(This is a sample cover image for this issue. The actual cover is not yet available at this time.

This article appeared in a journal published by Elsevier. The attached

copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

International Journal of Adhesion & Adhesives 40 (2013) 202–209

Contents lists available at SciVerse ScienceDirect

International Journal of Adhesion & Adhesives

journal homepage: www.elsevier.com/locate/ijadhadh

Evaluation of various cellulose ethers performance in

ceramic tile adhesive mortars
Jean-Yves Petit a,b,n, Eric Wirquin a
a
Université Lille Nord de France, F59000 Lille, France
b
Momentive Speciality Chemicals, F60772 Ribécourt, France

a r t i c l e i n f o a b s t r a c t

Article history: Eleven cellulose ethers (CE) were selected and tested in a cement based ceramic tile adhesive (CTA)
Accepted 12 September 2012 formulation designed to highlight the effect of the CE on the end-use values of the mortars. Results
Available online 26 September 2012 showed that the end-use values, either in the fresh or hardened state, are strongly influenced by the
Keywords: latex powder/CE choice. This is due to the competitive adsorption between these two components;
Composite mortars in the fresh state, it affects the CE concentration in the pore solution and hence the viscosity of the mortar,
Latex setting time, and skinning. In the hardened state, when CE films do not prevent evaporation at the tile–
Cement mortar interface, adhesion in hot curing conditions is lowered compared to dry tensile adhesion values. Test
Cellulose ether results showed that CTA formulations should be thought of in terms of the CE/latex powder couple, since
Ceramic tile adhesive
interactions between this couple and the cement, strongly influence end-use values of the mortar.
& 2012 Elsevier Ltd. All rights reserved.

1. Introduction and hydroxyethyl cellulose (HEC) (Fig. 1). The chemical nature of
HEC is determined mainly by two parameters i.e. the molecular
Adhesive mortars are widely used to fix ceramic tiles as weight (Mw) and the hydroxyethyl content (% OC2H4OH).
decorative finishes or for fixing polystyrene sheets onto building Conversely, HPMC and HEMC are characterised by three struc-
walls as part of external insulation systems; this is because their tural parameters i.e. the molecular weight, the methoxyl content
bond strength and flexural properties increase the cladding’s life- (% OCH3) the hydroxypropyl (in HPMC, % OC3H6OH) or the
time. To achieve the required adhesive properties, polymer disper- hydroxyethyl content (in HEMC, % OC2H4OH).
sions are added to the mortar formulation. Liquid polymer However, one negative aspect of the use of these macromole-
dispersions can be spray-dried and used as a powder component cules in mortar formulation is the delay in cement hydration
in dry-mix mortar formulations. When mixed with water, the [3–5]. Pourchez et al. highlighted various delays in cement
powder redisperses and forms films in the cement matrix, which hydration induced by cellulose ether (from 10 min up to several
further enhance the mechanical properties of cementitious mortars. hours) [6,7]. This delay seemed to depend mainly on the chemical
Cellulose ethers are commonly introduced into industrial structure of the molecule and in particular, on the degree of
mortar formulations in order to obtain some of the required substitution. It was shown that the adsorption of the polymers
properties of the adhesive mortar; from the fresh paste to the onto the cement clinker phases inhibits the formation of portlan-
hardened material [1]. These cellulose derivatives are molecules dite [8]. It was assumed that the CE strongly effects hydrated
which improve the water retention and workability of the fresh calcium silicate precipitation, leading to a decrease in the amount
material by a process of air entrapment, as noted in Fig. 1. They of initial hydrated calcium silicate nuclei, delaying the formation
also improve the adhesion to substrates [2]. Among the wide of a continuous hydrated calcium silicate shell around the
variety of existing cellulose ethers, four types are commonly used tricalcium silicate grain, and finally delaying the formation of a
in mortar manufacturing: methyl cellulose (MC), hydroxypropyl- thicker and more permeable hydrated calcium silicate layer.
methyl cellulose (HPMC), hydroxyethylmethyl cellulose (HEMC) When a CE is added to a cementitious matrix, a gradual reduction
of tricalcium aluminate dissolution rate is observed; this is also
associated with ettringite and calcium hydroaluminate precipita-
Abbreviations: CE, cellulose ether; MC, methyl cellulose; HPMC, hydroxypropyl- tion. Hydroxyethyl cellulose induces a stronger adsorption on
methyl cellulose; HEMC, hydroxyethylmethyl cellulose; HEC, hydroxyethyl cellu- calcium hydroaluminates and a longer tricalcium aluminate
lose; RDP, latex redispersible powder; CTA, ceramic tile adhesive
n
Corresponding author at: Université Lille Nord de France, F59000 Lille, France.
hydration delay than does hydroxypropylmethyl cellulose [9].
Tel.: þ33 3 44 75 86 47; fax: þ33 3 44 75 86 41. The lower the content of methoxyl groups, the larger will be the
E-mail address: [email protected] (J.-Y. Petit). retardation of cement setting time.

