Cement & Concrete Composites 21 (1999) 431±437

In¯uence of superplasticizer and curing on porosity and pore

structure of cement paste q
J.M. Khatib *, P.S. Mangat
School of Environment and Development, Sheeld Hallam University, Howard Street, Sheeld S1 1WB, UK

Abstract
Most concrete produced today contains admixtures. Superplasticizers (SP) are used for the purpose of improving workability and
reducing the water to cement ratio; therefore producing more durable concrete. SP cause better dispersion even at high water to
cement ratio. Although SP improves the dispersion of particles, it is not quite clear how the addition of SP a€ect the porosity and
pore size distribution of cement paste. The purpose of this study was to examine the in¯uence of one type of SP on porosity and pore
size distribution under di€erent curing regimes. Paste specimens with and without SP were prepared at constant water to cement
ratio of 0.45. Specimens were cured for 28 days and some for six months. Specimens were exposed to high temperature (45°C) and
normal temperature curing (20°C) and also subjected to di€erent relative humidities (100%, 55% and 25%). Curing at high
temperature was carried out to simulate temperature in hot climates. Tests on porosity and pore size distribution were conducted
using mercury intrusion porosimetry. The results show that the inclusion of SP decreases the total intruded pore volume of paste.
The dominant pore diameter, however, does not seem to be a€ected and the percentage of pores smaller than 100 nm increases in the
presence of SP. Ó 1999 Published by Elsevier Science Ltd. All rights reserved.

Keywords: Curing; Pore size distribution; Pore volume; Superplasticizer

1. Introduction presence of SP [5,7,12]. The rate of hydration was re-

ported to increase when SP is included [11], while others
The presence of superplasticizers (SP) in a concrete reported a retardation in cement hydration [9,10]. Table
mixture is quite advantageous, in that they assist in the 1 presents the data on porosity, pore structure and hy-
e€ective dispersion of cement particles and hence im- dration which were obtained by di€erent workers
proving the workability of concrete [1]. The water to [5,6,8,10±12]. Apparently it is not clear on how the ad-
cement ratio is reduced when SP is added to cement dition of SP a€ects the porosity, pore structure or hy-
paste, which leads to reduced permeability, increased dration of cement paste. The strength of concrete was
strength and producing durable concrete [2]. SP are reported to increase in the presence of SP [12±15]. An
broadly classi®ed into four groups, sulphonated mela- optimum dosage level of about 2.0±2.5% of SP was
mine-formaldeyde condensate (SMF), sulphonated found to give a maximum value of strength, beyond
naphthalene-formaldehyde condensate (SNF), modi®ed which level the strength begins to decline [13,14], which
lignosulphonates (MLS) and others including sulphonic- indicates that the rate of hydration increases when SP is
acid esters and acrylic esters [3,4]. present in the system (paste, mortar or concrete) up to
At a constant water to cement ratio, the addition of an optimum dosage of between 2% and 2.5%. However,
SP is variously reported to increase, to decrease or to others [16] reported that an increase in strength was
have no in¯uence on the porosity of cement paste, obtained up to an SP dosage level of about 1% (by mass
mortar or concrete [5±12], but many have agreed that a of cement). Beyond this level (i.e. 1%) a decline in
re®nement of the pore structure is obtained in the compressive strength was observed.
The aim of the present work is to further examine the
q
in¯uence of SP addition on porosity and pore structure
Paper presented at the Sheeld Infrastructure Conf., 1999, and
reviewed according to Journal procedures.
of cement paste at a constant water to cement ratio. The
*
Corresponding author. cement pastes were subjected to di€erent curing tem-
E-mail address: [email protected] (J.M. Khatib) peratures (i.e. high temperature and normal tempera-

0958-9465/99/$ - see front matter Ó 1999 Published by Elsevier Science Ltd. All rights reserved.
PII: S 0 9 5 8 - 9 4 6 5 ( 9 9 ) 0 0 0 3 1 - 1
432 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437

