Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes

Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes

Cement & Concrete Composites xxx (2014) xxx–xxx Contents lists available at ScienceDirect Cement & Concrete Composites journal homepage: www.elsevie...

409KB Sizes 1 Downloads 47 Views

Cement & Concrete Composites xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Cement & Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp

Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes Hela Bessaies Bey a,⇑, Julie Hot a, Robert Baumann b, Nicolas Roussel a a b

East Paris University, The French Institute of Science and Technology for Transport, Development and Networks, France DOW Construction Chemicals, Switzerland

a r t i c l e

i n f o

Article history: Received 11 December 2013 Received in revised form 29 April 2014 Accepted 9 May 2014 Available online xxxx Keywords: Superplasticizer Viscosity agent Cement paste Competitive adsorption Rheology Self Compacting Concrete

a b s t r a c t Self Compacting Concretes (SCC) are characterized by their high fluidity so they can be placed in sections with congested reinforcements and in restricted areas without vibration. Furthermore, SCC cement paste has to be viscous enough to avoid segregation and to maintain the stability of the suspension until the onset of hardening. To fulfil these rheological requirements, mix design engineers combine use of superplasticizers and viscosity agents. The mechanism of action of these chemical admixtures is very sensitive to their adsorption. The blending of these polymers generates a competitive adsorption on surface sites of cement particles, which influences their performances. For a better understanding of competitive adsorption, we measure here both the amount of adsorbed polymers and its consequences on the rheological behaviour of the system in terms of yield stress and plastic viscosity. Our results suggest that the competitive adsorption prevents some of the polymer molecules from adsorbing, thereby moderating the performances of adsorbing polymers and enhancing the effects of polymers potentially in solution on the rheological properties of cement paste. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Polymers are fundamental ingredients for formulating a selfcompacting concrete. Understanding the mechanism of action of these polymers is a key in selecting effective admixtures for various practical applications. A vast literature has studied the effect of these macromolecules on the rheology of cement pastes. These polymers either stay in the interstitial fluid, modifying its rheological behaviour, or adsorb at the surface of cement grains, modifying their colloidal interactions [1,2]. Most of these studies, however, have focused on individual polymers. In order to reach the rheological target requirements for self-compacting concretes, mix design engineers have to mix various organic molecules. When the polymers added to the system are both adsorbing species, a competitive adsorption may occur. The outcome of this adsorption competition between molecules shall dictate the rheological behaviour of the SCC. In this work, we studied the competitive adsorption of a so-called viscosity agent and of a plasticizer molecule. We measured both the amount of adsorbed polymers and its consequences on the rheological behaviour of the system in terms of yield stress and plastic viscosity. ⇑ Corresponding author. Tel.: +33 181668211. E-mail addresses: [email protected] (H.B. Bey), [email protected] (J. Hot), [email protected] (R. Baumann), [email protected] (N. Roussel).

Our results show that competitive adsorption prevents some of the polymer molecules from adsorbing. This process moderates the performances of adsorbing polymers and enhances the effects of polymers remaining in solution on the rheological behaviour of cement pastes. 2. Materials and procedures 2.1. Materials and mixing procedures The cement used in this study is a Portland cement CEM I 52 PMES CP2 (SAINT VIGOR LAFARGE). The polymers investigated here belong to two different families of polymers: a superplasticizer (polycarboxylate) referred as PCE and a cellulose derived viscosity enhancing agent (methyl-hydroxy-ethyl cellulose) referred as CE. They were used in liquid form. The preparation procedures of cement pastes for both rheological and adsorption measurements were identical. Water and cement were first homogenized by hand before a one minute high speed mixing phase using Turbo test Rayneri VMI mixer at 840 rpm. The cement paste was left at rest for 18 min before polymer addition in order to allow for the nucleation of the first hydration products without any interference with the organic molecules [3]. Polymer addition was followed by a 1 min high speed mixing phase. The mixture was then finally mixed at low speed for 18 min.

http://dx.doi.org/10.1016/j.cemconcomp.2014.05.002 0958-9465/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Bey HB et al. Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes. Cement Concrete Comp (2014), http://dx.doi.org/10.1016/j.cemconcomp.2014.05.002

2

H.B. Bey et al. / Cement & Concrete Composites xxx (2014) xxx–xxx

2.2. Adsorption measurements

Table 1 Langmuir parameters for adsorption of PCE and CE.

