Document [original]

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

Available online 1 March 2024

Influence of polycarboxylate superplasticizers with different molecular

structures on rheological properties of glass bead suspension at different

resting times

Yanliang Ji

, Alexander Mezhov

, Shukai Wang

, Dietmar Stephan

Department of Civil Engineering, Technische Universit¨

at Berlin, Berlin 13355, Germany

Division of Technology of Construction Materials, Bundesanstalt für Materialforschung und -prüfung (BAM), Berlin 12205, Germany

GRAPHICAL ABSTRACT

ARTICLE INFO

Keywords:

Glass bead

Rheological performance

PCE superplasticizers

Ion

Resting time

ABSTRACT

Using an inert glass bead suspension, the study explores the interaction between various PCE superplasticizers,

ions and investigates how it affects rheological performance during resting. To understand the underlying

mechanisms, surface charge properties, adsorption, aggregates amount, and size of PCE during resting were

measured. Results showed that yield stress increased with time, mainly when a high PCE dosage was used, and

the modulus from strain sweep experiments agreed with the yield stress variations. The increase in zeta potential

value and decrease in aggregate amount explain the decrease in yield stress in the non-PCE sample but are

inconsistent with the increase in yield stress in presence of PCE. The PCE clusters’ size increases during resting,

and increasing PCE adsorption does not disperse the glass bead particles. In addition, compared to the amount of

carboxylate group, the molecular weight of PCE is more dominant in increasing the yield stress during resting.

The increase in yield stress could be explained by the increase in the bridging and depletion forces due to the

increasing size of the PCE cluster. Another possible reason is that non-adsorbed PCE clusters may form a polymer

network that restricts the movement of the particles.

* Corresponding authors.

E-mail addresses: [email protected] (Y. Ji), [email protected] (D. Stephan).

Contents lists available at ScienceDirect

Colloids and Surfaces A: Physicochemical and

Engineering Aspects

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

https://doi.org/10.1016/j.colsurfa.2024.133583

Received 27 December 2023; Received in revised form 8 February 2024; Accepted 26 February 2024

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

1. Introduction

Rheological properties (i.e., yield stress, plastic viscosity, and thix-

otropy) of a cementitious material determine its applicability in prac-

tical engineering, such as mixing, filling, pumping, flowability, and even

long-term mechanical strength and durability [1–3]. During the last few

decades, various polycarboxylate (PCE) superplasticizers have been

developed and widely used to obtain adequate rheological properties [4,

5]. It is known that molecular weight [6], side-chain density [6,7],

functional group [8,9], and morphology of PCE superplasticizers

[10–12] can significantly affect rheological performance. Hence, related

studies on optimizing superplasticizers and their relations to rheological

behavior under various construction requirements have received

increasing attention from industry and academia.

A common problem in practice is that cementitious materials are

difficult to process after a resting period because they do not reach their

original flowability, even if they appear sufficiently flowable after

mixing [8,13,14]. The effect of the resting time on the rheological

behavior is generally attributed to the continuous formation of hydrates

and the microstructure development during the resting period [14,15].

For the cementitious materials mixed with PCE superplasticizers, the

continuously formed hydrates can bury or cover the adsorbed PCE on

the particle surface or over-adsorb the PCE [3,7], leading to decreased

dispersal capability and a loss of fluidity. On the one hand, to slow down

the hydration effect, retarders (i.e., dextrin, sodium gluconate, and

phosphoric acid [3,16,17]) are used in combination with PCE. However,

this may result in a lower mechanical strength of the cementitious ma-

terial, particularly in the early stage.

On the other hand, considering the adsorption during resting, PCE

with fewer carboxylate groups (R-COO

) and longer side chains have

been designed [13]. At high dosages, these types of PCEs have been

proven to maintain flowability better as more free PCE remains in pore

solution and can be gradually adsorbed on the cement surface. Another

type of PCE, namely slow-release PCE, synthesized with a functional

group of mercaptoacetic acid (HEA) or acrylamide (AM), can gradually

hydrolyze with OH

in the alkaline pore solution and release adsorbable

monomers of carboxylate polymers [8]. This property of the

slow-release PCE results in a more sustainable flowability of the cement

material during rest. It can be inferred that modifying the molecular

structure of PCE can help to achieve a sustainable rheological perfor-

mance from the point of view of cement hydration, although certain

questions remain unclear.

It is well known that cementitious material in the fresh state can be

considered as a solid suspension in which solid particles are suspended

in an ionic solution. The effect of resting time on the rheological per-

formance is related to the continual formation of hydrates and to the

liquid phase of this suspension, where different ions and PCE molecules

are present [15,18]. In the study of Hirata et al. [19], individual PCE

molecules in an ionic solution are observed to be smaller than those in a

non-ionic solution (deionized water). Additionally, the carboxylate

group of the PCE molecule will be complexed with cations (i.e., Ca

) [20], which may change the PCE morphology in the pore solution

during rest. As a result, the rheological performance during resting will

be affected. However, these effects are difficult to identify in a real

cementitious material because the hydration effect will be the dominant

factor to change the rheological performance during rest [6,17,21]. In a

previous study [15], a model system consisting of glass beads, PCE, and

ionic solutions was used to investigate the effect of resting time on

rheological properties. Compared to the cementitious material, this

model system has two obvious advantages. One is that the shape of the

glass bead particles is round, and the size range is similar to that of the

cement particles, eliminating the effect of shape change or packing on

the rheological results. The other is that there is little or no hydration in

this system, so the hydrates do not affect the PCE adsorption. Therefore,

without the hydration effect, this well-defined system allowed us to

focus on the PCE-ion interactions and their possible effects on the

rheological performance of the inert particle suspension during rest.

