Construction and Building Materials 65 (2014) 360–366

Contents lists available at ScienceDirect

Construction and Building Materials

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

Interaction effect of triisopropanolamine and glucose on the hydration

of Portland cement
Huang Hong, Shen Xiao-dong ⇑
State Key Laboratory of Materials-Oriented Chemical Engineering, College of Material Science and Engineering, Nanjing Tech University, 210009 Nanjing, China

h i g h l i g h t s

 The Glucose had an effect with TIPA that significantly promoted the hydration of paste after retardation.
 The heat generation of the paste with TIPA and glucose surpassed that of pastes with only TIPA or glucose.
 The addition of glucose promoted C4AF dissolution by TIPA.
 The types of hydrates were influenced by TIPA or glucose.
 The hydrate products were characterised by a high surface area when the glucose was present.

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

Article history: This study addresses the hydration mechanism of a P.II 52.5R cement in the presence of triisopropanol-
Received 9 January 2014 amine (TIPA), glucose or both. Calorimetry tests showed that the interaction effect of TIPA and glucose
Received in revised form 1 April 2014 significantly enhanced the degree of hydration of the cement after 7 days. X-ray diffraction analysis con-
Accepted 4 April 2014
firmed that the addition of glucose promoted C4AF dissolution by TIPA. The accelerated C4AF hydration
yielded a reduced CH content and an increased amount of chemically combined water. Due to the delayed
acceleration effect of glucose, the hydrate products were characterised by a high surface area. Further-
Keywords:
more, our results showed that the addition of TIPA promotes the formation of carboaluminate and the
Triisopropanolamine
Glucose
AFm phase when the limestone is in the presence of cement.
Interaction effect Ó 2014 Elsevier Ltd. All rights reserved.
Hydration
Cement

1. Introduction
Chemical additives are introduced to Portland cement during
manufacturing to improve the quality of the finished cement. Tri-
isopropanolamine (TIPA), a tertiary alkanolamine invented in
recent years, is widely used as a grinding chemical in the clinker
comminution process to reduce agglomeration in the ball mill
and to modify the particle distribution of the finished cement
[1,2]. TIPA also enhances the mechanical properties of cement.
According to various studies, the addition of TIPA to Portland
cement significantly improves strength development after 7 days
[3–5]. Gartner proposed that TIPA is a facilitated transporter that
chelates Fe3+ from the hydration product of ferrite and then releases
it into the aqueous phase, increasing the dissolution of ferrite and
promoting the overall silicate reaction [6]. An organic retarder, such
⇑ Corresponding author. Tel.: +86 25 83587234; fax: +86 25 83221690.
E-mail address: xdshen@njtech.edu.cn (X.-d. Shen).
as sugar, has been used as the concrete chemical admixture to con-
trol the set and to improve the rheological properties of the fresh
concrete mix [7,8]. Young attributed the retarding effect of sugar
to adsorption to both hydrating particles and hydration products
[9]. However, different investigations have yielded conflicting
experimental evidence regarding the site of adsorption during the
hydration process [10,11]. Glucose, a typical monosaccharide, is
the basic building block of sugar and other polysaccharides. The
retarding effect of glucose can also be explained by an adsorption
mechanism. Currently, glucose is mainly used as a set retarder.
In a recent publication by Huang et al., the authors employed
response surface methodology and found an interaction effect
between triisopropanolamine and a monosaccharide that strongly
contributed to the strength development of cement at 28 days
[4,13]. However, there was no interaction between monosaccha-
ride and triethanolamine (TEA), which is another popular alkanol-
amine used as a cement grinding aid [1,14]. In this work, glucose
was selected as the monosaccharide in order to investigate its

http://dx.doi.org/10.1016/j.conbuildmat.2014.04.077
0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366 361

potential interaction effect with TIPA on the hydration of Portland Table 1

cement. Analysis of 525 cement.

