Available online at www.sciencedirect.

com

Cement & Concrete Composites 30 (2008) 334–346

www.elsevier.com/locate/cemconcomp

Effect of DMA–MMA diblock copolymer on the properties

of Portland and composite cement
a,*
A. Olgun , N. Atar a, V. Bütün b, Y. Erdogan a

a
Department of Chemistry, Dumlupinar University, Kütahya, Turkey
b
Department of Chemistry, Eskisßehir Osmangazi University, Eskisßehir 26040, Turkey

Received 17 March 2006; received in revised form 29 March 2007; accepted 4 May 2007
Available online 18 May 2007

Abstract

In this paper, the effect of DMA–MMA diblock copolymer on the properties of ordinary Portland cement and cement containing
boron has been studied. Variation in setting time, compressive strength and volume expansion has been determined. Cement hydration
was monitored by X-ray diffraction (XRD) and by Fourier transforms infrared spectroscopy, in combination with the thermoanalytical
methods (TG/DTA). The microstructural observation of the hydrated cement paste was performed by scanning electron microscopy
(SEM). The result showed that DMA–MMA prolongs the early setting time of the Portland cement and composite cement, and had
no noticeable effect on the final setting time of the both cement types. DMA–MMA diblock copolymer reduced the water-to-cement
ratios (w/c) from 0.5 to 0.42, which improved the compressive strength of the mortars at all curing ages. The experimental results also
indicated that DMA–MMA did not only change the rate of cement paste hydration, but also the microstructure of calcium–silicate–
hydrate (C–S–H).
 2007 Elsevier Ltd. All rights reserved.

Keywords: Retardation; Microstructure; Strength; Chemical admixture; Cement paste; Diblock copolymer

1. Introduction mortar and retards the cement setting. The setting time
and early strength must be regulated by suitable activator
In recent years, investigation and production of com- to satisfy the demand of building engineering.
posite cements based on the partial replacement of The accelerating water-reducing admixtures give greater
Portland cement by waste materials have made consider- strengths during the earlier hydration period and faster set-
able progress [1–4]. The replacement materials participate ting times, which allow the finishing operation to be carried
in the hydraulic reactions, contributing significantly to out in a timely manner [5]. This product type is formulated
the composition and microstructure of the hydrated prod- using water-reducer ingredients (such as lignosulfonates
uct. Among these materials, are industrial by-products, and hydroxycarboxylic acid), and a high proportion of
such as borogypsum from boric acid production plants, accelerators (calcium chloride or calcium formate). How-
colemanite ore waste (CW), and tincal ore waste (TW) ever, one of the limitations of the use of water-reducing
from enrichment process in boron plants. These materials admixtures containing calcium chloride in reinforced con-
contain significant amount of boron impurity, which has crete is that when present in larger amounts, they promote
a positive contribution to the long-term strength develop- corrosion of the reinforcement. Hence, there is a contin-
ment of the mortar [3]. However, the presence of boron uing attempt at finding an alternative additive to calcium
in cement leads to a reduction in the early strength of the chloride, one equally effective but without these limitations.
Our earlier studies showed that increasing replacement
*
Corresponding author. Tel.: +90 274 2652051; fax: +90 274 2652056. of Portland cement by TW resulted in a strength reduction
E-mail address: aolgun@dumlupinar.edu.tr (A. Olgun). of concrete and retarded setting time of the cement [3].

