You are on page 1of 12


discussions, stats, and author profiles for this publication at:

Compatibility between superplasticizer

admixtures and cements with mineral

Article · January 2012


35 289

4 authors, including:

F. Puertas M. M. Alonso
Spanish National Research Council Spanish National Research Council


Some of the authors of this publication are also working on these related projects:

NORM4BUILD View project

carbonatacion del cemento aluminoso hidratado. Alteraciones microestructurales y relación con sus
propiedades mecánicas durables View project

All content following this page was uploaded by F. Puertas on 27 August 2015.

The user has requested enhancement of the downloaded file.

(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:
Author's personal copy

Construction and Building Materials 31 (2012) 300–309

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

journal homepage:

Compatibility between superplasticizer admixtures and cements

with mineral additions
Olga Burgos-Montes a, Marta Palacios b, Patricia Rivilla a, Francisca Puertas a,⇑
Eduardo Torroja Institute for Construction Sciences (IETcc-CSIC), Serrano Galvache, 4, Madrid 28033, Spain
Institute of Building Materials, ETH Zurich, Schafmattstrasse 6, 8093 Zurich, Switzerland

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

Article history: The incorporation of mineral additions to Portland cement reduces the amount of clinker required in
Received 26 April 2011 cement manufacture improving the eco-efficiency of this process. They also impact rheology and it
Received in revised form 13 December 2011 may affect the interaction between superplasticizers and cements. The presented study explores the
Accepted 23 December 2011
effect of limestone, fly ash and silica fume on Portland cement and the interaction of these additions with
naphthalene (PNS)-, melamine (PMS)-, lignosulfonate (LS)- and polycarboxylate (PCE)-based admixtures.
The adsorption isotherms, zeta potentials and rheological behaviour of the blended cements were found
and compared to the same parameters in non-blended cement. The results showed that cement–super-
plasticizers compatibility was altered by the physical (specific surface) and chemical (surface charge)
Superplasticizers admixtures characteristics of the mineral additions studied.
Limestone Ó 2011 Elsevier Ltd. All rights reserved.
Fly ash
Silica fume

1. Introduction properties of the end material, which is more fluid and less porous
than non-blended cement pastes [15–18].
The incorporation of mineral additions to Portland cement re- In light of its very fine particle size, silica fume has been studied
duces the amount of clinker required in cement manufacture as an addition to densify cement paste. As cements made with this
improving the eco-efficiency of this process by lowering both addition exhibit higher compressive strength than non-blended
greenhouse gas emissions and energy consumption [1]. These materials, it is used to manufacture high-strength concrete [19–
additions also enhance mainly the long-term strength and durabi- 21]. Moreover, it has been observed to act not only as filler or poz-
lity of the material obtained. Additions may be natural materials zolanic material, but also to improve the cement paste–aggregate
(limestone, pozzolans, schist) or industrial by-products (vitreous interface, which contributes to further improving mortar and con-
granulated blast furnace slag, fly ash or silica fume) with pozzola- crete strength. Its incorporation also affects fresh cement flowabil-
nic or hydraulic properties [2,3]. ity, however, substantially raising the water demand, and its high
Limestone-blended cements have been widely studied. In terms early age hydration reactivity increases the heat of hydration and
of durability, permeability and strength, they are similar to non- consequently intensifies total shrinkage [22–25].
blended cements, while exhibiting high compressive strength Cement paste, mortar and concrete microstructure, and in
when small amounts of limestone are added [4–7]. A final material particular the homogeneity, density and porosity of the reaction
with the desired properties can only be obtained, however, when products, must be closely monitored to attain the desired properties
the limestone used complies with certain quality, content, and in the hardened material. Fresh cement paste rheology is regarded
particle size and fineness requirements [8,9]. These characteristics to be closely associated with the development of mortar and concrete
allow the better packing of particles obtaining more compact and microstructure [26]. Given the direct impact of the presence of
hence less porous structures [10]. mineral additions on such microstructure, and consequently
The use of fly ash in cement compositions is very common behaviour rheology, an understanding of blended cement behaviour
because it raises product strength and reduces alkali silica reaction- in this regard is imperative. As a rule, the incorporation of mineral
induced expansion [11–14]. Fly ash fineness determines the additions raises yield stress and to a lesser extent, plastic viscosity,
especially in silica fume-blended cements [27–29].
The fresh characteristics of concrete made with Portland ce-
ment and rheology can be modified and controlled with super-
⇑ Corresponding author. Tel.: +34 913020440x206; fax: +34 913020700. plasticizers. Achieving the steepest decline in the water–cement
E-mail address: (F. Puertas).

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
Author's personal copy

O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309 301

ratio, greatest workability and decreasing the viscosity and the Table 1
yield stress is, however, contingent upon the compatibility Chemicals composition (wt%), Blaine and BET specific surface of Portland cements.
L.O.I.: loss on ignition. I.R.: insoluble residue. SS BET. Specific Surface Area determined
between the admixture chosen and the cement used. The presence by B.E.T.
of mineral additions such as limestone, fly ash and silica fume may
affect the interaction between the superplasticizer and the cement. wt% CEM I 52.5R CEM II/B-L CEM II/B-V CEM II/A-D