Oxide wt.% Mineral wt.%

2. Materials and testing methods CaO 63.89 C3S 64.60
SiO2 19.80 C2S 7.99
A P.II 52.5R Portland cement (Coded as 525 cement, composed of 90% clinker, Al2O3 4.43 C3A 6.53
5% natural gypsum and 5% limestone) was used. It was analysed by X-ray diffraction Fe2O3 3.08 C4AF 9.36
(XRD) and X-ray fluorescence (XRF) methods. The XRD pattern and other informa- SO3 3.77
tion for the cement are shown in Fig. 1 and Table 1, respectively. The TIPA used in K2O 0.067
this study was commercial grade (effective concentration P85%). TEA was also used Na2O 0.095
for a performance comparison in the calorimetry test. Glucose syrup was obtained Ignition loss 2.18
from a local sugar refinery. The FTIR spectrum of the glucose syrup is presented in
Fig. 2. The broad band at 3600–3200 cm 1 is attributed to OAH, and the OAH in-
plane bending vibration is observed in the 1440–1260 cm 1 band. The band at
2900–2840 cm 1 is due to the symmetrical CH stretching vibrations in ACH2, the
band observed at 1660–1600 cm 1 is due to C@O stretching vibrations, and the
band near 1000 cm 1 is attributed to a CAOH stretching vibration. All chemicals
were diluted in the mixing water for the paste preparation.
Approximately 6 g of paste with a w/c of 0.5 was prepared and loaded into an 8-
Channel TAM Air isothermal calorimeter to determine the rate of heat generation
during hydration in the first 72 h and the heat evolution for up to 7 days. Pastes
were also made with a w/c of 0.5 and stored at 20 °C in 50 ml sealed plastic vessels
for thermal analysis (TG–DSC and ignition method) and X-ray diffraction (XRD). The
hydration of the paste was stopped by solvent exchange using ethanol. Then, the
sample was dried in a vacuum drying oven at a temperature of 40 °C for 4 h.
An approximately 5 g powder sample was analysed by XRD using a Thermofish-
er ARL 9900 series X-ray workstation with Co Ka radiation (k = 0.1789 nm) oper-
ated in reflection geometry (h/2h) at room temperature. The analysis of the pure
clinker is shown in Table 1. The data were collected from 8° to 80° (2h) over
120 min with a 0.02° step size. The X-ray tube was operated at 40 kV and 40 mA.
Thermogravimetric analysis was performed on a NETZSCH STA 449C by heating
from 50 to 850 °C at 10 °C/min and purging with N2. The amount of Ca(OH)2 was
calculated from the weight loss of a dry sample from 400 to 500 °C. The chemically
combined water in the hydrated paste was determined by the ignition method. The
sample was dried in a vacuum drying oven at a temperature of 105 °C for 2 h to
remove the bound water. Then, the dried sample was loaded into a muffle furnace
and calcined at 1000 °C for 0.5 h. The amount of chemically combined water was Fig. 2. FTIR spectra of glucose syrup.
calculated from the weight loss of the sample from 105 to 1000 °C.

Table 2
3. Results Symbols and compositions of the samples.

Sample Composition
3.1. Heat evolution
1 100% 525 Cement
2 100% 525 Cement + 500 ppm TEA
To clarify the interaction of TIPA and glucose, another typical
3 100% 525 Cement + 500 ppm TIPA
alkanolamine, triethanolamine (TEA), was introduced for compari- 4 100% 525 Cement + 500 ppm glucose
son with the effect of TIPA on the heat evolution of the cement. The 5 100% 525 Cement + 500 ppm TEA + 500 ppm glucose
hydration heats of the pastes with different chemicals (Table 2) 6 100% 525 Cement + 500 ppm TIPA + 500 ppm glucose
was monitored by calorimetry, and the rate of heat generation dur-
ing hydration in the first 72 h and the cumulative heat evolution
for up to 14 days are shown in Figs. 3 and 4.
The addition of glucose to the pastes significantly delayed the
formation of the first main heat releasing peak by approximately
4–5 h, which is associated with the hydration of C3S [15]. Addition-
ally, the delayed peak was stronger than the peak in the paste
without glucose, which suggests that glucose promoted the hydra-
tion of C3S after the retardation. The heights and starting times of
the peaks in the pastes with or without TEA were almost identical,
which indicates that the addition of TEA did not affect the hydra-
tion of C3S. However, the addition of TEA contributed to the forma-
tion of a heat-releasing peak at 12 h, which is in agreement with
renewed AFt formation and indicates that the TEA–aluminate
interaction occurred during the early hydration period (Fig. 3a)
[15]. In the paste with both TEA and glucose, no peak occurred
after the first heat-releasing peak, indicating that the addition of
glucose inhibited the renewed aluminate reaction. After hydration
for 36 h, the rates of hydration in the different pastes were not sig-
nificantly different from each other. Fig. 3b shows that the addition
of TIPA did not affect the hydration of C3S, as did TEA. However,
TIPA promoted the formation of a shoulder peak at 24 h. The for-
mation of the shoulder peak is associated with the conversion of
AFt to AFm [15], which indicates that TIPA accelerated the sulphate
Fig. 1. XRD pattern of 525 cement. consumption. In the paste with both TIPA and glucose, the C3S
362 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366

Fig. 3. Heat flow of pastes with different chemicals. (a) TEA and glucose, (b) TIPA and glucose.

