Understanding the Effectiveness of Polycarboxylates as Grinding Aids.pdf

8/17/2019 Understanding the Effectiveness of Polycarboxylates as Grinding Aids.pdf

1/15

SP-288.16

Understanding the Effectiveness ofPolycarboxylates as Grinding Aids

by Ratan K. Mishra, Hendrik Heinz,

Jörg Zimmermann, Thomas Müller, and Robert J.

Flatt

Synopsis: Over recent years, polycarboxylate superplasticizers have found their way intogrinding aids used in cement production to reduce the electrical energy consumption. Theeffectiveness of these large molecules challenges the pre-existing theories concerning thefactors that govern the performance of grinding aids. This paper reports on moleculardynamics simulations to examine a physical property believed to control the effective-ness of grinding aids, namely their adsorption energy. The molecules selected are TIPA(Triisopropanol amine), TEA (Triethanol amine) and glycerine. The surfaces examined aredry and hydroxylated C3S surfaces, which are believed to be more representative of reality,since some humidity is always present during the grinding. Detailed results of this part ofthe work show that glycerine interacts relatively more with dry as well as hydroxylatedsurfaces of C3S both at 25°C, ambient temperature and 110°C, grinding temperature withrespect to TIPA and TEA. These result help to better understand the specific interaction ofthese molecules with cement surfaces. In the second part of this work oligomers of somePCE superplasticizers are examined with similar numerical tools on dry and hydroxylatedsurfaces of C3S. Results for different types of these oligomers, together with the previousresults, shed light onto the reasons why polycarboxylate superplasticizers have found toalso be effective grinding aids in cement production.

Keywords: adsorption; agglomeration; hydroxylated (Hyd); moleculardynamics (MD); TEA (Triethanol amine); TIPA (Triisopropanol amine);tricalcium silicate (C3S).

235


2/15

Ratan K. Mishra holds a BE degree in chemical engineering from Visveswaraya Tech-

nological University and worked as a research assistant at IISc Bangalore. He currently

pursues a Ph.D. degree at the Department of Polymer Engineering, University of Akron,

Akron, OH, with Prof. Hendrik Heinz. He investigates the interactions of polymers with

inorganic surfaces at the nanoscale and confined fluids using molecular simulation. He

received a best presentation award for his work on surface properties of C 3S at the 13th

ICCC Madrid 2011.

Hendrik Heinz received a MS in Chemistry and a PhD degree in Materials Science

from ETH Zurich in 2003. After postdoctoral work at the Air Force Research Laboratory

at Wright-Patterson Air Force Base, he joined the University of Akron, as an Assistant

Professor. His research interests include the simulation of cement materials, biomineral-

ization, and inorganic-organic interfaces using advanced simulation tools. He received

an NSF Career Award, an ACS HP Outstanding Junior Faculty Award, and the Giovanni

Novelli Prize of the Italian Clay Group.

Jörg Zimmermann obtained his PhD thesis at the Institut für Makromolekulare Chemie,

Albert-Ludwigs-Universität Freiburg (Germany). He is manager of the Polymer Synthesis

Department within the Construction Chemicals and Mortars division at Sika Technology

AG (Zürich, Switzerland).

Thomas Müller received a Diploma in Chemistry and a PhD degree in Chemistry from

FSU Jena, Germany in 2008. Directly after he joined Sika as Corporate responsible for

fundamental development of cement additives like grinding aids and quality improver.

Robert J. Flatt is Professor of Building Materials at ETHZ. Before that he was Principal

Scientist at Sika Technology AG and postdoctoral researcher at the Princeton Univer-

sity. He owns a master in Chemical Engineering and a PhD from EPFL. In 2003, he

received the RILEM Robert L’Hermite Medal for his contributions to the understanding

of chemical admixtures. In 2007, he received the Ross C. Purdy award from the basic

science division of the American Ceramic Society.