0143-7496/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijadhadh.2012.09.007
 Author's personal copy

J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209 203

Table 2
Cellulose ether employed in this investigation according to their type and use.

CE Type Code viscosity of the 2%

CE solution (mPa s)

MC MC1 5500
MC2 13,500

HEMC HEMC1 5000

HEMC2 14,000
HEMC3 25,000
HEMC4 25,000
HEMC5 30,000
HEMC6 44,500

Mixed HEMC/ HPMC Mixed1 6000

Mixed2 15,000
Mixed3 17,000

Fig. 1. Structure of cellulose ethers: (a) MC, (b) HEC, (c) HEMC, (d) HPMC. Table 3
Characteristics of the CEM I 52.5 N cement.
Table 1
Mix design for ceramic tile adhesive. Chemical characteristics (%, in mass) Physical characteristics

Component CTA mix design SiO2 20.0 specific gravity (g/cm3) 3.1
Al2O3 5.1 Blaine specific area (cm2/g) 4200
CEM I 52,5 N cement (mass % of dry mixture) 36 Fe2O3 3.3 Compressive strength (Mpa)
Silica sand (mass % of dry mixture) 58.36 CaO 63.9 2 days 24
Calcium carbonate (mass % of dry mixture) 2 MgO 0.8 7 days 37
Cellulose ether (mass % of dry mixture) 0.34 SO3 3.1 28 days 62
Calcium formate (mass % of dry mixture) 0.3 K2O 0.73 Vicat setting time (min)
RDP (mass % of dry mixture) 3 Na2O 0.32 Initial 160
Water/powder (mass %) 25 S– 0.03 final 240
Cl- 0.06
CO2 1.1
Free CaO 1.8
More generally, the properties of the CE depend on both the
Na2O eq 0.8
molecular weight of the polymer and on the degree of etherification. L.O.I 2.0
The mechanical properties of the ethyl cellulose are reported to be Clinker 97.0
influenced more by the former, whereas other physical properties other 3
such as solubility and water absorption are influenced more by the
latter [10]. Such information is not generally provided as it is
considered to be confidential by CE manufacturers; however, it of a 2% solution of each ranged from 5000 to 44500 mPa.s. Data
has been demonstrated that the average molecular weight of the on the cellulose ether are summarised in Table 2. The viscosities
polymer is linked to its solution viscosity, when it is dissolved in of the CE solutions were taken from the technical data sheets of
water at a set concentration and reference temperature [11]. the manufacturer. All viscosity measurements of the polymer
The present study aims at evaluating the effect of CE type on solutions (concentration at 2% by mass) were done using a
the fresh paste parameters of adhesive mortars for ceramic tile, Brookfield viscometer, with an RV type spindle turning at
formulated with a latex redispersible powder (RDP). Such data are 20 rpm and at a set temperature of 23 1C.
necessary to highlight some of the mechanisms affecting work- The cement used was a CEM I 52.5 N complying with the EN
ability of RDP mortars over time and which will affect their use on 196-1 standard, with a Blaine fineness of 4200 cm2/g and com-
construction sites. posed mainly of clinker. Physical and chemical characteristics of
the cement are summarised in Table 3.

2. Materials and mixture proportioning

3. Test methods
A typical C2TE ceramic tile adhesive formulation was used to
investigate the effect of various CEs on RDP mortars (Table 1). As mortar properties can be affected by the mixing procedure
C2TE refers to the fresh and hardened adhesive properties as and equipment and in order to have comparable results, the same
defined in EN 12004 and ceramic tile adhesive refers to a fine mixer and the same mixing sequence were used for each set of
inorganic aggregate Portland cement based mortar modified with mortars. The RDP and cellulose ether were incorporated into the
polymer material (redispersible powder) capable of adhering two cement prior to mixing. The mortars were prepared in batches of
other materials together by means of surface attachment. 2 l using a helicoidal mixer rotating 62 rpm. The mixing consisted of
In this investigation, a redispersible powder based on a adding the water into the mixer and then introducing the powders
modified vinyl acetate copolymer was used. This polymer had (binder, redispersible powder, cellulose ether and sand) gradually
neither carboxylic nor hydroxyl groups on its surface, minimising over 30 s with the mixer off. The mortar was then mixed during 30 s
any possible interactions between the RDP and the CE. Polyvinyl at a speed of 62 rpm. After a rest period of 30 s, the mixing was
alcohol was added at the end of the production process as a resumed for an additional 90 s at 62 rpm. Prior to sampling, the
drying aid to assure the redispersibility of the powder. mixture was left at rest for 5 min before a final 15 s homogenisation
Eleven sets of CEs were selected: 2 MC-based, 6 HEMC-based, mixing. The raw materials were conditioned to assure the targeted
and 3 composed of a mixtures of HEMC and HPMC. The viscosities test temperature of the mixed mortars. In order to avoid heat loss or
 Author's personal copy

