CHEN90022- Chemical Engineering Research Project

RESEARCH PROJECT REPORT

Project: Synthesising Carboxylic Superplasticizer for Geopolymer

Name: Ky Phong Ngo Student number: 286362 Supervisors: Jiangtao Xu, John L. Provis, Greg G. Qiao

Research Project Report 2011

Executive summary
Geopolymer material is considered as the most practical solution for CO2 reduction in cement production. However, it is a relatively new material of which the behaviour is not fully understood. In addition, compared to traditional ordinary Portland cement, geopolymer has totally different structure and hydration mechanism. As a result, commercial admixtures such as superplasticizer do not have any effect on geopolymer. This necessitates new polymers to be synthesized. During two months of this research project, one linear and five graft polycarboxylic superplasticizers were successfully synthesized. As linear polymer was found to degrade back to its original form, it was concluded that the ring opening reaction is not suitable for synthesizing the linear polymers. The remaining five graft polymers have similar structure that contains long Poly Ethylene Oxide (PEO) side chains and charged polyoxyethylene acrylate (PA) side chains. By varying the side chain length, charge, charge density and molecular weight, the effect of polymers structure on their performance can be investigated Comparing the performance of the remaining five structures, it was found that carboxylic with PEO side chain outperforms polymers with shorter PEO side chain. This can be attributed for increasing in steric hindrance effect. Unlike in normal Ordinary Portland cement (OPC) where high molecular weight polymers perform better, in geopolymer cement, too high molecular weight polymers performs worse than low molecular weight. It is proposed that large polymer chain would have the tendency to form large spherical structure that can possibly trap the charge side chains within and render their effects useless. In addition, polymers with too low molecular weight do not give PA side chain enough protection to prevent it from being neutralized in geopolymer cement solution with high PH conditions. It is suggested that molecular weight of an ideal polymer would be at mid range. Up until now, molecular weight of 14 000 give the best result. At the present, trends on how the structure affects performance of polymers are not clear. It is recommended that more polymers should be synthesized to obtain clearer trends.

Research Project Report 2011

Table of Contents

Executive summary ................................................................................................................................. 2 1. Introduction ........................................................................................................................................ 4 2. Theory ................................................................................................................................................. 5 2.1. Free Radical Polymerization......................................................................................................... 5 2.2. Raft Polymerization ...................................................................................................................... 6 3. Literature Review ................................................................................................................................ 7 3.1. Geopolymer ................................................................................................................................. 7 3.1.1. Mechanism of Alkali Activation ............................................................................................ 8 3.2. Superplasticizer ............................................................................................................................ 9 3.2.1. Working Mechanism of Superplasticizers ............................................................................. 9 3.3. Carboxylic superplasticizer......................................................................................................... 10 4. Material and Method ........................................................................................................................ 12 4.1. Precipitation ............................................................................................................................... 12 4.2. Proton nuclear molecular resonance ......................................................................................... 12 4.3. Gel Permeation Chromatography GPC ...................................................................................... 13 5. Procedure and result of synthesizing polymers ................................................................................ 14 5.1. Synthesising monomer PEG 2K MA........................................................................................... 14 5.2. Polymers synthesizing ............................................................................................................... 17 5.2.1. Synthesizing C50 (43)-1:1 ................................................................................................... 19 5.2.2. Synthesising C14 (43)-1:5 ................................................................................................... 22 5.2.3. Synthesizing linear structure C5 (110)-1:1 Linear .............................................................. 25 5.2.4. Synthesising C8 (22)-1:5 ..................................................................................................... 27 5.2.5. Synthesising A6 (22)-1:5...................................................................................................... 29 5.2.6. Synthesising A42 (22)-1:5................................................................................................... 31 6. Rheology Test Result and Discussion ................................................................................................ 34 7. Future work ...................................................................................................................................... 36 8. Conclusion ......................................................................................................................................... 36

Research Project Report 2011

1. Introduction
Concrete is the most commonly used constructional material. Traditionally, concrete is made using the Ordinary Portland Cement (OPC) as binders. However, OPC production is energy intensive and involves high level of carbon dioxide. Therefore, many studies and researches have been done to mitigate the effect of cement industry, among which, geopolymer cement is the most promised solution. Geopolymer is commonly formed by alkali activation of industrial aluminosilicate waste material such as coal ash and blast furnace slag. Nevertheless, compared to traditional OPC, which has hundred years under research and study, the structure and behaviour of geo-polymer is not fully understood. As a result, in order to utilize geopolymer to its full potential, many researches and studies are required. Superplasticizer is the most important admixture that was discovered in concrete industry in last decades. It refers to a wide range of synthetic water soluble polymers such as sulfonated melamine formaldehyde (SMF) condensate, sulfonated naphthalene formaldehyde (SNF) condensate and modified sugar-free lignosulfonate (MLS) and newly developed polycarboxylate (PC) based polymers. Superplasticizer (SP) has several effect on normal OPC including increasing strength and durability by water reducing, improve workability of cement, reduce both water and cement content in cement [1]. However, it is found that traditional superplasticizers have no effect on geopolymer. There are no published literatures that showed any success in synthesizing a new type of superplasticizer for geopolymer. Therefore, it is within the interest of Zeobond, leading producer in geopolymer cement, to develop the first commercial superplasticizer for geopolymer. Funded by Zeobond and University of Melbourne, the only aim of this project is to synthesize a new type of polymer that has enhancement effects on geopolymer cement. New living radical polymerization technique, RAFT, was utilized in this research to tailor polymers with distinguished characteristics that believed to have positive effects on the geopolymer cement. These polymers were then analysed using proton nuclear molecular resonance spectroscopy (1HNMR) and gel permeate chromatography (GPC) to confirm the structure and properties. Finally, different geopolymer slags were used for rheological test to study the effect of the synthesized polymer. Through this project, it is hoped that the structure and behaviour of geopolymer material can be further understood. In addition, mechanism of how polymer and geo-material particles interact can be predicted in this research project and this will set a foundation for any future projects in the field.

Research Project Report 2011

2. Theory
There are mainly two types of polymerization mechanism; chain polymerization and step polymerization. These two techniques have distinct mechanism which results in different properties in the polymer formed. The most significant difference is that high molecular weight polymers are formed immediately in chain polymerizations [2]. Once free radical anionic or cationic centre is formed, monomer will be added rapidly to into the chain to form high molecular weight polymer. As a result, the reaction mixture contains only monomer, high polymers and the growing chains. During the polymerization, the molecular weight of the polymer is unchanged, although the overall conversion increases proportional with reaction time. In the other hand, in step polymerization, any two molecules in the mixture can react with each other to form dimers, trimers, and so on. This leads to faster disappearance of monomer species. However, the high molecular weight polymer is not obtained until the end of the polymerization. Therefore, step reaction polymerization requires high reaction time to achieve high molecular weights and high conversion rate. Between these two polymerizations, chain polymerization especially free radical is one of the most widely used in the processes for commercial production of high molecular weight polymer for several reasons [3]. Firstly, it can be used with a large variety of monomers including (meth)acrylates, styene, (meth)acrylamides, butadiene and vinyl acetate. The second reason is that free radical polymerization can be used for a wide range of functional group and reaction condition such as bulk, solution, emulsion, miniemulsion, suspension. Finally, compared to other similar technologies, it is inexpensive and relatively simple to be carried out. However, conventional process of radical polymerization does not allow the degree of control over composition, weight distribution and structure of the polymer [4]. Living radical polymerization, first discovered by Szwarc in 1950, offers polymers of controlled composition, architecture and molecular weight distribution while retaining the versatility of conventional radical polymerization [2]. In this research, RAFT method, a relatively new method of living radical polymerization, is utilized to architect brush-liked carboxylic polymer. In order to fully understand RAFT method, the conventional process of radical polymerization is also included in this section.