INTRODUCTION

Grinding of clinker is a highly energy intensive process during cement production. Onits own, it consumes about one third of the entire electrical energy used in cement produc-tion.1 Apart from this, it should be noted that mainly due to the decarbonation of limestone,but also to the major use of carbon based fuels, cement production contributes 5%-7% ofglobal CO2 emissions.2,3 Presently the most used approach to mitigating this is the partialreplacement of cement by supplementary cementitious materials. This however, generallyleads to a lower initial reactivity that may be compensated by finer grinding.4 In order toreduce energy consumption in the cement production, grinding can therefore be expectedto be of growing importance.5,6 In particular, there is certainly a greater potential for savingreasonable amount of energy using chemical grinding. These products limit agglomerationof fine particles, which enhances the separator efficiency as less fine material gets fed back

236 Mishra et al.


3/15

into the mill.7,8 They also prevent the milling media from getting covered by a compactedlayer of powder, which also reduces grinding efficiency.

Despite the reasonable amount of research work in the area of chemical admixtures, stillthe understanding of molecular level mechanisms of grinding aids remains poorly under-stood. As of now, a few mechanisms have been suggested to explain the action of variousgrinding aids. These include7-10:

(a) Preventing agglomeration of the freshly cleaved mineral surfaces(b) Preventing minerals particles from sticking to the milling media and mill wall(c) Adsorption-induced mobility(d) Reduction in surface free energy due to adsorption on freshly cleaved surface.In order to critically examine some of these hypotheses, there would be a need to access

physio-chemical properties such as adsorption energy on freshly cleaved surfaces, agglom-eration energy in the presence of grinding aids and driving forces for adsorption. Unfortu-nately, these properties are very difficult to access experimentally, which is why we find thenecessity of using molecular modeling to get access to some of them.

In particular, we are interested in comparing the behavior of classical grinding aids asTIPA, TEA and glycerol with the more recently introduced polycarboxylates. These arelarge comb copolymers that are largely used in concrete for their ability to improve work-ability, strength and durability of concrete through mechanisms that have largely beenstudied.5,11-14 Their effectiveness as grinding aid is more surprising as their volatility is low,which is one parameter that (in the past) has been considered important for grinding aids.However, mobility may also take place on the surface and can be expected to be affectedby the strength of the bond formations between the grinding aids and the cement particlesurfaces. This is why we focus in the present study on determining adsorption energies andexamining the nature of these bonds.

More specifically, we used a molecular model of tricalcium silicate (3CaO·SiO2, notedC3S) to examine some important questions of direct practical relevance such as adsorp-tion energy of grinding aids on C3S in the presence of grinding aids.15 The C3S wasselected among the various clinker phases because it is a primary constituent of cementclinker (50-70%) and also due to its abundance and main role in strength development,C3S frequently serves as a model system for both alite and cement.16 Kundu et al.17 alsodescribed the atomistic simulation approach to simulate the surface of another calciumsilicate, wollastonite (CaSiO3). They studied the adsorption of water and organic additives,finding that the simulations predicted the well-established trend of experimental findings.Our present study is limited to the hydroxylation of oxides ions of C3S surface due to pres-ence of humidity in the grinding process, but in this study we have not considered the extraamount of water, which may be present at clinker surfaces as a result of the dehydrationof gypsum.

RESEARCH SIGNIFICANCE

The present research work methodically uses atomistic simulations to understand theworking mechanisms of grinding aids. It proposes new insights into the working mecha-nisms of these admixtures and accesses information that is not or barely accessible experi-mentally. Thereby it contributes to the global efforts devoted to improving grinding effi-

Effectiveness of Polycarboxylates as Grinding Aids 237


4/15

ciency. This is needed in order to compensate for the loss of early age reactivity in cementcontaining larger amounts of SCMs (Supplementary Cementitious Materials).

Approaching this problem by a molecular level understanding of grinding aids is a clearoriginality of this work. It makes it possible to obtain information that would not be acces-sible otherwise. For example it is shown for the first time that grinding aid adsorptioninvolves hydrogen bonds to the C3S surface. We also examine the differences in behaviorbetween classical grinding aids as TIPA or TEA with respect to polycarboxylate polymersmore recently introduced in this area.