204 J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209

gain during mixing, all equipment used for mixing and testing was drop in viscosity is seen, with an increase in the viscosity of the
maintained at a constant temperature of 23 1C. 2% CE solution from 5500 to 13,500 mPa.s (MC1 to MC2). As for
Viscosity and specific gravity measurements of the mortars the HEMC based mixtures, with the exception of the HEMC1 mix
were done on a 290 mL sample. Due to the high viscosity of the which exhibits the lowest 2% CE solution viscosity, the lowest
tested samples, a Brookfield viscometer turning at constant mortar viscosity and the highest density, all collected data were
rotating speed of 20 rpm and using an H spindle was used. The in the same range in terms of density and viscosity of the mortar
viscosity was taken after a stabilisation period of 1 min. for mixtures prepared with HEMC2 to HEMC6. As for HEMC/
Monitoring of rheological parameters was done during the plastic HPMC mortars, no trend can be seen in the viscosity of the mortar,
state, 5 and 20 min after the end of the batching sequence. The but the density of the mixture decreases with the increase in the
measurements were done using an Anton Paar MCR 301 rheometer 2% CE solution viscosity.
with a plane/plane geometry. The upper and lower plates were Setting time measurements are highly dependent on the cellu-
roughened to avoid any slippage between the tested sample and the lose ether type, as shown if Fig. 3. For MC-based CE or the mixed
surface of the measuring device. The test protocol consisted in a equivalent, the lower the viscosity of the 2% CE solution, the shorter
linear ramp of shear rate (g1) from 0 to 2.5 s  1 during 30 s, followed is the setting time. This implies that the adsorption of the cellulose
by a 30 s constant shear rate of 2.5 s  1. Typical measurements ether onto cement grains increases with increasing viscosity of the
during the linear ramp show a rapid increase in shear stress followed 2% CE solution and hence with the increase in the cellulose ether
by a descent. The peak value of shear stress corresponds to the yield molecular weight. The same observations are made for mixtures
stress (t0), e.g. when the sample switches from a deformation state prepared with HEMC/HPMC–based cellulose ether. On the contrary,
to a flow. Viscosity measurements at constant shear rate were done no particular trend is seen for HEMC mortars.
at the constant shear rate of 2.5 s  1, after a stabilisation period of Patural et al. [13] showed that in a CTA like mortar, made without
30 s. In order to respect the open time test protocol, the sample a RDP and with a water over powder ratio of 0.30, the higher the
tested at 20 min. was prepared at the end of the batch and not average molecular weight of the cellulose ether, (i.e. the higher the
covered. This protocol is adapted from one currently used in the food viscosity of the 2% solution) the lower is the viscosity of the mortar.
industry for ‘solid’ samples i.e. for non Newtonian fluids and time This is in contradiction with the results obtained with the
dependant materials [12]. After each rheological test, the sample HEMC2- to HEMC6-based mixtures which exhibit the same level
which had been already sheared was discarded to ensure the same of mortar viscosity, with no relation between mortar viscosity and
shear history between replicate test samples.
For ceramic tile adhesive mixtures, dry, wet and heat adherence
tests were performed and open time measurements at 20 and 30 min 14
Initial setting time
were done according to the EN 1348 and EN 1346, respectively. The Final setting time
12
open time refers to the tensile joint strength of the adhesive mortar
Setting time (hrs)

when tiles are embedded in the applied ceramic tile adhesive 20 or 10

30 min after application of the CTA onto a concrete slab.
8

6
4. Results and discussion 4

4.1. Effect of various cellulose ethers on fresh properties of ceramic 2

tile adhesive mortars
0

3
1

2
ed
C

ed

As shown in Fig. 2, a switch from one cellulose ether to ed

M

EM

EM

EM

EM

EM

EM

ix
ix

ix
m
m

m
H

another, at the same dosage by mass, produces sharp changes in

either viscosity or density of CTA mortars. While the two MC- Fig. 3. Effect of various cellulose ethers on the setting times of ceramic tile
based cellulose ethers gave similar results in terms of density, a adhesive mortars.