Fig. 4. Heat released in the pastes with different chemicals. (a) TEA and glucose, (b) TIPA and glucose.

hydration peak was delayed slightly. However, a shoulder peak fol-
lows, the starting time of which was the same as that of the peak in
the paste with TIPA. This suggests that the addition of glucose did
not affect the AFt to AFm conversion.
The cumulative heat released reflects the degree of hydration of
the paste. When the paste in the presence of glucose was hydrated
for 24 h, the cumulative heat release was 159 J/g, which is nearly
identical to the heat released by the plain paste (158 J/g). This indi-
cates that the glucose-incorporated paste reached the same level of
hydration even though the C3S hydration was delayed. From 24 h
to the end of the calorimetry test, the degree of hydration of the
paste with glucose was much higher than that of the plain paste.
Similar results have been reported in other studies, in which the
degree of hydration of paste containing 1% sugar surpassed the
control when the pastes were hydrated for a certain period
[10,12]. In the paste with TEA, the total heat release was slightly
higher than that of the plain paste before 1 day (Fig. 4a). After
1 day, there was no difference between the two pastes. This sug-
gests that TEA is only effective for the first day of hydration. The
released heat in the paste containing both TEA and glucose was
lower than that of the blank paste at 6 days, indicating that the
hydration of the paste with both TEA and glucose was inhibited.
The addition of TIPA significantly enhanced the heat generation
after 6 days compared to the blank paste, and the heat generation
of the paste with TIPA surpassed that of the paste with only
glucose after 8 days (Fig. 4b). The interaction effect of TIPA and glu-
cose was strong after hydration for 7 days. The heat generation of
the paste with TIPA and glucose surpassed that of the paste with
only TIPA after approximately 2 days and continued to be greater
until the end of the calorimetry test. The comparison of the alka-
nolamines with glucose revealed that glucose had no interaction
effect with TEA but did have an interaction effect with TIPA, which
significantly promoted the hydration of the paste after retardation.
3.2. Hydrate products
The thermal analyses of the pastes with TIPA, glucose or both
compared with the plain paste are shown in Fig. 5. The endother-
mic peak at approximately 100 °C refers to the decomposition of
ettringite or the C–S–H gel, and the peak at approximately 150 °C
refers to the decomposition of carboaluminate or of the AFm phase
[16].
According to Fig. 5, there was no obvious carboaluminate or
AFm, other than the ettringite forms, at 1 day. At 3 days, the forma-
tion of carboaluminate or AFm occurred. After 7 days, the forma-
tion of carboaluminate or AFm in the pastes with TIPA or with
glucose and TIPA became obvious, and the endothermic peak at
100 °C remained present, which suggests that ettringite was pres-
ent in the pastes. However, compared with the plain paste or with
the paste containing only glucose, the features of the endothermic
 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366 363

Fig. 5. Thermal analysis of the hydrated pastes with TIPA, glucose, or both.

peaks of carboaluminate and AFm were not significantly different.
Thus, the findings validated the assignment of carboaluminate, and
AFm could not be clearly distinguished by thermal analysis.
To further identify the types of hydrates, different pastes were
investigated by XRD analysis, as shown in Fig. 6.
The main differences in the XRD patterns were observed at low
angles, where the AFm phase, carboaluminate, and C4AF are found.
In Fig. 6, the diffraction peaks in the range of 12–14° correspond to
the AFm phase and to carboaluminate. However, there are some
differences in the two patterns that can be attributed to the effects
of TIPA and glucose on the course of hydration. When TIPA was
present, AFm and hemicarboaluminate were found in the pastes.
Monosulphoaluminate (2h:13.6°; d:7.6) was found in the plain
and glucose-incorporated pastes. Meanwhile, the difference in
the height of the C4AF peak was marked, which indicates that
the C4AF content varied from sample to sample. In this investiga-
tion, the change in the relative intensity of the C4AF peak in the
XRD pattern was used rather than the Rietveld method to evaluate
the ferrite consumption. This method was chosen because when
cement hydrates, the increased amount of the amorphous phase