COMPUTATIONAL MODELS AND METHODS

Computer simulation provides a direct route from the microscopic details of a system (themass of the atoms, interactions between atoms, molecular geometry etc.) to macroscopicproperties of experimental interest (the equation of state, transport coefficients, structuralorder parameters and so on).18 The simulation of interfacial processes can be done at thesmall scale (


5/15

Fig. 1—Structure of tricalcium silicate (1 x 2 x 1 supercell).

Fig. 2—Single step dissolution mechanism for C 3S based on Barret et al.26,27

Fig. 3—Adsorption energy approach: molecules adsorbed on

the surface and reference state away from surface.28



6/15

Semiempirical classical simulation

Classical molecular dynamics simulations were performed using the polymer consistentforce field. The adsorption energy calculations of grinding aids on the dry and hydroxyl-ated C3S surfaces were done with 12-16% surface coverage. Here the surface coveragedefines that how much cross-sectional area is occupied by grinding aids molecules on theC3S surface. We carried out MD simulations in the NVT (N = number of atoms, V =volume of the simulation box and T = temperature of the system) ensemble with a time stepof 1 fs, and a temperature of 298.15 K maintained by the Andersen thermostat with highaccuracy (10–6 kcal/mol) for Coulomb interactions of the Ewald summation in periodicboundary condition in all three directions. For the adsorption energy calculations, a simula-tion approach was applied using the concept developed by Heinz28 as shown in (Fig. 3) andalso formulated in Eq. (2). Adsorption energies ( E ads) were calculated using the Discoverprogram24 and LAMMPS29 by comparing the energy ( E close) of adsorbed molecule close tothe surface (on the surface) and energy ( E far ) of adsorbed molecule far from the surface (20nm away from the surface) at two temperature 298.15 K and 383.15 K. Simulation run foradsorption energy calculation was total of 4 ns time where the first 2 ns time were used forequilibrium of the system and the remaining 2 ns data were used for calculation.

E ads = E close – E far (2)

Simulations with standard grinding aids

The classical grinding aids that we have considered are triethanol amine (TEA), triiso-propanolamine (TIPA) and glycerine as shown in Fig. 4. They were deposed on the dryC3S surface, (040) crystal plane and adsorption energies in vacuum were calculated forthe dosage referred to above. This was done both at ambient temperature and at 110°C,which is representative of average conditions in a ball mill. At such a temperature and fora material coming out of the kiln (1450°C) the humidity is low. However, it is not fully drybecause of small amounts of water available from different sources such as the grindingaid solution, cooling water (used occasionally to avoid overheating), some humidity, partialdehydration of gypsum etc. For this reason, we also examined the situation in which theC3S surface is partially hydroxylated as has been discussed in the molecular model section.

Fig. 4—Chemical structure of grinding aids used.

240 Mishra et al.


7/15

The simulations were performed with amounts of grinding aids per unit surface compat-ible with what is fine ultimately in cement. Values are obtained considering that typicaldosages are of about 500 g/ton of solutions containing 40% of grinding aid by solid mass.For specific surface ranging from 0.3 m2 /g by Blaine to 1 m2 /g by BET, this gives onaverage an effective grinding aid dosage 0.44 mg/m2.

Simulations with polycarboxylate oligomers

As a reminder, polycarboxylate esters are composed of comb copolymers with anadsorbing linear anionic backbone (typically a polyacrylate or polymethacrylate) as wellas non-adsorbing side chains (typically polyethylene oxides).

In order to reduce computational time, we have reduced the size of the PCEs and talkrather of oligomers. We have used a methacrylate backbone of 696 and 2056 g/mol withtwo side chain lengths of 220 and 440 g/mol. The notations used for these oligomers followthe one introduced by Gay and Raphaël’s model30 for comb-homopolymers and extendedby Flatt et al.31 for comb copolymers. In this notation, n is the number of side chains in thepolymer, P is the number of monomers in a side chain, and N is the number of backbonemonomers per side.