In this study, the glass bead suspensions were prepared with three

lab-synthesized PCE with different molecular weights (also various

amounts of carboxylate groups) and cement artificial pore solutions. The

rheological properties, including shear yield stress and oscillatory strain

sweep, were measured at different resting times (within 120 min). To

investigate the mechanisms, the zeta potential and TOC (total organic

carbon) were used to measure the surface charge properties and

adsorption during rest, respectively. In addition, dynamic light scat-

tering (DLS) was used to study the morphology of PCE clusters in the

artificial pore solution. The change in the number of agglomerates in the

glass bead suspension at different resting times was observed using a

light microscope and image analysis software. Furthermore, based on

the experimental results, a detailed discussion of the PCE-ion interaction

and its possible mechanism regarding yield stress increase during resting

is given.

2. Materials and experiment

2.1. Materials

Glass beads (GBs) were purchased from Sigmund Lindner GmbH

(Germany), and the main chemical composition is SiO

(72.3 wt%) and

O (13.3 wt%) as measured in a previous study [22]. To avoid

possible effects of contamination, the glass beads were processed in

several steps, including a magnetically stirred process, a long-time

(48 h) sedimentation and deagglomeration [22]. This process signifi-

cantly reduced the ions and smaller crushed particles from the original

glass beads. For instance, the Na

ions in solution (glass beads in

deionized water) are decreased from 34 mmol/l to 0.07 mmol/l [15,22].

The particle size distribution as measured by a laser diffraction scat-

tering (Malvern Mastersizer S) after the cleaning process is shown in

Fig. 1a.

The recipe for the artificial pore solution is from [18,23]. Potassium

hydroxide (KOH=7.12 g), sodium sulfate (Na

=6.96 g), potassium

sulfate (K

=4.76 g), and calcium sulfate dihydrate

(Ca

⋅2 H

O=1.72 g) are dissolved in 1000 mL of deionized water.

The measured ionic strength and pH are 36.7 m/s and 12.7, respec-

tively. The three types of PCE (PCE24–4, PCE24–6, and PCE40–4) used

in this study are lab-made from Tsinghua University (the chemical for-

mula see Fig. 1b), and the main characteristics are listed in Table 1.

Detailed information and characterization on these three PCE

superplasticizers can be found in (Ref [24]). It can be seen that the

molecules of PCE40–4 and PCE24–4 have a higher molecular weight but

a lower carboxylate amount, while PCE24–6 owns a smaller molecular

weight but a higher carboxylate amount.

2.2. Materials design

To investigate the time effect on the glass beads suspension, a fixed

water to glass beads ratio (0.28) was kept, and PCE dosage was used at a

low (0.1%) and high (0.4%) dosage (solid content) by glass beads

weight. The detailed mixed design is shown in Table 2.

The weighted glass beads, PCE, and artificial pore solution were

mixed by a hand mixer (HM-753BG, Qilive, France) for 2 min to prepare

the glass beads suspension. Besides, it must be noted that the rheological

measurement for inert materials will be influenced by the sedimentation

effect caused by the density difference (i.e., glass beads and liquid

phase). Therefore, a 30 s remixing procedure was applied for all mea-

surements at 30, 60, and 120 min resting times.

2.3. Experiments

2.3.1. Rheological measurement

A temperature-regulated rheometer, HAAKE RheoStress 600

(Thermo Electron Corporation), was used in this study. The geometry

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

selected for the rheological measurements was “bob in cup” (the graph is

shown in Fig. 2a), and the inner radius of the bob is 20 mm, and the

radius of the cup is 21.7 mm. All measurements were performed at a

constant temperature of 22 ◦C. The shear rate was increased from 1.0 s

−1

to 20 s

−1

, then maintained at a shear rate of 20 s

−1

for 1 min, and at the

end, the shear rate was gradually reduced to 1.0 s

−1

. The shear rate and

shear stress were analyzed using the Herschel-Bulkley model to obtain

the rheological properties (i.e., yield stress). Fig. 2 shows a typical set of

rheological data and the corresponding results obtained through fitting

with the Herschel-Bulkley model.

An amplitude sweep oscillation measurement was then performed.

The sweep frequency was kept constant at 1.0 Hz, and the shear strain

was swept logarithmically from 10

−5

to 20% during the measurement.

The oscillatory strain sweep test can be used to obtain the storage

modulus (G

′

) and the loss modulus (G

′′

), which reflect rheological per-

formance at different resting times.

2.3.2. Zeta potentials

In this study, the surface potential of the particles in the sample was

evaluated using a zeta potential device (DT-310, Dispersion Technology

USA). Based on the concentration and composition of the salts, the liquid

density was calculated. Glass bead suspensions were measured at resting

times of 0, 60 and 120 min, and all experiments were at a constant

temperature of 22℃ after the samples were prepared and loaded into the

measuring mold. In addition, the solutions (PCE in artificial pore solu-

tions according to the proportion in Table 2) without the glass beads

were also prepared to measure the ionic vibration current (IVI) at

different resting times. The measurements were conducted three times,

and the average zeta potential was calculated.

2.3.3. PCE adsorption during resting

The prepared samples were centrifuged at 3000 rpm for 5 min at

resting times of 0, 60, and 120 min. The supernatant was then filtered

through a 450 nm membrane. The prepared solution was diluted 20

times and then measured using a TOC analyzer (TNM-L, Shimadzu) to

determine the amount of PCE in the solution and the adsorbed amount in

the glass bead suspension.

2.3.4. Light microscope and image analysis

In this study, the agglomerates in the GB suspension were observed

using a light microscope (Stemi SV 11, Zeiss Microscope, Germany). The

image analysis software for computing the aggregates amount is Image-

Pro plus 6.0 (Media Cybernetics, Rockville, USA). The GB suspension

was poured onto a glass plate (27 ×47 mm

) after 0, 60, and 120 min

resting time. A thin and semi-transparent layer on the sample is then

covered with a plastic film to prevent evaporation and possible impu-

rities (i.e., dust) from air on the sample surface, which may influence the

results. The sample is then observed under a microscope at 0.6 X and 1.2

X magnification. Fig. 3 shows, as an example, a typical measurement of a

GB suspension on a glass plate using the optical microscope and image

analysis process.