Fig. 6. XRD patterns of hydrated pastes in the presence of TIPA and glucose.
364 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366

will impact the accuracy and precision of the Rietveld method,

especially for those crystal phases whose contents are low in the
hydrate system. Fig. 7 shows the relative contents of the ferrite
phase in the different pastes. The data indicate that all the ferrite
had participated in hydration after 1 day; however, the consump-
tion of ferrite depended greatly on the chemical type.
The addition of glucose alone inhibited the hydration of C4AF
after 1 day compared with that in the plain paste. The intensity
of the C4AF peak in the paste with TIPA decreased after 1 day, sug-
gesting that TIPA participates in the hydration process and that the
solid C4AF is consumed gradually. When TIPA and glucose were
both added, the consumption of C4AF between 1 and 4 days was
typically inhibited compared with the paste with only TIPA. This
results from the retarding effect of glucose during early hydration,
which was validated by the results of the calorimetry test shown in
Fig. 3b. After the paste had hydrated for 4 days, the intensity of the
C4AF peak dramatically decreased, indicating that the hydration of
C4AF was significantly accelerated.
Fig. 8. CH content in different pastes.
The CH content calculated from thermal analysis is presented in
Fig. 8, and the amount of chemically combined water in the
hydrated pastes as determined by the ignition method is shown
in Fig. 9. Fig. 9 shows that the hydration degrees of the pastes with
TIPA or with TIPA and glucose after 3 days were higher than that of
the plain paste and that of the glucose-only paste, in agreement
with the calorimetry test. However, thermal analysis showed that
the CH contents in the pastes with TIPA or TIPA and glucose were
lower than those of the others. This result indicates that CH partic-
ipates in specific chemical reactions and is consumed. In addition,
the amount of chemically combined water in the paste with both
TIPA and glucose was more than that in the paste with only TIPA
after 3 days and was accompanied by a higher consumption of
CH. This confirms that the addition of glucose strengthened the
effect of TIPA on hydration.

3.3. Pore structure

The gel pore structures of the pastes were determined by a BET

test, the result of which is shown in Fig. 10 and Table 3.
Fig. 10 and Table 3 show that the addition of TIPA and glucose
significantly influenced the gel pore structures of the pastes. Gen-
erally, the pastes with glucose had a higher total pore volume at
3 days, which was most likely due to the retardation effect inhibit-
ing the hydration of the cement. On the other hand, compared with
the plain paste, the paste with TIPA had a relatively small specific
Fig. 9. Chemically combined water in different pastes.
surface area at 3 days, and the specific surface area of the paste
with only glucose was also higher. The specific surface area of
the paste with both TIPA and glucose was the smallest, indicating
that the addition of TIPA reduced the specific surface area at
3 days. When the pastes were cured for 28 days, no significant dif-
ferences in pore volume were observed among the pastes. How-
ever, the specific surface area of the paste with TIPA was still
smaller than that of the plain paste, whereas the paste with added
glucose had a high specific surface area. The paste with both TIPA
and glucose had the highest specific surface area. In addition, a
higher specific surface area tended to correspond with finer pores,
especially in the pastes with glucose at 28 days. We therefore pro-
pose that the addition of glucose reduces the pore diameter and
results in an enhanced specific surface area at later stages of
hydration.