0958-9465/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cemconcomp.2007.05.002
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 335

Therefore, there is a need to find out a chemical, which will Table 1

increase early strength of the mortar and not retard setting Chemical characteristics of used material
time of the cement. The aim of the present study was to Chemical analysis (wt%)
determine the effect of the addition of DMA–MMA Clinker Tincal waste Gypsum
diblock copolymer on the setting characteristics of the SiO2 21.22 10.95 0.94
cement paste and strength development of the mortar. Al2O3 4.28 0.91 0.54
After analysis of the morphology of hydrated cement Fe2O3 5.05 1.47 0.37
phases, the content of calcium hydroxide was determined. CaO 66.04 14.90 32.46
MgO 1.47 7.57 1.71
The infrared spectra of DMA–MMA and cement paste SO3 1.20 0.43 42.56
were compared. Techniques for microstructural character- Na2O 0.22 12.00 0.47
ization, such as infrared spectroscopy, differential thermal K2O 0.63 1.31 0.25
analysis, thermogravimetric analysis and scanning electron B2O3 – 9.70 –
microscopy, were also employed. Loss on ignition 0.16 39.06 20.78

2. Experimental procedure
2.2. Cement mixtures
2.1. Raw materials
Two series of mixtures noted as P(96%PC + 4%G) and
Raw materials were Portland cement clinker (PC), gyp- T(93.3%PC + 3.7%G + 3%TW) were prepared according
sum (G) from Denizli Cement Plant (Denizli, Turkey), and to TS-EN 197-1 [8]. The raw materials mixed in the
tincal ore waste from Eti-Boron Plant (Eskisßehir, Kırka, required proportion were ground using a laboratory type
Turkey). The chemical compositions of main raw materials grinding mill to a fineness of 27% mass residue on 32-
are listed in Table 1. The hydrophilic–hydrophobic DMA– lm-size mesh. Particle size analysis was carried out accord-
MMA [2-(dimethylamino) ethyl methacrylate]-[methyl ing to TS EN 196-6 [8]. The fineness of the cement was
methacrylate] diblock copolymer used in this study was determined by Blaine specific surface area equipment.
synthesized by group transfer polymerization. The number
of average molecular wight (Mn) and molecular weight 2.3. Preparation of cement paste
distribution (PD) determined by gel-permeation chromato-
graph was 40,000 and 1.10, respectively. The DMA content The required water of standard consistency, setting time,
of the DMA–MMA diblock copolymer was 92 mol% as and volume expansion were examined according to TS-EN
determined by 1H NMR spectroscopy. The details of the 196-3 [8]. The mixing of cement pastes was carried out with
synthesis are reported elsewhere [6,7]. the standard water of consistency as given in Table 2. Two
reference pastes were prepared from the cements designated
CH3 CH 3 as P and T with deionized water only. DMA–MMA was
CH2 C CH 2 C added to the mixing water before mixing operation. The
C=O 0.92 C=O 0.08
dosages of DMA–MMA were 0.1 and 0.5 mass% of the
O O cement. The cement pastes were obtained using a mixer
CH2 CH3 for 1 min at low speed (60 rpm) and 4 min at high speed
CH2
N
(120 rpm). The paste was then poured into moulds creating
H 3C CH3 100 · 10 · 10 mm prisms. The samples were cured at
20 ± 2 C and 90 ± 2% relative humidity. The samples were

Table 2
Water percent, setting time, and physical characteristics of cementitious mixes
Symbol Cement mixes Water (%) Setting time (min) Fineness (wt%) Volume expansion (mm) Density (kg/m3) Specific surface
(m2/kg)
Initial Final +32 l 90 l Cold Hold Total
P 25 190 240 27 1 1 1 2 3.15 352.7
P1 P + 0.1%DMA–MMA 25.5 200 250 0 0 0
P2 P + 0.2%DMA–MMA 26.0 210 260 0 0 0
P3 P + 0.3%DMA–MMA 26.5 220 270 0 0 0
P4 P + 0.4%DMA–MMA 27.0 230 270 0 0 0
P5 P + 0.5%DMA–MMA 28.0 240 270 0 0 0
T 26 300 380 27.2 1.0 1 1 2 3.13 352.0
T1 T + 0.1%DMA–MMA 26.5 310 390 0 0 0
T2 T + 0.2%DMA–MMA 27.0 320 400 0 0 0
T3 T + 0.3%DMA–MMA 27.5 330 400 0 0 0
T4 T + 0.4%DMA–MMA 28.0 340 400 0 0 0
T5 T + 0.5%DMA–MMA 28.2 350 400 0 0 0
336 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