2.3.5. Dynamic light scattering (DLS)

A dynamic light scattering (DLS) instrument from Malvern In-

struments Ltd was used to measure the hydrodynamic diameter of the

original PCE in artificial pore solution at different times. The dosage of

PCE in artificial pore solution used for this measurement is the non-

adsorbed PCE (considering the adsorption results) in glass beads at

0 min. The detector was positioned at 173◦for the optical arrangement.

The experiment followed the same procedure as the glass bead suspen-

sion; the PCE and artificial pore solution were mixed using the hand

mixer at 0, 60, and 120 min. Subsequently, a 1.0

m membrane was

used with a syringe to remove any impurities, such as large particles,

from the PCE-containing solution, which could have originated during

synthesis, transport, or storage processes. A typical PCE size distribution

is shown in Fig. 4.

In Fig. 4, it can be seen that the measurement results exhibit two

peaks, i.e., Peaks_S and Peak_B. Peaks_S (from 5 nm to 50 nm) corre-

sponds to the single polymers or cluster consisting of a few polymers,

Fig. 1. Particle size distribution of the processed glass beads (a) and the chemical formula of the PCEs used in this study.

Table 1

Characteristic parameters of the synthesized PCE superplasticizers (Ref[24]).

PCE-

type

Theoretical ratio of

a:b

Mw/mol/

PDI [COO−] /mmol/g

polymer

PCE24–4 1:4 156000 2.886 2.33

PCE24–6 1:6 99500 2.481 2.85

PCE40–4 1:4 120000 2.014 1.50

Table 2

Mix design of the glass bead suspensions.

Sample type Glass beads

(g)

Artificial pore solution (g) PCE dosage

(g)

GB-PCE_Ref 100 28 0

GB-PCE24–4-DS0.1 100 28 0.1

GB-PCE24–4-DS0.4 100 28 0.4

GB-PCE24–6-DS0.1 100 28 0.1

GB-PCE24–6-DS0.4 100 28 0.4

GB-PCE40–6-DS0.1 100 28 0.1

GB-PCE40–6-DS0.4 100 28 0.4

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

and Peaks_B (from 100 to 1000 nm) are considered to be related to the

PCE clusters formed by numerous PCE molecules, connected by ions

from the artificial pore solutions. In this study, we will analyze the peak

position and the percentage of the peak areas of these two peaks with

time. Signals beyond the size distribution range (between 5 nm and

1000 nm) are not considered in this study. The measurements were

performed five times for each time point, and the average value was

taken as the hydrodynamic diameter. Data with a deviation exceeding

40% from the mean were discarded.

3. Experimental results

3.1. Yield stress with the resting time

Fig. 5 shows the calculated yield stress of glass bead suspensions

prepared with artificial pore solutions and different types of PCEs as a

function of resting time (within 120 min). The bars in the graph repre-

sent the fitting errors when applying the Herschel-Bulkley model to fit

the shear stress and shear rate (in the range of 1.0 s

−1

and 20 s

−1

In Fig. 5, at a resting time of 0 mins, the yield stress in all cases

decreased when PCE was added to the glass bead suspensions. At a low

dosage (PCE =0.1 wt%), the yield stress values decreased more than in

the case of a high dosage (PCE =0.4 wt%). In addition, it can be

observed that without the PCE addition (Ref sample), the yield stress of

Fig. 2. Bob in cup geometry (a) used in this study and a typical rheological data and Herschel-Bulkley model fit (b) (the fitted results on shear rate and shear stress of

the glass beads suspension made by PCE24-4 at 0, 30, 60 and 120 mins).

Fig. 3. A typical measurement of GB suspension film on a glass sheet with the Light microscope (a) and image analysis process(b) (in the image, the white part is

well-distributed glass beads, and the dark part is considered as glass beads agglomerates).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

glass bead suspensions decreases within the first 60 m, followed by a

slight increase at 120 m. These results are consistent with our previous

study[15].

However, with the PCE addition, the yield stress increase with the

resting time, and results also indicated that with more PCE, a higher

yield stress of the glass bead suspensions could be observed at 60 and

120 min. More interesting, it is observed that the evolution of the yield

stress of glass bead suspensions increased to various extents when

different PCEs were used. When the PCE dosage is 4 wt% of the glass

bead weight, the glass bead suspensions containing PCE24–4 or

PCE40–4 has a higher yield stress than that of PCE24–6.

3.2. Storage modulus (G

′

) and loss modulus (G

′′

)

Fig. 6 displays the storage modulus (G

′

) and loss modulus (G

′′

) of the

glass bead suspensions at various resting times, measured using sweep

oscillation over different strain sweep ranges (from 10

−5

to 20%). Fig. 5a

shows the storage modulus (G

′

) and loss modulus (G

′′

) curves of the

reference sample (no PCE is used). In addition, Fig. 6b-d shows the

measurement results of glass beads suspension using PCE24–4,

PCE24–6, and PCE40–4 (PCE =0.4 wt%), respectively. The storage

modulus (G

′

) and loss modulus (G

′′

) curves of glass bead suspensions at

low dosage (0.1 wt%) are summarized in supplementary materials

(Figure S1).

In Fig. 6, we can find that the storage modulus (G

′

) and loss modulus

′′

), in general, are decreased with the applied sweep strain in the glass

bead suspensions. The storage modulus (G

′

) and loss modulus (G

′′

) of

samples without PCE addition slightly change at different resting times.