2.1. Free Radical Polymerization

The mechanism of free radical polymerization usually consists of three steps, namely, initiation, propergation and termination [5]. As shown in Figure 1, initiation step involves a reactive species R* which produced from initiator. It adds to a monomer molecule by opening -bond to form a new radical centre. This process is repeated as many monomers are added to the propagating chain. The polymerization process terminates when reactive centres are irreversible destroyed by coupling or dispropornation. Coupling or disproportionation referred to self-reactions of two reactive centres which lead to dead polymers with no ability to reinitiate the polymerization.

Research Project Report 2011

Figure 1: Chain polymerization

2.2. Raft Polymerization

Living radical polymerization is a new technique which allows better control over the synthesized polymer. It offers control over composition, architecture and molecular weight distribution. Therefore, it has been a useful tool to synthesize narrow dispersity end-funtional polymers, synthesizing high purity block copolymer, stars and other polymers with complex structure[4]. In an ideal living polymerization, all chains are initiated at the beginning, propagating at the same rate and no termination occurs. This cannot be achieved by free living radical which involves radical-radical termination. Living polymerization, irreversible terminations are suppressed or eliminated by reversible deactivation or reversible chain transfer. There are three techniques in living radical polymerization, which are nitroxide-mediated polymerization (NMP), atom transfer radical polymerization (ATRP) and reversible addition fragmentation chain transfer (RAFT). In this project, RAFT method was used. RAFT polymerization involves the use of a chain transfer agent. Basically, RAFT is a normal free radical chain with the presence of chain transfer agent to prolong the life of propagating species. The mechanism of RAFT polymerization is depicted in Figure 2 in sequence of addition, fragmentation equilibria. In the 1st step, the initiation of the polymerization process occur same as in conventional radical polymerization where initiator gives rise to radical I* and react with monomer to form propagating radicals (Pm*). The propagating radicals will then react with the chain transfer agent to form polymeric thiocarbonylthio compound [PnS(Z)C=S] and a new radical R*. The reaction of this new radical with monomer forms a new propagating radical (Pm*). Equilibrium between the active propagating radical and the dormant polymeric thicarbonylthio compounds allows equal propability for all chain to grow and allows for the production with narrow ranged molecular weight polymer. When the polymerization is complete, most of chains retain the thiocarbonylthio end group and can be isolated as stable materials.

Research Project Report 2011

Figure 2: mechanism if RAFT polymerization [6]

RAFT polymerization can be easily performed simply by adding a suitable chain transfer agent to a conventional free radical polymerization. As a result, it is a simple solution to control the architecture and molecular weight distribution of the synthesized polymers. By using RAFT polymerization, narrow weight distributions can be achieved. In a ideal RAFT polymerization, molecular weights of the synthesized can be predicted using equation below ([ ] [ ])

Where Mo is the number of mole of the initial monomer and Mt is the molar mass of the monomer after the polymerization.

3. Literature Review
3.1. Geopolymer
Geopolymer refers to a synthetic alkali aluminosilicate, a material produced from reaction between solid aluminosilicate and highly concentrated aqueous alkali hydroxide or silicate solution. In some literatures, it is sometimes referred to as inorganic polymer [7]. Compared to ordinary Portland binder which is attributed for 5% global man-made emission [8], geopolymer exhibits a comparable performance with much lower carbon emission. Therefore, it has emerged as an 7

Research Project Report 2011

alternative to ordinary Portland binder. Though there have been many researches and studies, the structure of geopolymer and its mechanism of alkali activation is not fully understood. This section outlines the most up to date mechanism of alkali activation.

3.1.1. Mechanism of Alkali Activation The mechanism of alkali activation is complicated and not fully understood. Glukhovsky proposed the first model in 1950s. According to his proposal, alkali activation process consists of three main steps: an initial phase of silica dissolution, followed by the phases of transportation and finally polycondensation [7]. The first step of activation process involves dissolution of solid aluminosilicate. This results in breaking down of covalent bonds of Si-O-Si and Al-O-Si in high PH around 11 and lead to liberation of aluminate and silicate species. These species are then transported into aqueous phase in which may already contain silicate depending on the activation solution. The accumulation of these species in solution leads to complicated interactions between themselves and forms a large gel network. The network is cross-linked of SiO4 and AlO4 tetrahedrons united together by oxygen bridges [9]. The time required for saturated aluminosilicate to form gel network is highly dependent on the raw material, synthesis conditions. The gel network will keep rearranging and reorganizing to increase the number of connectivity and result in a three dimension aluminosilicate network. Physically, it is observed that this process results in a condense structure of geopolymer. The structure of geopolymer is shown in Figure 3.

Figure 3: Geopolymer proposed structure

Research Project Report 2011

3.2. Superplasticizer
Superplasticizer (SP) is the most important admixture that was discovered in concrete industry in last decades. It refers to a wide range of synthetic water soluble polymers including sulfonated melamine formaldehyde (SMF) condensate, sulfonated naphthalene formaldehyde (SNF) condensate, modified sugar-free lignosulfonate (MLS) and newly developed polycarboxylate (PC) based polymers [10]. Superplasticizers (SP), also called high-range water-reducers, allow same degree of workability at up to 30% lower water/cement (w/c) ratio. Also, due to this property, adding SP at same (w/c) can be utilized to produce flowing concrete which is self-levelling and can be placed by pumping or gravity feed through vertical pipe. Compared to normal water reducer, SP can be added at higher concentration without causing excessive retardation or air entrainment [11]. 3.2.1. Working Mechanism of Superplasticizers

SPs increase the fluidity of the cement mixture by improving the dispersion of cement particles in water hence decrease or prevent flocculation. This effect is attributed for the electrostatic properties of the polymer. An example of SNF superplasticizer is illustrated in Figure 4 below. SNF polymers neutralize the electrostatic attraction, which causes agglomeration of cement particles. In order to achieve this effect, the structure contains a side chain with negative charge to create an electrostatic repulsion while sulphonic group responsible for the adsorption into the particles surface. Zeta potential was used as a measure for the effectiveness of superplasticizers [10, 12]. However, researches on different type of superplasticizer polymer have proven that zeta potential is not the only factor which influents the performance of superplasticizers.

Figure 4: charge neutralizing effect

Polycarboxylate based superplasticizer has negligible zeta potential change effect and yet still be able to prevent cement particles to flocculate. In 1996, Uchikawa et.al have first proved existence of steric repulsive force by measuring the interactive forces between cement clinker particles in present of different type of admixtures [13].As shown in Figure 5, the mechanism of action involves adsorption of a charged side group into the surface of cement particles and forms an electric double 9

Research Project Report 2011

layer which creates steric hindrance effect. As the result, cement particles cannot come to contact with each other hence prevents the flocculation.