The performance of additions could be also influenced by admix- L.O.I. 2.35 14.83 3.58 2.56
tures. Conventional superplasticizers such as naphthalene (PNS)-, SiO2 20.51 13.45 28.44 32.58
Al2O3 5.37 3.53 14.11 4.25
melamine (PMS)- or lignosulfonate (LS)-based disperse the Fe2O3 2.10 1.54 2.93 1.76
particles due to a electrosteric mechanism while the PCE form a MnO 0.02 0.01 0.02 0.02
steric obstacle to any direct inter-particle contact [35–41]. The MgO 3.86 2.79 2.93 3.13
compatibility of mineral additons with, and affinity for, a given CaO 57.05 57.49 40.79 48.47
Na2O 0.64 0.56 0.55 0.64
admixture must therefore be determined [3,30–34].
K2O 1.44 1.02 1.44 1.23
The studies of limestone-blended cements and the superplasti- TiO2 0.16 0.12 0.45 0.14
cizers have focused primarily on PCE admixtures [42–45]. P2O5 0.13 0.12 0.45 0.12
Magarotto et al. [43] and Banfill [44] determined the importance SO3 6.37 6.16 4.33 5.09
of the molecular weight and the structure of the PCE on the I.R. 0.26 1.48 2.00 2.33
CaOfree 1.27 0.10 0.93 0.96
rheological behaviour and the water-reduction of the limestone-
Blaine (m2/kg) 501.7 524 472.1 559.0
blended cement. Mikanovic and Jolicoeur [46] studied the relation- SS BET (m2/g) 1.22 2.11 1.53 2.98
ship between particle–superplasticizer interactions, rheology and
paste stability on the one hand, and blending, sedimentation and
consolidation on the other. Their findings showed that the mecha-
Table 2
nisms involved in superplasticizer action on limestone and cement Chemicals composition (wt%), Blaine and BET specific surface of mineral additions.
type I varied depending on whether the admixture was PCE or PNS.
wt% L CV HS
They also observed that while the dispersion effect of the two was
similar in water–limestone pastes, the presence of Ca(OH)2 L.O.I. 43.56 6.76 3.96
SiO2 0.34 46.32 94.29
improved the effectiveness of PCE.
Al2O3 0.04 31.01 0.24
The use of fly ash has been reported to improve cement rheol- Fe2O3 0.11 4.50 0.11
ogy and lower the dose of superplasticizer needed (PCE and PMS) MnO 0.01 0.05 0.02
to obtain the desired reological properties [45,47]. Due to the MgO 0.93 1.29 0.24
nearly spherical shape and the size of fly ash particles, these ce- CaO 54.56 4.90 0.46
Na2O 0.36 0.34 0.11
ment pastes demand a lower concentration of admixture than
K2O – 1.34 0.37
limestone blended cements [45]. TiO2 0.01 1.53 –
Silica fume blended cements behave differently from other P2O5 0.08 0.98 0.05
blended cements due to the mineral surface [48] of this addition. SO3 – 0.98 0.15
Sireact. 0 36.4 92.7
The use of silica fume reduces paste workability. The improvement
SS BET (m2/g) 4.38 2.70 20.29
in reological properties depends on the type of superplasticizer. Dv (lm) 1 1.7 14.7 0.31
The PMS dose needed by silica fume-containing cements is similar 2 14.7 36.2 10.48
to the dose required by cement type I to induce the same reological
behaviour [47–49]. The molecular architectures of the PCE are crit-
ical to the dispersion of the silica fume particles [48].
Table 3
Despite the importance of admixture compatibility, the interac- Physical and chemical characteristics of superplasticizers.
tions between blended cements and superplasticizers have been
Admixture LS PNS PMS PCE
scantly studied, and the interactions between limestone, fly ash
and silica fume particles and superplasticizers have barely been re- Solid content (%) 40.1 39.6 41.9 40.9
Mw (Da) 39,230 136,995 78,828 59,596
searched at all.
Mn 16,915 25,695 7315 35,923
The present study aimed to establish the effect of mineral addi- Viscosity (mPa.s) 24.28 51.11 31.50 118.20
tions on fresh type II Portland cement pastes containing limestone %C 37.08 43.78 18.65 51.67
(CEM II/B-L), fly ash (CEM II/B-V) and silica fume (CEM II/A-D), and %S 5.91 9.13 10.65 0.30
to determine the compatibility between these additions and four %H 4.89 4.53 3.98 8.14
%N 1.46 0.80 22.17 0.17
superplasticizers: naphthalene (PNS), melamine (PMS), lignosul- Na (ppm) 41,840 31,400 55,280 2820
phonate (LS) and polycarboxylate (PCE) polymers. The control used K (ppm) 390 340 0.2 10
throughout was non-blended CEM I 52.5-R. pH 8 8.5 8 4.5

2. Materials and methods

2.2. Methods
2.1. Materials
2.2.1. Adsorption isotherms
The characteristics of the commercial CEM I 52.5 R Portland cement used are Admixture adsorption isotherms were determined for cements and mineral
given in Table 1. The three type II cements studied were prepared in the laboratory: additions. The suspensions were prepared as described by Perche [50]. The admix-
CEM II/B-L, containing (30 wt%) limestone, CEM II/B-V, containing (30 wt%) fly ash tures were dissolved in water and 40 g of each solution were mixed with 20 g of
and CEM II/A-D, containing (10 wt%) silica fume. The cement constituents were cement. The dosage ranged from 0 to 40 mg of polymer/g of cement for LS, PNS
blended in a turbula mixer for 2 h. The chemical composition and specific surface and PMS and from 0 to 2.5 mg of polymer/g of cement for the PCE. The suspen-
of these blended-cements are given in Table 1. Table 2 lists the chemical composi- sions were stirred magnetically for 30 min at 25 °C and subsequently centrifuged
tion and B.E.T. specific surface of the mineral additions. for 3 min to separate the supernatant from the solid. The amount of admixture
Four commercial superplasticizers were used: a lignosulphonate (LS) derivative, present in the supernatant was determined with a SHIMADZU TOC-VCSH/CSN to-
a naphthalene-based compound (PNS), a melamine-based material (PMS) and a tal organic carbon analyzer. The amount of admixture consumed was obtained as
polycarboxylate superplasticizing admixture (PCE). The physical–chemical charac- the difference between the amount initially added and the quantity remaining in
teristics of the admixtures are given in Table 3. the supernatant.
Author's personal copy

302 O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309

120 IIs. The zeta potential values were determined using the Smoluchowski approxima-
tion. Thirty grams of each cement and addition were suspended in 160 g of water.
The suspensions were stirred magnetically for 15 min, subsequently dispersed with
100 a sonicator for 5 min and finally mixed by magnetic stirring for a further 5 min. They
were then placed in a measuring cell to determine the zeta potential after this 25-
min contact between material and water. The superplasticizers were added to the
Shear Stress (Pa)