However, in samples with PCE added, the storage modulus (G

′

) and loss

modulus (G

′′

) curves at 60 and 120 min are noticeably different from

those at 0 min. In all cases, a higher storage modulus (G

′

) and loss

modulus (G

′′

) can be found at 120 min, consistent with the results ob-

tained from the calculated yield stress at different resting times. It is

worth noting that under a sweep strain of 0.001, it is challenging to

discern differences between the storage modulus (G

′

) and loss modulus

′′

) due to the turbulent data. Nonetheless, it is still apparent that

PCE24–4 and PCE40–4 exhibit more significant changes in storage

modulus (G

′

) and loss modulus (G

′′

) curves over resting time compared to

samples with PCE24–6.

Overall, it was found that different types of PCE used with glass bead

Fig. 4. A typical size distribution of PCE in the artificial pore solutions (Peak_S

is the defined peak which distributes from 2 to 50 nm, and Peak_B is assigned to

the peak which distributes from 100 to 1000 nm).

Fig. 5. The yield stress evolution as a function of resting time of glass bead suspensions made by various molecular structures of PCEs (a: PCE24-4, b: PCE 24–6, c:

PCE40-4 and the water to glass beads ratio is kept at 0.28).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

Fig. 6. The storage modulus (G

′

) and loss modulus (G

′′

) as a function of sweep strain (from 10 to 5–20%) at 0, 60, and 120 min, for selected GBs suspensions made by

artificial pore solution without PCE (a), with PCE24-4-DS0.4(b), PCE24-6-DS0.1(c) and PCE40-4-DS0.4(d).

Fig. 7. Zeta potentials change as a function of resting time (within 2 h) of glass bead suspensions made by PCEs (a: PCE24-4, b: PCE 24–6, and c: PCE40-4).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

suspensions produce different yield stress curves and the storage and

loss modulus (G

′

and G

′′

) behavior at different resting times. These

changes are related to the surface change properties, PCE adsorption,

and PCE size and aggregations of glass beads at different resting times,

which are investigated and shown in the following sections.

3.3. Zeta potentials during resting

The zeta potential values of glass bead suspensions at different

resting times, 0, 60, and 120 min, were measured and plotted in Fig. 7.

In Fig. 7, it is shown that all zeta potential values are negative, which

suggests that the surface of GB in artificial pore solution is negatively

charged. The results also suggest that regardless of PCE type, during a 2-

hour resting period, the surface of GB becomes more negatively charged.

It can also be seen that PCE24–4 and PCE24–6 result in more significant

changes in zeta potential values during resting time compared to

PCE40–4. It must be noted that the larger values of zeta potentials mean

the presence of higher electrostatic force during resting, which explains

the decrease in yield stress over time of the reference sample (no PCE),

but this is not consistent with the increase in yield stress (in Section 3.1)

when PCE is used.

3.4. PCE adsorption over resting time

Fig. 8 shows the amount of PCE adsorbed in glass bead suspensions at

different resting times. Generally, the more PCE is used in the glass bead

suspensions, the higher is the adsorption amount. During the rest, the

amount of PCE adsorption increases as a function of resting time. Spe-

cifically, at both dosages, initially, the PCE24–6 owns a higher adsorp-

tion amount, while at a resting time of 60 and 120 min, the PCE40–4 has

a higher amount remaining in the glass beads suspensions than the other

types of PCEs.

3.5. Hydrodynamic radius of PCE cluster in solutions

The information (Peak-S and Peal-B) regarding the hydrodynamic

radius of different types of PCEs (PCE24–4, PCE26–4, and PCE40–4) in

artificial pore solutions at different resting times, as measured DLS, is

presented in Fig. 9. The results for PCE at high dosage (considering the

adsorption amount) are shown. The data for the PCE at low dosage

(DS=0.1%) are summarized in supplementary materials (Figure S2).

Fig. 9 shows that the peak value (hydrodynamic radius of PCE) and

peak area percentage (Peak-S, assigned to single distributing in solu-

tions) generally decrease with resting time. In both cases, the PCE40–4

exhibits lower values of the hydrodynamic radius of PCE and peak area

percentage. As for the Peak-B, which corresponds to the PCE clusters,

both the hydrodynamic radius and peak area percentage increase with

the resting time, showing that the PCE size in the artificial pore solutions

increases with the resting time. These results suggest that during 60 and

120 min, PCE can no longer be seen as a single polymer but as a cluster

of many PCE molecules. For PCE40–4, the peak area percentage even

reaches nearly 100% at 120 minutes, which is higher than the other two

types of PCE (PCE24–4 and PCE24–6), suggesting that more clusters are

being formed in artificial pore solution with PCE40–4.

3.6. Aggregates amount of glass beads over time

Fig. 10 shows the evolution of the normalized values of aggregates

amount (area to image ratios) in glass bead suspension changing with

resting time. Yet, it must be admitted that very significant standard

deviations are obtained with the method. Besides that, with the increase

of PCE dosage, there is no regularity in the area-to-image ratio of GB

suspension, and the extent of decreasing agglomerates amount after 60

and 120 min could be various, which is independent of the used PCE. It

is also observed that the area of aggregates to the image ratios all

decrease with the resting time (except for the sample of PCE24–4-DS0.4

at 120 min), implying that fewer agglomerates are present in the glass

beads suspensions during rest. Nevertheless, the results in Fig. 10 sug-

gest that the increase in PCE dosage and varying the PCE types did not

contribute to the formation of macro agglomerates in the glass bead

suspensions system.

4. Discussions

In this section, we will further discuss the reasons for the increase in

yield stress over time when different PCEs are used in such an inert

system as glass bead suspensions. The analysis and discussion will be

based on the experimental results of yield stress, surface changes, PCE

adsorption, PCE size changes in pore solution, and the amount of ag-

gregations. The influencing mechanisms of resting time on the rheo-

logical performance can be summarized in Fig. 11.

