PROCESS ENGINEERING

Performance and Cost Optimisation in Ceramic Tile Production:

Achieving Whiteness and Opacity through Zircon Opacifier
S. Cook, J. Masbate, Ch. Barrington

In global ceramic tile production zircon remains, in the opinion of the Zircon Industry Association (ZIA),
the material of technical choice for imparting opacification and whiteness. Despite the price volatility
over the period 2011–2012, the crucial role of zircon in ceramic tile production appears to have been
sustained, although substitution or thrifting of zircon by alternative whiteners, notably calcined alumina,
became more prevalent. As with most raw materials, thrifting and substitution of zircon is incentivised
when supply-demand dynamics push prices above trend levels. However, following the peak in 2012, a
lower and more stable price level for zircon has been evident.
With this in mind, ZIA believes that it is now time for ceramic tile producers to re-evaluate the advan-
tages offered by zircon in tile manufacture, most importantly, based on its relative economic value-in-
use. Against this background, ZIA commissioned the Spanish Instituto de Tecnología Cerámica (ITC) to
undertake a comparative study of the performance of zircon against alternative materials in ceramic tile
applications, namely bodies, glazes, frits, and engobes. This paper presents the key findings, particularly
the effect of opacifier particle size on performance in porcelain tile bodies.

Introduction
Micronised zircon is used as an opacifier in
the production of engobes, opaque glazes,
and porcelain tile bodies [1–2]. The price
of zircon, as for other materials, is largely
driven by supply-demand factors. Elevated
price levels, as seen in 2011–2012 (Fig. 1),
tend to incentivise thrifting and substitution
of zircon by alternative materials. In most
cases, substitution impairs product proper-
ties (mainly whiteness and opacity). How-
ever, for economic reasons these quality
trade-offs have been accepted. Fig. 1 Indicative zircon price trend, 2007 – 2015 (Source: Iluka/TZMI)
In porcelain tile body manufacture, zircon is
used in compositions with varying degrees
of whiteness, accounting for a significant fiers (including alumina), together with vari-
proportion of the porcelain tiles made. An ous additives to offset the change in perfor- Simon Cook, Jorge Masbate, Chris Barrington
earlier zircon price spike in 2006 led to a mance caused by the different properties of Zircon Industry Association
considerable reduction in zircon content the opacifiers. Camberley GU15 1AB
in body compositions (in certain cases up As with ceramic tile bodies, engobe and Great Britain
to 40 %) with accompanying reduction of glaze producers have sought to reduce
whiteness. To offset the consequent loss in costs by using cheaper materials, as a re- Corresponding author: Ch. Barrington
whiteness, efforts have been made to im- sult of which zircon has been substituted in E-mail: [email protected]
prove whiteness by using higher quality raw some applications, typically by alumina.
materials, reducing clay content, and using A clear exception is the use of zircon in www.zircon-association.org
feldspars and sands with very low chromo- opaque frits for glazes – in this case, re-
phore oxide contents [3–5]. search and development work has shown Keywords: zircon, whiteness, opacifier,
Whitening raw material mixes have also that there are no viable substitutes for value-in-use, ceramic tile
emerged, made up of one or more opaci- zircon in this type of frit. Some potential

cfi/Ber. DKG 92 (2015) No. 10-11 E1

 PROCESS ENGINEERING

Fig. 2 Cross-sections of the porcelain tile specimens: B-ZR (l.), B-AL1 (r.); Identified phases: ZrSi) micronised zircon (white particles);
Q) quartz; and Al) alumina (grey particles)

Tab. 1 Deflocculation behaviour

Composition B-ZR B-AL1 B-AL2

Solids content
71 71 71
[%]
Minimum
420 550 550
viscosity [cP]
Deflocculant
0,29 0,25 0,26
content [%]
Thixotropy
14 22 30
[diff. ºG]