80 cement suspensions with an automatic titrator at a rate of 0–20 mg polymer/g bin-

der and to the addition suspensions at 0–40 mg polymer/g addition. The zeta poten-
tial values were corrected for the presence of the background solution.
2.2.3. Rheological behaviour
Yield stress was evaluated for the cement pastes with a Haake Rheowin Pro RV1
rotational viscometer fitted with a roughened cylindrical rotor. The inner cylinder
has of 37 and 55 mm of diameter and height, respectively, and the outer cylinder
20 has 44 and 70 mm of diameter and height, respectively. In both cylinders the height
of the grooves is 1 mm. The cement pastes were prepared by mixing 100 g of ce-
ment and 45 ml of water for 3 min in a mechanical blade stirrer. Superplasticizer
0 doses ranging from 0 to 8 mg of polymer per g of cement were added with the
0 20 40 60 80 100 mixing water.
The cement paste rheology test consisted of exposure to pre-shear stress at
Shear Rate (s ) 100 s1 for 1 min, after which the rotor speed was reduced to 0 s1, subsequently
ramped up to 100 s1 in 12 min and then lowered to 0 s1, likewise in 12 min.
Fig. 1. Hysteresis cycles of the binder paste. The cement cycles obtained in the rheological tests exhibited virtually no hystere-
sis, an indication that the binder pastes were perfectly floc-free (Fig. 1). All the rheo-
logical curves followed the same pattern, with the decline in shear rate conforming
to the Bingham equation (Eq. (1)), in which the y-intercept defines the yield stress
mg polymer consumed/g cement

1.00 and the slope of the fitted line, paste plastic viscosity

s ¼ s0 þ g ðcÞ ð1Þ

(a) 0.50 PCE

20 Table 4
Experimental admixture adsorption on cements in the adsorption plateau values.
mg polymer consumed/g cement

Cement Admixtures Experimental adsorption
16 0.00 LS
0 1 2 3 4 5
mg polymer/g cement mg polymer/g binder mg polymer/m2 binder
CEM I LS 14.28 11.70
12 PMS PNS 10.07 8.25
PMS 11.53 9.45
10 PNS PC 0.62 0.50
8 CEMII/BL LS 18.27 8.66
PNS 9.56 4.53
PMS 16.35 7.75
4 PC 0.65 0.31
CEMII/BV LS 10.86 7.10
PNS 11.08 7.24
0 PMS 17.84 11.66
0 10 20 30 40 50 PC 0.73 0.48

mg polymer/g cement CEMII/AD LS 22.10 7.42

PNS 13.63 4.57
PMS 20.01 6.71
(b) 5
CEM I 52.5 R PC 0.85 0.29

Zeta Potential (mV)

0 5 10 15 20 25 30 70 CEM I 52,5R
mg polymer/g cement
-5 LS
PNS 50
Yield Stress (Pa)

-15 20

Fig. 2. Superplasticizer adsorption isotherms (a) and variation in the zeta potential
(b) on CEM I (a). PNS
2.2.2. Effect of superplasticizers on the zeta potential of cements and additions
0 2 4 6 8 10
The effect of different dosages of four superplasticizers (LS, PNS, PMS and PCE) mg polymer/g cement
on the zeta potential of the suspensions of non-blended and blended cements as
well as on suspensions additions was obtained on a Colloidal Dynamics Acoustosizer Fig. 3. Variation in yield stress in cement pastes CEM I.
Author's personal copy

O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309 303

mg polymer consumed/g binder


(a) 0.75
20 PCE
mg polymer consumed/g binder

18 0.25

16 PCE

Zeta Potential (mV)

0 1 2 3 4 5 PMS 0
mg polymer/g binder
14 LS 0 5 10 15 20 25 30
12 mg polymer/g binder
10 PNS -5
0 -15
0 10 20 30 40 50
mg polymer/g binder

(b) 20 (d) Limestone

mg polymer consumed/g addition

mg polymer consumed/g cement

18 1.00 10
16 0.75 Zeta Potential (mV)
mg polymer/g addition
14 0
0 10 20 30 40 PCE 50
LS -5
0 1 2 3 4 5
10 mg polymer/g cement -10
8 -15
6 PMS -20
4 -25 PMS

2 -30
0 -35
0 10 20 30 40 50
mg polymer/g addition

Fig. 4. Superplasticizer adsorption isotherms and variation in the zeta potential on CEMII/B-L (a and c) and Limestone (b and d).

3. Results and discussion ies [3,30,37,51] determine that the PCE-based admixture exhibited
a substantially larger radius of gyration therefore larger thickness
3.1. Control cement (CEM I 52.5 R) of the polymer layer than the other polymers, required signifi-
cantly smaller dosages to reach the adsorption. The adsorption pla-
Cement paste behaviour is conditioned by the physical–chemi- teau value of PCE corresponds to 0.62 mg of admixture/g of
cal interaction between the particles and the medium. The inclu- cement.
sion of superplasticizers, which are adsorbed onto the particles, Information on both superplasticizer adsorption and the varia-
prevents flocculation and water entrapment in the flocs, reducing tion in the charge on particle surfaces were obtained by studying
the amount of water needed to obtain similar fluidity in the pastes. the variation in the zeta potential of the particles with the amount
However admixture effectiveness depends on its compatibility and
affinity with the cement and mineral addition particles. This calls Table 5
for a detailed study of the surface interaction between these poly- Experimental admixture adsorption on additions in adsorption plateau values.
mers and the various types of particles comprising the cements. Addition Admixtures Experimental adsorption
The behaviour of the non-blended cement in the presence of the
mg polymer/ mg polymer/
four superplasticizers is shown in the adsorption isotherms in
g addition m2 addition
Fig. 2a. Admixture adsorption on the cement exhibited exponential
Lime LS 14.77 3.37
behaviour, linear at low concentrations of the admixture, followed
PNS – –
by a plateau adsorption period in which the inclusion of larger PMS 6.06 1.38
amounts of superplasticizer led to no further consumption by the PC 0.91 0.21
cement. The traditional admixtures had high adsorption values Fly ash LS 13.44 4.98
for CEM I: 14.28, 11.53 and 10.07 mg of admixture/g of cement PNS – –
for LS, PMS and PNS, respectively (see Table 4). These values are PMS 10.67 3.95
in keeping with the molecular weight of the admixtures (see Ta- PC 0.30 0.11

ble 3), for adsorption was most intense in the superplasticizer with Silica fume LS 33.21 1.64
the lowest molecular weight (LS), and least intense in the polymer PNS – –
PMS 19.95 0.98
with the highest (PNS). Molecular weight may be also related to PC 0.78 0.04
the radious of gyration of the polymer in the solution. Many stud-
Author's personal copy

304 O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309

Table 6
Theoretical and normalize adsorption plateau data.