For the reference sample (without PCE addition), the decrease in

yield stress could be well explained by the results of zeta potentials and

the amount of aggregation in the glass bead suspensions. According to

the electrostatic force (also known as coulomb force [11,25]), the in-

crease of surface charge results in increased repulsion forces among

particles.

Fe=2

πεε

2dκe−κh

1+e−κh(1)

The zeta potential is an indicator for the surface potential of the

particle [4]. Higher zeta potentials were found in this study because the

surface of glass beads undergoes a dissociation process. The solution

used here is cement artificial pore solution (pH=12.7). The silanol

groups (SiOH) on the glass bead surface gradually react with OH

, and a

more negatively charged surface constituted by the SiO

will be exposed

Fig. 8. PCE adsorption amount in glass bead suspensions with resting time (within 2 h) made with different PCEs (initial PCE dosages are 0.1 (a) and 0.4 (b)).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

in the system [26,27]. If we assume that the potentials of the solution

remain unchanged during the rest, then the increase in the absolute

value of zeta potential indicates an increase in surface charge density or

potentials. As a result, as shown in Fig. 11(i), this would contribute to a

higher electrostatic repulsion force (F

), which helps overcome van der

Waals attractive force among particles during the movement of particles

[18,28]. Regarding the amount of glass bead aggregates, for the refer-

ence sample (GB-Ref), a decrease in aggregate amount means the par-

ticles are well distributed in the system. This will also lead to fewer

obstacles or jamming effects for the movement of particles [29–31],

resulting in a gradual decrease in yield stress over time. Regarding the

decreasing aggregate amount, there are two possible reasons. The first

reason could be that the additional mixing (at 30, 60 and 120 mins)

continuously destroys these macro aggregates [31,32]. The other

possible reason is that after mixing, aggregates in the solution dissolve

over time, potentially decomposing the aggregate structure as well.

For glass bead suspensions mixed with PCE, it must be noted that the

zeta potentials (absolute values) increased, and fewer aggregates were

found during resting, as shown in Fig. 7 and Fig. 10, respectively. This

implies that the yield stress increase will be dominated by the PCE and

its properties rather than the surface charge properties and aggregation

of glass bead particles during rest. Considering this, we analyze the

relationship between the PCE molecular structure (molecular weight as

measured by SEC and amount of -COO

) and the yield stress increase

during resting. The analyzed results are described in Fig. 12.

In Fig. 12, we can find that the difference in yield stress is generally

positively correlated to the molecular weight, and the amount of R-COO

amount of the PCEs has no apparent relation to the increase of yield

stress. These results are unexpected from the perspective of bridging

forces. It has been widely accepted [20,33] that R-COO

on the PCE

backbone and cations (i.e., Ca

) in pore solution will have a complex

effect, which may bridge the solid particles together and result in higher

yield stress. Therefore, a higher amount of R-COO

theoretically means

higher yield stress during rest, which is not observed in Fig. 12.

In this regard, the morphology of the PCEs in the artificial pore so-

lution and its associated effects (i.e., bridging and depletion effect [34])

should be why a higher yield stress is observed during resting. In Fig. 9b

and d, the size of the PCE cluster will increase with the resting time,

leading to two possible effects: the bridging effect and the depletion

force. For the PCE bridging effect (as shown in Fig. 11 (ii)), we consider

that the size of the PCE cluster increases, resulting in a higher possibility

for PCE to bridge the glass bead particles [15,35]. To consider this effect

from the PCE size, we need to take the initial concentration, peak per-

centage, and the size of the PCE size into account. For convenience, a

factor (

size) that considered the above factors has been defined in this

study.

size =n0⋅af⋅(fclu,t−fclu,0)(dt−d0)(2)

Where n

is the PCE used in the glass beads suspension, a

is the free PCE

amount in the solutions at resting time of 0 min, f

clu,t

and f

clu,0

are the

peak area percentage of PCE in artificial pore solution at 0 min and at

different resting times (60 and 120 min), and d

and d

are the hydro-

dynamic radius of PCE cluster at 0 and at different resting time(60 and

120 min), respectively. It can be seen from Eq. 2 that the differences in

peak area percentage and hydrodynamic radius all reflect the size

change during resting. In Fig. 13, we analyze the relations of the size

factor (

size) of PCEs to the adsorption increase during the rest, as well as

the increase of yield stress.

Results in Fig. 13 indicate that the size factor of the PCE cluster is

linearly correlated with the increase in adsorption amount and yield

stress. The positive linear relationships in Fig. 13 could be due to the

following reasons. As shown in Fig. 9, the PCE can no longer be seen as

individual polymers. The adsorption of these enlarged PCE clusters will

not contribute to the plasticizing effects for glass bead suspensions, i.e.,

electrostatic repulsion and steric hindrance [36,37]. On the contrary,

Fig. 9. The changes of hydrodynamic radius of PCE in artificial pore solution as measured by dynamic light scattering measurement (Figs. a and b are for the radius

and peak area of Peak-S, while Figs. c and d are for the radius and peak area of Peak-B).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

the PCE cluster size is getting bigger in the solutions, which will bridge

the glass bead particles and are considered to bind the PCE molecules

themselves on the surface.

Besides, it must be noted that adsorption measurement depends on

how much non-adsorbed PCE can escape from the glass bead suspension

during the centrifuging process. The PCE clusters with increasing size

will be more possibly stuck or wrapped (on particles), and consequently,

a higher adsorption amount could be observed. On the other hand, in a

previous study [15], we proposed that the size of the free PCE increase

will also lead to an increase in depletion force, which can be evaluated

by Eq. 3.