After the mixing protocol described above, the cement pastes were centrifuged at 1000 g during 10 min in order to extract the interstitial fluid. The carbon content was determined using a Total Organic Carbon analyser (TOC) manufactured by Shimadzu. The results were corrected for the carbon contained in cement powder mostly because of the presence of grinding aids. The quantity of polymer adsorbed was then computed from the difference between the carbon content in the interstitial solution extracted by centrifugation solution before and after contact with cement [4]. 2.3. Rheological measurements Rheological experiments were conducted using a Bohlin C-VOR shear rheometer equipped with a Vane geometry [5]. The Vane tool diameter is 25 mm, the outer cup diameter is 50 mm whereas its depth is 60 mm. The cup of the rheometer was filled with the tested cement paste and the measurement sequence was started. In order to bring the cement paste to a reference structural state, it was first pre-sheared at a shear stress equal to 90 Pa during 120 s. A decreasing shear rate was then directly applied from 100 s1 to 1 s1 (with a logarithmic distribution of shear rates) during 200 s. 3. Experimental results 3.1. Single adsorption isotherms of CE and PCE on cement The measured adsorption isotherms for both polymers are reported in Fig. 1. Both adsorption equilibrium data are well fitted by the Langmuir representation.



Cm K L C e 1 þ K LCe

where C (g/kg of cement) is the adsorbed polymer amount per mass of cement, Cm is the maximum adsorbed, KL (L/g) is the Langmuir constant and Ce (g/L) is the concentration of polymer remaining in the suspending fluid. Determination of the maximum adsorbed amount Um and the Langmuir constant KL for these polymers on cement particles is carried out for comparison purpose. These fitting parameters are given in Table 1. The plateau value for PCE is around three times

Polymer

KL (L/g)

Um (g/kg)

PCE CE

2.5 1.24

0.7 0.24

higher than CE. The Langmuir constant of CE is around two times lower than the Langmuir constant of PCE. 3.2. Competitive adsorption between CE and PCE Adsorption of PCE and CE when simultaneously added to the cementitious system are measured and compared to their individual adsorptions in Fig. 2, in which the ratio between dosage of PCE and CE is constant (PCE/CE = 0.3). As the TOC technique does not allow for a distinction between the two polymers, the total amounts of adsorbed polymer and polymer remaining in solution are expressed in Carbon/kg of cement and Carbon/L respectively. We also plot in Fig. 2 the resulting sum of the polymers individual adsorption. The abscissa or ordinate of this curve, for a given amount of PCE and CE, are respectively computed by simply summing the abscissa or ordinate of both PCE and CE single adsorption. It can be noted that the measured total adsorption is systematically lower than the sum of the individual adsorptions, indicating that competitive adsorption seems to prevent some of the polymer molecules from adsorbing. This reduction in adsorption seems to increase when the amount of adsorbed species becomes closer to the saturation plateau. 3.3. Rheological consequences In order to get additional information on the competitive adsorption between the two polymers, we now measure the rheological changes in the system due to a change in polymer proportions. The apparent viscosity is plotted as a function of shear rate for some of the tested cement pastes (W/C = 0.4) in Fig. 3. Each flow curve is fitted with a Bingham model in order to compute the value of yield stress and plastic viscosity. The dependence of yield stress and plastic viscosity on dosage and type of admixture is shown in Figs. 4a and 4b. We observe, as expected, a decrease in plastic viscosity and yield stress with the dosage of PCE. The addition of CE, however, increases both yield stress and plastic viscosity. The effects of these two polymers on cement paste rheology can be explained by their mechanism of action from literature. On one

0,8

0,5 0,4

PCE CE Langmuir fitting PCE Langmuir fitting CE

0,3 0,2 0,1 0 0

0,5

1

1,5

2

2,5

3

3,5

Ce (g/L) Fig. 1. Adsorption isotherms of PCE and CE on cement paste.