Cement Admixtures Theoretical adsorption Normalize adsorption

mg polymer/g binder mg polymer/m binder mg polymer/g binder mg polymer/m2 binder
CEMII/BL LS 14.43 9.20 1.27 0.94
PNS – – – –
PMS 9.89 7.03 1.65 1.10
PC 0.70 0.41 0.92 0.75
CEMII/BV LS 14.03 9.68 0.77 0.73
PNS – – – –
PMS 11.72 7.80 1.58 1.50
PC 0.52 0.38 1.40 1.25
CEMII/AD LS 16.17 10.69 1.37 0.69
PNS – – – –
PMS 12.37 8.60 1.62 0.78
PC 0.64 0.45 1.34 0.64

of admixture included (Fig. 2b). The slightly positive (+0.5 mV) zeta physical–chemical properties of the cement and its behaviour in
potential of the cement particles favoured adsorption of the anion terms of superplasticizer adsorption according to the values shown
group-bearing superplasticizers. The main effect of traditional in Table 4 on admixture adsorption in the plateau region (in mg of
superplasticizers LS, PMS and PNS induce the dispersion of cement admixture/g of cement). Since these adsorption plateau values
particles by an electrosteric mechanism favouring mutual repul- were closely related to the physical characteristics of the cement,
sion and hindering flocculation [3,31]. Taking into account of the however, adsorption was also found per m2 of particle. This param-
zeta potential values observed in Fig. 2b, the PMS have the highest eter defines the amount of admixture adsorbed per particle area,
electrical contribution followed by PNS and, lastly, LS. This pro- i.e., the affinity of the polymer molecules for the particle studied,
gression concurred of concentration of percentage of sulphur in irrespective of the specific surface of the material (see results in
the admixture (see Table 3) that corresponds to the number of Table 5).
R  SO 3 groups originally present in the sulphonates and the poly-
mer molecules directly involved in adsorption. The PCE-based
admixtures, in turn, induced a steric obstacle to any direct inter-
(a) 70
particle contact [3,51]. The zeta potential values of close to 0 for
the PCE admixture showed that the electrostatic contribution to
the cement particle stability will be negligible. 50
Yield Stress (Pa)

The preparation of cement pastes, mortars and concretes calls

for high concentrations of solids. Therefore, in addition to the 40
interactions between the particles and admixtures studied, others, LS
such as particle–particle or particle–admixture–particle interac- 30
tions, take place in these products. In this connection, Fig. 3 shows
the results of the yield stress of the cement pastes with and with- 20
out superplasticizers to determine their rheological behaviour. In
the non-blended CEM I cement pastes, the inclusion of superplas- 10 PC
ticizers reduced yield stress, i.e., inter-floc interaction, since these
polymers stabilized the particles, hindering their agglomeration. 0
Here also, the behaviour of the PCE-based admixture deviated sig- 0 2 4 6 8 10
nificantly from the pattern recorded for the traditional polymers, mg polymer/g cement
due to the differences in their stabilization mechanisms. Similar
observations have been reported by other authors, who found that
at least twice as much PNS as PCE was needed to achieve similar (b)
Yield stress CEMI/Yield stress CEM II/BL

yield stress values [33,7,51]. Yield stress was observed to decline 4.0 LS
rapidly by as much as 90% at low dosages of PCE (lower that PNS
2 mg PCE/g cement), whereas much larger amounts of the other 3.5
admixtures were required to obtain a significant difference in these 3.0
values. Behaviour was observed to be linear in the presence of tra-
ditional admixtures up to values of 6 mg of admixture/g of cement, 2.5
while concentrations of 8 mg of admixture/g of cement were
needed to lower the yield stress. The steepest such reduction 2.0
was obtained with PNS (82%), followed by PMS (44%) and finally
LS (22%).
3.2. Effect of limestone (CEM II/B-L)
The inclusion of 30% limestone in cement pastes (Fig. 4a) raised
LS and PMS adsorption, indicating that these admixtures had a 0 2 4 6 8
greater affinity for CEM II/BL than for CEM I, while PNS and PCE be- mg polymer/g binder
haved in much the same way in the limestone blend as in the non-
blended cement. The presence of mineral additions changes the Fig. 5. Variation in shear stress in cement pastes CEM II/B-L (a) and Limestone (b).
Author's personal copy

O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309 305


(c) 5
mg polymer consumed/g binder

16 PCE

Zeta Potential (mV)

14 0 5 10 15 20 25 30
12 PNS mg polymer/g binder

10 mg polymer consumed/g binder LS -5 LS

1.00 PNS
0.50 PCE -10

2 0.00
0 1 2 3 4 5

PCE mg polymer/g binder
0 10 20 30 40 50
mg polymer/g binder

(b) 20 Fly Ash

(d) 15
mg polymer consumed/g cement

Fly Ash
mg polymer consumed/g addition

18 0.75
16 0.50
5 mg polymer/g addition
Zeta Potential (mV)

14 PCE
0 1 2 3 4 5 0 10 20 30 40 LS 50
mg polymer/g cement PMS -5
10 -10 PMS
8 -15
6 -20
4 -25
2 -30
0 -35
0 10 20 30 40 50
mg polymer/g addition

Fig. 6. Superplasticizer adsorption isotherms and variation in the zeta potential on CEMII/B-V (a and c) and Fly ash (b and d).