Fd,t= −

s,t(1−ϕ(t))kBT

dGB(2Rg,t−h)(3)

Where,

s,t is the number density of free PCEs (single polymer case) in

the bulk solution at 0 min. Φ(t) is the fraction of the amount of non-

adsorbed, bonded PCE molecule pairs to the total amount of non-

adsorbed PCE molecules, d

is the mean diameter of the glass beads,

is the Boltzmann constant (1.38⋅10

−23

J/K), T is the absolute Tem-

perature (295 K) and Rg,t is the mean hydrodynamic radius of the non-

adsorbed PCE at resting time t. According to Eq. 3, the increasing size

of none adsorbed PCE clusters (Rg,t) in the solutions will contribute to a

higher depletion force at a longer distance during resting (as shown in

Fig. 11 (iii)). As a result, the yield stress increases. In addition to this

depletion force, another possible reason for the linear relationship of

PCE size and yield stress increase is that the non-adsorbed PCE may form

a polymer network connected by the RCOO

and cations of artificial pore

solutions. The formation of this polymer network in the liquid phase has

a lower mobility, and meanwhile, this network may also refrain from the

movement of glass beads and result in higher yield stress (shown in

Fig. 11 (iv)).

Here, we can re-examine the properties and features of the whole

system of glass bead suspension and the experiment regimes we have

conducted to discuss the possible guiding values of this research for

industrial practice. First, the glass bead suspension has a relatively low

yield stress at 0 min (most of them <1.0 Pa), suggesting the GBs sus-

pension is well dispersed at first, and most of the attractive forces (i.e.,

Van der Waals force) have been eliminated by the dispersion of PCE.

Then, the suspended glass beads have a negatively charged surface, and

a large portion (over half PCE) of the PCE polymers are free in the pore

solutions. Moreover, additional mixing procedures are applied to the

glass bead suspensions to reduce the sedimentation impact. The glass

beads system reminds us that in practical engineering, a high dosage of

PCE was occasionally used in concrete (i.e., self-compacting concrete

[13,38,39]) to achieve long-lasting workability, in which a similar

physical or chemical condition could be achieved compared to that of

glass beads suspension. The results in this study suggest that a higher

molecular weight of PCE will result in higher yield stress during resting,

which may help us to make wise decisions when choosing PCE type. In

the recent experiment, we also find that the yield stress does not increase

with the resting time (see Figure S3), which highlights that the resting

time effect may also depend on the properties of the solid phase (the new

glass beads have been characterized by a FTIR spectra and a laser par-

ticle size analyser, see Figure S4 and S5). Therefore, the resting time

effect of PCE resulting from its size change and its interaction with the

solid phase (i.e. glass beads or cement) on the rheological performance

of cement-based materials at resting time needs to be further

investigated.

Fig. 10. The normalized aggregates amount (area to image ratios) of glass bead particles changes with resting time (The normalized value was divided by the area of

the image at resting time of 0 min).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

5. Conclusions

In this study, an inert system consisting of glass beads-PCE-artificial

pore solution was used to investigate the effect of resting time on the

yield stress, storage, and loss modulus. The surface charge properties,

adsorbed amount of PCE, PCE cluster size in artificial pore solution, and

the amount of aggregates during the resting time were measured by

various techniques, and the following conclusions can be drawn:

1) The yield stress increases with resting time when a high dosage of

PCE is used, and PCEs with higher molecular weight contribute to

higher yield stress. The storage and loss moduli from strain sweep

experiments align well with the yield stress changes over time.

2) The zeta potentials increase with resting time, suggesting a higher

surface charge density and an increased electrostatic force. The PCE

adsorption amount increases with resting time, suggesting that the

later PCE adsorption does not disperse the glass bead particle but

restrains the particle movement.

3) The amount of glass bead aggregates is found to decrease with time.

The decreased amount of aggregates and the increased zeta potential

could explain the decrease in yield stress in the sample without PCE

Fig. 11. The influencing mechanism of resting time on rheological properties of glass bead suspension mixed without PCE and with PCE.

Fig. 12. The relationship between the molecular weight, amount of polycarboxylate group (COO

), and the increase of yield stress between 0 and 60 min, and

0 and 120 min.

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

but conflicts with the increase of yield stress with time when PCE is

used.

4) The DLS results indicated that the PCE superplasticizer in artificial

pore solution has two obvious peaks, and results indicated that the

size of PCE clusters gets bigger during resting, and the signal of the

PCE cluster peak intensifies.

5) It is found that the molecular weight has a more dominant influence

on increasing the yield stress compared to the amount of carboxylate

group. The increase in yield stress during the resting could be

explained by the increase in bridging and depletion force due to the

growth of PCE cluster size. Another possible reason is that non-

adsorbed PCE may form a polymer network linked by the carbox-

ylate group and cations.

CRediT authorship contribution statement

Yanliang Ji: Writing – review & editing, Writing – original draft,

Supervision, Methodology, Investigation, Formal analysis, Data cura-

tion, Conceptualization. Shukai Wang: Software, Methodology, Inves-

tigation. Dietmar Stephan: Writing – review & editing, Writing –

original draft, Supervision, Project administration, Conceptualization.

Alexander Mezhov: Writing – review & editing, Writing – original

draft, Validation.

Declaration of Competing Interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence

the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgment

The authors gratefully thank the German Research Foundation (DFG)

for funding the Priority Program DFG SPP 2005 Priority Program “Opus

Fluidum Futurum - Rheology of reactive, multi-scale, multiphase con-

struction materials” (project number 387092747) and the group of Prof.

Xiangming Kong in Tsinghua University for providing the poly-

carboxylate superplasticizers. In addition, the authors would like to

appreciate the assistance from Halim Choo, Jessica Conrad, and Dr.

Christian Lehmann at Technische Universit¨

at Berlin for performing the

experiments in this study.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the

online version at doi:10.1016/j.colsurfa.2024.133583.

References

[1] K.H. Khayat, W. Meng, K. Vallurupalli, L. Teng, Rheological properties of ultra-

high-performance concrete- an overview, Cem. Concr. Res 124 (2019) 105828,

https://doi.org/10.1016/j.cemconres.2019.105828.