4

Carbon adsorbed (/Kg cement)

0,6

Γ (g/Kg)

CE single adsorption PCE single adsorption PCE+CE adsorption Sum of PCE and CE single adsorption

1400

0,7

1200 1000 800 600 400 200 0 0

1000

2000

3000

4000

5000

6000

Carbon remaining in solution (/L water) Fig. 2. Measured and computed adsorption of PCE and CE.

Please cite this article in press as: Bey HB et al. Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes. Cement Concrete Comp (2014), http://dx.doi.org/10.1016/j.cemconcomp.2014.05.002

3

H.B. Bey et al. / Cement & Concrete Composites xxx (2014) xxx–xxx

0,4

10

Plastic viscosity (Pa.s)

Apparent viscosity (Pa.s)

0,35

1

0,3 0,25 0,2 0,15 CE

0,1

PCE

0,05 Cement Cement+0,5g PCE/L water Cement+1g PCE/L water

0

Cement+0,5g CE/L water Cement+1g CE/L water Cement+1,5g CE/L water

0

0,5

100

Shear rate (s-1)

Fig. 3. Viscosity versus shear rate for various dosages of CE and PCE (W to C ratio = 0.4).

hand, adsorbed PCE on cement particles provides steric hindrance that decreases the intensity of attractive colloidal forces between particles, decreasing therefore the yield stress [6,7] and increasing the fluidity of the corresponding concrete. By decreasing yield stress, PCE also changes the flocculation state of cement pastes. Plastic viscosity decreases as flocculation modifies the way shear concentrates in the fluid layers between flocs or cement particles [8]. On the other hand, CE adsorbs on several cement particles at the same time. It increases the yield stress through polymer bridging [1]. As CE adsorbs only slightly at the surface of cement particles, a significant amount of polymer remains in solution increasing the viscosity of the interstitial fluid and thus increasing the plastic viscosity of the cement paste [1,9]. The decrease in mobility of the interstitial fluid allows for an improvement of the paste stability and its resistance to bleeding. For the following and for simplicity sake, we assume that the effect of PCE and CE mixture on the rheology of cement paste can be seen as the sum of the contribution of each adsorbed polymer fraction. This assumption makes it possible to extrapolate the yield stress and plastic viscosity of a cement paste containing a mixture of PCE and CE, which would not compete to adsorb at the surface of cement particles. Calculation of the expected yield stress and plastic viscosity is done as follow:

FðA%PCE þ B%CEÞ ¼ Fðreference:cementÞ  

1,5

2

Dosage (g polymer/L water)

0,1 10

1

FðA%PCEÞ Fðreference:cementÞ

FðB%CEÞ Fðreference:cementÞ

Fig. 4b. Effect of CE and PCE on the plastic viscosity at different dosages.

Table 2 Polymer mixtures. Mixture number

%PCE (by weight of water)

%CE (by weight of water)

1 2 3

0.25 0.25 1

0.5 1 1.5

where F is yield stress or plastic viscosity, A% and B% are respectively the dosages of PCE an CE. F(reference.cement) is the yield stress or plastic viscosity of the reference paste without any polymers. We measured the yield stress and plastic viscosity of three cement pastes containing different mixtures of PCE and CE, as indicated in Table 2, and we compare the measured and the computed values in Figs. 5a and 5b. We note that the expected yield stress and plastic viscosity are different from the measured values. These differences suggest once again that the relative fractions of each adsorbed polymer change when they are mixed together. Fig. 5a shows that the measured yield stress is lower than the expected yield stress for mixture 1 and 2. Assuming that PCE adsorption in presence of CE cannot be higher than PCE adsorption when alone in the system, this suggests that the decrease in measured yield stress cannot come from an increased amount of adsorbed PCE but shall come from a decreased amount of adsorbed CE and therefore a decrease in the bridging effect. Fig. 5b shows that the measured plastic viscosity is higher than the expected plastic viscosity. This variation could be explained by 30

30 25

Yield stress (Pa)

Yield stress (Pa)

25 20 15 10 CE 5

20 15 10 Expected (if no competitive adsorption)

5

Measured

PCE 0 1

0

0

0,5

1

1,5

2

2

3

Mixture number

Dosage(g polymer/L water) Fig. 4a. Effect of CE and PCE on the yield stress at different dosages.

