses that of all the commerciallyavailable 5 micron products with 12 wt per cent opacifieradditions. Between 17 and 33 per cent less ZT product istherefore necessary to obtain the same level of whiteness ona ceramic tile compared to the commercially available 5-micron prime grade zircon opacifier products.

The upgrading of an inferior grade zircon to superior opacifier for sanitary wareJournal

Paper

▲461The Journal of The South African Institute of Mining and Metallurgy VOLUME 105 NON-REFEREED PAPER AUGUST 2005

Table II

Typical analyses of a South African premium gradezircon opacifier product

Composition Unit ZP – 5 µm

ZrO2 (+HfO2) % 66.2SiO2 % 32.6TiO2 % 0.11Fe2O3 % 0.05Al2O3 % 1.20Cr2O3 % <0.01MgO % 0.01CaO % <0.05P2O5 % 0.11U + Th mg/kg <500

Table III

CIE L* a* b* parameters of zircon opacifier products

Sample L* a* b*

ZT 90.11 1.57 3.44ZP1 88.62 2.08 5.65ZP2 88.20 2.25 5.77ZP3 88.10 2.26 6.14

The upgrading of an inferior grade zircon to superior opacifier for sanitary ware

Milling tests were performed on the 325 mesh zircon flourZT and ZS samples in order to evaluate the influence of thecalcination step on the milling characteristics of the respectivezircon products. Figure 2 graphically displays the comparisonin the mean particle size (d50 value) at different timeintervals. After the calcination step in the presence of theselected mineralizers, a slight increase in the mean particle to14.8 micrometre was measured in comparison with the ZSflour product of 12.3 micrometre. This effect is due to theagglomeration of certain particles during the calcining step.After a milling time of two hours, the particle size analysesalready shows a difference in the particle size of the ZTsample compared to the ZS sample. After 8.5 hours ofmilling, the mean particle size of the ZT sample measured 5.6micrometre compared to a particle size of 5.5 micrometreachieved after 15 hours of milling on the ZS sample.

SEM micrographs of the ZT sample are reported inFigures 3 and 4, and that of the ZS sample in Figure 5respectively. The definite role played by the presence of theselected mineralizers during the calcination stage is quite

evident in Figure 3 (x 8 500 magnification) and Figure 4 (x20 000 magnification). The most apparent feature in thesemicrographs is the evidence of cracking of the zircon particlesand ‘pinhole’ effects on the surface of the particle that wascaused by the presence of the selected mineralizers during thecalcination stage. In comparison with the ZT product, theuntreated 325 mesh ZS product, shown in Figure 5, has arelatively smooth surface appearance with no evident cracksin the individual zircon particles. The surface effects that arevisible in the ZT sample, Figure 3 and Figure 4, thereforeimprove the milling characteristics of the treated zirconsample quite dramatically compared to the ZS sample. This isalso evident from the results of the milling tests that arereported in Figure 2.

Conclusions

A superior zircon opacifier product was prepared by calciningan inferior grade of zircon concentrate with a relatively highFe2O3 and Al2O3 content, which is generally not suitable forhigh quality ceramic applications, in the presence of selected

▲

462 AUGUST 2005 VOLUME 105 NON-REFEREED PAPER The Journal of The South African Institute of Mining and Metallurgy

Figure 1—L* parameters of different zircon opacifier wt% additions

Figure 2—Milling comparison tests of ZS and ZT samples

mineralizers. The product was evaluated against threecommercially available prime grade zircon opacifier productsand the advantages of the treatment process have beenestablished. The CIE L* a* b* parameters showed that the L*value (whiteness) increased substantially after the treatmentprocess compared to the L* values of the three commerciallyavailable zircon opacifier products. Moreover, the obtainedresults suggest that between 17–33 per cent less of thetreated zircon opacifier product is necessary to obtain thesame level of whiteness after application on a ceramic tilecompared to the commercially available zircon opacifierproducts.

It was also found that the milling characteristics of thetreated zircon product, compared to the untreated zirconmaterial, improved dramatically after the calcining step in thepresence of the selected mineralizers. It was established inthis study that a milling time of 8.5 hours is necessary forthe treated material compared to that of 15 hours necessaryfor the untreated material to achieve the same mean particlesize of about 5.6 micrometre. This factor is significant whenit is considered that excluding feedstock, electric power is thehighest cost element in opacifier production, followed by thecost of grinding media.

The zircon treatment process, as suggested in this study,is therefore a very affective method to improve the opacifyingproperties of an inferior grade of zircon concentrate due tothe large decrease of the two main impurities, namely ironand alumina, that adversely affect the opacifier quality.Substantial cost savings can potentially be realized throughshorter milling times and consumption of the opacifier. Amore detailed analysis to take the cost of the calcinationprocedure into account is required in follow-up work toestablish the true economic evaluation of this proposedtreatment process.

Acknowledgements

The authors would like to express their gratitude to the SouthAfrican Nuclear Energy Corporation (Necsa) and theInnovation Fund of the National Research Foundation ofSouth Africa for the funding of the project. The valuablecontributions of Eggerding South Africa during this study aregratefully acknowledged. The authors would also like tothank TICOR South Africa and ASTRON for the chemicalanalyses of the zircon samples.

References

1. BOOTH, F.T. and PEEL, G.N. The principals of glaze opacification withzirconium silicate. Br. Ceram. Soc. Trans. 1959. vol. 58. pp. 532–564.

2. CASTILONE, R.J., SRIRAM, D., CARTY, W.M., and SNYDER, R.L. Crystallization ofzircon in stoneware glazes. J. Am. Ceram. Soc. 1999. vol. 82, no. 10. pp. 2819–2824.

3. BUCHANAN, R., TAYLOR, R.J., and HELLIWELL, J.G. Application of Mie theoryto opacification of zircon glazes. Br. Ceram. Trans. 1998. vol. 97, no. 2. pp. 83–86.

4. PEARSON, K. Grinding & bearing it—a zircon market insight. IndustrialMinerals. 1999. pp. 25–35.

5. REGGAL, H. Opacifiers, best results with dry processing. 17th IndustrialMinerals International Congress. Barcelona, Spain. 2004. pp.1–9.

6. The global zircon industry—a five year outlook with a special focus onChina. TZ MINERALS international PTY LTD. 2001.

7. POYNTON, C.J. Process for treating zircon. United States Patent 4,395,384:Appl. 6 April 1982: Acc. 26 July 1983. ◆

The upgrading of an inferior grade zircon to superior opacifier for sanitary wareJournal

Paper

▲463The Journal of The South African Institute of Mining and Metallurgy VOLUME 105 NON-REFEREED PAPER AUGUST 2005

Figure 3—SEM image of ZT sample (x8 500 magnification)

Figure 4—SEM image of ZT sample (x20 000 magnification)

Figure 5—SEM image of ZS sample (x7 500 magnification)

▲

464 AUGUST 2005 VOLUME 105 NON-REFEREED PAPER The Journal of The South African Institute of Mining and Metallurgy