1.6 45
40
1.55
35
1.5 30
µ (Pa.s)
Density

25
1.45
20
1.4 15
10
1.35
5
1.3 0
MC1

MC2

HEMC1

HEMC2

HEMC3

HEMC4

HEMC5

HEMC6

mixed1

mixed2

mixed3

Density Viscosity at 20 rpm

Fig. 2. Effect of various cellulose ethers on the density and viscosity of ceramic tile adhesive mortars.
 Author's personal copy

J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209 205

the viscosity of the CE 2% solution, i.e. the molecular weight of the 20

cellulose ether. As the redispersible powder also adsorbs onto the 18
surface of the cement grain [14,15], it is assumed that a competi- Initial Final
16
tion between the cellulose ether and the redispersible powder

Setting time (hrs)

takes place. Consequently, more cellulose ether is released into 14
the interstitial solution, acting as a viscosity enhancing admixture 12
and explaining the high viscosity level of the HEMC2- to HEMC6- 10
based mortars. This is also observed in the decrease in initial
8
setting time, as the CE preferentially absorbs onto aluminates.
A decrease in absorption of cellulose ether onto cement grains can 6
be translated by a decrease in the setting time due to less 4
inhibition of nucleation sites. 2
To verify such a hypothesis of redispersible powder/cellulose
0
ether competition, a set of mortars were made with only one of no RDP / no CE RDP / no CE no RDP / CE RDP / CE
each of the following; the HEMC1 as cellulose ether, the redis-
persible powder and the 0.15% of polyvinyl alcohol that can be Fig. 5. Initial and final setting time (Vicat) on mortars depending on components
of the binder.
released into the interstitial solution. One mortar was made
without any of these additions and a final complete mortar
including redispersible powder and cellulose ether. All these Since it was demonstrated that RDP can also adsorb onto the
mixtures were prepared at constant water over powder ratio surface of cement grains, it can be concluded that a competition in
(w/p, by mass) to permit comparisons after rheology measure- adsorption between cellulose ether and redispersible powder takes
ments. Results are summarised in Table 4. place in the interstitial solution of the mortars. In the case of the
Results clearly demonstrate that part of the CE is adsorbed MC1-based mixture, it can be assumed that this competition is in
onto the surface of the cement grains and the remainder is favour of the redispersible powder, explaining the high result
released into the interstitial solution, explaining why both the obtained for the viscosity of the mortar: most of this cellulose
viscosity and setting time of the mixture including cellulose ether ether should be kept back into the interstitial solution, resulting in
and redispersible powder are lower than the values obtained for increasing the viscosity of the paste. As MC1 2% solution viscosity
CE-only mortar (Figs. 4 and 5). In the latter, as no competition is relatively low (low molecular weight) a lower absorption rate of
occurs, cellulose ether can freely adsorb onto the cement, extend- this CE was expected, compared to high molecular weight cellulose
ing the setting time compared to the mortar with neither ethers; in accordance with literature examples [13].
cellulose ether nor redispersible powder. A smaller amount of The selective competition onto the cement grain of the
cellulose ether is released into the interstitial solution, explaining cellulose ether/ redispersible powder will affect cement hydra-
why the viscosity of this mortar is slightly lower than that tion, impacting mechanical interlocking of the cementitious
measured on the CE-only solution. Note that such adsorption hydrates within the surface porosity of the tile. It will also impact
competition has already been suggested in a previous study [16]. CE and RDP distribution at the tile’s interface. As these two
polymers are known to form films between the inorganic matrix
of the mortar and the surface of the tile, a change in the
distribution of these two materials can lead to drastic changes
Table 4 in the adhesion properties brought by the polymers. In particular,
Effect of the different components on mortar. as the surface tension of the cellulose ether is higher than that of
the RDP’s or the polyvinyl alcohol’s, bringing CE to the interface
Mix design including monitored No RDP RDP No RDP
may favour the formation of hydrogen bonds between the
characteristics of the fresh paste no CE no CE RDP CE
CE substrate and the CE film. This can be verified by open time and
joint strength measurements.
Specific gravity (g/cm3) 2.04 1.98 1.67 1.59
Viscosity of the mortar at 20 rpm (Pa s) 1.2 2.2 41.2 36 4.2. Effect of various cellulose ethers on open time of ceramic tile
Initial setting time (h) 4 5.75 12.25 9.25
Final setting time (h) 5.25 6.75 18 9.50
adhesive mortars