Adsorption analysis of the mineral additions revealed, firstly, stone to the cement led to greater consumption of traditional
that due to the dispersion of the addition particles in the presence admixture, in particular PMS, per gram of cement, whereas the
of PNS, the solid particles could not be separated from the superna- amount of PCE consumed was similar with and absence of the
tant. Consequently, the amount of free admixture present in the li- addition. The surface adsorption rate (mg of admixture/m2 of ce-
quid, i.e., not adsorbed onto the addition particles, could not be ment) was lower in LS and PCE after the inclusion of limestone,
quantified for this admixture. Superplasticizer adsorption on lime- however, while for PMS this parameter was essentially unaffected
stone was similar to its adsorption on limestone-blended cement: by the addition.
the highest consumption was observed for LS, followed by PMS and The variations in the zeta potential of cement CEM II/B-L
PCE, in that order (Fig. 4b). The amount of admixture consumed by (blended with limestone) in the presence of superplasticizers are
the limestone particles was similar to the value observed for CEM shown in Fig. 4c and d. Initially, the cement containing limestone
II/B-L cement, although the amount of polymer adsorbed onto the behaved very much like the unadditioned material, with PMS
surface was substantially lower: i.e., affinity was much smaller for affording the greatest electrostatic contribution to stabilization.
limestone than for cement particles. Nonetheless, the presence of Limestone particle behaviour was the outcome of a high zeta po-
the filler led to a higher total consumption of admixture as a result tential (+13 mV), which favoured the adsorption of the negatively
of its greater specific surface (See Table 5). Mikanovic and Jolicoeur charged (with values of over 20 mV) traditional admixtures. Of
[46] observed that twice as much PNS as PCE was adsorbed and these, PMS had the highest negative charge. The adsorption of
that adsorption of both polymers was twice as high on cement as the PCE superplasticizer leads to zeta potential values close to 0
on limestone particles. (2 mV), in the limestone particles.
Table 6 gives the theoretical adsorption data in both mg of The cement pastes containing limestone exhibited only slightly
admixture/g of cement and mg of admixture/m2 of cement, given higher yield stress than the control (Fig. 5a). Their rheological
the proportions of each component in the blend and their adsorp- behaviour in the presence of the superplasticizers was more dis-
tion plateau, assuming that the cement and addition particles did tinctly differentiated, however. The yield stress values for the lime-
not interact. The normalize adsorption has been calculated by the stone cement were observed to decline with even very small
comparation of the theoretical and experimental data, in order to quantities of superplasticizer. PNS proved to be the most effective
determine the existence or otherwise of synergies between the traditional admixture, for at concentrations of 6 mg/g of cement,
addition and the cement. This exercise showed that adding lime- the value dipped to 17.5 Pa, whereas 8 mg of PMS/g of cement
Author's personal copy

306 O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309

were required to attain similar levels. Admixture LS lowered yield

(a) 70 CEM II/BV
stress least effectively. To 1.6 mg of PCE/g of cement the yield
stress has reduced in a 78% for limestone cement whereas a 53% 60
for non blended cement. PCE was more effective in CEM II/B-L than
in the control cement, with a steeper decline in yield stress ob- 50

Yield Stress (Pa)

served at low dosages of the admixture lower that 2 mg PCE/g
cement. 40
Fig. 5b shows the normalized yield stress values for the lime-
stone-blended cement with respect to the control non-blended ce- 30
ment. Values of less than 1 denote a decline in yield stress, i.e., an
improvement in paste rheological properties. The addition of lime- 20
stone induced higher yield stress values when the admixture dos-
age was low or non-existent. At values of 8 mg of admixture/g of 10 PMS
cement (the optimal concentration found for CEM I), however, PCE
the limestone cement pastes exhibited better rheological proper- 0
ties than the non-blended material. 0 2 4 6 8 10
mg polymer/g binder
3.3. Effect of fly ash (CEM II/B-V)
(b) LS
The adsorption isotherms for superplasticizers on cement con- CEMII/BV

Yield stress CEMI/Yield stress CEM II/BV

4.0 PNS
taining fly ash (Fig. 6a) showed that LS consumption was much
lower than in the control cement, while a striking rise was ob-
3.5 PCE
served in PMS consumption (11.66 mg of PMS/g of cement). Actual
adsorption was greater than forecast in the theoretical calculations 3.0
(Table 6), an indication of the existence of interaction between the
cement and fly ash particles. The difference between the calculated 2.5
and experimental data was higher here than for limestone-blended
cement, with normalized adsorption values of nearly 1.6 and 1.4
for PMS and PCE, respectively. Moreover, the addition of fly ash 1.5
to the cement led to the adsorption of more admixture molecules
per unit of area. In other words, PMS and PCE consumption was 1.0
greater both overall (per gram of cement) as on the surface (per
m2 of particle surface area). The adsorption of superplasticizers
on fly ash particles confirmed that PMS was consumed more inten- 0.0
sely than on limestone particles, with values close to the findings 0 2 4 6 8
recorded for LS (Fig. 6b). The amount of admixture consumed per mg polymer/g binder
unit area was generally higher than in the limestone cement, in
particular for PMS, for which the values doubled the limestone Fig. 7. Variation in shear stress in cement pastes CEM II/B-V (a) and Fly ash (b).