then demolded and placed in deionized water. The cement A cement prism was cut into cubes of approximately
paste specimens were cured for 2, 7, and 28 days, and then 10 mm square, one side of which was ground flat. The
they were taken out of the water. The hydration process was hydrated samples were soaked in acetone to stop hydration
stopped by grinding the hydrated samples with acetone and reactions. After drying the sample of specimens were
by washing the residue several times with more acetone. The coated with gold, and then the SEM images of samples
samples were dried at 65 C. were obtained using a JEOL JXA 840A scanning electron
microscope.
2.4. Compressive strength test
3.2. X-ray diffraction
The specimen preparation for strength tests was per-
formed at room temperature. Mortar specimens were A piece of hydrated cement paste prism was taken and
produced by mixing one part of the cement with three parts was ground to a fine powder of <63 lm. A Rikagu Miniflex
of sand, using two different water-to-cement ratios (w/c) of X-ray diffractometer using mono-chromatic CuKa radia-
0.50 and 0.42. DMA–MMA diblock copolymer was added tion operating at a voltage of 30 kV and current of
to mixing water during the mixing operation. The dosages 15 mA was used. A scanning speed of 22h/min and a step
of DMA–MMA diblock copolymer were 0.1, 0.2, 0.3, 0.4, size of 0.02 were used to examine the samples in the range
and 0.5 mass% of cement. The cement–water mixtures were of 5–752h.
stirred at low speed for 30 s, and then, with the addition of
sand, the mixtures were stirred for 4 min. The mortars pre- 3.3. Infrared spectroscopic test
pared were cast into 40 · 40 · 160 mm moulds for strength
tests. After 24 h of curing at 20 C, the samples were The FT-IR spectra of both the 28-day hydrated neat
demolded, and then immediately immersed in a water- cement and DMA–MMA diblock copolymer modified
curing tank. The temperature of the water was maintained cement pastes were recorded on a Bruker Vertex 70 FT-
at 20 ± 1 C during the curing period. The compressive IR spectrometer equipped with the harric MVP2-unit in
strength test was carried out at the ages of 2, 7, and 28 days the range of 4000–400 cm1 region.
according to TS EN 196-1 [8]. The strength value was the
average of three specimens. 3.4. Differential thermal analysis

3. Characterization DTA and thermo gravimetric analysis (TG/DTA) of the

selected 28-day hydrated sample were obtained using a
3.1. Scanning electron microscopy Shimadzu, DTG-60H Simultaneous DTA-TGA apparatus.
The samples were heated from 35 to 1100 C at a constant
For scanning electron microscopy (SEM) studies, rate of 10 C/min in nitrogen gas dynamic atmosphere
selected cement paste samples cured for 28 days were used. (100 cm3/min). All the weight loss data are expressed as a

50
P
45
P1
40 P2
P3
Compressive strength, MPa

35
P4
30 P5

25

20

15

10

0
2 Day 7 Day 28 Day
Age, Days

Fig. 1. Effects of DMA–MMA diblock copolymer on the compressive strength of the PC mortar (water to cement ratio, 0.50).
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 337

function of the ignited weight of the sample as suggested by are the molecular weights of portlandite and water,
Taylor [9]. The amount of portlandite CH (%) was respectively.
obtained directly from TG curves by Eq. (1):
4. Results and discussion
MWCH
CHð%Þ ¼ WLCHð%Þ ð1Þ
MWH
4.1. Setting characteristics and volume expansion

where WLCH(%) corresponds to the weight loss, in percent, The initial and final setting time of the Portland cement
occurring during CH hydration, and MWCH and MWH and TW cement are given in Table 2. The initial and final

50
T
45 T1
T2
40
T3
T4
Compressive strength, MPa

35
T5
30

25

20

15

10

0
2 Day 7 Day 28 Day
Age, Days

Fig. 2. Effects of DMA–MMA diblock copolymer on the compressive strength of the TW mortar (water to cement ratio, 0.50).