[2] N. Roussel, H. Bessaies-Bey, S. Kawashima, D. Marchon, K. Vasilic, R. Wolfs, Recent

advances on yield stress and elasticity of fresh cement-based materials, Cem.

Concr. Res. 124 (2019) 105798, https://doi.org/10.1016/j.

cemconres.2019.105798.

[3] J. Plank, E. Sakai, C.W. Miao, C. Yu, J.X. Hong, Chemical admixtures — chemistry,

applications and their impact on concrete microstructure and durability, Cem.

Concr. Res. 78 (2015) 81–99, https://doi.org/10.1016/j.cemconres.2015.05.016.

[4] D. Lowke, C. Gehlen, The zeta potential of cement and additions in cementitious

suspensions with high solid fraction, Cem. Concr. Res. 95 (2017) 195–204, https://

doi.org/10.1016/j.cemconres.2017.02.016.

[5] J. Zhang, H. Ye, X. Gao, W. Wu, Adsorption and desorption of polycarboxylate

ether superplasticizer in fresh cementitious materials blended with mineral

admixtures, J. Mater. Res. Technol. 17 (2022) 1740–1751, https://doi.org/

10.1016/j.jmrt.2022.01.145.

[6] Q. Ran, J. Liu, Y. Yang, X. Shu, J. Zhang, Y. Mao, Effect of molecular weight of

polycarboxylate superplasticizer on its dispersion, adsorption, and hydration of a

cementitious system, J. Mater. Civ. Eng. 28 (2016), https://doi.org/10.1061/

(ASCE)MT.1943-5533.0001460.

[7] P.G. Ng, C.B. Cheah, E.P. Ng, C.W. Oo, K.H. Leow, The influence of main and side

chain densities of PCE superplasticizer on engineering properties and

microstructure development of slag and fly ash ternary blended cement concrete,

Constr. Build. Mater. 242 (2020) 118103, https://doi.org/10.1016/j.

conbuildmat.2020.118103.

[8] X. Lin, Study on adsorption behavior of sustained-release polycarboxylate ether

molecules using molecular dynamics simulation, J. Phys. Conf. Ser. 2539 (2023)

012006, https://doi.org/10.1088/1742-6596/2539/1/012006.

[9] S. Zhang, P.-H. Chuang, Y. Ke, Y. FangConformation and Adsorption Properties in

Aqueous Solution for Polycarboxylate Superplasticizers with Different Main Chain

Lengths, in: 2020 International Conference on Artificial Intelligence and

Electromechanical Automation (AIEA), IEEE, 2020: pp. 670–674. https://doi.org/

10.1109/AIEA51086.2020.00149..

[10] R.J. Flatt, P. Bowen, Yodel: a yield stress model for suspensions, J. Am. Ceram. Soc.

89 (2006) 1244–1256, https://doi.org/10.1111/j.1551-2916.2005.00888.x.

[11] Y.W. Wang, H.X. Zhao, X. Shu, Y. Yang, Q.P. Ran, Modelling the equilibrium sizes

of comb-shaped MPEG-type polycarboxylate-type superplasticizers in dilute

solution and their apparent molecular weight in conventional size exclusion

chromatography, Acta Polym. Sin. (2017) 1008–1018, https://doi.org/10.11777/j.

issn1000-3304.2017.16297.

[12] S.D. Bauchkar, H.S. Chore, Effect of PCE superplasticizers on rheological and

strength properties of high strength self-consolidating concrete, Advances in

Concrete, Adv. Conc. Construc. 6 (2018) 561–583, https://doi.org/10.12989/

acc.2018.6.6.561.

[13] A.M. Ley-Hern´

andez, D. Feys, Resting time effect on the rheological behavior of

cement paste in presence of superplasticizer, Cem. Concr. Res. 142 (2021) 106347,

https://doi.org/10.1016/j.cemconres.2020.106347.

[14] Y. Ji, S. Becker, Z. Lu, A. Mezhov, R. von Klitzing, S. Wolfram, D. Stephan, Effect of

resting time on rheological properties of glass bead suspensions: depletion and

bridging force among particles, J. Am. Ceram. Soc. (2023), https://doi.org/

10.1111/jace.19469.

[15] A. Boukhsib, L. Khamar, M.S. Kadiri, K. Khallaki, L. Omari, Rheological behaviour

modelling of phosphoric acid produced from Moroccan phosphate rocks, Indian

Chem. Eng. 65 (2023) 14–24, https://doi.org/10.1080/00194506.2022.2068078.

[16] R. Costa, T. Cardoso, M. Degen, L. Silvestro, E. Rodríguez, A.P. Kirchheim,

Influence of retarder admixtures on the hydration, rheology, and compressive

strength of white Portland cements under different temperatures, CEMENT 11

(2023) 100057, https://doi.org/10.1016/j.cement.2023.100057.

[17] H. Bessaies-Bey, M. Palacios, E. Pustovgar, M. Hanafi, R. Baumann, R.J. Flatt,

N. Roussel, Non-adsorbing polymers and yield stress of cement paste: effect of

depletion forces, Cem. Concr. Res. 111 (2018) 209–217, https://doi.org/10.1016/

j.cemconres.2018.05.004.

[18] T. Hirata, P. Branicio, J. Ye, J. Zheng, Y. Tomike, A. Lange, J. Plank, M. Sullivan,

Atomistic dynamics simulation to solve conformation of model PCE

superplasticisers in water and cement pore solution, Adv. Cem. Res. 29 (2017)

418–428, https://doi.org/10.1680/jadcr.16.00137.