blend figures. In contrast, PCE was adsorbed less profusely on fly

ash than on limestone.
The variations in the zeta potential with the amount of admix- affinity of the fly ash for the traditional admixtures observed on
ture are shown in Fig. 6c. CEM II/B-V behaved in much the same the adsorption isotherms translated into better rheological behav-
way as the cements discussed above, except that slightly higher iour in these pastes than in the limestone cement material. This
zeta potential values were recorded in the presence of PMS, which same effect was reported by Artel and Garcia [45]. Moreover, sig-
is consistent with the more intense consumption of this admixture nificant improvement was observed in the cement paste in the
observed on the adsorption isotherms. Fly ash particles, with a zeta presence of PMS.
potential of +2.5 mV (see Fig. 6d), favour polymer adsorption. The
zeta potential was lower than found for the limestone particles, 3.4. Effect of silica fume (CEM II/A-D)
however, which translated into lower zeta potential values in the
presence of the admixtures. Despite this lower surface charge Fig. 8a and b shows the admixture adsorption isotherms on sil-
and the smaller specific surface in fly ash particles, adsorption ica fume-containing cement and silica fume particles. No adsorp-
per unit of area was higher than in limestone (see Table 5). Surface tion plateau was attained in any of the cases studied, due to the
morphology and the adsorption sites available for the polymer may large specific surface of silica fume and the concomitantly high de-
have changed the polymer adsorption configurations and thus mand for admixture. CEMII/A-D consumed more LS and PMS than
effective surface coverage [44]. PNS and, as in the preceding cements, the behaviour exhibited by
The rheological behaviour of cement CEM II/B-V (Fig. 7a) con- the PCE polymer was completely different. A comparison of the
firmed the greater affinity of fly ash particles for PMS observed ear- theoretical and experimental adsorption data revealed the exis-
lier. The cement pastes with PMS and PNS exhibited a 75 % decline tence of interaction between the cement and silica fume particles,
in the yield stress at dosages of 6 mg of admixture/g of cement. As reflected in the higher consumption of admixture. The silica fume
in the preceding cases, LS accounted for the smallest decline in particle adsorption isotherms were practically linear (see Fig. 8b).
yield stress in cement pastes. Fig. 7b shows that at low doses of Silica fume was, moreover, the addition that consumed the great-
admixture or in its absence, the fly ash-blended cement behaved est quantity of admixture (see Table 5 and Fig. 8b) due to its high
like the CEM I cement under the same conditions, rheologically specific surface (Table 2), five- and ten-fold larger than in fly ash
speaking. At concentrations of over 1 (for PCE) or over 5 (for the and limestone, respectively. That notwithstanding, in terms of
traditional polymers) mg of admixture/g of cement, however, con- admixture consumed per unit of area, silica fume was observed
siderable improvement was observed in yield stress. The greater to have lower affinity for the admixture molecules although the
Author's personal copy

O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309 307


(c) 5
mg polymer consumed/g binder

18 LS
16 PCE

Zeta Potential (mV)

14 0 5 10 15 20 25 30
12 mg polymer/g binder
10 -5
mg polymer consumed/g binder
1.00 LS
8 0.75
6 0.50
2 0 1 2 3 4 5 PMS
mg polymer/g binder
0 -15
0 10 20 30 40 50
mg polymer/g binder

(b) 20 Silica Fume

(d) 15 Silica Fume
mg polymer consumed/g addition

mg polymer consumed/g addition

18 0.75 10
16 0.50
Zeta Potential (mV)

14 0
0.00 0 10 20 30 40 50
12 0 1 2 3
mg polymer/g addition
4 5
mg polymer/g addition
10 -10
8 -15 PCE
6 -20
4 -25
2 PCE -30 LS
0 -35 PMS
0 10 20 30 40 50
mg polymer/g addition

Fig. 8. Superplasticizer adsorption isotherms and variation in the zeta potential on CEMII/A-D (a and c) and Silica fume (b and d).

affinity could be higher due to the adsorption plateau has not been stress declined by only 60%. Steeper declines in the yield stress
reached. In other words, the considerable rise in consumption per (79%) were observed with the PCE admixture, but concentrations
gram of cement observed with respect to the control was due to of over 1 mg of PCE/g of cement were required to attain a rapid
the large surface area available, which offset the effect of the lower drop in shear stress. While cement containing silica fume had yield
affinity of this addition for the admixtures. stress values similar to the control (see Fig. 9b), larger amounts of
The variations in the zeta potential for the silica fume-blended admixture were needed to bring it down. The rheological behav-
cement (see Fig. 8c) reached higher absolute values (greater elec- iour of cement pastes containing silica fume was impacted by
trostatic stability) than in the other cements. As with CEM II/B-V, the physical and chemical properties of silica fume particles, which
a greater difference in potential values was obtained for PMS than raised superplasticizer and water consumption during cement
in CEM I and CEM II/B-L, a result that is consistent with the adsorp- paste preparation, similar results were determined by Plank el al.
tion findings. These higher values were due to the high negative [48], Vikan and Justnes [49] and Zhang and Han [52].
surface charge (16 mV) obtained by silica fume particles, which As a rule, the inclusion of mineral additions, especially lime-
was responsible for lower zeta potential on the blended than the stone and fly ash, was observed to improve the rheological behav-
non-blended cements. Nonetheless, this surface charge does not iour of cements with superplasticizers. Such superplasticizers
favour any adsorption affinity for likewise negatively charged stabilized the cement, limestone and fly ash particles electrosteri-
superplasticizers therefore the adsorption observed could be due cally in the case of the traditional admixtures and sterically in the
to a entropic factor [53] and to their higher specific surface area case of PCE, hindering flocculation. CEM II/A-D behaviour was
(see Fig. 8d). This finding corroborated the adsorption isotherm characterized by the high specific surface of silica fume and con-
data: the amount of admixture consumed by the silica fume was comitant high admixture consumption, which nonetheless failed
due to its high specific surface, not to its affinity for the to translate into cement particle stabilization. Higher admixture
admixtures. dosages were therefore required on silica fume-cements to attain
Fig. 9a shows the variation in yield stress for silica fume-con- values similar to the results observed for the other blended ce-
taining cement. This cement required higher doses of admixture ments. While smaller amounts were adsorbed and it was the only
than the preceding materials to achieve a decline in yield stress. superplasticizer studied here that did not stabilize the particles
As in the other three cases, the most effective traditional admixture electrosterically, PCE exhibited the best results at low doses, con-
was PNS, although at doses of 8 mg of admixture/g of cement, yield firming the enormous effectiveness of the steric stabilization
Author's personal copy

308 O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309

70 4. In the case of silica fume, with a high negative zeta potential

(16 mV) the physical characteristics are dominant. Due to
60 the high specific surface of silica fume, CEM II/A-D pastes
demand high doses of superplasticizers to improve the rheolog-
50 ical behaviour of fresh pastes.
Yield Stress (Pa)