60
P
P1
P2
50
P3
P4
Compressive strengths, MPa

P5
40

30

20

10

0
2 Day 7 Day 28 Day
Age, Days

Fig. 3. Effects of DMA–MMA diblock copolymer on the compressive strength of the PC mortar (water to cement ratio, 0.42).
338 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

setting times of the neat cement and cement containing MMA diblock copolymer, the initial setting times of the
boron were found to be 190 and 240 min, and 300 and both cement and cement containing boron are gradually
380 min, respectively. Many workers have reported that prolonged. Similarly, the inclusion of DMA–MMA
the presence of boron in the cement retards the cement set- diblock copolymer into the cement up to 0.3% (weight of
ting [1,3]. J. Bensted et al. observed this retardation and cement) extends the final setting time of the both cements
attributed it to a coating effect and the inhibition of cal- in the same extends. It is interesting that DMA–MMA
cium hydroxide nucleation [10]. Apagyi and Csetenyi also diblock copolymer slightly increases the final setting time
noted that the boron in the cement increased portlandite of the PC and TW cement, and the retarding ability dimin-
solubility [11]. ish with the increase of the polymer content. The present
Table 2 shows that the initial setting times of cements study is not concerned to reveal the nature of the mecha-
are prolonged by the addition of DMA–MMA diblock nism of retardation of DMA–MMA diblock copolymer.
copolymer. With the increase of the amount of DMA– However, many investigators have concluded that the

60
T
T1
50 T2
T3
Compressive strength, MPa

T4
T5
40

30

20

10

0
2 Day 7 Day 28 Day
Age, Days

Fig. 4. Effects of DMA–MMA diblock copolymer on the compressive strength of the TW mortar (water to cement ratio, 0.42).

DTA
uV

20.0
P
10.0

0.0
P1
-10.0

-20.0
P5
-30.0

-40.0

-50.0

-60.0
0.0 200.0 400.0 600.0 800.0 1000.0
Temp [C]

Fig. 5. DTA/TG of the hydrated Portland cement pastes (with and without DMA–MMA diblock copolymer).
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 339

retarding action of the admixture is, in most cases, due to The effects of diblock copolymer on the volume
its effect on C3S hydration and adsorption of the admixture expansion of cement paste are given in Table 2. The
on the surface of the cement particles [12,13]. It seems that results indicate that DMA–MMA diblock copolymer
the retardation process in the presence of DMA–MMA slightly decreases expansion of cement paste compared
diblock copolymer is due to its tendency to decrease the to control cement paste containing no polymer. In
surfaces areas available for the hydration of C3S, and polymer-modified systems, the possible factor behind
C3A particles. the reduction in volume may be attributed to the

DTA
uV
30.0

20.0 T

10.0

0.0 T1

-10.0

-20.0 T5

-30.0

-40.0

-50.0

-60.0
0.0 200.0 400.0 600.0 800.0 1000.0
Temp [C]

Fig. 6. DTA/TG of the hydrated TW cement pastes (with and without DMA–MMA diblock copolymer).

13.0

P
T
12.5

12.0
CH content (%)

11.5

11.0

10.5
0.0 0.1 0.5
Polymer content (%)