This means that the total number of monomers in the backbone is given by: n × N and theused notation of PCE has been shown in Fig. 5(a).

The simulations were performed with these PCEs on dry and hydroxylated C3S surfaces.We have used four PCE for the calculation of adsorption energy. Details of all the PCEs arelisted in Table 1. We have shown the chemical structure of PCE1 and PCE2 in the Fig. 5 (b)and (c). Other structures have been drawn in similar way.

We have used a similar method to calculate the adsorption energy of PCE, which isshown in Fig. 3. Here we have just selected the grinding temperature for computation. Dueto the different backbone lengths, two different sizes of the simulation box have been used.For the smaller (PCE1 and PCE3) and larger backbone (PCE2, PCE4 and PCE5), we usedsurface area of 2.5 x 2.5 nm2 and 4.8 x 7.5 nm2 respectively. In this case we get surfacecoverage of approximately 30%, which is higher than the coverage of smaller moleculesused above.

Fig. 5—(a) Structure notation shown with an example of PCE with four side chains; and

(b) chemical structure of PCE1 and (c) PCE2 grinding aids. The figures (b) and (c) were

prepared in Materials Studio.24



8/15

SIMULATION RESULTS

This section is divided into two parts. The first one deals with standard grinding aidsand the second with polycarboxylates. In the first part, in addition to adsorption energy, wepresent results about the mechanism that controls the grinding aid adsorption.

Classical grinding aids

We introduced classical grinding aids term earlier in this paper. These are the standardand older products, which are used in cement industry. Below we have discussed theadsorption energy results and mechanism of the triethanol amine (TEA), triisopropanol-amine (TIPA) and glycerine on different types of C3S surface.

Adsorption energy

The results of these calculations are summarized in Fig. 6. Values are given per unitmass because, in practice, grinding aids are compared at equal mass dosage and not equalnumber of molecules. The main conclusions that can be drawn are that:

Table 1—Details of PCE-based grinding aids

Abbreviation N × n No. of Ca++ P n Molecular weight, g/mol

PCE1 8 3 5 2 1315

PCE2 8 2 5 4 1774

PCE3 8 3 10 2 1756

PCE4 8 2 10 4 2655

PCE5 24 9 10 6 5236

Fig. 6—Adsorption energies data for TEA, TIPA, and Glyc-

erine on the dry C 3S surface and hydroxylated C 3S surface at

two temperatures 298 K and 383 K.

242 Mishra et al.


9/15

1. Adsorption energies are lower at 110°C, but remain nevertheless quite high, in therange of 20-50 kcal/mol.

2. Surface hydroxylation reduces adsorption energies, but only by about 20%.3. The trend of adsorption energy of grinding aids is as follows: Glycerine > TIPA ~

TEA. This does not follow the ranking of boiling points.4. Hydrogen bonding is responsible for better adsorption on the surface.5. Adsorption energies of TIPA and TEA on C3S surface show different trends, when

expressed per mole or per unit mass. For example, at 298 K, adsorption energies of TIPAand TEA on hydroxylated C3S surface are respectively –39 and –32 kcal/mol whenexpressed per mole and –0.20 and –0.21 kcal/g when expressed by mass.

Adsorption mechanism

We have found that the adsorption is hydrogen bond driven and does not follow theboiling points trend of grinding aids. TIPA (boiling point = 306°C) and TEA (boiling point= 335.4°C) are much more sensitive to temperature (~50%) than glycerine (~15%). Thepercentages given refer to the change in adsorption energy of TIPA and TEA between thegrinding temperature and room temperature. It was also found that the adsorption energyvaries within the range of 1 to 2 kcal/mol per molecule with different position of grindingaids on the surface. Glycerine (boiling point = 290°C) has strong interactions with thehydroxylated surface compared to TIPA/TEA, due to difference in hydrogen bond lengths.As an example, we show in the (Fig. 7) that hydrogen bonds form between alcoholic groupsof glycerine with silicate ions, and hydroxide ions on the hydroxylated C3S surface.