4. Discussion

Fig. 7. Relative peak intensity of C4AF in the hydrated pastes.
The initial portion of this study compared the effect of TIPA with
that of TEA and glucose on the heat evolution of a P.II 52.5R cement
containing 5% limestone. Tests showed that these two alkanolam-
ines behave differently during the course of hydration: TEA pro-
motes a renewed aluminate reaction, whereas TIPA accelerates
the AFt to AFm conversion. Another difference between TEA and
TIPA is that the former does not contribute significantly to the total
heat release. In contrast, TIPA tends to markedly enhance heat
 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366
Fig. 10. Pore size distribution of pastes with TIPA, glucose or both. (a) At 3 days, (b) At 28 days.
Table 3
Pore structures of pastes with TIPA, glucose or both.
Test item 3d
Blank TIPA Glucose
as,BET (m2/g) 29.84 27.99 33.80
Total pore volume (%) 0.107 0.119 0.138
Average pore diameter (nm) 14.29 16.08 16.32
release after 7 days, which suggests that it promotes the hydration
of cement. Gartner attributed the difference in behaviour of TEA
and TIPA to their adsorption effects. The successful chelation of
Al/Fe ions at high pH conditions was validated for TEA and TIPA.
However, the TEA-complex tended to be easily absorbed by hydra-
tion products, whereas the TIPA-complex was able to maintain a
high iron concentration in the aqueous phase for a longer time
[6]. A previous study showed that TIPA was able to accelerate the
dissolution of C4AF, which exposed more surfaces of the adjacent
phases to more rapid reaction. Thus, the overall reaction of cement
was accelerated, resulting in enhanced cement strength [6].
Our study also confirmed that monocarboaluminate and
ettringite were present, but no AFm phase existed in the plain
paste or the paste with glucose. This is because the CaCO3–C3A
reaction inhibits the AFt to AFm conversion after sulphate deple-
tion. In pastes with TIPA or with glucose, hemi-carboaluminate
and the AFm phase were identified, which suggests that TIPA and
limestone participated in hydration and that the acceleration of
ferrite hydration promoted the AFt to AFm conversion and carboa-
luminate hydration. This is in agreement with Ichikawa’s findings
[5].
Furthermore, we found that the interaction effect of TIPA and
glucose was able to enhance the degree of hydration of the cement
at 28 days. The chemical nature of glucose most likely affected the
course of hydration. Ramachandran confirmed that some retarders
can be adsorbed on the hydrate products of aluminate, thus inhib-
iting its further hydration [17]. Thomas and Birchall concluded that
sugars increase the solubility of silicate due to their ability to che-
late Ca2+ ions, thereby increasing the availability of silicate ions in
solution [12]. Juenger and Jennings [10] proposed that sugar
adsorbs onto the nuclei of hydrates and then poisons their growth,
thus promoting the formation of more nuclei. Once the sugar was
consumed, a period of heterogeneous growth followed, producing
higher surface areas due to the large number of nuclei. Peterson
and Juenger validated the concept that sugar acts as a delayed
accelerator, causing a longer induction period but an increased rate
365
28 d
TIPA + glucose Blank TIPA Glucose TIPA + glucose
26.34 36.16 26.98 46.76 50.94
0.129 0.103 0.096 0.105 0.103
19.56 11.43 14.29 9.01 8.074
and extent of formation of the hydrated phases once hydration has
begun [11]. The addition of sugar has also resulted in a higher dif-
fusion coefficient, indicating that the C–S–H formed was more per-
meable and had a higher specific surface area [11].
In the above investigation, compared with the paste containing
only TIPA, the cement with both TIPA and glucose was character-
ised by greater heat release and by more chemically combined
water in the course of hydration after 3 days. XRD analysis showed
that the presence of glucose inhibited the dissolution of C4AF
before 3 days. Afterwards, the dissolution of C4AF was greatly
strengthened in the paste with TIPA, which we attribute to the
‘‘delayed acceleration effect’’ of the retarder. The presence of glu-
cose also altered the gel pore structure, which was characterised
by a finer pore diameter and a higher specific surface area. The
‘‘facilitated iron-transport’’ mechanism of TIPA was promoted
because of the more permeable hydrates, which resulted in higher
degrees of hydration of the silicates, indirectly explaining why the
interaction effect of TIPA and glucose enhanced the late-stage
strength of cement in previous studies [4,13].
5. Conclusions
 The addition of glucose to cement delayed the C3S hydration for
4–5 h, and the rate of hydration was more rapid after the retar-
dation. TEA did not influence the hydration of C3S but did influ-
ence that of C3A, resulting in the accelerated renewed formation