Figure 5: Steric hindrance effect

3.3. Carboxylic superplasticizer

Studies of carboxylic superplasticizer were reported in many literatures [12, 14-16]. However, due to high possibility of its commercialization, there is limited number of literatures that describe in detail the process of synthesizing superplasticizer. Fundamentally, there are five categories of carboxylic superplasticizer : polyacrylic acid (salt or ester), polymaleic acid (anhydride), polyethylene glycol, acrylamid, methallyl-sulfonic (salt or ester)[17, 18]. Major researches that has been done in the field recently are the study of Kazuo Yamada et el. on series of comb structure polyacrylic acid acid-based superplasticizer [19], study of Chong-Zhi et al on block and graft structure of polyethelene oxide chain superplasticizer [12], study of Heon-Young Cho et al on superplasticizer made from Ethylene glycol and methyl ether [15] and study of Tung-Sheng Liao et al on the effects of methyl acrylate-sulfonic acid copolymers structure of on properties [16]. However, these studies have been done for OPC, not geopolymer. Therefore, only two studies with similar structure to this research are reviewed in this section as a mean of comparison. Kazuo Yamada researched on the superplasticizer with comb-liked structure shown in Figure 6 [14]. By varying the ratio of a: b: c at different molecular weight and side chain length m, several conclusions were made. Longer PEO side chain gave more fluidity at same dosage, shorter main back 10

Research Project Report 2011

bond chain length give more fluidity at the same dosage and slightly longer settling time, did not affect the loss of fluidity with time very much, higher sulfonic group content give highly fluidity, settling time of cement paste depended roughly on the ionic functional group concentration in the aqueous phase of the cement paste. However, best ratio of the ionic group and PEO side chain is not given in article.

Figure 6: Kazuo Yamada studied superplasticizer species

A similar research was done by Ma Xiuxing [20]. In his paper, superplasticizers with chemical shown in Figure 7 are studied. The paper showed that the longer XPEG side chain, the more effect the polymers have on cement. This can be explained by the increasing in steric hindrance effect the polymer has in cement solution. However, his study showed that too long side chain lead to decrease in fluidity conversion ability of the cement. The best ratio reported in his paper is XPEGMAnMAS = 241.5 with 1600 XPEG molecular weight. Effectively, the ratio between charge and long side chain is approximately 1: 5.

Figure 7: Ma Xuixing studied superplasticizers species

It is noted that there have been no papers that reported superplasticizers specially made for geopolymer. As the mechanism of hydration in conventional OPC and geopolymer are different, it is expected that the effects of superplasticizers on geopolymer are different compared to OPC. Currently, in the market, there is no superplasticizer that is specially made for geopolymer cement. Therefore, the main objective of the experiment described below is synthesising a new type of carboxylic superplasticizer to accommodate the growing need for Green concrete.

11

Research Project Report 2011

4. Material and Method
4.1. Precipitation
Precipitation is a common purifying method; it is sometime referred to as salting out method. This method is used to crystallize the solute by altering its solubilities. The solute is crystallized or sometime called crush out by the addition of a second solvent (anti solvent) in which the solute is relatively insoluble. This anti solvent is immiscible with the primary solvent hence result a decreasing solubility of the solute in the mixture [21]. One main advantage of this method is that it can be carried out in ambient temperature. This allows the process to be carried out conveniently and economically. In addition, this method can be used for heat sensitive product therefore is used intensively in pharmaceutical purification process. However, the main disadvantage of this method in large scale production is the need to separate the binary solvent for recycling; this will add a significant amount to the involved cost. In the lab scale, depend on the properties of the product and amount of anti-solvent added, using this method for purification can result in loss of final product. As the solubility of the solute is not absolute, that is if too much anti-solvent is added, an amount of solute can solubilize in anti-solvent and do not crystallized, therefore, cause a loss in final product. Also it is suggested by Karpinski and Nyyt in 1983 that the quality of the final product can be improve by a programmed addition rate of anti-solvent, proportional to the instantaneous crystal surface area [21] . In this research, the primary solvent can be DCM, THF or dioxane. The anti-solvent used is Diethyl Ether. The antisolvent was chosen from a small scale test where the product solvent is added to the anti-solvent. It was found that all the product solutions in purifying process can be purified using diethyl Ether. The best ratio of anti-solvent/product solution is found to be 10. Therefore, in order to reduce the scale of equipment and the amount of anti-solvent used in precipitation process, the primary solvent in the production solution were usually reduced to a less than 40 ml by evaporation in a rotavapor.

4.2. Proton nuclear molecular resonance

In this research, proton nuclear molecular resonance (1HNMR) can be used to confirm the structure of synthesized material. Proton NMR is an applicant of NMR spectroscopy. This technology allows detection of hydrogen-1 nuclei in the molecules of the substances to determine the structure of the molecule. For sole purpose of understanding the result, detail on Nuclear Molecular Resonance spectroscopy and how it works are not discussed in great detail in this report but rather an outline on which type of result will be generate using this method of analysing [22]. A nucleus of 1H (proton) behaves as though it is a magnet spinning around an axis. When a compound containing proton is placed in a very strong magnetic field and simultaneously stimulated under electron magnetic energy at an appropriate frequency, nuclei of the compound adsorb energy through a process called magnetic resonance. The absorption of energy is quantized in a NMR graph 12

Research Project Report 2011

and can be interpreted to provide useful information about the chemical structure of the compounds. The number of signal in the spectrum represents how many different set of protons there are in the molecule of the compound. The position of the signals in the spectrum along the x-axis provides information about the magnetic environment of each set of proton which has different electron density in their environment [23]. The areas under the spectrum tell us how many protons there are in the set being measured. Usually, the1H NMR would not give the exact number of proton under the spectrum. However, the relative number of proton in a special position can be obtained to specify the structure of the studied compound. Coupling, sometime referred to as signal splitting, can provide useful data about the structure of a compound. This is a magnetic effect where the energy absorbed by nearby hydrogens interfering with the energy level of observed hydrogens whose signal are being observed. As a result, the signal is split into multiple peaks. Signals with 1 peak are called singlet while signals with two and three peaks are call doublet and triplet respectively. The NMR spectra are measured in a solution. In order to avoid the interfering between the solvent and the investigating species spectra, it is important that the solvent used in NMR does not contain any 1H atom. As the result, solvents used for 1HNMR either do not contain any hydrogen at all or hydrogens atoms are replaced by its isotope such as deuterium. All the 1HNMR spectroscopy in this experiment use deuterated chloroform (CDCl3) as the solvent.

4.3. Gel Permeation Chromatography GPC

Gel Permeation chromatography (GPC) is a useful analyse tool to observe the molecular weight distribution in a synthesize polymer and average molecular weights. GPC involves passing diluted polymer solution through a stationary column packed with finely divided solid gel particles with same pore size [24]. The small monomers with low molecular weight will enter the pores of gel particle and result in longer effective path compared to larger polymer molecules. Consequently, the large polymer molecules will follow the solvent and reach the detector ahead smaller monomer molecules [25]. A plot of elution time or elution volume verse the amount of polymer present can be obtained using GPC. This can be calibrated to determine the molecular weight of the polymer. There are several way to calibrate a GPC graph such as using polystyrene and light sensor. However, they are complicated and unrelated to the topic in this report. Therefore, instead of describing the process of GPC graph calibration, this report only gives final molecular weight of synthesized polymers. In this report, dimethylfloro (DMF) was used as solvent for GPC test. Compared to water GPC, using DMF GPC gives more accurate molecular weight analysis. GPC tests were only carried out for the polymers before quartenization or hydrolysis step, in which end groups of short side chains are modified to obtain final desired polymers. The first reason for this is that quatenization or Hydrolysis of the polymers will alternate their structure and hence change their solubility in DMF. As a result, DMF cannot be used as solvent after quaternization or hydrolysis. Even though, water GPC could be utilized to carry out GPC test for final polymers, it is not advisable since water GPC gives 13

Research Project Report 2011

incorrect molecular weight and not comparable with DMF GPC result obtained in the previous steps. Moreover, modifying the end groups does not change the polymers molecular weight and their distribution significantly. Finally, the success of quaternization and hydrolysis step can be confirmed using other analyse method such as 1HNMR. In addition to molecular weight, analysis on GPC graph also gives the polydispersity index (PDI) which is the measure of the distribution of molecular mass in a given polymer sample. The PDI calculated is the weight average molecular weight divided by the number average molecular weight.