Fig. 7. Effect of DMA–MMA diblock copolymer content on the calcium hydroxide content (CH) of cement pastes.
340 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

attractive forces produced between polymer and hydrated the polymer beyond 0.1%, while the strength of the PC
cementitious particles. mortar slightly increased (Figs. 1 and 2). As curing time
expanded to 7 days, the compressive strength of the
4.2. Effect of DMA–MMA diblock copolymer on the DMA–MMA diblock copolymer activated TW mortars
strength development of mortar was higher than that of TW mortar containing no polymer,
and the strength increased gradually with the increase of
The effects of DMA–MMA diblock copolymer on the polymer content. The effect of polymer on strength of PC
strength development of PC and TW mortars (w/c = 0.5) mortar showed similar trend as 2 days curing age. At 28
are shown in Figs. 1 and 2. At 2 days of curing age, the days, the compressive strength of the polymer activated
compressive strength of the PC mortar and TW mortar TW mortar was lower than that of TW mortar without
without DMA–MMA diblock copolymer were observed DMA–MMA diblock copolymer. This observation shows
as 19.0 and 15.2 MPa, respectively. The strength of the that incorporation of DMA–MMA diblock copolymer
TW mortar considerably increased with the addition of in mortar clearly influences the hydration process. The

442
P
T
440

438
THC ( C)
O

436

434

432

430
0.0 0.1 0.5
Polymer content (%)

Fig. 8. Effect of DMA–MMA diblock copolymer content on dehydration temperature (TCH) of cement pastes.

Fig. 9. XRD patterns of PC cement cured for 2 days (with and without DMA–MMA).
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 341

possible mechanism, which may lead to decreased strength increased with the increasing DMA–MMA diblock copoly-
in DMA–MMA diblock copolymer activated TW speci- mer content. This observation shows that compressive
mens, may be related to the fact that the presences of boron strength of the mortars can be improved at all curing ages
in cement affects the solubility of portlandite that alters the by the addition of DMA–MMA diblock copolymer with a
degree of the attractive forces produced between polymers reduction water–cement ratio.
and hydrated cementitious particles.
The effect of DMA–MMA diblock copolymer on the 4.3. DTA analysis
compressive strength of PC and TW mortars (w/c = 0.42)
cured up to 28 days is given in Figs. 3 and 4. Apparently, Differential thermal analysis and TGA were used to
the trend in Fig. 3 is similar to that in Fig. 4. The compres- study the effects of DMA–MMA diblock copolymer on
sive strength of the DMA–MMA diblock copolymer mod- the hydration of different cement phases. Three major
ified mortars was higher than that of the mortars without endothermic reactions were occurred during the heating
DMA–MMA diblock copolymer. The strength values of the cement pastes hydrated for 28 days. As Bensted

Fig. 10. XRD patterns of PC cement cured for 7 days (with and without DMA–MMA).

Fig. 11. XRD patterns of PC cement cured for 28 days (with and without DMA–MMA).
342 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

and Varma [14], and Odler and Abdul-Maula [15] sug- show a slight decrease in the dehydration temperature of
gested, the decomposition peaks of CSH (at 120 C) and CH. When more DMA–MMA diblock copolymer is pres-
ettringite (at 145 C) overlap beyond 4 h of hydration. ent (0.5%), dehydration temperature of CH increases for
Therefore, it is possible to assign the first peaks to CSH both cement pastes (Fig. 8). Fig. 7 shows that DMA–
hydration and ettringite decomposition (Figs. 5 and 6). MMA diblock copolymer (0.5%) considerably increases
An endothermic peak at about 410–460 C corresponds the quantity of CH phases in the pastes. It is not that the
to the decomposition of calcium hydroxide to calcium amount of CH in PC cement paste without DMA–MMA
oxide and water. An endothermic peak at about 650 C is diblock copolymer is lower than that of TW cement paste.
related to the decomposition of carbonate phases. This result indicates that the presence of boron impurity in
As shown in Figs. 5 and 6, DMA–MMA diblock the cement paste has remarkable effects on the development
copolymer influences the DTA curves of PC and TW of CH in the cement. The addition of a higher level of
cement pastes. Cement pastes with 0.1% DMA–MMA DMA–MMA diblock copolymer limited the development

Fig. 12. XRD patterns of TW cement cured for 2 days (with and without DMA–MMA).