Polycarboxylate grinding aids

The results of PCE adsorption energy calculations are summarized in Fig. 8. These showthat the oligomer is more affected by changes in the charge density on the backbone ( N )

Fig. 7—Chemical Snapshot of glycerine on the Hyd C 3S

surface, showing the hydrogen bonds between hydroxyl

groups and silicate ions as well as hydroxyl group and

hydroxide ion at 298 K.



10/15

than on the length of the side chains (P). The reasons for this are discussed more exten-sively at the end of the next section.

Data also show that the PCEs have higher adsorption energy on dry C3S with respect tohydroxylated C3S surface. Counter ions also has an effect. Calcium ions tend to stay withthe oligomer, so that to some extent it gets repelled from the surface as shown in the Fig. 9.Longer simulation time will give a much better picture of side chain conformations as wellas backbone orientation towards surface. In the longer side chains, we also see excludedvolume interactions. There is also a minor surface reconstruction on the top of the surfacedue to the adsorption of PCE on the surface. We also found that the silicate ion also comesout from the top layer of surface.

DISCUSSION

Various types of chemicals are or have been considered as grinding aids. The most widelyused today are triethanolamine (TEA), and triisopropanolamine (TIPA). These productsidentified as efficient grinding aids due to their volatility and resulting ability to diffusefaster from one freshly cleaved surface to another through the gas phase.7,32-34 Diffusionoccurs due to the adsorption-induced mobility of grinding aids. While this concept isappealing, it lacks solid evidence and also suffers from the fact that less volatile productsas polycarboxylate comb-copolymers with much lower volatility have also been shownto be very effective grinding aids.5,6,11,12 In this paper we have used the possibility offeredby molecular modeling to calculate adsorption energies of grinding aids and thereforeexamine the claim concerning the importance of their gas phase mobility.

The results of adsorption energies on classical grinding aids show that values are higheven on hydroxylated surface and at the highest temperature. This is a first point that ques-tions the notion that volatility is an essential factor for the performance of these prod-ucts. In addition, TIPA and TEA show similar adsorption energies but not perform equally

Fig. 8—Adsorption energy of PCEs on dry and hydroxylated

C 3S surface.

244 Mishra et al.


11/15

as grinding aid. On the other hand both TIPA and TEA are reported to be clearly bettergrinding aids than glycerol, which has higher adsorption energy. In contrast, this wouldsupport the argument about the importance of gas phase mobility.

The proven efficiency of polycarboxylates as grinding aids also challenges the classicalprevailing theory that stands that mobility in the gas phase is important. However, surfacemobility may play an important role and this may be relevant for polycarboxylates. In moregeneral terms, grinding aid mobility, must involve the coexistence of diffusion (surface orgas phase) with the breaking and making of bonds, which are hydrogen bonds for classicalgrinding aids and electrostatic interactions for PCEs. In the case of classical grinding aids,we must critically examine the making and breaking of hydrogen bonds. For this we canrefer to the extensive studies by Luzar and Chandler35,36 on hydrogen bond dynamics inliquid water as well as water-dimethyl sulfoxide systems. They showed that diffusion canpersist at random with average lifetimes of hydrogen bond making and breaking. For purewater, the hydrogen bond lifetime is 1 ps. In the present study, we are not concentratingon the hydrogen bond lifetimes, as our main intention is to know the criterion of diffusionof adsorbed molecules on the C3S surface and the strength of the hydrogen bonds thatlead to adsorption. The length and strength of hydrogen bond are discussed by Steiner37 and Smallwood et al.38 for weakly and highly polar groups respectively. For highly polargroups, where short and strong hydrogen bonds form, their lengthening by 0.5 Å leadsto decrease of approximately 6 kcal/mol in the strength of the ionic hydrogen bond.39 Such issues would require further investigation in order to properly assess the mobility ofgrinding aids on the surface of cement particles.

Fig. 9—(a) Snapshot of the system PCE2-Hyd C 3S; and (b) snapshot of the system PCE4-

Hyd C 3S after 700 ps simulation time.