5. Procedure and result of synthesizing polymers

RAFT chain transfer agents were chosen base on complex mechanism of kinetic which are not be discussed in this report. Generally, the ratio between initiator and chain transfer agent usually 1:1, 1:2 or 1: 3 depend on the chosen chain transfer agent [4]. Solvents in these reactions are DCM, THF or dioxane. The polarity of the solvent must allows good solubility of reactants. Also, boiling point of the chosen solvent must comfortably accommodate the reaction temperature. For example, for reaction of synthesizing monomer Poly ethylene glycol methacrylate molecular weight 2000 that is described in this section, reaction was carried out at room temperature. As a result, DCM with relatively low boiling point of 35.6 oC can be used. This also allows ease in removing the solvent under vacuum using rotavapor. Moreover, DCM is non-polar which allow phase separation with water in the purifying process but still able to provide a good degree of solubility for non-polar reactant such as PEG methyl ether and Man.

5.1. Synthesising monomer PEG 2K MA

In order to synthesis high molecular weight polymers, poly (Ethylene Glycol) Methacrylate molecular weight of 2000 (PEG 2K MA) is required. Unfortunately, this monomer is not commercially available. Therefore, it has to be synthesized before any polymerization reactions can be carried out. The reaction for synthesizing PEG 2K MA is shown in Figure 8. The amount of substances used is shown in Table 1. It is noticed that the molar mass of MAn has to be in great excess compared to PEG 2K (five times the required amount in this experiment). This is to ensure that all PEG 2K is consumed in the reaction. If there is trace of PEG 2K left in the final reaction mixture, it cannot be separated from the desired product since they both have relatively same molecular weight as well as physical properties and structure. The reaction was carried out in well stirred condition at room temperature for 24 hours to ensure sufficient reaction time. As there were great excess MAn in the reaction mixture, low concentrated sodium hydroxide was injected to convert MAn into acid which could be easily removed in

14

Research Project Report 2011

purification process. The reaction was then left for another 6 hours for sodium hydroxide to react with MAn. After the reaction has stopped, the product solution was mixed with 150 ml saturated NaHCO3 in a separating funnel in order to absorb all acid which was originally from the reaction between MAn and sodium hydroxide. Phase separation occurred in the funnel as water is immiscible in DCM which contain the final desired product. As the density of DCM is greater than water, the lower phase is DCM while water is on the top. The DCM phase was extracted while water phase that contains impurity from the reaction between NaHCO3 and Man was discarded. The process was repeated to ensure all Man is removed from the DCM phase. Since DMAP is a catalyst for the reaction, it was not consumed during the reaction hence need to be removed in purification process. Since DMAP is slightly basic, diluted HCl (0.01M) was used to remove it. The solution in above step was washed with 100 ml diluted HCl and separate the DCM phase. During the purification process, it is unavoidable that some water has mixed in the DCM phase. This can be clearly seen by the cloudy solution due to immiscible water in the DCM phase. Consequently, the solution was washed with 100 ml NaCl twice. This would mostly remove water in DCM phase. However, to ensure all trace of water was removed, Na2SO4 salt was added to the solution and leaved overnight. Na2SO4 with absorbed water was removed using filtration. It was noticed that the collected solution was clear which indicate no water left in the solution. Final PEG2K MA product was collected from the DCM solvent using precipitation process. The solution was then filtrated to collect PEG2K MA powder.
Table 1: Synthesizing PEG2K MA

Substances Poly( Ethylene Glycol ) methyl ether (PEG 2K) Methacrylic Anhydride (MAn) 4-Dimethylaminopyridine (DMAP) Dichloromethane (DCM)

MW 2000 154.16 22.17

Mass 20g

Vol

7.6 ml 0.131 g 80 ml

DMAP

Figure 8: Reaction for synthesizing PEG 2K MA

15

Research Project Report 2011

A yield of 52% was achieved in this polymerization; 10.43 g of polymer product was collected. HNMR was used to confirm the structure of the monomer. Figure 9 shows the 1HNMR result for PEG 2K MA. The chemical shifts of characteristic protons in the expected structure are added to the graph. Table 2 illustrates the NMR analysis. Since each proton will generate peak at different chemical shift, using handbook of NMR [23], chemical shift of each proton in the structure can be found. Since the synthesized monomer structure is relatively simple, most the characteristic peaks such as a,b,c,d and f can be identified. However, in the repeat unit where the magnetic environment is similar, chemical shift protons are overlapped. As a result, all peaks in chemical shift range of 3.356-3.39 ppm are denoted as e. It can be seen clearly in Table 2, that integration under the peaks are match with expected number of protons. As a result, the synthesized product has the desired structure and can be used to synthesize polymers.
1

Table 2: 1HNMR analysis for PEG 2K MA

Proton a b c d e f

Chemical shifts 6.105 5.55 1.924 4.274 3.356-3.39 3.351-3.354

Number of proton 1 1 3 2 N/A 3

Integration 0.99 1.18 2.82 1.9 N/A 3

Figure 9: HNMR for monomer PEG-2K MA

16

Research Project Report 2011

5.2. Polymers synthesizing
All polymers were synthesized using RAFT polymerization. The synthesized polymers have a general structure depicted in Figure 10. The graft polymers have long PEO side chains with charges attach on the short PA chains. Assumed an ideal polymerization that all the monomers are polymerized, the ratio of the charge and long side chain in a polymer should be equal to the monomer ratio initially hence can be controlled. The molecular weight of ideal RAFT polymerization can be predicted using 5-1 hence can be controlled accordingly. Overall, in this research project, by manipulating n:m:p ratio, different polymer structures can be obtained. The effect of these polymers on geopolymer material can be studied using rheology test.

[ [ ]

5-1

Synthesizing the polymers consists of two steps shown in Figure 10. In the first step, polymerization of Poly ethylene glycol (PEG) and a co-monomers is carried out to obtain the desired structure. The co-monomer is tert-butyl Methacrylate if anionic polymer was synthesized. 2(Dimethylamino)ethyl acrylate (DMAEA) was used if cationic polymers was desired. The second step involves quaternization or hydrolysis of the synthesized polymer to convert the end group of short side chain to charged end group. Positive charge was created by reaction with polymer to methyl iodine (MeI). Tertiary amine group in DMAEA side chain is a reactive nucleophilic group; can be easily charged with compound of good living group group such as iodine and create cationic compound. Unlike tertiary amine end group, t-butyl end group is unreactive; it acts as a protection group that prevent any reactions to occur. Therefore, in order to create anionic end group, trifluoroacetic acid (TFA) was used to remove the protection group. This is a popular method and has been reported in many articles [26, 27]. This hydrolysis process is sometime referred to as deprotection process. Six polymers were synthesized in a period of two months time. The structure of these polymers are coded and categorized in Table 3 whereas C or A letter stand for anionic or cationic, number after A (or C) is the average molecular weight of the polymer, number in the bracket is the repeating units of ethylene oxide in Poly ethylene oxide side chain (PEO); the number after the short dash represents the ratio of the charged monomer PA to long side chain PEO.