[19] H. Tian, X. Kong, X. Miao, L. Jiang, X. Pang, A new insight into the working

mechanism of PCE emphasizing the interaction between PCE and Ca2+in fresh

cement paste, Constr. Build. Mater. 275 (2021) 122133, https://doi.org/10.1016/

j.conbuildmat.2020.122133.

[20] M. Palacios, D. Sanz-Pont, A. Kunhi Mohamed, F. Boscaro, L. Reiter, D. Marchon,

S. Mantellato, R.J. Flatt, Heating cement to slow down its hydration: the

unexpected role of PCE interpolymer bridge formation, Cem. Concr. Res. 156

(2022), https://doi.org/10.1016/j.cemconres.2022.106765.

[21] Z. Lu, S. Liu, D. Stephan, Effects of cations on the yield stress of a highly

concentrated suspension of glass beads with the addition of polycarboxylate

Fig. 13. The relationship between the size factor of the PCE cluster, the in-

crease of adsorption amount, and the increase of yield stress (refers to the

difference in the yield stress between the values measured at 60 min (also

120 min) and 0 min).

Y. Ji et al.

Colloids and Surfaces A: Physicochemical and Engineering Aspects 688 (2024) 133583

superplasticizer, Colloids Surf. A Physicochem Eng. Asp. 575 (2019) 176–183,

https://doi.org/10.1016/j.colsurfa.2019.05.015.

[22] L. Zhang, X. Miao, X. Kong, S. Zhou, Retardation effect of PCE superplasticizers

with different architectures and their impacts on early strength of cement mortar,

Cem. Concr. Compos 104 (2019), https://doi.org/10.1016/j.

cemconcomp.2019.103369.

[23] A. Amiri, G. Øye, J. Sj¨

oblom, Influence of pH, high salinity and particle

concentration on stability and rheological properties of aqueous suspensions of

fumed silica, Colloids Surf. A Physicochem Eng. Asp. 349 (2009) 43–54, https://

doi.org/10.1016/j.colsurfa.2009.07.050.

[24] B. Wu, B.-W. Chun, L. Gu, T.L. Kuhl, Effect of Ca2+ion concentration on

adsorption of poly(carboxylate ether)-based (PCE) superplasticizer on mica,

J. Colloid Interface Sci. 527 (2018) 195–201, https://doi.org/10.1016/j.

jcis.2018.05.016.

[25] S.H. Behrens, D.G. Grier, The charge of glass and silica surfaces, J. Chem. Phys. 115

(2001) 6716–6721, https://doi.org/10.1063/1.1404988.

[26] P. Leroy, C. Tournassat, O. Bernard, N. Devau, M. Azaroual, The electrophoretic

mobility of montmorillonite. Zeta potential and surface conductivity effects,

J. Colloid Interface Sci. 451 (2015) 21–39, https://doi.org/10.1016/j.

jcis.2015.03.047.

[27] A. Chougnet, T. Palermo, A. Audibert, M. Moan, Rheological behaviour of cement

and silica suspensions: particle aggregation modelling, Cem. Concr. Res. 38 (2008)

1297–1301, https://doi.org/10.1016/j.cemconres.2008.07.001.

[28] C.G. Robertson, S.R. Vaikuntam, G. Heinrich, A nonequilibrium model for particle

networking/jamming and time-dependent dynamic rheology of filled polymers,

Polymers 12 (2020) 190, https://doi.org/10.3390/polym12010190.

[29] R.D. Ferron, S. Shah, E. Fuente, C. Negro, Aggregation and breakage kinetics of

fresh cement paste, Cem. Concr. Res. 50 (2013) 1–10, https://doi.org/10.1016/j.

cemconres.2013.03.002.

[30] P. Claisse, J.P. Lorimer, M.H. Omari, Workability of cement pastes Workability of

Cement Pastes P A Claisse, 2001. 〈http://www.claisse.info/Publish.htm〉.

[31] H. Bazin, A. Bouchu, G. Descotes, M. Petit-Ramel, Comparison of calcium

complexation of some carboxylic acids derived from D-glucose and D-fructose, Can.

J. Chem. 73 (1995) 1338–1347, https://doi.org/10.1139/v95-165.

[32] Alex Lange, Tsuyoshi Hirata, Johann Plank, The role of non-adsorbed PCE

molecules in cement dispersion: experimental evidence for a new dispersion

mechanism. SP-288: Tenth International Conference on Superplasticizers and

Other Chemical Admixtures, American Concrete Institute, 2012, https://doi.org/

10.14359/51684246.

[33] H. Huang, E. Ruckenstein, The bridging force between colloidal particles in a

polyelectrolyte solution, Langmuir 28 (2012) 16300–16305, https://doi.org/

10.1021/la303918p.

[34] Y. Ma, C. Shi, L. Lei, S. Sha, B. Zhou, Y. Liu, Y. Xiao, Research progress on

polycarboxylate based superplasticizers with tolerance to clays - a review, Constr.

Build. Mater. 255 (2020) 119386, https://doi.org/10.1016/j.

conbuildmat.2020.119386.

[35] L. Ferrari, J. Kaufmann, F. Winnefeld, J. Plank, Interaction of cement model

systems with superplasticizers investigated by atomic force microscopy, zeta

potential, and adsorption measurements, J. Colloid Interface Sci. 347 (2010)

15–24, https://doi.org/10.1016/j.jcis.2010.03.005.

[36] R. Mandal, S. Kumar Panda, S. Nayak, Rheology of self-compacting concrete: a

critical review and future perspective, Mater. Today Proc. (2023), https://doi.org/

10.1016/j.matpr.2023.07.180.

[37] D. Lowke, Thixotropy of SCC—a model describing the effect of particle packing and

superplasticizer adsorption on thixotropic structural build-up of the mortar phase

based on interparticle interactions, Cem. Concr. Res. 104 (2018) 94–104, https://

doi.org/10.1016/j.cemconres.2017.11.004.

Y. Ji et al.