Table 1
In¯uence of SP addition on porosity and pore structure for data obtained by di€erent workers

Year Ref No. SP type SP dosagea W/C Age Pore Pore Hydration
(%) volume structure

1984 [5] SMF or 1 0.28 28 d Increased Re®ned Accelerated

SNF
1986 [7] SNF 0.35 365 d Decreased Re®ned
1990 [6] 0.5 112 d No change
1992 [10] SMF 0.5 0.5 90 Decreased Re®ned Retarded
1992 [10] SNF 0.5 0.5 90 Increased Coarsened Retarded
1994 [9] SMF or 1.5 0.3 1±28 d Increased Coarsened Retarded
SNF
1995 [11] SMF 4 0.28 1±3 yr Decreased Accelerated
a
Percent by mass of cement.

ture) and di€erent relative humidities (high and low). Table 3

Only one type and one dosage of SP were used in this Properties of superplasticizer
investigation. Speci®c gravity 1.08 at 20°C
Viscosity Approx. 8cP
Air entrainment Nil
Chloride content Nil
2. Experimental programme Freezing point ÿ17°C
Ash content 9%
2.1. Materials Appearence Dark brown liquid

Ordinary Portland cement was used, the composition

2.2. Curing
of which is given in Table 2. The water used was dis-
tilled. The water to cement ratio (W/C) was kept con-
Cube specimens …100  100  100 mm3 † were cast in
stant at 0.45. Pastes with and without SP were prepared.
steel moulds. Five di€erent curing regimes were adopt-
The dosage of SP was 1.2% by mass of cement. The SP
ed. Specimens were subjected to high temperature (to
was the sodium salt of a sulphonated napthalene-form-
simulate temperature in hot climates) and normal tem-
aldehyde condensate. The SP conformed to Types F and
perature. Details of the curing methods are described
G of ASTM C 494. The main features of the SP used are
below and given in Table 4.
listed in Table 3.
1. After casting, specimens in their moulds were
subjected to air curing at 45°C and 25% R.H. After
Table 2 demoulding specimens continued to be cured in air at
Composition of the cement 45°C and 25% R.H. for a total period of 28 days. Curing
SiO2 (%) 20.4
was called air curing at 45°C and 25% R.H. and was
Al2 O3 (%) 4.89 designated as AH.
Fe2 O3 (%) 3.18 2. Specimens were covered in plastic sheeting and wet
CaO (%) 64.02 burlap after casting and put in an environmental
MgO (%) 2.56 chamber at 45°C and 25% R.H. After 24 h demoulding
SO3 (%) 2.85
Na2 O (%) 0.07
took place and specimens were quickly covered with wet
K2 O (%) 0.53 burlap and plastic sheeting for further 13 days at the
Insoluble residue (%) 0.37 same environmental conditions. Thereafter the covers
Loss on ignition (%) 0.98 were removed and specimens were exposed to air at
Free lime (%) 2.37 45°C and 25% R.H. for 14 days (i.e. the total curing
C3 S (%) 47.82
C2 S (%) 23.14
period was 28 days). This method of curing was named
C3 A (%) 7.56 moist/air curing at 45°C and 25% R.H. and was coded
C4 AF (%) 9.69 as CH.
Speci®c surface area (m2 /kg) 367.76 3. The specimens were cured in the same way as in
Residue retained on 45 lm sieve (%) 15.16 (b). The environmental conditions, however, were dif-
Initial set (min) 84.20
Final set (min) 136.50
ferent. The temperature and relative humidity were re-
spectively 20°C and 55%. Curing was named moist/air
 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437 433