Fig. 13. XRD patterns of TW cement cured for 7 days (with and without DMA–MMA).
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 343

of CH as demonstrated by mix P5, where the extent of CH copolymer content. Calcium carbonate was formed after
formation was significantly lower than for other mixtures. 7 days of curing age (Fig. 10). The relative amount of cal-
cium carbonate in these XRD patterns can be attributed to
4.4. X-ray diffraction the carbonation of Ca(OH)2. In the case of C3S phases, the
peaks of maximum intensity overlap with those of C2S,
The development in time of the two cement systems which makes its evolution difficult to determine. Therefore,
(with and without polymer), as indicated by XRD, is their trend can be interpreted jointly. The amount of these
shown in Figs. 9–14. They show the X-ray patterns phases decreased with increasing time and DMA–MMA
obtained after 2, 7, and 28 days of curing. The XRD diblock copolymer content. There is a marked difference
patterns indicating the hydration of the neat cement and in the XRD patterns of the cement cured for 7 days. Cal-
polymer activated cement are shown in Fig. 9. The cium hydroxide was formed in greater amounts in control
expected crystalline hydration products are evident. Cal- system, and the amount decreased with increasing DMA–
cium hydroxide was formed with appreciable amounts in MMA diblock copolymer content.
two days, and remained a good crystalline reaction product The X-ray traces for the TW cement paste are shown in
throughout the period of investigation. The amount of Figs. 12–14. The phases identified were: calcium hydroxide,
Ca(OH)2 increased with increasing DMA–MMA diblock C3S, C2S, and calcium carbonate. The amount of the

Fig. 14. XRD patterns of TW cement cured for 28 days (with and without DMA–MMA).

Fig. 15. Fourier-transform infrared spectrum of Portland cement pastes (with and without DMA–MMA diblock copolymer).
344 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

Fig. 16. Fourier-transform infrared spectrum of TW cement pastes (with and without DMA–MMA diblock copolymer).

calcium hydroxide present in cement increase slightly dur-

ing the curing ages. The peak of calcium carbonate
appeared at 7 days of hydrated cement (Fig. 13) and the
intensity of it increased with increasing curing time
(Fig. 14).
The results show some qualitative differences in the
hydration rate of the two different cement systems due to
the incorporation of DMA–MMA diblock copolymer.
With regards to the formation of calcium carbonate, it
can be observed that the incorporation of DMA–MMA
diblock copolymer leads to a fast increase in calcium car-
bonate content in PC cement. The rate of the formation
of calcium hydroxide in PC cement is also higher between
2 and 7 days, indicating a faster growth of the C–S–H
phase.

Fig. 17. SEM micrographs of fracture surface of cement paste after 28

4.5. IR spectroscopy study days of hydration (without DMA–MMA diblock copolymer).

The IR spectra of 28-day hydrated cements (with and

without admixture) are shown in Figs. 15 and 16. The
spectrum of Portland cement without DMA–MMA
diblock copolymer shows band at 3639, 1419, 1121, 966,
868 661, 511, and 451 cm1. The band observed at about
3639 cm1 is due to the OH band from Ca(OH)2. The
bands observed at 1419 cm1 show the presence of calcium
carbonate in the hydrated cement past, and the peak inten-
sity decreases with increasing DMA–MMA diblock
copolymer content for both cement pastes. The band
observed at 868 cm1 can be attributed to the presence of
anhydrated dicalcium silicate (C2S) in the pastes. Regard-
less of cement type, the intensity of the band gradually
decreases with increasing DMA–MMA diblock copolymer
content. This result indicates that the presence of DMA–
MMA in the cement paste has positive contribution to
the hydration rate of C2S.
Previous studies have shown that the strongest Si–O
stretching (v3) bands appeared at 925 cm1, Si–O bending Fig. 18. SEM micrographs of fracture surface of cement paster after 28
(v4) vibration at 524 cm1, and Si–O bending (v5) vibra- days of hydration (with 0.1% DMA–MMA diblock copolymer).
 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346 345