a reject of 2,0 % was reached on a 40-µm

sieve. The resulting suspensions were dried
Fig. 3 Evolution of the L* coordinate with the opacifier particle size for porcelain tile under infrared lamps and dry milled.
body compositions with 5 mass-% zircon and 7,5 mass-% alumina, respectively
Data on deflocculation curves of the com-
positions tested are shown in Tab. 1. The
substitutes do exist, but they are unsuitable behaviour and properties of porcelain tiles composition with zircon provided the best
for either technical or economic reasons, or when the zirconium silicate in the body deflocculation behaviour, with the lowest
both. composition was entirely replaced with viscosity and thixotropy. The data obtained
Since the price spike of 2011–2012, prices alumina. For this study, a standard body with the aluminas were quite similar to
for zircon opacifiers have fallen sharply to composition (B-ZR) consisting of [mass-%] each other.
a much more stable level which has been 30 % Ukrainian clay, 50 % Turkish feldspar, Cylindrical test specimens (4 cm in diam-
sustained ever since, as shown in Fig. 1. 15 % quartz, and 5 % zircon (d50 = 1,4 µm) eter and approximately 7 mm thick) were
This should prompt ceramic producers to re- was used as reference. The zircon was en- formed by uniaxial pressing, at a pressing
visit the advantages and, most importantly, tirely replaced with two types of calcined moisture content of 5,5 % (dry weight ba-
the relative value-in-use of zircon opacifiers alumina commonly used in the manufacture sis) and a pressing pressure of 30 MPa. The
relative to alternatives, such as calcined of porcelain tiles: AL1 (d50 = 5,2 μm) and test specimens were dried at 110 °C to con-
alumina-based whiteners. AL2 (d50 = 5,1 µm). Secondly, composi- stant weight and then fired with a fast firing
The ITC study aims to determine perfor- tions with different levels of alumina were cycle and 6-min hold at peak temperature
mance of micronised zircon relative to tested in order to determine the alumina in an electric laboratory kiln. The heating
calcined alumina in porcelain tile body, en- content that provided the same whiteness rate was 25 °C/min.
gobe, and glaze applications. (L* coordinate) as the zircon-containing ref- Tab. 2 summarises the properties of these
erence composition. compositions (bulk density, linear shrink-
Porcelain tile body compositions To conduct the study, the different raw age, water absorption, and chromatic co-
The aim of this section was firstly to de- materials were mixed in the appropriate ordinates) at the maximum densification
termine the influence on the processing percentages and wet milled in water until temperature (Tmax), this being the firing

E 2 cfi/Ber. DKG 92 (2015) No. 10-11

 PROCESS ENGINEERING

Tab. 2 Properties at maximum densification temperature (Tmax) Tab. 3 Properties of the fired
pressed engobe compositions;
Composition B-ZR B-AL1 B-AL2
water absorption = 0 %
Tmax [ºC] 1219 1224 1227
Composition E-ZR E-AL3 E-AL4
Bulk density [g/cm ]3
2,492 2,474 2,474
Temperature
Linear shrinkage [%] 8,5 8,8 8,7 1080 1090 1100
[ºC]
Water absorption [%] <0,1 <0,1 <0,1 Bulk density
2,305 2,130 2,119
L* 84,2 82,0 82,3 [g/cm3]
a* 0,5 0,8 0,9
b* 8,9 9,3 9,0 From Fig. 3, it can be observed that
Bending strength [kg/cm ] 2
700 ± 10 720 ± 10 690 ± 10 5,0 mass-% loading of zircon (d50 = 1,5 µm)
corresponds to about 7,5 mass-% loading
of alumina (d50 = 4 µm) to provide porcelain
tile bodies with a similar L* value (84,6).
Furthermore, the slope of the L* vs particle
size lines are quite different for zircon and
alumina. For zircon, the slope is significantly
steeper, so that there is a significant benefit
in performance if the opacifier is milled to a
smaller size. For alumina the benefit of finer
milling is much smaller.
Consequently, for finer zircon (d50 = 1,2 µm)
the alumina loading required to match the
same level of whiteness is higher, close
to 8,5 %. Therefore, to achieve the same
L* value, the alumina added to the porce-
lain tile body formulation must be between
Fig. 4 Variation of the L* chromatic coordinate with temperature for engobed tiles 50–70 % higher than when finer micron-
ised zircon is used, depending on the zircon
temperature at industrial scale. The main In conclusion, when 5 mass-% calcined alu- particle size.
inferences to be drawn from these results mina was added to the porcelain tile body,
were as follows: the differences between zircon and alu- Engobe compositions
•The use of alumina increased maximum mina were not critical. However, the need Tests were performed substituting the
densification temperature and shrinkage, to increase the alumina content in order to zircon in an engobe composition used in
which modified final tile size. The higher achieve a similar whiteness to that of the the production of wall tiles. The reference
shrinkage values could be related to the zircon-containing composition would aug- engobe composition (E-ZR) contained
higher porosity in the dry compacts when ment these differences. [mass-%]: 35 % opaque frit, 20 % sodium
the two aluminas were used. To determine the effect of the opacifier feldspar, 25 % white clay, 12 % quartz, and
•Both aluminas provided the tiles with a particle size and to obtain porcelain tiles 8 % zircon. Two commercial alumina sam-
lower L* coordinate than that of the zir- with alumina, with a whiteness similar ples were used as zircon substitutes: AL3
con-containing composition. The a* and to that provided by the 5 mass-% zircon- (d50 = 45 μm) and AL4 (d50 = 15 μm).
b* values were a little higher. containing composition, further tests were The engobe compositions were character-
•The zircon-containing composition ex- conducted with other body compositions, in ised by determining their firing behaviour
hibited the lowest closed porosity, while which the zircon and alumina particle size (pressed engobe compositions), firing range
the compositions with the two aluminas and content were changed. Fig. 3 summa- (variation of aesthetic characteristics of en-
displayed similar closed porosity. SEM rises some of the results obtained. gobed tiles with temperature), and thermal
observation (Fig. 2) revealed significant Typically, zircon for porcelain tile bodies expansion.
differences between the zircon and the (“five” quality, green circle on the chart) Microstructural characterisation was also
alumina particles: the zircon particles were has a particle size (d50) of about 1,5 µm, al- performed, which consisted of identifying
very small (<5 µm), whereas the alumina though many commercially available micro- the arising crystalline structures (XRD) and
particles were much larger in some cases nised zircon are finer, with d50 in the range observing the engobe layers by scanning
(<10 µm). 1,0–1,3 µm. On the other hand, commer- electron microscopy (SEM).
•Stain resistance is related to closed poros- cially available calcined alumina for zircon The determination of the firing behaviour
ity: consequently, the zircon-containing substitution in porcelain tile bodies has a (Tab. 3) showed that replacing zircon with
composition provided slightly better stain particle size (d50) between 3–5 µm (orange alumina led to increased engobe refracto-
resistance. circle on the chart).

