Scholarly article on topic 'Resistance of Three-Component Cement Binders in Highly Chemically Corrosive Environments'

Resistance of Three-Component Cement Binders in Highly Chemically Corrosive Environments Academic research paper on "Civil engineering"

Share paper
Academic journal
Procedia Engineering
OECD Field of science
{mortars / "silica fume ;blast furnace slag" / "durability ;sulfate ammonium"}

Abstract of research paper on Civil engineering, author of scientific article — Maria Fiertak, Teresa Stryszewska

Abstract The paper presents the results of a study of three-component binders involving Portland cement, pulverised blast-furnace slag and silica fume and their chemical resistance to ammonium sulphate. The study involved samples with a range of different mass-ratios between the components. The research programme on samples exposed to highly chemically corrosive environments included compression testing after 28 and 90 days of ageing and corrosion testing, including measurements of weight change and linear distortion. Samples were also analysed for the microstructural effects of ammonium sulphate using a scanning microscope with an EDS sound.

Academic research paper on topic "Resistance of Three-Component Cement Binders in Highly Chemically Corrosive Environments"

Available online at

SciVerse ScienceDirect Procedia


Procedia Engineering 57 (2013) 278 - 286 =

www. el sevi er. com/1 ocate/procedi a

11th International Conference on Modern Building Materials, Structures and Techniques,

MBMST 2013

Resistance of Three-Component Cement Binders in Highly Chemically

Corrosive Environments

Maria Fiertaka, Teresa Stryszewskab*

a bFaculty of Civil Engineering, Cracow University of Technology, 24 Warszawska Street, 31-155 Cracow, Poland


The paper presents the results of a study of three-component binders involving Portland cement, pulverised blast-furnace slag and silica fume and their chemical resistance to ammonium sulphate. The study involved samples with a range of different mass-ratios between the components. The research programme on samples exposed to highly chemically corrosive environments included compression testing after 28 and 90 days of ageing and corrosion testing, including measurements of weight change and linear distortion. Samples were also analysed for the microstructural effects of ammonium sulphate using a scanning microscope with an EDS sound.

© 2013 TheAuthors. Published by ElsevierLtd.

Selection and peer-review/ under responsibility of the Vilnius GediminasTechnicalUniversity Keywords: mortars; silica fume;blast furnace slag; durability;sulfate ammonium.

1. Introduction

Concrete is one of the most popular construction materials. Its composition can be modified to achieve properties suited to a wide range of operating conditions desired. These range from common uses in housing and road building to the most demanding ones with corrosive environments, for example use in waste water treatment plants, industrial chimney sand cooling towers [1-3].

Existing studies on the durability of cement materials in corrosive environments reveal that binding matrixes containing a significant component of pulverised blast-furnace slag display high resistance to corrosion [4-5]. Cement matrixes with slag have a number of useful properties when compared to pure Portland cement due to the replacement of a portion of the Portland clinker by the slag component [6-7]. This causes a significant change in the cement's chemical composition, primarily by reducing the content of calcium oxide. Also the CaO/SiO2molar ratio is lowered and, consequently, also the CH concentration in the hydrated matrix [8].

Taking into account the properties of binders containing pulverised slag it could have been expected that the introduction of another mineral additive, the silica fume, would offer additional benefits. Silica fume (SF) has a beneficial effect on a number of properties of cement binders due to its chemical and phase composition and its large specific surface area (SSA) [9-12]. Silica fume plays the role of a micro filler and active pozzolana in cement binders. The small size of the particles helps fill the empty spaces between the large grains of clinker and slag, thus sealing the physical structure of the binder. It also enters in a pozzolan reaction with CH resulting in additional products that fill in the gaps in the cement matrix [9], [13]. A combination of Portland cement, granulated slag and silica fume produces very densely packed binders [10], [14].

Considering the benefits of a partial replacement of Portland clinker with pulverised blast-furnace slag and the properties of silica fume it could be expected that the resulting three-component binder would have increased utility parameters [15], [16]. This should be particularly true as far as its durability in environments particularly corrosive to concrete is concerned [17-18], [1].

* Corresponding author.

E-mail address:


1877-7058 © 2013 The Authors. Published by Elsevier Ltd.