Table 4
Details of the curing methods

Curing After casting and After demoulding

before demoulding

Up to 24 h After 24 h and up to 14 After 14 days and up to After 28 days and up to

days 28 days 6 months

AH Exposed to air at 45°C Exposed to air at 45°C Exposed to air at 45°C and 25% N/A
and 25% R.H. and 25% R.H. R.H.
CH Covered with plastic Covered with plastic Exposed to air at 45°C and 25% N/A
sheeting and wet burlap sheeting and wet burlap R.H.
at 45°C and 25% R.H. at 45°C and 25% R.H.
CN1 Covered with plastic Covered with plastic Exposed to air at 20°C and 55% N/A
sheeting and wet burlap sheeting and wet burlap R.H.
at 20°C and 55% R.H. at 20°C and 55% R.H.
CN2 Covered with plastic Covered with plastic Exposed to air at 20°C and 55% Exposed to air at 20°C
sheeting and wet burlap sheeting and wet burlap R.H. and 55% R.H.
at 20°C and 55% R.H. at 20°C and 55% R.H.
MH Top surface sprayed The other ®ve surfaces Exposed to air at 45°C and 25% N/A
with curing membrane were sprayed with R.H.
at 45°C and 25% R.H. curing membrane at 45°C
and 25% R.H.

curing at 20°C and 55% R.H. and was designated as 2.4. Mercury intrusion porosimetry
CN1.
4. Specimens were cured in the same way as in (c) up The porosity and pore size distribution were deter-
to a curing period of 28 days. The specimens continued mined using mercury intrusion porosimetry technique.
to be cured in air at 20°C and 55% R.H. for a total The instrument has a pressure capacity of 200 MPa. The
curing period of six months. Curing was coded as data were computed by using the Washburn equation as
CN2. follows
5. After casting specimens were put in the environ-
D ˆ ÿ…4d cos h†=P ;
mental chamber at 45°C and 25% R.H. Before water at
the top surface (trowelled face) of the cube had evapo- where D is the pore diameter, P the applied pressure, d
rated, the top surface was sprayed with a curing com- the surface tension which was taken as 484 dynes/cm
pound. After 24 h specimens were demoulded and the and h is the contact angle which was taken as 142°.
other ®ve faces of the cube were sprayed with curing
membrane at the same temperature and relative hu-
midity (i.e., 45°C and 25%). Specimens continued to be 3. Result and discussion
cured for further 27 days. Curing was called membrane
curing at 45°C and 25% R.H. and was called as MH. The e€ect of addition of SP on intruded pore volume
for paste subjected to di€erent curing regimes is shown
2.3. Drying in Fig. 1. Although the water to cement ratio is 0.45 in
this investigation, which is considered to be high for
After the predetermined curing period of 28 days paste, the intruded pore volume for the paste containing
from casting (except specimens subjected to CN2 curing SP is consistently lower than the paste without SP. This
where the total of curing period was six months), spec- seems to be the case regardless of the curing regime
imens for porosity and pore size distribution measure- adopted and the age of curing. The drop in intruded
ments were taken from the side surface of the cubes pore volume due to the addition of SP can vary between
(samples were not taken from either the top surface or 8.5% and 26% depending upon the curing condition.
the bottom surface of the cube). The mass of each The intruded pore volume of the paste increases under
sample ranged from 0.8 to 2.5 g. Samples were dried in dry curing (AH) compared with pastes which were
an oven at 100  5 C until constant mass. Prior to oven subjected to some initial moist curing (CN1 and CH).
drying, samples were introduced to a hot environmental The membrane cured specimens (MH) yielded higher
chamber at 45°C and 25% R.H. The samples were then intruded pore volumes than initially moist cured pastes
transferred to an air-tight container until the time of the and lower volume than air cured specimens (AH). The
mercury intrusion porosimetry test. intruded pore volume decreases with the age of curing as
434 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437

Fig. 1. In¯uence of superplasticizer on intruded pore volume of paste.

demonstrated by pastes initially subjected to moist

curing at 20°C for 14 days (CN1 and CN2) and cured
for a total periods of 28 days and six months.
The incremental pore volume versus the mean pore
diameter for paste subjected to the di€erent exposure
conditions is presented in Figs. 2±6. An observation

Fig. 3. In¯uence of superplasticizer on pore structure for moist/air

cured specimens 45°C and 25% R.H. (CH).