5. While the effects generated by the traditional admixtures are

LS probably the results of an electrosteric mechanism, the PCE-
based superplasticizer stabilizes cement and addition particles
30 by a sterical mechanism. In order to know accurately the stabil-
PMS ization mechanism a nanoscale studies are necessary. PCE is
20 adsorbed less intensely and lowers yield stress more effectively
at lower dosages than the other admixtures. PNS is the most
10 PCE effective traditional admixture in all the type II cements stud-
ied, and LS the polymer that delivers the poorest rheological
0 results.
0 2 4 6 8 10
mg polymer/g binder

Yield stress CEMI/Yield stress CEM II/BV

4.0 CEM II/AD This research was funded by the Spanish Ministry of Science
PMS and Innovation under projects BIA2007-61380 and BIA2010-
PCE 15516. Olga Burgos-Montes participated in the project under a
3.0 Spanish National Research Council post-doctoral grant. The
authors wish to thank Marie-Gabrielle Patry for her collaboration.

2.0 References

1.5 [1] Roskovic R, Bjegovic D. Role of mineral additions in reducing CO2 emission.
Cem Concr Res 2005;35:974–8.
1.0 [2] Taylor HFW. Cement chemistry. London: Academic Press; 1990.
[3] Leás. In: Hewkett Peter C, editor. Chemistry of cement and concrete. 4th ed.
0.5 New York; 1998.
[4] Livesey P. Strength characteristics of Portland-limestone cements. Constr Build
0.0 Mater 1991;3(5):147–50.
0 2 4 6 8 [5] Tsivilis S, Chaniotakis E, Badogiannis E, Pahoulas G, Ilias A. A study on the
parameters affecting the properties of Portland limestone cements. Cem Concr
mg polymer/ g binder Compos 1999;21:107–16.
[6] Tsivilis S, Chaniotakis E, Kakali G, Batis G. An analysis of the properties of
Fig. 9. Variation in shear stress in cement pastes CEM II/A-D (a) and Silica fume (b). Portland limestone cements and concrete. Cem Concr Comp 2002;24:371–8.
[7] Ramezanianpour AA, Ghiasvand E, Nickseresht I, MAhdikhani M, Moodi F.
mechanism. Of the admixtures that induced stabilization electro- Influence of various amounts of limestone powder on performance of Portland
limestone cement concretes. Cem Concr Compos 2009;31:715–20.
statically, LS conveyed the lowest zeta potential to the particles [8] Vuk T, Tinta V, Gabrovsek R, Kaucic V. The effect of limestone addition, clinker
and was also the least effective, even though more of this than type and fineness on properties of Portland cement. Cem Concr Res
the other polymers was consumed in the process. As a rule, the 2001;31:135–9.
[9] Voglis N, KAkali G, Chaniotakis E, Tsivilis S. Portland-limestone cements. Their
PNS-based admixture lowered the yield shear stress most effec- properties and hydration compared to those of other composite cements. Cem
tively, generating more fluid cements. This finding confirmed, from Concr Compos 2005;27:191–6.
the adsorption trial results, the most effective admixture is PNS. [10] Pipilikaki P, Beazi-Katsioti M. The assessment of porosity and pore size
distribution of limestone Portland cement pastes. Constr Build Mater
According to the adsorption isotherms, cements containing fly 2009;23:1966–70.
ash showed considerable affinity for PMS, which resulted in rheo- [11] Feldman RF, Cerette GG, Malhotra. Studies on mechanism of development of
logical behaviour similar to the behaviour found for these cements physical and mechanical properties of high-volumne fly ash-cement pastes.
Cem Concr Compos 1990;12:245–51.
in the presence of PNS.
[12] Papadakis VG. Effect of fly ash on Portland cement systems. Part I. Low-
calcium fly ash. Cem Concr Res 1999; 29: 1727–1736.
4. Conclusions [13] Li G, Wu X. Influence of fly ash and its mean particle size on certain
engineering properties of cement composite mortars. Cem Concr Res
The main conclusions to be drawn from the present study are: [14] Yilmaz B, Olgun A. Studies on cement and mortar containing low-calcium flu
ash, limestone, and dolomitic limestone. Cem Concr Compos
1. The rheological behaviour of blended cements depends on the [15] Fu X, Wang Z, Tao W, Yang C, Hou Dong Y, Wu X. Studies on belnded cement
physical (specific surface) and chemical (surface charge and with a large amount of fly ash. Cem Concr Res 2002;32:1153–9.
reactivity) characteristics of the mineral addition used. These [16] Lee SH, Kim HJ, Sakai E, Daimon M. Effect of particle size distribution of fly ash-
cement on the fluidity of cement pastes. Cem Concr Res 2003;33:763–8.
characteristics are affected by the presence of superplasticizers. [17] Chindaprasirt P, Jaturapitakkul C, Sinsiri T. Effect of fly ahs fineness on
2. Limestone has considerable affinity for the polymer molecules, compressive strength and pore size of blended cement paste. Cem Concr
which adsorb onto the surface of its particles. It tends to adsorb Compos 2005;27:425–8.
[18] Chindaprasirt P, Jaturapitakkul C, Sinsiri T. Effect of fly ahs fineness on
PCE admixtures more intensely than the other additions stud-
microstructure of blended cement paste. Constr Build Mater
ied. The steepest decline in yield stress is obtained in the pres- 2007;21:1534–41.
ence of PCE and PNS. [19] Toutanji HA, El-Korchi T. The influence of silica fume on the compressive
3. Fly ash exhibits greater affinity for PMS than the other cements strength of cement paste and mortar. Cem Concr Res 1995;25(7):1591–601.
[20] Rao GA. Role of water-binder ratio on the strength development in mortars
and mineral additions studied in this work, yielding similar incorporated with silica fume. Cem Concr Res 2001;31:443–7.
results for PMS and PNS. [21] Jkflgjsdgjdfdhdfl.
Author's personal copy