As shown in Table 5 and Fig. 6, apart from the HEMC-based

mixture, the better the transfer on the glass time, the better the
45
joint strength at either 20 or 30 min. In a fresh state, water
40 evaporates quickly on the surface of the CTA ridges. A thin
cellulose ether film forms after 5 min, preventing further eva-
Viscosity at 20 rpm (Pa.s)

35
poration. More water migrates from the centre of the ridges,
30 partially dissolves the initial film, and eventually evaporates. The
25 water flux also transports additional CE to the surface of the
adhesive where it concentrates with time, forming a more solid
20 skin. This skin will no longer dissolve; thus preventing the
15 adhesive from wetting the tile used to perform tests [17]. One
drawback is that this skin also diminishes adhesive properties at
10 the mortar/tile interface with time. Consequently, an increase in
5 setting delay due to a better and longer adsorption of the cellulose
ether on the cement grain, as the one observed on HEMC1- or
0
no RDP / no CE RDP / no CE no RDP / CE RDP / CE
HEMC2-based mixes, decreases the migration of CE to the surface
for skin formation. As previously mentioned, this will impact on
Fig. 4. Viscosity at 20 rpm on mortars depending on components of the binder. the adhesion properties as the surface tension of the cellulose
 Author's personal copy

206 J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209

Table 5
Results on fresh mortars.

CE code MC1 MC2 HEMC1 HEMC2 HEMC3 HEMC4 HEMC5 HEMC6 mixed1 mixed2 mixed3

Viscosity of the mortar 42 26 28 36 36 40 36 40 28 40 25

at 20 rpm, (Pa s)
Open time (MPa)
at 20 min 1.64 1.05 0.86 1.78 0.94 0.87 0.71 0.75 1.45 0.90 0.85
at 30 min 0.89 0.38 0.49 1.33 0.70 0.52 0.37 0.30 1.47 0.60 0.55
Initial setting time (h) 9 10.75 10 11.75 8.75 9.5 5.25 7.75 5.75 7.75 10.5
Final setting time (h) 10.5 12 11.75 13 10.25 10.75 7.25 9.25 8 9.75 12.25

Open time @ 20 min Table 6 regroups the measured yield stress at 20 and 30 min
and the corresponding open time. The Ds30 corresponds to the
2.00 Open time @ 30 min variation between the two peaks, when available. Note that,
except for the MC2-based mixes, high Ds30 values correspond
Joint strength (MPa)

to low open time values at 30 min, for mixtures exhibiting two

1.50
peaks in yield stress at 30 min. This can be explained as follows: a
high Ds30 value should correspond to a thick skin or a dry mortar
1.00 surface, which will result in a bad open time value. A low Ds30
value should correspond to a thin protective skin that will easily
dissolve when tiling, resulting in good open time performances.
0.50
Note that rheological measurements allow the detection of
skin control problems as the plate/plate geometry of the rhe-
0.00 ometer used in this investigation mimics the tiling procedure:
C1 C2 C1 C2 C3 C4 C5 C6 ed1 ed2 ed3 when the upper plate is put into contact of the sample, a normal
M M M M M M M M ix ix ix force is applied to get the necessary 1 mm width of the sample.
HE HE HE HE HE HE m m m
This procedure is similar to the one described in the EN 12004
Fig. 6. Open time measurements at 20 and 30 min. (according to EN 12004). standard in which a 2 kg load is applied for 30 s on the tile to
embed it into the mortar.

ether is higher than that of the RDP or the polyvinyl alcohol.