which can be drawn from these ®gures is that the

dominant pore diameter did not signi®cantly change
with the addition of SP. The dominant pore size is the
diameter whose size occupies the largest proportion of
pores compared with other pore diameters (sizes). The
dominant pore size shifts to a smaller value for pastes
subjected to some initial moist curing compared with
paste exposed to dry curing. The data from these ®gures
were used to determine the threshold diameter which is
shown in Fig. 7. The threshold diameter is the diameter
on the cumulative pore volume curve before which the
intruded pore volume rises sharply. Further details
about the determination of threshold diameter are given
elsewhere [17]. The addition of SP decreases the
threshold diameter at all the exposure conditions, indi-
cating that there is a re®nement in the pore structure.
With regards to the curing regime, initially air cured
specimens at high temperature (AH) exhibited a larger
Fig. 2. In¯uence of superplasticizer on pore structure for membrane threshold diameter than any other curing methods.
cured specimens 45°C and 25% R.H. (MH). Initially moist curing specimens at both normal
 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437 435

Fig. 4. In¯uence of superplasticizer on pore structure for moist/air Fig. 5. In¯uence of superplasticizer on pore structure for air cured
cured specimens 20°C and 55% R.H. (CNI). specimens 45°C and 25% R.H. (AH).

temperature (20°C) and high temperature (45°C) yielded presence of SP agrees with results reported elsewhere
a lower threshold diameters compared with other curing [3,5,9]. The proportion of large pores for paste subjected
regimes. For pastes which were initially cured at 20°C to dry curing (AH) is substantially higher than that of
(CN1 and CN2), the threshold diameter decreases with paste exposed to initially moist curing. An increase of up
the age of curing as expected. to 25% in coarse pores can be observed for initially air
It was suggested that the pores in a cement paste may cured paste at 45°C compared with initially moist cured
divided into small pores which have a diameter smaller specimens at the same temperature. The availability of
than 100 nm and large pores which have a diameter more water in the moist cured pastes produces more
larger than 100 nm [18]. Based on this assumption the hydration products (i.e. C±S±H gel), which has a pore
pores of the paste under the various curing conditions ®lling e€ect, resulting in a denser matrix, reduction in
were divided into small and large pores, which are pre- pore volume and ®ner pore structure. Contrary, in dry
sented in Table 5. It is quite noticeable that there is a cured pastes, less water is available for the formation of
reduction in large pores due to the presence of SP. This hydration products, thereby larger pore volume and
reduction in large pores is present under the various coarser pore structure are obtained.
curing regimes adopted and at di€erent periods of cur-
ing. The magnitude of this reduction varies between 2%
and 10%. As a consequence of reduction in large pores, 4. Conclusion
the volume of small pores is increased. This further in-
dicates that pore structure is ®ner due to the inclusion of The following conclusions are based on the work of
SP. The re®nement of the pore structure due to the this study:
436 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437

Fig. 7. In¯uence of superplasticizer on threshold diameter of paste.

paste specimen is has a substantial in¯uence on po-

rosity and pore structure. Less pore volume and ®ner
pore structure are obtained when cement paste is sub-
jected to initial moist curing as compared with initially
dry curing.

Acknowledgements

The ®rst author would like to thank Hariri Founda-

tion for initially supporting the project and Sheeld
Fig. 6. In¯uence of superplasticizer on pore structure for moist/air
cured specimens 20°C and 55% R.H. (CN2). Hallam University for allocating the time to write this
work.

The inclusion of SP in cement paste leads to a re-

References
duction in the total pore volume and to a re®nement of
the pore structures. The dominant pore size is unaf- [1] Macias A, Goni S. Characterization of admixtures as plasticizer
fected and the threshold diameter is reduced in the or superplasticizer by de¯occulation test. ACI Materials Journal
presence of SP. The initial curing regime to which a 1999;96(1):40.