17

Research Project Report 2011

Figure 10: General synthesizing process

Table 3: Summarized of synthesized polymers' structure

Linear polymer Low MW polymer High MW polymer PA/PEO 110 22 43 22 43 Anionic 1:1 A42 (22)-1:5 1:5 A6 (22)-1:5 cationic 1:1 C5 (110)-1:1 C8 (43)-1:1* C51 (43)-1:1 1:5 C8 (22)-1:5 C14 (43)-1:5 *C8 (43)-1:1 result was borrowed from supervisor to complete the trend in the analysis

18

Research Project Report 2011

5.2.1. Synthesizing C50 (43)-1:1

Table 4

Substances PEG 2K MethaAcrylate DMAEMA DEGMA AIBN CTA * THF

MW 2000 188.22 157.21 279 164.21

Mass Synthesizing polymer 2g 0.753 mg 0.157 g 116.16 mg 3.28 mg Quaternization

Vol

8 ml

Polymer 2.155 g MeI 0.15 g DCM *Sulfur 2-Methyl-2-[(dodecylsulfanylthiocarbonyl)sulfanyl]propanoic acid

7 ml

Figure 11: reaction of synthesizing C50 (43)-1:1

PEG2K Acrylate- DMAEA copolymer was synthesized. Ratio of PEG2KA: DMAEA: DEGMA of 1:1:4 was used in this experiment. The reaction is shown in Figure 11. RAFT polymerization in THF solvent was used ultilized to synthesize the polymer. As the chain transfer agent is sensitive to moisture and sensitive to basic condition, the mixture was bubbled with argon in half of an hour to eliminate the trace of air in the round bottom flash. The reaction was then carried out at 65oC for 24 hours. Besides, all liquid monomers are usually prevented from seft-polymerization during transport and storage by adding inhibitors. These inhibitors had to be removed before polymerization was carried out. At this small scale, monomers are de-inhibited by passing through a pipette packed with aluminium oxide powder at neutral condition. To accelerate the de-inhibiting process, a teat is used to applied pressure to force the monomers passing through the powder; inhibitors are filtered back by the powder Aluminium oxide. The amounts of reactants in the reaction are shown in Table 4. At this small scale, the desired polymer can be separated by centrifuged after precipitated in diethyl Ether. However, in any larger scale, filtration is required to extract the polymer powder. It is noticed that the chain transfer agent is originally pink. If the reaction is carried out successfully, reaction mixture will have dark pink colour. Otherwise, changing reaction mixture colour to brown or yellow implicates that disintegration of chain transfer agent had happened. Disintegration of chain transfer agent could be due to contamination in the reaction which leads to basic conditions in the 19

Research Project Report 2011

reaction mixture. Quaternization was carried out by reacting the polymer in the step above with Methyl iodine at room temperature for overnight. THF was used as solvent; the product was then precipitated in diethyl ether and extracted by centrifugation. GPC test were done for the product before quaternization. GPC graph shown in Figure 12 describes the molecular weight distribution. There are three peaks shown in GPC graph. Sharp peek at low retention time represents polymers. As discussed in GPC section, polymers have higher molecular weight. As a result, they go through the column in GPC machine and reach the sensor before monomers hence recorded at lower retention time. The sample polymer shows a sharp peak, which represent a narrow range of molecular weight hence low PDI. However, there is a significant of unreacted monomers in the sample, which is represented by peak at retention time around 2.25 to 2.5. The intensity of the peak is more than half of the polymers peak. However, the amount polymer in sample was deemed sufficient to continue with quaternization step. The remaining peaks are the solvent peaks. By analysing the GPC graph, average molecular weight of 49 000 was obtained with PDI of 1.25.
C50 (43)-1:1 before Quaternization 1.2 1 0.8 0.6 0.4 0.2 0 0 -0.2 -0.4
Figure 12: GPC graph for C50(43)-1:1

10

20

30

40

Proton NMR was done for the sample in prior of quaternization. 1HNMR result is illustrated in Figure 13. From Figure 13, it is impossible to confirm the structure the polymer due to overlapping of proton in characteristic position. Protons in methyl group in methacrylate end group,b, c, d are overlapped with proton of carbon attach to amine group a. However, it is expected that the peak would be clearer after the quaternization due to change in chemical shift after modifying the amine end group to be positive charge. The analysis of 1HNMR graphs is shown in Table 5; after quaternization, due to changing in magnetic condition, the chemical shift for protons of carbon attach to tertiary amine will be shifted to chemical shift of approximately 3.15 ppm and no longer overlap with any other protons. By comparing proton 1HNMR for sample before (Figure 13) and after quaternization (Figure 14), the structure of the synthesized polymer can be confirmed. Quaternization step changes modified amine

20

Research Project Report 2011

group to become positive change with three methyl group attached to it. As a result, this increases the number of proton in the amine end group to nine. The ratio of PEO and amine end group side chain can be confirmed using the 1H NMR for polymer after quaternization. The ratio of PEO:PA:Ethylene side chain in the final product is 1: 0.95: 4.52 which is close to the desire structure ratio 1:1:4. Moreover, it is noticed that the integration under e peak before quaternization is 11.72 of which the value changed to 13.56 after quaternization. This could be error originated from the integration error. As the base line is not horizontal and peaks are so closed together, it is difficult to obtain an exact value of integration. Overall, the experiment was a successful in creating a polymer with molecular weight of 50 000 with ratio PEO: PA of 1:1

Figure 13: HNMR for C50(43)1:1 before quaternization

21

Research Project Report 2011

Figure 14:1HNMR for C50 (43)-1:1 after quaternization Table 5: HNMR analysis for C50 (43)-1:1
1

Proton

a b c d e f Ratio PEO/PA/EO

Before quaternization Number of Actual NMR proton integration before quaternization 6 NA 3 NA 3 NA 3 NA 3 11.72 3 3 N/A

After quaternization Number of proton NMR integration after quaternization 9 3 3 3 3 3 8.56 NA 13.56 NA 13.56 3

5.2.2. Synthesising C14 (43)-1:5

Table 6

Substances PEG 2K Acrylate DMAEA AIBN CTA* dioxane

MW 2000 142.21 164.21 92.12

Mass Synthesizing polymer 4g 57 mg 32 mg 34 mg

Vol

25 ml 22

Research Project Report 2011

Polymer MeI THF
*thioglycolic Acid

Quaternization 3.4 g 0.2 g 15 ml

Figure 15: Reactions synthesize C14 (43)-1:5

Similar to synthesize C40 (43)-1: 1, RAFT polymerization method was used in this synthesize.. Thiolglycolic acid was used as chain transfer agent. Reactants consist of PEG2K Acrylate, DMAEA with AIBN initiator PEG 2K: DMAEA: AIBN: CTA= 5 :1 :0.5 :1. Exact amount is shown in Table 6. After degassing for 30 minutes, the reaction was carried out at 65oC in dioxane solvent for overnight. Purification was done by precipitating in diethyl ether. The polymer was extracted by centrifuge and dry in vacuum for overnight. Quaternization was carried out using same method described in sections above. GPC test was done for polymer before the quaternization step. It shows that the conversion rate is not very high as there is still a significant amount of monomer in the synthesized sample which represented by the lower peak at high retention time, around 23 minutes (Figure 16). From GPC graph, molecular weights of 14.85 with relatively low PDI of 1.35 were obtained.