4.3. Effect of cellulose ether on adhesion performance of ceramic tile
A decrease in cellulose ether concentration at the interface will
adhesive mortars.
lower the rate of hydrogen bond formation between the substrate
and the CE film.
The interfaces and adhesion properties between ceramic tiles and
Skinning is essential to prevent surface drying of the mortar
polymer modified Portland cement mortar are derived from several
ridges applied onto a concrete slab, but skin control is also
physical and chemical phenomena that take place when they are
indispensable as
formed. The interfacial resistance is affected by a number of factors,
such as ceramic tile water absorption, the amount of cement and its
 A too thin skin that dissolves permits surface evaporation and, composition, the amount and type of polymer used as cement
hence, local decohesion of the matter.
modifier, installation procedures, and water to cement ratio [16].
 A too thick skin due to high residual CE concentration in the pore For example, ceramic tile water absorption influences the migration
solution does not allow local remixing of the matter while tiling,
of pore solution into the tile porosity allowing the precipitation of
with dissolution of the cellulose ether film into the mortar.
cement hydrates inside the pores and promoting a mechanical
anchoring [17]. Moreover, the ceramic tile/ mortar interface is not
a static system but an evolving process depending on weathering,
As the skin control is of importance for the adhesive properties
mortar drying and hydration shrinkage, degree of hydration of
between the mortar and the tile and as it is related to cellulose
cement, tile size and location on the construction site [18].
ether film formation at the surface of the CTA, rheological
measurements made at 20 and 30 min should allow an estimation
of open time, e.g. the joint strength when tiling is done 20 or 4.3.1. Dry adhesion: tensile test performed after 28 days curing
30 min after the application of the CTA onto a concrete slab. at 23 1C and 50% relative humidity.
As previously mentioned, the yield stress is measure at the shear Data presented on Fig. 8 showed that no particular relation-
stress peak or change in slope value in the shear stress response of ship can be observed between the viscosity of the 2% CE-solution
the material to a linear increase in shear rate. However, as shown on and the mortar adhesion stress to ceramic tile in dry curing
Fig. 7(b), shear stress response, at 30 min, to a linear increase in conditions. During the hardening process, cement hydration and
shear rate exhibits a double peak on most of the tested mixtures. latex film formation continuously increase the strength of the
One of the two peaks should correspond to the destructuration of bulk system. One critical parameter during hardening is the water
the skin, allowing a surface flow in the plane/plane rheometer. concentration gradient, which is generated by capillary transport
The second peak should match the beginning of the flow inside the into the concrete substrate, by evaporation at the tile grouts and
mortar, which is related to the yield stress. However, this double by chemical drying due to cement hydration. Due to the early
response does not allow the identification of which peak corre- drying, the cement hydration can be drastically reduced, and,
sponds to the yield stress and which to the skin destructuration. thus, differs from the degree of hydration of systems without this
Nevertheless, it is assumed that the thicker the skin the greater the early loss of water [18]. The water retention ability of cellulose
shear stress value corresponding to its destructuration. ether and the concentration of the finest-sized pores retard
 Author's personal copy

J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209 207

MC2 t20
600 2500

500
2000

Shear stress (Pa)

Viscosity (Pa.s)
400
1500
Shear Stress
300 Viscosity
1000
200

500
100

0 0
0 0.5 1 1.5 2 2.5
Shear rate (1/s)

MC2 t30
double peak in shear stress
1600 10000

1400 9000
8000
1200
7000
Shear stress (Pa)

Viscosity (Pa.s)
1000 6000
Shear Stress
800 Viscosity 5000

600 4000
3000
400
2000
200 1000
0 0
0 0.5 1 1.5 2 2.5
Shear rate (1/s)

Fig. 7. Plate/plate rheological response of MC2-based mortar at 20 min (a) and at 30 min (b), Yield evaluation on shear stress response.

Table 6
Yield stress and open time values at 20 and 30 min.

CE code MC1 MC2 HEMC1 HEMC2 HEMC3 HEMC4 HEMC5 HEMC6 mixed1 mixed2 mixed3

Yield stress (Pa)

at 5 min 857 257 745 655 590 609 545 653 557 541 439
at 20 min. 1300 407 1070 1150 1660 858 627 1010 822 637 442
at 30 min. 1720/2540 1230/1480 2120 1742/3010 2187/2527 1700/2120 1818/2998 1349/2971 2130 2790/3054 1342
Ds30 (%) 47.7 20.3 – 72.8 15.5 24.7 64.9 120.2 – 9.5 –

Open time (Mpa)

at 20 min 1.64 1.05 1.78 0.86 0.94 0.87 0.71 0.75 1.45 0.9 0.9
at 30 min 0.89 0.38 1.33 0.49 0.7 0.52 0.37 0.3 1.47 0.6 0.41

Double values means double peak for Yield stress measurements.

drying. Moreover, enrichment of polyvinyl alcohol and cellulose the CE-RDP competition, in the porosity of the tile. As the
ether at the tile/ substrate interface contributes to the adhesive rheological properties of the bulk paste are mainly governed by
strength. However, skinning due to cellulose ether migration at the cellulose ether (air entrapment due to high surface tension),
the surface prevent entrapped air to be mixed and sheared into the yield stress measurements at the time of the tiling (5 min)
the paste while tiling. Instead, the entrapped air remains at the (Table 6) show that different amounts of CE are released into the
interface with the tile and decreases the contact area between tile pore solution after CE/ RDP competition. As cellulose ether will
and mortar, thus reducing adhesion properties [19,20]. Conse- affect hydrated calcium silicate, ettringite and calcium hydroalu-
quently, adhesive performances of CTA mortars are closely related minate precipitation, it will impact on the interlocking and the
to skin formation that may occur at the time of the tiling; 5 min mechanical anchoring of cementitious hydrates in the porous
applying the mixture onto a concrete slab. It explains the surface network of the tile. Further, the ratio of cellulose ether to
diversity of the results and the absence of correlation between RDP films at the interface should be linked to the release of
the CE properties and joint strength as skinning also depends on cellulose ether into the interstitial solution, affecting hydrogen
 Author's personal copy