Fig. 5a. Comparison between expected and measured yield stress at different dosages of PCE and CE.

Please cite this article in press as: Bey HB et al. Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes. Cement Concrete Comp (2014), http://dx.doi.org/10.1016/j.cemconcomp.2014.05.002

4

H.B. Bey et al. / Cement & Concrete Composites xxx (2014) xxx–xxx

suggest that the competitive adsorption prevents some of the polymer molecules from adsorbing, thus potentially moderating the performances of adsorbing polymers such as PCE and enhancing the effects of polymers remaining in solution such as CE on the rheological properties of cement paste. The knowledge gained allows for a better prediction of the rheological consequence of competitive adsorption. However, there is still a need for a deeper understanding of the microscopic behaviour of polymers competing to adsorb at the surface of cement particles. The future goal of our on-going work is to evaluate in a more quantitative manner the adsorbed amount of each polymer and to determine the key parameters dictating the behaviour of admixture in combination.

0,4

Plastic Viscosity (Pa.s)

0,35 0,3 0,25 0,2 0,15 0,1 Expected (if no competitive adsorption)

0,05

Measured 0 1

2

3

Mixture number Fig. 5b. Comparison between expected and measured plastic viscosity at different dosages of PCE and CE.

the same above effect. The desorption of the CE would indeed lead, along with a lower yield stress, to a higher concentration of CE in the interstitial fluid and therefore to a higher viscosity. Simultaneously, a decrease in PCE adsorption because of the competition with CE could also lead to an increase in viscosity through an increase of the flocculation state of the system. These two trends suggest that PCE is able to desorb some of the CE molecules. 4. Conclusion In this work, the competitive adsorption of a PCE and a CE were studied. The amounts of adsorbed polymers and their consequences on the rheological behaviour of the system in terms of yield stress and plastic viscosity were measured. Our results

References [1] Brumaud C. Origines microscopiques des consequences rhéologiques de l’ajout d’éthers de cellulose dans une suspension cimentaire. PhD thesis, Université Paris-Est, Paris, France; 2011. [2] Roussel N, Lemaître A, Flatt RJ, Coussot P. Steady state flow of cement suspensions: a micromechanical state of the art. Cem Concr Res 2010;40:77–84. [3] Flatt RJ, Houst YF. A simplified view on chemical effects perturbing the action of superplasticizers. Cem Concr Res 2001;31:1169–76. [4] Perrot A, Lecompte T, Khelifi H, Brumaud C, Hot J, Roussel N. Yield stress and bleeding of fresh cement pastes. Cem Concr Res 2012;42:937–44. [5] Nguyen QD, Boger DV. Direct yield stress measurement with the Vane method. J Rheol 1985;29(3):335–47. [6] Houst Yves F, Bowen Paul, Perche François, Kauppi Annika, Borget Pascal, Galmiche Laurent, et al. Design and function of novel superplasticizers for more durable high performance concrete (superplast project). Cem Concr Res 2008; 38:1197–209. [7] Flatt RJ, Schober I, Raphael E, Plassard C, Lesniewska E. Conformation of adsorbed. Comb copolymer dispersants. Langmuir 2009;25(2):845–55. [8] Hot H, Roussel N. Influence of adsorbing polymers on the macroscopic viscosity of concentrated cement pastes. In: Malhotra VM, editor. Proceedings tenth international conference on superplasticisers and other chemical admixtures in concrete, Prague, ACI, SP 288; 2012. p. 223–33. [9] Khayat KH. Viscosity-enhancing admixtures for cement-based materials. Cem Concr Res 1998;20:171–88.

Please cite this article in press as: Bey HB et al. Consequences of competitive adsorption between polymers on the rheological behaviour of cement pastes. Cement Concrete Comp (2014), http://dx.doi.org/10.1016/j.cemconcomp.2014.05.002