GPC C18 (43)-1:5 before Quaternization 0.014 0.012 0.01 0.008 0.006 0.004 0.002 0 -0.002 -0.004 -0.006 0 5 10 15 20 25 30 35 40

23

Research Project Report 2011

Figure 16: GPC for C18 (43)-1:5

HNMR was performed to confirm the structure of the polymer. Compared to C40 (43) 1:1, this sample was synthesized using acrylate monomer instead of methacrylates monomer. As a result, there is no overlapping in peak of PEO side chain represented by c in Figure 17. The structure of polymer before quaternization can be confirmed. From Figure 17 and Figure 18, analysis is done in Table 7. It shows that before quaternization, the ratio of PEO/PA chains in the polymer is 5, which is expected. However, after quaternization, this ratio reduces to 3.6. This happens due to some amine end group did not convert to positive charge. Overall, the experiment was successful.

Table 7: HNMR analysis for C14 (43) 1:5

a b C Ratio of PEO/PA

Before Quaternization Number of protons Integral 3 3 2 2.11 6 1.22

After Quaternization Number of protons Integral 3 3 2 1.79 9 2.5

Figure 17: HNMR for C14 (43)-1:5 before quaternization

24

Research Project Report 2011

Figure 18: HNMR for C14 (43)-1:5 after quaternization

5.2.3. Synthesizing linear structure C5 (110)-1:1 Linear

Table 8: Synthesizing C5 (110)-1:1 linear

Substances PEG 5K-OH Succinic anhydride THF DMAP

MW 5000

Mass 10g 2g 49 mg

Vol

50 ml

Figure 19: reaction of synthesizing linear structure

To investigate the effect of steric hindrance to the polymers performance, linear polymer with long PEG chain has been synthesized. The polymer was simply synthesized by modifying the end group of PEG-OH to be acid using ring opening mechanism in presence of DMAP catalyst. Reaction is 25

Research Project Report 2011

illustrated in Figure 19. The reaction was carried out at well stirred and temperature of 50oC for overnight. THF was used as solvent in this reaction. As PEG5K-OH cannot be removed during purification process, succinic anhydride was used in excess to ensure all PEG5K-OH are all reacted. Similar to synthesising monomer PEG 2K MA, distilled water or dilute sodium hydroxide was added to the reaction solution after the reaction had finished converting all excess anhydride to acid and allowing it to be removed during purification process. The purification process is exactly the same with purifying PEG 2K MA in section 5.1. As a result, it would not be described in this section again. As 1HNMR result of the product compound is very clear, the conversion of this reaction can be analysed. Figure 20 shows the result for 1HNMR test. Modifying the hydroxide group of PEG-OH changes the nuclear magnetic environment which consequently causes the peak of surrounding protons to shift. As can be seen in Table 9, the starting material contains no peaks at a and b. After the reaction, there are no other peaks observed except for a, b, c and the repeat unit. This indicates that the product is pure and no impurities present in the sample. Furthermore, the integration shows the expected number of protons in each position. The insignificant difference can arise from the error in integration. Therefore, it can be seen that that the final product is almost pure with expected structure.
Table 9: HNMR analysis for linear acid
1

a b c

Starting material Number of proton 0 0 3

Product Number of proton 4 2 3

Integral 3.84 2.13 3

Figure 20: HNMR linear-acid

Interestingly, the result using MALDI technique to analyse the substance shows two substances in the samples with slightly different molecular weight. A reason for this can be disintegration of the final acid product in under disturbed conditions. Because MALDI technique uses laser to analyse the sample, it is believed that the product disintegrated under the relatively high energy of the laser 26

Research Project Report 2011

beam. It deemed that the synthesized compound is not stable and has tendency to go disintegrate to dispose of acid group. As a result, it is not competent to be a test sample.

Figure 21: Mass Spectroscopy for Acid-Linear

5.2.4. Synthesising C8 (22)-1:5

Table 10: Synthesizing C8 (22)-1:5

Substances PEG 1K MA DMAEMA AIBN CTA* THF Polymer MeI THF

MW 1100 157.21 164.21 364.63

Mass Synthesizing polymer 4g 0.114 g 0.06 g 0.265 g Quaternization 3.5 g 0.2 g

Vol

10 ml

15 ml

*Sulfur 2-Methyl-2-[(dodecylsulfanylthiocarbonyl)sulfanyl]propanoic acid

Figure 22: reaction of synthesizing C8 (22)-1:5

27

Research Project Report 2011

Reaction mechanism for this polymer is exactly the same with C 50(43)1:1 except no DEGMA copolymer was used. Molar ratio in the reaction is PEG1KMA: DMAEMA: AIBN: CTA=1 : 5 : 0.5 : 1. Shorter PEG monomer with 22 repeat units was used for this reaction. Reactions, quaternization and purification method are similar to synthesizing C50 (43) 1:1. GPC for polymer before quaternization (Figure 23) shows a sharp peak of polymer. Since the monomer has low molecular weight, compared to C50 (43)-1:1 or C14 (43)-1:5, monomer molecules would take much longer to pass through the column in GPC machine. As the result, the distance between monomer peak and polymer peak is further. By analysing GPC graph, molecular weights of 8000 and PDI 1.25 were obtained.

GPC C8 (22)-1:5 before Quarternization 1.2 1 0.8 0.6 0.4 0.2 0 0 -0.2
Figure 23: GPC C8 (22)-1:5 before Quaternization
1

10

20

30

40

50

HNMR for the compound after quaternization (Figure 24) can be use to confirm the structure of the compound. The ratio of PEO/ PA side chains is almost 1:5 as shown in Table 11
Table 11: HNMR analysis for C8 (22)-1:5
1

Peak location a b c Ratio PEO /PA chain

Expected proton 9 3 2.5

Integration 2.05 3 2

28

Research Project Report 2011

Figure 24: HNMR after quaternization

5.2.5. Synthesising A6 (22)-1:5

Table 12

Substances PEG 1K methacrylate t-BuMA AIBN CTA THF Polymer TFA DCM
*thioglycerol

MW 1100 142.2 164.21 108.2

Mass Synthesizing Polymer 4g 0.104 59.2 78.1 Hydrolysis 3.8 g

Vol

10 ml

8 ml 5 ml

Figure 25: reaction of synthesizing A6 (22)-1:5

In order to synthesize anionic polymer, side chain with t butane end group was used. Reaction took place at 65oC. Thioglycerol chain transfer agent was used. To obtain the desired molecular 29

Research Project Report 2011

weight and structure, molar ratio of PEG 1K: t-BuMA: CTA:AIBN= 5: 1: 1: 0.5 was used. Instead of quaternization using CH3I to obtain cationic polymer, anionic polymer was obtained by hydrolysis. This step was carried out by let the polymer reacted with 10 ml of TFA in DCM solvent for 24 hours under room temperature. Figure 26 shows GPC analysis for compound before hydrolysis. It can be seen that effectively no trace of monomer left in the polymer sample. Sharpe peak shows narrow distribution of polymer molecular weight with only small monomer peak circled in Figure 26. From GPC, average molecular weight of 6000 and PDI of 1.261 were obtained.