Table 5
In¯uence of superplasticizer on large and small pores

Curing Age of SP % bwc Pore volume (mm3 /g) % Pores

curinga (%)
Total volume Large pores Small pores Large Small
(>100 nm) (<100 nm)

MH 28 days 0 182.6 125.4 57.2 68.7 31.3

MH 28 days 1.2 167.0 111.2 55.9 66.5 33.5
CH 28 days 0 168.4 81.8 86.1 48.6 51.4
CH 28 days 1.2 135.2 52.3 82.9 38.7 61.3
CN1 28 days 0 169.5 78.6 90.9 46.4 53.6
CN1 28 days 1.2 125.9 46.5 79.4 36.9 63.1
AH 28 days 0 199.1 138.9 60.1 69.8 30.2
AH 28 days 1.2 178.5 116.8 61.7 65.5 34.5
CN2 6 months 0 141.4 64.1 77.3 45.34 54.7
CN2 6 months 1.2 119.6 45.6 73.9 38.2 61.8
a
By weight of cement.
 J.M. Khatib, P.S. Mangat / Cement & Concrete Composites 21 (1999) 431±437 437

[2] Collepardi M. A holistic approach to concrete durability ± role of [11] Durekovic A. Cement pastes of low water to solid ratio: an
superplasticizers. In: Swamy RN, editor. Proceedings of Interna- investigation of the porosity characteristics under the in¯uence of
tional Conference. University of Sheeld, 28 June±2 July 1999:15. a superplasticizer and silica fume. Cement and Concrete Research
[3] Ramachandran VS. Concrete admixtures handbook: properties 1995;25(2):365.
science and technology. Noyes Publication, 1985:211. [12] Syal SK, Kataria SS. Development and interaction of a concrete
[4] Concrete Society, Guide to Chemical Admixtures for Concrete: additive for improved performance and durability. Cement
Report of a Joint Working Party, Concrete Society Technical Concrete and Aggregates 1982;4(2):110.
Report No. 18, 1980:8. [13] Skirkavand M, Baggott R. E€ect of superplasticizer on work-
[5] Mor A, Mehta PK. E€ect of superplasticizing admixtures on ability and ¯exural strength of autoclaved calcium silicates.
cement hydration. Cement and Concrete Research 1984;14(5):754. Cement and Concrete Research 1995;25(7):1512.
[6] Reinhardt HW, Gaber K. From pore size distribution to an [14] Krstulovic R, Smikic A, Dabic P. Examination of reaction
equivalent pore size of cement mortar. Materials and Structures between the NSF superplasticizer and cement. Cement and
1990;23(2):3. Concrete Research 1994;24(5):948.
[7] Kumar A, Roy DM. Pore structure and ionic di€usion in [15] Ghosh RS, Malhotra VM. Use of superplasticizers as water
admixture blended portland cement system. In: Proceedings of the reducers. Cement Concrete and Aggregates 1979;1(2):56.
Eighth International Conference on the Chemistry of Cement, [16] Gagne R, Boisvert A, Pigeon M. E€ect of superplasticizer dosage
vol. V. Rio de Janeiro 1986:73. on mechanical properties, permeability and freeze-thaw durability
[8] MacInnis C, Racic D. The e€ect of superplasticizers on the of high-strength concrete with and without silica fume. ACI
entrained air-void system in concrete. Cement and Concrete Materials Journal 1996;93(2):111.
Research 1986;16:345. [17] Khatib JM, Wild S. Pore size distribution of Metakaolin paste.
[9] Ping G, Ping X, Beaudoin JJ, Jolicoeur C. Investigation of the Cement and Concrete Research 1996;26(10):1545±53.
retarding e€ect of superplasticizers on cement hydration by [18] Mehta PK, Manmohan D. Pore size distribution and perme-
impedance spectroscopy and other methods. Cement and Con- ability of hardened cement pastes. In: Proceedings of the
crete Research 1994;24(3):432. Seventh International Congress on Chemistry of Cements
[10] Singh NB, Sarvahi R, Singh NP. E€ect of superplasticizers on the 1980;3:(VII):1.
hydration of cement. Cement and Concrete Research
1992;22(2):725.