tions at 460 cm1 in dry cement [16]. In our study, the shift- hydrated for 28 days are presented in Figs. 17–22. The
ing of Si–O asymmetric stretching (v3) vibration to higher microstructure of the PC cement paste without polymer
wave numbers (961–967 cm1) indicates the polymerization shown in Fig. 17 indicates a comparatively dense micro-
of silicate units SiO4
4 with the formation of the CSH phase structure, typical of a sound Portland cement paste. A very
[17]. Relative intensity of the Si–O stretching (v3), Si–O large Ca(OH)2(CH) crystal and a porous composite mass
bending (v4) vibration at 524 cm1, and Si–O bending (v5) of calcium silicate hydrate (CSH) are also observed.
vibration bands also undergoes significant changes as a Fig. 18 shows that the morphology of crystallite size of
result of polymerization of the SiO4
4 in PC and TW cement CH phase is significantly affected by polymer. The crystal-
paste. The intensity of Si–O stretching (v3) vibration bands lites are smaller and thinner with the presence of polymer
for PC cement paste gradually decreased with the increas- than with the PC paste containing no admixture. In the
ing DMA–MMA diblock copolymer content (Fig. 15). PC cement containing 0.5% DMA–MMA diblock copoly-
mer, the CH has been significantly converted the hydration
4.6. SEM observation of microstructure products, and a very dense structure of CSH has been
formed (Fig. 19). A polymeric film partially covering the
Scanning electron micrographs of selected cement paste hydrated phases was also observed.
(with and without DMA–MMA diblock copolymer)

Fig. 21. SEM micrographs of fracture surface of cement paste containing

Fig. 19. SEM micrographs of fracture surface of cement paste after 28 tincal ore waste after 28 days of hydration (with 0.1% DMA–MMA
days of hydration (with 0.5% DMA–MMA diblock copolymer). diblock copolymer).

Fig. 20. SEM micrographs of fracture surface of cement paste containing Fig. 22. SEM micrographs of fracture surface of cement paste containing
tincal ore waste after 28 days of hydration (without DMA–MMA diblock tincal ore waste after 28 days of hydration (with 0.5% DMA–MMA
copolymer). diblock copolymer).
346 A. Olgun et al. / Cement & Concrete Composites 30 (2008) 334–346