12/15

Other factors may however play more important roles. For example, grinding aids mayplay their greatest role in preventing agglomeration. An approach to evaluate this wouldbe to calculate the re-agglomeration energy, but other approaches may prove more rele-vant. The first step is therefore to determine the most relevant property in influencing thegrinding performance. After that it is possible to use molecular modeling to understandhow specific factors of the molecular structure affect this property.

Concerning the adsorption of PCE oligomers, we found that the PCE1 and PCE3 (twoside chains, n = 2) has higher adsorption than PCE2 and PCE4 (four side chains, n = 4) dueto higher anionic charge of backbone respectively. Plank et al.39 also described the similartrend of polycarboxylates as a superplasticizers. They also reported that the higher anioniccharge density of PCE gives stronger adsorption. In the case of PCE2 (P = 5) and PCE4(P = 10), the only difference is in the length of side chains, which also plays an importantrole due to the interactions between side chains. PCE4 has slightly less adsorption energythan PCE2 due to the steric hindrance effect, caused by the extension of graft chains ofPEO (polyethylene oxides) as shown in Fig. 9. As the length of the backbone increasesfrom n × N = 8 to 24 then adsorption of oligomers also increases due to the increase inanionic charge of backbone. We found that the adsorption of PCE on C3S surface happensmainly due to electrostatic interactions and negligible contributions from van der Waalsinteractions.

CONCLUSIONS

We calculated the adsorption energy of TIPA (Triisopropanol amine), TEA (Triethanolamine) and glycerine on the dry and hydroxylated C3S surfaces at room temperature 298 K(25°C) and grinding temperature 383 K (110°C). Simulation results show that the adsorp-tion of grinding aid is weaker on the hydroxylated surface than on the dry C3S surface. Thetrend of the adsorption energy of grinding aids is as follows: Glycerine > TIPA ~ TEA.We have also shown that the adsorption is hydrogen bond driven and it does not followthe boiling points trend of grinding aids. Adsorption strength depends on the lengths ofthe hydrogen bond. Interactions of glycerine with C3S have shorter hydrogen bond lengthsthan TIPA and TEA. Glycerine is much less sensitive to temperature than TIPA and TEAon the hydroxylated surface. From the adsorption energy results of PCEs on the hydroxyl-ated surface, we found that some of the PCE oligomers have identical adsorption energyvalues like TIPA and TEA. This could explain that PCE are also effective grinding aid,as they would maintain more mobility on the surface that might be expected. However, itshould be pointed out that PCEs used are generally larger and that the effect of PCEs ofmore realistic structures would have to be investigated. On the other hand we also can notethat PCEs adsorption energy reduces on hydroxylated surfaces in comparison to dry C3S.PCEs with higher anionic charge backbone have higher adsorption energy on C3S as couldbe expected. It is also found that there is a dependence on the length of grafted chains,but to a much lower extent. We hope that this work will motivate further research relatingatomistic simulations to problems of industrial relevance where clear opportunities forinnovation can be seen.

246 Mishra et al.


13/15

REFERENCES

1. Tokyay, M., “Effect of Chemical Composition of Clinker on Grinding Energy Require-ment,” Cement and Concrete Research, V. 29, 1999, pp. 531-535.

2. Skinner, L. B.; Chae, S. R.; Benmore, C. J.; Wenk, H. R.; and Monteiro, P. J. M.,“Nanostructure of Calcium Silicate Hydrates in Cements,” Physical Review Letters, V. 104,2010, 195502-195504.