O. Burgos-Montes et al. / Construction and Building Materials 31 (2012) 300–309 309

[22] Mazloom M, Ramezanianpour AA, Brooks JJ. Effect of silica fume on [38] Palacios M, Puertas F, Bowen P, Houst YF. Effect of PCs superplasticizers on the
mechanical properties of high-strength concrete. Cem Concr Compos rheological properties and hydration process of slag-blended cement pastes. J
2004;26:347–57. Mater Sci 2009;44:2714–23.
[23] Al-Amoudi OSB, Abiola TO, Maslehuddin M. Effect of superplasticizer on [39] Palacios M, Houst YF, Bowen P, Puertas F. Adsorption of superplasticizer
plastic shrinkage of plain and silica fume cement concretes. Constr Build Mater admixtures on alkali-activated slag pastes. Cem Concr Res 2009;39:670–7.
2006;20:642–7. [40] Banfill PFG, Daimon M, Roy DM. A discussion of the papers rheological
[24] Jo B, Kim C, Tae G, Park J. Characteristics of cement mortar with nano-SiO2 properties of cement mixes. Cem Concr Res 1979;9:795–6.
particles. Constr Build Mater 2007;21:1351–5. [41] Ferrari L, Kaufmann J, Winnefeld F, Plank J. Interaction of cement model
[25] Kadri E, Aggoun S, Schutter G. Interaction between C3A, silica fume and systems with superplasticizers investigated by atomic force microscopy, zeta
naphthalene sulphonate superplasticiser in high performance concrete. Constr potential, and adsorption measurements. J Colloid Interface Sci 2010;347:15.
Build Mater 2009;23:3124–8. [42] Sahmaran M, Christianti HA, Yaman IO. The effect of Chemicals admixtures
[26] Chindaprasirt P, Hatanaka S, Chareerat T, Mishima N, Yuasa Y. Cement paste and mineral additives on the properties of self-compacting mortars. Cem
characteristics and porous concrete properties. Constr Build Mater 2008;22: Concr Compos 2006;28:432–40.
894–901. [43] Magarotto R, Torresan I, Zeminian N. Influence of the molecular weight of
[27] Cry M, Legrand C, Mouret M. Study of the shear thickening effect of policarboxylate superplasticizers on the rheological properties of fresh cement
superplasticizers on the rheological behaviour of cement pastes containing pastes, mortar and concrete. In: Proceeding 11th ICCC. 2003; 2: 514.
or not mineral additives. Cem Concr Res 2000;30:1477–83. [44] Banfill PFG. Additivity effects in the rheology of fresh concrete containing
[28] Ferraris CF, Obla KH, Hill R. The influence of mineral admixtures on the water-reducing admixtures. Constr Build Mater 2011;25:642.
rheology of cement paste and concrete. Cem Concr Res 2001;31:245–55. [45] Artelt C, Garcia E. Impact of superplasticizer concentration and of ultra-fine
[29] Park CK, Noh MH, Park TH. Rheological properties of cementious materials particles on the rheological behaviour of dense mortar suspensions. Cem Concr
containing mineral admixtures. Cem Concr Res 2005;33:842–9. Res 2008;38:633–42.
[30] Ramachandran S, Malhotra VM, Jolicoeur C, Spiratos N. Superplasticizers: [46] Mikanovic N, Jolicoeur C. Influence of superplasticizers on the rheology and
properties and applications in concrete. Canadá: Ottawa; 1998. stability of limestone and cement pastes. Cem Concr Res 2008;38:907.
[31] Spiratos N, Pagé M, Mailvaganam NP, Malhotra VM, Jolicieur C. [47] Erdogdu S, Arslantürk C, Kurbetci S. Influence of fly ash and silica fume on the
Superplasticizers for concrete. Fundamentals, technology and practice. 2nd consistency retention and compressive strength of concrete subjected to
ed. Canadá: Quebec; 2006. prolonged agitating. Constr Build Mater 2011;25:1277–81.
[32] Vikan H, Justenes H, Winnefeld F, Figi R. Correlating cement chracteristics with [48] Plank J, Schroefl C, Gruber M, Lesti M, Sieber R. Effectiveness of
rheology of paste. Cem Concr Res 2007;37:1502–11. polycarboxylate superplasticizers in ultra-high strength concrete: the
[33] Zhang T, Shang S, Yin F, Aishaha A, Salmaiah A, Ooi TL. Adsorptive importance of PCE compatibility with silica fume. J Adv Concr Technol
behaviour of surfactants on surface of Portland cement. Cem Concr Res 2009;7:5–12.
2001;31:1009–15. [49] Vikan H, Justnes H. Rheology of cementitious paste with silica fume or
[34] Stranel O, Sebok T. Relationships between the properties of ligninsulphonates limestone. Cem Concr Res 2007;37:1512–7.
and parameters of modified samples with cement binders. Part I. [50] Perche F. Adsorption de polycarboxylates et de lignosulfonates sur poudre
Characterizing ligninsulphonates and studying their sorption properties. modèle et ciments. PhD; École Polytechnique Fédérale de Lausanne; 2004.
Cem Concr Res 1997;27:185–91. [51] Alonso MM, Palacios M, Puertas F, de la Torre AG, Aranda MAG. Influencia de la
[35] Golaszewski J, Szwabowski J. Influence of superplasticizers on rheological estructura de aditivos basados en policarboxilato sobre el comportamenito
behaviour of fresh cement mortars. Cem Concr Res 2004;34:235–48. reológico de pastas de cemento. Mater Constr 2007;57(286):65–81.
[36] Puertas F, Santos H, Palacios M, Martinez-Ramirez S. Polycarboxylate [52] Zhang X, Han J. The effect of ultra-fine admixture on the rheological property
superplasticiser admixtures: Effect on hydratation. Microstructure and of cement paste. Cem Concr Res 2000;30:827–30.
rheological behaviour in cement pastes. Adv Cem. Res. 2005;17:77. [53] Aschauer U, Spagnoli D, Bowen P, Parker SC. Growth modification of seeded
[37] Mollah MYA, Adams WJ, Schennach R, Cocke DL. A review of cement– calcite using carboxylic acids: atomistic simulations. J Colloid Interface Sci
superplasticizer and their model. Adv Cem Res 2000;12(4):153–61. 2010;346:226–31.

View publication stats