208 J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209

2.5 5. Conclusions
dry wet heat
2 The effect of various cellulose ethers, with viscosities of their 2%
Joint strength (MPa)

solution varying from 5000 to 44,500 mPa s, on the end-use value

of ceramic tile adhesive mortars were evaluated. Based on the
1.5
results presented herein, the following conclusions are warranted:

1 1. Cellulose ether type strongly influence fresh properties of

ceramic tile adhesive mortars in terms of viscosity, at a fixed
0.5 dosage of cellulose ether. This is due to a competition between
cellulose ether and the redispersible latex powder in adsorbing
0 onto cement grain.
2. The cellulose ether/redispersible powder competitive adsorption
1

3
C

ed

ed

ed
M

M
EM

EM

EM

EM

EM

EM

ix also affects setting time and skinning of ceramic tile adhesive

ix

ix
M

M
H

mortars, with obvious effects on open time at 20 and 30 min.

Fig. 8. Joint strength in dry, wet and hot curing conditions on the tested mortars. 3. A too thin or too thick skin formation at either 20 or 30 min
can be detected using plate/plate rheology. This methodology
forecasts the open time behaviour of the mortar monitored at
bonding and the weak cohesion between interfacial layers (air 28 days by tensile tests.
bubbles release at the tile/ mortar interface). This explains the 4. Joint strength is also affected by the redispersible powder/
various joint strengths obtained which are presented in Fig. 8. cellulose ether couple: when cellulose ether films do not
prevent evaporation at the tile–mortar interface, adhesion in
hot curing conditions is lowering compared to dry tensile
4.3.2. Wet adhesion: tensile test performed after 7 days curing
adhesion values.
at 23 1C and 50% relative humidity, and 21 days storage in water.
5. A ceramic tile adhesive mix design should be thought in terms
As previously stated, polyvinyl alcohol and cellulose ether at
of cellulose ether/redispersible powder couple as interactions
the tile/ substrate interface components can form films that
between this couple and cement strongly influence end-use
contribute to the adhesive strength [17], with a major role of
values of the mortar.
cellulose ether compared to the low rate of polyvinyl alcohol in
the pore solution of the mortar. However, these films are weaker
than latex films and, most of all, water soluble. As a consequence,
water can dissolve these polyvinyl alcohol and cellulose ether References
films, explaining weaker adherence strength in wet conditions. In
addition, water can also weaken the latex film adhesive strength. [1] Ruot B, Goto T, Pourchez J. Some aspects of cellulose ethers and latexes
influence on the properties of cement-based materials—examples of results
This explains the lowering of the adhesive strength between dry obtained within the CEReM. In: Proceedings of the VII SBTA (71 Symposio
adhesion and wet adhesion tests, independently of the cellulose Brasileiro De Tecnologia Das Argamassas); 2007.
ether in use (Fig. 8). Note that the mechanical anchoring by the [2] Bertrand L, Maximilien S, Guyonnet R. Wedge splitting test: a test to measure
the polysaccharide influence on adhesion of mortar on its substrate. In:
cementitious hydrates in the porous surface network of the tile Maultzsch M, editor. Proceedings of the 11th international congress on
can also be affected as hydration is a time dependant reaction and polymers in concrete. ICPIC/BAM pub. Berlin (Germany); 2004. p. 569–76.
the storage of the sample in water can affect the migration of pore [3] Khayat KH. Viscosity-enhancing admixtures for cement-based materials—an
overview. Cem Concr Compos 1998;20(2–3):171–88.
solution into the tile porosity, allowing the precipitation of [4] Peschard A, Govin A, Grosseau P, Guilhot B, Guyonnet R. Effect of poly-
cement hydrates inside the pores and promoting a mechanical saccharides on the hydration of cement paste at early ages. Cem Concr Res
anchoring. This also explains the lowering of the adhesive 2004;34(11):2153–8.
[5] Betioli AM, Gleize PJP, Silva DA, John VM, Pileggi RG. Effect of HMEC on the