3. Hendriks, C. A.; Worrell, E.; Price, L.; Martin, N.; Meida, L. O.; de Jager, D.; andRiemer, P., “Emission Reduction of Greenhouse Gases from the Cement Industry,”Proceedings of the Fourth International Conference on Greenhouse Gas Control Technolo-

gies, Interlaken, Switzerland , Aug. 30-Sept. 2, IEA GHG R&D Programme, UK, 1998.4. Schneider, M.; Romer, M.; Tschudin, M.; and Bolio, H., “Sustainable Cement Produc-

tion—Present and Future,” Cement and Concrete Research, V. 41, 2011, pp. 642-650.5. Jost, P., Schrabback, J. M., “Creative Grinding Solutions,” International Cement

Review, Sept. 2007, pp. 70-72.6. Schrabback, J. M., “Polycarboxylate Polymer-Powered Grinding Efficiency,” Global

Cement Magazine, July/Aug. 2009, pp. 14-16.7. Sohoni, S.; Sridhar, R.; and Mandal, G., “The Effect of Grinding Aids on the Fine

Grinding of Limestone, Quartz and Portland Cement Clinker,” Powder Technology, V. 67,1991, pp. 277-286.

8. Moothedath, S. K., and Ahluwalia, S. C., “Mechanism of Action of Grinding Aids inComminution,” Powder Technology, V. 71, 1992, pp. 229-237.

9. El-Shall, H., and Somasundaran, P., “Physico-chemical Aspects of Grinding: A Reviewof Use of Additives,” Powder Technology, V. 38, 1984, pp. 275-293.

10. Rehbinder, P. A., “On the Influence of Changes in the Surface Energy on the Cleavage,Hardness and Other Properties of Crystals,” P. A. Rehbinder. Selected Transactions. Physi-cochemical Mechanics (Nauka, Moscow) V. 2, 1979, pp. 1-142.

11. Sakai, E.; Kawakami, A.; and Daimon, M., “Dispersion Mechanism of Comb-TypeSuperplasticizers Containing Grafted Poly(ethylene oxide) Chains,” MacromolecularSymposia, V. 175, 2001, pp. 367-376.

12. Fuerstenau, D. W.; Venkataraman, K. S.; and Velamakanni, B. V., “Effect of ChemicalAdditives on the Dynamics of Grinding Media in Wet Ball Mill Grinding,” International Journal of Mineral Processing, V. 15, 1985, pp. 251-267.

13. Flatt, R. J., and Houst, Y. F., “A Simplified View on Chemical Effects Perturbing theAction of Superplasticizers,” Cement and Concrete Research, V. 31, 2001, pp. 1169-1176.

14. Sakai, E.; Yamada, K.; and Ohta, A., “Molecular Structure and Dispersion-Adsorp-tion Mechanisms of Comb-Type Superplasticizers Used in Japan,” Journal of AdvancedConcrete Technology, V. 1, No. 1, 2003, pp. 16-25.

15. Mishra, R. K.; Flatt, R. J.; and Heinz, H., “Cleavage Energy of Tricalcium Sili-cate and Interactions with Amine Additives,” Proceedings of the XIII ICCC InternationalCongress on the Chemistry of Cement , Madrid, Spain, 3-8 July, 2011.

16. Taylor, H. F. W., Cement Chemistry, second edition, Academic Press, London, 1997.17. Kundu, T. K.; Rao, K. H.; and Parker, S. C., “Competitive Adsorption on Wollas-

tonite: An Atomistic Simulation Approach,” The Journal of Physical Chemistry B, V. 109,2005, pp. 11,286-11,295.



14/15

18. Allen, M. P., and Tildesley, D. J., Computer Simulation of Liquids, Oxford UniversityPress, 2006.

19. Mumme, W. G., “Crystal Structure of Tricalcium Silicate from a Portland CementClinker and Its Application to Quantitative XRD Analysis,” Neues Jahrbuch fur Miner-alogie, V. 4, 1995, pp. 145-160.

20. Jeffery, J. W., “The Crystal Structure of Tricalcium Silicate,” Acta Crystallographica,V. 5, 1952, pp. 26-35.

21. Heinz, H.; Koerner, H.; Anderson, K. L.; Vaia, R. A.; and Farmer, B. L., “Force Fieldfor Mica-Type Silicates and Dynamics of Octadecylammonium Chains Grafted to Mont-morillonite,” Chemistry of Materials, V. 17, No. 23, 2005, pp. 5658-5669.