of AFt. The presence of TIPA also favoured the aluminate reac-
tion, reflected by the accelerated AFt to AFm conversion. No
interaction effect between TEA and glucose was detected in
the calorimetry test, but the combination of TIPA and glucose
strongly promoted the hydration of cement after 7 days, which
resulted in more heat release than for the pastes with only
either TIPA or glucose.
 The hydration products were determined by the presence of
TIPA or glucose. In the cement with glucose alone, the AFt and
monocarboalunimate were identical to the hydration products
366 H. Huang, X.-d. Shen / Construction and Building Materials 65 (2014) 360–366
in the plain paste. For the pastes containing TIPA, the hydration
products comprised AFt, AFm and hemicarboaluminate. Glucose
alone inhibited the hydration of C4AF, whereas TIPA accelerated
it compared with the plain paste. In the cement with both TIPA
and glucose, the hydration of C4AF was inhibited before 4 days;
however, it was significantly promoted afterwards compared
with TIPA alone.
 The interaction effect of TIPA and glucose promoted the hydra-
tion of cement up to 28 days, as validated by calorimetry and
thermal analysis. The promoted hydration resulted from the
addition of glucose, which strengthened the ‘‘facilitated trans-
portation’’ effect of TIPA and accelerated ferrite dissolution.
The pore structure of the cement paste was modified by this
interaction effect, as well. TIPA alone tended to reduce the spe-
cific surface area of the paste and to increase the average pore
diameter. The addition of glucose, on the contrary, enhanced
the specific surface area and reduced the average pore diameter.
The interaction effect of TIPA and glucose was more pro-
nounced, and the corresponding paste had the highest specific
surface area and the lowest average pore diameter at 28 days.
Acknowledgements
The authors are grateful for the support from the National Basic
Research Program of China (2009CB623100) and from the Post
Doctorate Fund in Jiangsu Province (1301037B). The authors also
offer thanks to the Program for Changjiang Scholars and the Inno-
vative Research Team in University (PCSIRT, No. IRT1146) and A
Project Funded by the Priority Academic Program Development
of Jiangsu Higher Education Institutions (PAPD).
References
[1] Katsioti M, Tsakiridis PE, Giannatos P, et al. Characterization of various cement
grinding aids and their impact on grindability and cement performance. Constr
Build Mater 2009;23:1954–9.
[2] Somasundaran P, Shrotri S. Grinding aids: a review of their use, effect and
mechanisms. Selected topics in mineral processing. Wiley Eastern Limited;
1995. p. 47–70.
[3] Sandberg PJ, Doncaster F. On the mechanism of strength enhancement of
cement paste and mortar with triisopropanolamine. Cem Concr Res
2004;34:973–6.
[4] Huang Hong, Shen Xiaodong. Statistical study of cement additives with and
without chloride on performance modification of Portland cement. Progr Nat
Sci-Mater Int 2011;21:246–53.
[5] Ichikawa M, Kanaya M, Sano S. Effect of triisopropanolamine on hydration and
strength development of cements with different character. In: Proceedings of
the 10th international congress on the chemistry of cement, Gothenburg,
Sweden; 1997 [Paper 3iii005].
[6] Gartner E, Myers D. Influence of tertiary alkanolamines on Portland cement
hydration. J Am Ceram Soc 1993;76:1521–30.
[7] Dodson VH, editor. Concrete admixture. New York: Van Nostrand Reinhold;
1990.
[8] Francis Young J, Darwin David, Mindess Sidney. Concrete. 2nd ed. New
Jersey: Prentice Hall; 2002.
[9] Young JF. A review of the mechanisms of set-retardation in portland cement
pastes containing organic admixtures. Cem Concr Res 1972;2(4):415–33.
[10] Garci Juenger MC, Jennings HM. New insights into the effects of sugar on the
hydration and microstructure of cement pastes. Cem Concr Res
2002;32(3):393–9.
[11] Peterson VK, Juenger MCG. Hydration of tricalcium silicate: effects of CaCl2
and sucrose on reaction kinetics and product formation. Chem Mater
2006;18:5798–804.
[12] Thomas NL, Birchall JD. The retarding action of sugars on cement hydration.
Cem Concr Res 1983;13(6):830–42.
[13] Huang Hong, Shen Xiaodong, Zheng Jiaoling. Modeling, analysis of interaction
effects of several chemical additives on the strength development of silicate
cement. Constr Build Mater 2010;24(10):1937–43.
[14] Ramachandran VS. Hydration of cement – role of triethanolamine. Cem Concr
Res 1976;6(5):623–31.
[15] Taylor HFW. Cement chemistry. London: Thomas Telford; 1997.
[16] De Weerdt K, Ben Haha M, Le Saout G, et al. Hydration mechanisms of ternary
Portland cements containing limestone powder and fly ash. Cem Concr Res
2011;41(3):279–91.
[17] Ramachandran VS, Lowery MS. Conduction calorimetric investigation of the
effect of retarders on the hydration of Portland cement. Thermochim Acta
1992;195:373–87.