strength between dry and wet adhesion tests and this is inde-
consolidation of cement pastes: isothermal calorimetry versus oscillatory
pendent of the cellulose ether used (Fig. 8). rheometry. Cem Concr Res 2009;39(5):440–5.
[6] Pourchez J, Peschard A, Grosseau P, Guyonnet R, Guilhot B, Vallée F. HPMC and
HEMC influence on cement hydration. Cem Concr Res 2006;36(2):288–94.
4.3.3. Hot adhesion: tensile test performed after 14 days curing at [7] Pourchez J, Grosseau P, Guyonnet R, Ruot B. HEC influence on cement
23 1C and 50% relative humidity, and 14 days curing at 70 1C and 50% hydration measured by conductometry. Cem Concr Res 2006;36(9):1777–80.
[8] Weyer HJ, Müller I, Schmitt B, Bosbach D, Putnis A. Time-resolved monitoring
relative humidity. of cement hydration: Influence of cellulose ethers on hydration kinetics. Nucl
Cement hydration and capillary transport towards the substrate Instrum Methods Phys Res B 2005;238:102–6.
that take place after the tile inlaying is affected by drying due to [9] Pourchez J, Grosseau P, Ruot B. Current understanding of cellulose ethers
impact on the hydration of C3A and C3A-sulphate systems. Cem Concr
hot curing conditions. If cellulose ether and latex films do not Compos 2009;39(8):664–9.
prevent enough evaporation, hydration at the tile interface will be [10] Lorand EJ. Cellulose ethers: variations of physical properties with composi-
incomplete, resulting in a decrease in joint strength in hot curing tion. Ind Eng Chem 1938;30(5):527–30.
[11] /http://www.ashland.com/Ashland/Static/Documents/AAFI/From%20East%20
conditions. This is the case for HEMC2, HEMC3, HEMC4 mortars Database/22001.pdfS.
and for Mixed3 CTA, which exhibit lower joint strength in hot [12] Whittingstall P. Measuring the viscosity of non-newtonian fluids, viscosity of
curing conditions than in dry curing conditions (Fig. 8). On the liquids, solutions, and fine suspensions. Current protocols in food analytical
chemistry. Hoboken, New Jersey: John Wiley & Sons, Inc.; 2001, p. 375-384.
contrary, when film formation of latex, polyvinyl alcohol and [13] Patural L, Marchal P, Govin A, Grosseau P, Ruot B, Deve s O. Cellulose ethers
cellulose ether can prevent evaporation from the bulk material influence on water retention and consistency in cement-based mortars.
and under the tile/ mortar interface, hydration of the cement paste Cem Concr Res 2009;41(1):46–55.
[14] Plank J, Gretz M. Study on the interaction between anionic and cationic latex
is enhanced, promoting mechanical anchorage of the cementitious
particles and Portland cement. Colloids Surf A 2008;330(2–3):227–33.
hydrates in the pores of the tile by increasing the degree of [15] Rottstegge J, Arnold M, Herschke L, Glasser G, Wilhelm M, Spiess HW, et al.
hydration, resulting in an increase in joint strength, as shown on Solid state NMR and LVSEM studies on the hardening of latex modified tile
Fig.9 on MC2, HEMC6, Mixed1 and Mixed2 mixtures. For these mortar systems. Cem Concr Res 2005;35(12):2233–43.
[16] Mansur AAP, Do Nascimento OL, Mansur HS. Physico-chemical characteriza-
mortars, it is also assumed that the temperature promotes not only tion of EVA-modified mortar and porcelain tiles interfaces. Cem Concr Res
mechanical interlocking, but also polymer bonding. 2009;39:1199–208.
 Author's personal copy

J.-Y. Petit, E. Wirquin / International Journal of Adhesion & Adhesives 40 (2013) 202–209 209

[17] Jenni A, Holzer L, Zurbriggen R, Herwegh M. Influence of polymers on [19] Jenni A, Herwegh M, Zurbriggen R, Aberle T, Holzer L. Quantitative microstruc-
microstructure and adhesive strength of cementitious tile adhesive mortars. ture analysis of polymer-modified mortars. J Microsc 2003;212(2):186–96.
Cem Concr Res 2005;35(1):35–50. [20] Wetzel A, Zurbriggen R, Herwegh M. Spatially resolved evolution of adhesion
[18] Dimmig A. Einflqsse von Polymeren auf die Mikrostruktur und die properties of large porcelain tiles. Cem Concr Compos 2010;32:327–38.
Dauerhaftigkeit kunststoffmodifizierter Mfrtel (PCC). PhD thesis. Weimar:
Bauhaus-Universität; 2002.