cfi/Ber. DKG 92 (2015) No. 10-11 E3

 PROCESS ENGINEERING

Fig. 5 Cross-sections of the pressed engobe compositions (Left: E-ZR 1080ºC – Right: E-AL3 1100ºC). Identified phases: Q) quartz; Ab)
albite; ZrSi) micronised zircon; ZrSid) zircon devitrified from the opaque frit; and Al) alumina.

Fig. 6 Variation of the L* chromatic coordinate (l.), and gloss (r.) with temperature for glazed tiles

Tab. 4 Characteristic temperatures of the riness (i.e. a higher firing temperature was The same was the case with microstruc-
test glazes [ºC]
needed to reach the same water absorp- ture: apart from the presence of alumina
tion). It would be possible to correct this by particles instead of zircon particles, no sig-
Characteristic increasing the engobe flux content, in order nificant changes in microstructure (XRD and
G-ZR G-AL3 G-AL4
Temperature thus to reduce the firing temperature to the SEM) were observed (Fig. 5).
appropriate level (this was not done in this
Shrinkage start
915 1090 925 study). Glaze compositions
temperature
Engobed tiles (Fig. 4) with the zircon- Tests were performed substituting zircon
Shrinkage end containing engobe gave rise to higher L* in a glaze composition used in the produc-
1170 1215 1200
temperature coordinate values (i.e. higher opacity and tion of glazed porcelain tiles. The refer-
whiteness) when zero water absorption ence glaze composition (G-ZR) contained
Softening was reached (1100 ºC). Alumina produced [mass-%]: 30 % matt frit, 20 % wollas-
1185 1260 1225
temperature whiter engobes at lower temperatures, tonite, 26 % nepheline syenite, 5 % alumi-
when zero water absorption was not na, 5 % quartz, 8 % kaolin, and 6 % zircon.
Half-sphere
1245 1300 1265 reached. This phenomenon is typically due Two commercial alumina samples, the same
temperature
to the higher opacity of the engobe layer ones as those used in the engobe compo-
Melting with zircon. With respect to thermal expan- sitions (AL3 and AL4), were used as zircon
1275 1315 1285
temperature sion, no significant changes were observed substitutes. The glaze compositions were
when zircon was replaced with alumina. referenced G-AL3 and G-AL4.

E 4 cfi/Ber. DKG 92 (2015) No. 10-11

 PROCESS ENGINEERING

Fig. 7 Cross-sections of the glazes: G-ZR 1180 ºC (l.), and G-AL4 1180 ºC (r.); identified phases: Al) alumina; ZrSi) micronised zircon;
Q) quartz; Wo) wollastonite; Gh) gahnite; Na/Al/Si) a phase rich in sodium, aluminium, and silicon; and Al/Si/Ca) a reaction layer richer in
aluminium, silicon, and calcium