A6 (22)-1:5_ Before Hydrolysis 1.2 1 0.8 0.6 0.4 0.2 0 0 -0.2 10 20 30 40

Figure 26: GPC for A6 (22)-1:5 before hydrolysis

HNMR before and after hydrolysis are shown in Figure 27 and Figure 28. Analysis was done in Table 13. As can be seen, the ratio of PEO and short t-butane side chains is 4.77 which is close to anticipated ratio of 5. During hydrolysis step, t-butane group is cleaved out. As a result, it is expected that peak a which represent protons in t-butane group to be disappeared. The result in Table 13 shows that peak a has in fact disappeared from the 1HNMR graph. Therefore, it can be confirmed that the polymer obtained from this synthesizing is A6 (22)-1:5.

Table 13: HNMR analysis for A6 (22)-1:5

a b Ratio PEO/

Before hydrolysis Number of protons Integral 9 9.22 3 15

After hydrolysis Number of proton Integral 0 0 3 3 N/A

30

Research Project Report 2011

Figure 27: H NMR for A6 (22)-1:5 before hydrolysis

Figure 28: HNMR for A6 (22)-1:5 after hydrolysis

5.2.6. Synthesising A42 (22)-1:5

Table 14

Substances PEG 1K MA t-BuMA AIBN CTA * THF

MW 1100 142.2 162.1 364.63

Mass Synthesizing polymer 4g 0.104g 100 mg 44.4 mg

Vol

10 ml 31

Research Project Report 2011

Polymer TFA DCM
*thioglycerol

Hydrolysis 3.5g 10 ml 5 ml

Figure 29: Reactions synthesizing A42 (22)-1:5

This polymer is synthesized by carried out exactly the same step as A6 (22)-1:5. However, the ratio of reactant is modified to obtain higher molecular weight. Molar ratio of PEG 1K: tBuMA: CTA:AIBN is 30:6:1:0.25. GPC result shows in Figure 30, only a small amount of low molecular weight monomer is unreacted. Molecular weight of the polymer is 42300 with PDI around 1.607. 1HNMR results analysis for polymer before hydrolysis (Figure 31) and after hydrolysis (Figure 32) is shown in Table 15. The ratio of PEO and t-butane before hydrolysis is 4.75 which is closed to 5. Furthermore, the analysis shows that hydrolysis step has successfully removed all t-butane groups which indicated by disappearance of peak at chemical shift 1.367 ppm.

A42 (22)-1:5_Before Hydrolysis 1.2 1 0.8 0.6 0.4 0.2 0 0 -0.2 -0.4
Figure 30: GPC for A42 (22)-1:5 before hydrolysis

#19_BeforeQu

10

20

30

40

32

Research Project Report 2011

Table 15: HNMR analysis for A42 (22)-1:5
1

a b Ratio PEO/tbutane

Before hydrolysis Number of protons Integral 9 9.46 3 15

After hydrolysis Number of proton Integral 0 0 3 3 N/A

Figure 31: Polymer for A42 (22)-1:5 before hydrolysis

33

Research Project Report 2011

Figure 32: Polymer for A42 (22)-1:5 after hydrolysis

6. Rheology Test Result and Discussion

Rheology tests were done for geopolymer mixture, which comprises of slag, superplasticizer, water and activator. Activator can be waterglass (sodium metasilicate Na2SiO3) or sodium hydroxide solution. Activator adjusts PH of the mixture to approximately 11 at which the dissolution of slag can be initiated effectively to release to silicate and aluminium species. Rheology tests were done by different research group. Therefore, rather than describing in detail on how rheology tests were conducted, only the result of rheology test are analysed in this section of the report. Trends are obtained to reveal the best structure of superplasticizer for geopolymer cement. The result depicted in Figure 33. The effect of polymer to geopolymer cement can be quantified in term of yield stress reduction or viscosity reduction. Cement with lower viscosity or yield stress has more fluidity. As result, polymers with more yield stress reduction are better. The dispersing effect is attributed for both steric hindrance and electrostatic repulsion of the side chain. The steric hindrance is created by hydrophilic PEO side chain floating in liquid phase and prevent cement particle from contacting and flocculate. Electrostatic repulsion force is accounted for the electrostatic adsorption of charged side chains to the cement particles and creates a repulsive force between cement particles. Therefore, all discussion is related to these two properties.

Figure 33: Rheology test result

Linear polymer with long backbone chain has negative effect on yield stress reduction. This is because the dis-integration of linear polymer in stimulated conditions that reverses the reaction back to its original substance PEG 5K OH. This was discussed in section 5.2.3. Moreover, linear polymer has completely different structure compared to others samples hence it is not comparable 34

Research Project Report 2011

to others. It is synthesized only to investigate the effect of long chain length on polymer performance. As mention in section 5.2.3, the synthesized linear structure is not competent for testing. Therefore, result shown in Figure 33 is not reliable and does not eliminate chances of linear structures working better than grafted structure. The effects of structure in yield stress reducing are investigated by compared four sets of polymers: C14 (43)-1:5 and C8 (22)-1:5; A6 (22)1:5 and C8 (22):1:5, A6 (22)1:5 and A42 (22)-1:5; C51 (43) 1:1 and C8 (43): 1:1. The PEO side chain length can be investigated by comparing C14 (43)-1:5 and C8 (22)-1:5. These two polymers have similar molecular weight, charge and charge density. The only difference between these polymers is the PEO side chain length. The slight different in molecular weight is due to different in side chain. However, the length of backbone is relatively unchanged. It can be clearly seen in Figure 33, C14 (43)-1:5 results in greater reduction in yield stress. This implies that polymers yield better effect with longer PEO side chains, this is because the longer side chain had a more steric hindrance effect, which is favourable to the dispersion of cement particles. In the liquid phase, the oxygen atoms of the ether (C-O-C) group in the side chains can bind with water molecules form strong hydrogen bond producing a large hydrophilic steric protection barrier that greatly contribute to dispersing stability. As A6 (22)1:5 and C8 (22):1:5 have exactly the relatively same structure except the charged end group, they can be compared to investigate if cationic polymer is better. At the beginning of the research project, it is tested that geopolymer cement mixture has negative zeta potential. Therefore, it is expected that cationic polymers would have better effects than anionic. However, the result shows the same performance for these two polymers. The fact that C14 (43)-1:5 is the best sample so far shows that cationic cannot be bad for the polymer. Therefore, there must be a mechanism to prevent cationic polymer from working. It is proposed that, electrostatic repulsion effect is insignificant at low molecular weight. Since geopolymer material is activated in relatively high PH, it is thought that at this condition, positive charges in polymers chains are neutralised before come into contact with the particles hence eradicate the effect of electrostatic repulsion. It is expected that longer the back bond chain will increase the dispersing effect. At high molecular weights, with the same PEO/PA ratio, there are more charges and anchor points which result in increasing adsorption area and enhance the effect of electrostatic repulsion. In this case, analysing two sets of polymers (A6 (22)1:5 and A42 (22)-1:5; C51 (43) 1:1 and C8 (43): 1:1) shows that the low molecular weight polymers gives better result. However, it is realised that two sets that were used in this analysis are at two extreme i.e. one is at very high and the other is at very low molecular weight. Therefore, it is unable to confirm this trend only with these two sets. However, it is thought that structure with very long back bond will have tendency to wrap itself in spherical form instead of dispersing in the liquid phase hence not effective. Also, for too high molecular weight, there is high chance of charges being wrapped by polymers back bond chain. This would prevent the adsorption of polymers to the surface of particles. Overall, the analysis shows the ideal polymers would have these following characteristics: Cationic.