The micrograph of TW cement paste is shown in Fig. 20 Acknowledgements

at the same magnification with Fig. 17 to allow direct com-
parison. The microstructure is notably different from that The authors thank Denizli Cement Plant (Denizli, Tur-
developed in the PC cement paste, where very dense struc- key) for providing facilities of mechanical testing of mate-
tures of CSH were observed. The morphology of the TW rials. V. Bütün thanks Turkish Academy of Science for an
cement paste with polymer (0.1%) was similar to that of encouragement award for science scholarship. We also
the PC cement containing same amount of polymer, but wish to thank Mrs. Birgül Musß for many helpful
the cement hydration products seemed to be localized in discussions.
TW cement paste (Fig. 21). Compared to PC cement paste
with 0.5% polymer (Fig. 22), TW cement paste has a denser References
deposit of CH encapsulating CSH dense structure. This
may be attributed to the effect of boron impurity on the [1] Targan S ß , Olgun A, Erdoğan Y, Sevinç V. Effects of supplementary
cementing materials on the properties of cement and concrete.
hydration rate of cement.
Cement Concrete Res 2002;32:1551–8.
[2] Yan P, Yang W. The cementitious binder derived with flurogyp-
5. Conclusions sum and low quality of fly ash. Cement Concrete Res 2000;30:
275–80.
In this study, the effect of DMA–MMA diblock copoly- [3] Kula I, Olgun A, Sevinc V, Erdogan Y. An Investigation on the use
mer on the properties of ordinary Portland cement and of tincal ore waste, fly ash and coal bottom ash as Portland cement
replacement materials. Cement Concrete Res 2002;32:227–32.
cement containing boron has been investigated. It is worth [4] Hill J, Sharp JH. The mineralogy and microstructure of three
mentioning that this is a preliminary study on the use of composite cement with high replacement levels. Cement Concrete
DMA–MMA diblock copolymer as a chemical admixture. Comp 2002;24:191–9.
Moreover, based on the experimental result, the following [5] Rixom MR. Chemical admixtures for concrete. London: Spon;
conclusions are confirmed: 1999.
[6] Baines FL, Billingham NC, Armes SP. Synthesis and solution
properties of water-soluble hydrophilic–hydrophobic block copoly-
1. The replacement of Portland cement clinker by tincal mers. Macromolecules 1996;29(10):3416–20.
ore waste decreases the compressive strength of the [7] Bütün V, Armes SP, Billingham NC. Synthesis and aqueous solution
mortar at early ages, but the strengths of the mortar properties of near-monodisperse tertiary amine methacrylate homo-
are comparable to that of the PC mortar at 28 days of polymers and diblock copolymers. Polymer 2001;42:5993–6008.
[8] TS EN-196-1, TS EN-196-3, TS EN-196-6, TS-EN 197-1, Turkish
curing age. National Standards. Ankara, Turkey: TSE Turkish Standard
2. The addition of DMA–MMA diblock copolymer up to Institute.
0.3% (weight of cement) to the mortar containing PC [9] Taylor HFW. Studies on the chemistry and microstructure of cement
cement slightly decreases the compressive strength of pastes. Pro Br Ceram Soc 1984;35:65–82.
the samples at all curing ages. However, the use of [10] Bensted JB, Callaghan IC, Lepre A. Comparative study the efficiency
of various borate compounds as set-retarders of class G oilwell
DMA–MMA diblock copolymer in PC cement signifi- cement. Cement Concrete Res 1991;21:663–8.
cantly increases 7-day strength of the mortars, and [11] Apagyi Z, Csetenyi LJ. Phase equilibrium study in the CaO–K2O–
decreases 28-days strength. B2O3–H2O system at 25 C. Cement Concrete Res 2001;31:1087–91.
3. DMA–MMA diblock copolymer allows a significant [12] Yilmaz VT, Glasser FP. Influence of sulphonated melamine formal-
dehyde superplasticizer in cement. Adv Cement Res 1989;2:111–9.
reduction in the water to cement ratio, which accelerates
[13] Barnes P. Structure and performance of cement. USA: Routledge;
the strength development of PC and TW mortars at all 2001.
curing ages. [14] Bensted J, Varma SP. Some applications of infrared and Raman
4. The addition of up to 0.3% (weight of cement) DMA– spectroscopy in cement chemistry: Part 3. Hydration of Portland
MMA diblock copolymer to the PC and TW cement cement and its constituents. Chem Technol 1974(September/
pastes retard the initial setting time and have no notice- October):440–50.
[15] Odler I, Abdul-Maula S. Possibilities of quantitative determination of
able effect on the final setting time of the cement. Aft-(ettringite) and Afm (monodulphate) phases in hydrated cement
5. The hydration products formed in the two systems stud- pastes. Cement Concrete Res 1984;14:133–41.
ied were corresponds to the effected products, although [16] Perraki Th, Kakali G, Kontoleon F. The effect of natural zeolites on
they differed in magnitude. Calcium hydroxide is formed the early hydration of portland cement. Micropor Mesopor Meter
initially in all systems, remaining as a good crystalline 2003;61:205–12.
[17] Mollah MYA, Yu Wenhong, Schennach Robert, Cocke David L. A
reaction product throughout the period of investigation. fourier transform infrared spectroscopic investigation of the early
6. The DMA–MMA diblock copolymer modified cement hydration of Portland cement and the influence of sodium lignosul-
pastes shows a dense microstructure compared to that fonate. Cement Concrete Res 2000;30:267–73.
of cement pastes without activator.