Tab. 5 Resistance to wear by pedestrian higher characteristic temperature values. In alumina as opacifiers in porcelain tile bod-
traffic [ºC] general, the glazes prepared with alumina ies, engobes, and glazes.
exhibited higher whiteness, but much lower In porcelain tile bodies it was necessary
Glazed Abrasion gloss (Fig. 6). However, no change in ther- to increase the alumina loading rate by
Result Class
Tiles Stage mal expansion was observed. 50–70 mass-% to obtain the same de-
Change The use of alumina increased the formation gree of whiteness. These results allow a
in visible 2500 revo- of crystalline phases (Fig. 7). proper economic comparison to be made.
G-ZR H4
appear- lutions Microstructural characterisation showed The use of alumina impaired the rheology
ance (*)
that, in this type of glaze, the use of alumina of the body suspension, increased the fir-
Change instead of zircon favoured the formation of ing temperature, and reduced final tile size.
in visible >10 000 new crystalline phases like anorthite, labra- For polished products, alumina increased
G-AL3 H6
appear- revolutions
dorite, and gahnite. As the quantity of such the closed porosity slightly; as a result, the
ance (*)
crystalline phases increased, the refractori- zircon-containing composition provided
Change ness and opacity of the glaze increased and slightly better stain resistance.
in visible 7500 revo-
G-AL4 H5 glaze gloss decreased. The results obtained as a function of parti-
appear- lutions
This contribution to opacity depended on cle size in porcelain tile body showed that,
ance (*)
the size and refractive index of the new for zircon, there was a significant benefit in
(*) Change of appearance owing to surface
staining crystals. Gahnite has a high refractive in- performance when the opacifier was milled
dex (1,80), while the values for anorthite to a smaller size. For alumina the benefit of
The glaze compositions were characterised and labradorite are lower, namely 1,58 finer milling was much smaller.
by determining their fusion behaviour (hot and 1,56, respectively. It should be taken The replacement of zircon in engobe and
stage microscope), firing range (variation into account that alumina, with a refractive glaze compositions led mainly to increased
of aesthetic characteristics of glazed tiles index of 1,77, will also contribute to glaze refractoriness.
with temperature), and thermal expansion. opacity. In the case of engobe compositions, in order
Microstructural characterisation was also The presence of crystalline phases in a glaze to use alumina as opacifier, reformulation
performed, consisting of the identification increases wear resistance, so that the use of of the composition would be necessary. In
of crystalline structures (XRD) and the ob- alumina will also increase wear resistance, the case of the glazes, the use of alumina
servation of the glaze layers by scanning as evidenced by the results of the determi- gave rise to a significant decrease in glaze
electron microscopy (SEM). nation of the resistance to wear by pedes- gloss because it favoured the formation of
Finally, the glazed tiles were characterised trian traffic (UNE EN ISO 10545-7: 1999) crystalline phases that made it difficult to
by determining their chemical resistance shown in Tab. 5. obtain glazes with a non-matt finish.
and resistance to wear by pedestrian traffic. This study equips ceramic tile producers
The replacement of zircon with alumina Conclusions with performance data for zircon and its
increased glaze refractoriness (Tab. 4); the This paper reports on a study to compare alternatives, making possible value-in-use
glazes prepared with alumina displayed the use of micronised zircon and calcined comparisons.

cfi/Ber. DKG 92 (2015) No. 10-11 E5

 PROCESS ENGINEERING

Acknowledgements
The authors gratefully acknowledge the
assistance and cooperation of Instituto de
Tecnología Cerámica/ES in the preparation
of this paper.

References
[1] Zirconium Concentrate: Global Industry Markets
& Outlook. Roskill 2014 report. http://www.
roskill.com/reports/industrial-minerals/zirco-
nium/leaflet. 17.07.2015
[2] Iluka Ceramics Tile Study: 07.03.2014. www.
iluka.com/docs/default-source/mineral-sands-
briefing-papers/iluka-ceramics-tile-study-%287-
mar-2014%.29; 17.07.2015
[3] Ferrari, S.; Gualtieri, A.F.: The use of illitic clays in
the production of stoneware tile ceramics. Appl.
Clay Sci. 32 (2006) [1–2] 73–81
[4] Salem, A.; Jazayeri, S.H.; Rastelli, E.; Timellini,
G.: Effect of nepheline syenite on the colorant
behavior of porcelain stoneware body. J. Ceram.
Process. Res.10 (2009) [5] 621–627
[5] Ferrari, S.; Gualtieri, A.F.; Bellarmi, T.; Pasquali,
C.: The use of talc in super-white porcelain tile
bodies: mineralogy, chemistry and microstruc-
ture. C+CA (2004) [3–4] 69–83

E 6 cfi/Ber. DKG 92 (2015) No. 10-11