22. Heinz, H., and Suter, U. W., “Atomic Charges for Classical Simulations of PolarSystems,” The Journal of Physical Chemistry B, V. 108, 2004, pp. 18,341-18,352.

23. Heinz, H.; Vaia, R. A.; Farmer, B. L.; and Naik, R. R., “Accurate Simulation ofSurfaces and Interfaces of Face-Centered Cubic Metals Using 12-6 and 9-6 Lennard-JonesPotentials,” The Journal of Physical Chemistry C , V. 112, 2008, pp. 17,281-17,290.

24. Materials Studio and Discover program, Accelrys, Inc. 2006.25. Fu, Y. T., and Heinz, H., “Cleavage Energy of Alkylammonium-Modified Montmo-

rillonite and Relation to Exfoliation in Nanocomposites: Influence of Cation Density, HeadGroup Structure, and Chain Length,” Chemistry of Materials, V. 22, 2010, pp. 1595-1605.

26. Barret, P.; Menetrier, D.; and Bertrandie, D., “Mechanism of C3S Dissolution andProblem of the Congruency in the Very Initial Period and Later On,” Cement and Concrete Research, V. 13, 1983, pp. 728-738.

27. Taylor, H. F. W. et al., ““The Hydration of Tricalcium Silicate,” RILEM Committee68-MMH, Task Group 3,” Materials and Structures, V. 17, 1984, pp. 457-468.

28. Heinz, H., “Computational Screening of Biomolecular Adsorption and Self-Assemblyon Nanoscale Surfaces,” Journal of Computational Chemistry, V. 31, 2010, pp. 1564-1568.

29. Plimpton, S. J., “Fast Parallel Algorithms for Short-Range Molecular Dynamics,” Journal of Computational Physics, V. 117, 1995, pp. 1-19.

30. Gay, C., and Raphaël, E., “Comb-Like Polymers Inside Nanoscale Pores,” Advancesin Colloid and Interface Science, V. 94, 2001, pp. 229-236.

31. Flatt, R. J.; Schober, I.; Raphael, E.; Plassard, C.; and Lesniewska, E., “Conforma-tion of Adsorbed Comb Copolymer Dispersants,” Langmuir , V. 25, 2009, pp. 845-855.

32. Serafin, F. G., “Amide Grinding Aid,” U.S. Patent, 3,459,570, 1969.33. Gartner, E., and Myers, D., “Influence of Tertiary Alkanolamines on Portland Cement

Hydration,” Journal of the American Ceramic Society, V. 76, No. 6, 1993, pp. 1521-1530.34. Ramachandran, V. S., “Hydration of Cement—Role of Triethanolamine,” Cement

and Concrete Research, V. 6, 1976, pp. 623-632.35. Luzar, A., and Chandler, D., “Hydrogen-Bond Kinetics in Liquid Water,” Nature, V.

379, 1996, pp. 55-57.36. Luzar, A., and Chandler, D., “Structure and Hydrogen Bond Dynamics of Water-

Dimethyl Sulfoxide Mixtures by Computer Simulations,” The Journal of ChemicalPhysics, V. 98, No. 10, 1993, pp. 8160-8173.

37. Steiner, T., “The C—H····O Hydrogen Bond in (dicyanomethyl)ammonium p-tolu-enesulfonate,” Acta Crystallographica, V. C57, 2001, pp. 775-776.

248 Mishra et al.


15/15

38. Smallwood, C. J., and McAllister, M. A., “Characterization of Low-Barrier HydrogenBonds. 7. Relationship between Strength and Geometry of Short-Strong Hydrogen Bonds.The Formic Acid-Formate Anion Model System. An ab Initio and DFT Investigation,” Journal of the American Chemical Society, V. 119, 1997, pp. 11,277-11,281.

39. Plank, J., and Hirsch, C., “Impact of Zeta Potential of Early Cement HydrationPhases on Superplasticizer Adsorption,” Cement and Concrete Research, V. 37, 2007, pp.537-542.


Understanding the Effectiveness of Polycarboxylates as Grinding Aids.pdf

Documents