35

Research Project Report 2011

Moderate back bone chain that is long enough to protect the charge group but not too long.to form a spherical structure. Long PEO chains.

7. Future work
Up until now, C14 (43)-1:3 sample has shown the best performance. However, the result is still not good enough for real life production. In order to commercialise this product, there are still lots of work need to be done. At the moment, the trend is rather confusing. Therefore, in the future, more polymers should be synthesized to obtain a clearer trend of how molecular weight and charge density affect the performance. Microscopic study on how geopolymer and superplasticizer interact with each other can be done to further understand the mechanism as well as reconfirm the trend observed in this research. In the future, if it is believed that the current structure would not yield any better result, the research scope can be broaden by looking into different structure such as linear, star polymers. At this stage of the project, since it is at experiment scale, cost of synthesizing polymers is not a problem to be worried about. However, if an ideal superplasticizer that gives perfect result could be synthesized, more researches have to be done to find a cheap method for large-scale production.

8. Conclusion
Six different polycarboxylic superplasticizers were successfully synthesized in this research project. Among these six, five samples are graft polymers while the remaining has linear structure. It was found that linear structure is unstable and has tendency to reverse back to original material hence does not yield correct result for rheology test. Furthermore, as it has totally different structure, it is not comparable with the other five samples. Five graft polymers were obtained by modifying electrostatic property, side chain length, molecular weight (back bond length) and charge density (ratio PEO/PE). By analysing the rheology, several conclusions were made. First, polymers with long side chain length yield better result due to steric hindrance effect. The longer the polymers side chain, the harder the cement particle can come into contact and flocculate. Though it is first believed that cationic polymer would yield better result, analysis on samples shows similar result for cationic and anionic polymers. With only two samples, this trend cannot be confirmed. However, it is proposed that the short back bond chains cannot give enough protection to the charges to prevent them from being neutralized in strong basic condition within the cement paste mixture. Finally, it is found that high molecular does not enhance the performance of the polymer but rather eradicate the electrostatic effect of polymer. It is believed that high molecular weight would form a spherical ball that trap the charge ends inside and prevent them from adsorbing to the surface of geopolymer cement particles.

36

Research Project Report 2011

References 1. Collepardi, M., Admixture-enhancing concrete performance, in 6th international congress, Global COnstruction, Ultimate Concrete Opportunities. 2005: United Kingdom. 2. Odian, G., Principle of Polymerization. 4th ed. 2004. 3. Moad, G. and D.H. Solomon, The chemistry of Radical Polymerization 1995. 4. Moad, G., E. Rizzardo, and S.H. THang, Living Radical Polymerization by the RAFT process. CSIRO Molecular Science, 2005. 5. Seymour, R.B. and C.E. Carraher, Polymer Chemistry an Introduction. 2nd ed. 1988. 6. Kowollik, C.B., Handbook of RAFT polymerization. 2008. 7. Torgal, F.P., J.o.C. Gomes, and S. Jalali, Alkali-activated binders: A review Part 1. Historical background, terminology, reaction mechanisms and hydration products. Construction and Building Materials, 2007. 22: p. 1305-1314. 8. Roy, D.M., Alkali-activated cements Opportunities and challenges. Cement and Concrete Research, 1999(29): p. 249-254. 9. Duxson, P., et al., The role of inorganic polymer technology in the development of green concrete. Cement and Concrete Research, 2007. 37: p. 15901597. 10. Collepardi, M., Admixtures-Enhancing Concrete Performance in 6th International Congress, Global Construction, Ultimate Concrete Opportunities. 2005: Dundee U.K. . 11. Taylor, H.F.W., Cement Chemistry. 2nd ed. 1997. 12. Li, C.-Z., N.Q. Feng, and Y.-D. Li, Effects of Polyethlene Oxide Chains on the Performance of Polycarboxylate-type Water- reducers. Cement and Concrete Research, 2005. 35(5): p. 847873. 13. Uchikawa, H., S. Hanehara, and D. Sawaki, THE ROLE OF STERIC REPULSIVE FORCE IN THE DISPERSION OF CEMENT PARTICLES IN FRESH PASTE PREPARED WITH ORGANIC ADMIXTURE. Cenment and Concrete research, 1995. 27: p. 37-50. 14. Yamada, K., et al., Effects of the chemical structure on the properties of polycarboxulate-type superplasticizer. Cement and Concrete Research, 1999. 30: p. 197-207. 15. Cho, H.-Y. and J.-M. Suh, Effects of the Synthetic Conditions of Poly{carboxylate-g- (ethylene glycol) methyl ether} on the Dispersibility in Cement Paste. Cement and Concrete Research, 2005. 35(5): p. 891-899. 16. Liao, T.-S., C.-L. Hwang, and Y.-S. Ye, Effects of a Carboxylic Acid/sulfonic Acid Copolymer on the Material Properties of Cementitious Materials. Cement and Concrete Research, 2006. 36(4): p. 650-655. 17. Hui, W., G. Huiling, and L. Jiaheng, Research on Synthesis of Molecular Structure Design of Polycarboxylate Superplasticizer. Journal of WuhanUniversity of Technology, 2006. 28(09): p. 18-20. 18. Rongguo, Z., et al., The Synthesis Technique of Polyacrylic Acid Superplasticizer. Journal of Wuhan University of Technology, 2008. 23: p. 830-834. 19. Yamadaa, K., et al., Effects of the chemical structure on the properties of polycarboxylatetype superplasticizer. Cement and Concrete Research, 1999. 30: p. 197-207. 20. Xiuxing, M., Synthesis of New Polyether Polycarboxylate Superplasticizer. Journal of Wuhan University of Technology-Mater, 2010: p. 799-802. 21. Nvlta, J., et al., Scale-up of agitated crystallizers; Cooling batch crystallization of potassium aluminium sulphate. 1982. 47: p. 1189-1194. 22. Solomons, T.W.G. and C.B. Fryhle, Organic Chemistry. 2010. 23. Mistry, B.D., A handbook of Spectroscopic Data: Chemistry (UV,IR,13CNMR and Mass Spectrocopy). 2009. 24. Striegel, A.e.M., et al., MODERN SIZE-EXCLUSION LIQUID CHROMATOGRAPHY: Practice of Gel Permeation and Gel Filtration Chromatography. 2nd ed. 2009. 25. Wilhelm, P., Modern Polymer Spectroscopy. 2008. 37

Research Project Report 2011

26. Behrend, L.F., et al., Removal of t-Butyl and t-Butoxycarbonyl Protecting Groups with Trifluoroacetic acid. International Journal of Peptide and Protein Research 1978. 12(5): p. 258-268. Hughes, A.B., Amino Acids, Peptides and Proteins in Organic Chemistry: Protection Reactions, Medicinal Chemistry, Combinatorial Synthesis.

27.

38