Selection and peer-review under responsibility of the Vilnius Gediminas Technical University doi: 10.1016/j.proeng.2013.04.038

2. Description of research

2.1. Corrosive environment

A 10% solution of ammonium sulphate was used due to its highly corrosive effect on concrete. It is a salt that is easily soluble in water and its hydrolysis lowers the solution's pH by ca. 3. Products of the hydrolysis involve three types of reaction, including as an acid (pH< 3), sulphate and ammonium. Ion concentrations include SO42- at 72000 mg/dm3 and NH4+ at 28000 mg/dm3.

2.2. Materials tested

The testing involved mortar with standard sand and binder in a 3:1 ratio. Corrosion tests were performed on small beams with a size of 2^2^16 cm, while compressive strength tests were carried out on beams with a size of 4x4x16 cm. The compositions of three-component cement binders are shown in Table 1, while oxide composition is shown in Table 2.

Table 1. Mixture of cement paste

Table 2. Oxide composition of cement binders

Portland Slag Addition Identification

Clinker Furnace % Silica Fume of Samples Chemical Binder

100 0 C/0 composition [%] C/0 C/60 C/90

100 - 3 C/0/3 CaO 63,05 51,94 45,9

100 - 5 C/0/5 MgO 1,32 3,7 5,59

100 - 10 C/0/10 SiO2 18,74 30,8 35,4

40 60 0 C/60 Al2O3 4,90 6,46 6,86

40 60 3 C/60/3 Fe2O3 3,17 1,76 1,27

40 60 5 C/60/5 Na2O 0,13 0,11 0,12

40 60 10 C/60/10 K2O 0,89 0,75 0,79

10 90 0 C/90 SO3 2,8 1,77 1,85

10 90 3 C/90/3 Cl- 0,01 0,048 0,09

10 90 5 C/90/5

10 90 10 C/90/10

Table 3. Compressive strength of binders

Binder fc [MPa]

After After After After

7 days 14 days 28 days 90 days

C/0 23 27 30,2 30,3

C/0/3 25 27 30,2 30,4

C/0/5 26 28,7 34,0 35,5

C/0/10 28 35 35,5 38,0

C/60 19,5 26 28,1 29,8,

C/60/3 21,5 29 32,0 35,0

C/60/5 23 29,5 32,0 36,1

C/60/10 28,2 30 36,1 39,7

C/90 5,5 9,5 9,5 10,2

C/90/3 14 14,5 16,6 16,4

C/90/5 18,8 19,5 19,5 20,0

C/90/10 18 19,5 21,6 23,2

The mortar samples were matured for 90 days in laboratory conditions, and then exposed to a 10% solution of ammonium sulphate for 22 weeks, during which they were weighed for mass change and measured for length every seven

Additionally, after 28 and 90 days of ageing, the 4x4x16cm elements were broken into cubes and tested for compressive in accordance with the PN-EN 196-1:2006standard [19]. The results are shown in Table 3 as averages of 10 measurements.

The compressive strength of all mortars increases along with increase of SF content. The highest compressive strength (35,5 and 39,1 MPa) obtained after 28 days of setting is related to C/0 and C/60 containing 10% SF.

The lowest compression strength value, in each setting period were obtained in case of mortars based on binder C/90. Tests results obtained prove the lack of compressive strength increases versus time. After 14 days there were not significant changes of compressive strength observed.

3. Test results

3.1. Scanning microscopic tests

Microscopic observation was performed on the samples both prior to and after the exposure to corrosion. Microstructural scanning images were taken of selected spots, while an X-ray analysis was performed using an EDS sound at strategic points on the samples. The original texture was determined, as were changes resulting from the exposure to a corrosive environment. The images and the EDS analysis were then used to describe changes in the morphology and structure of the CSH phase, both in the presence of the mineral additives and resulting from sulphate-ammonia corrosion. Other corrosion products were also identified.

The following conclusions were drawn from the microstructure observations and EDS analysis: • In C/0/0 binders a relatively porous CSH phase dominates with primary ettringite in the pores and in the contact zone, see Fig. 1.

Fig. LMicrostructure C/0, mag.2000x

• In C/60/0 binders the CSH phase is more compact, but primary ettringite is also present in the pores and the contact zone

(Fig. 2).

Fig. 2. Microstructure C/60, mag.2000x

• In the C/90/0 binders with the highest concentration of blast-furnace slag, the gel CSH phase is dominant (Fig. 3) and the free spaces are not entirely filled up with the products of hydration.

Fig. 3. Microstructure C/90, mag.2000x

• The presence of both mineral additives, as in C/60/10, reduces the amount of primary ettringite and portlandite, while the gel CSH phase has negligible porosity and low calcium ion concentration, see Fig. 4. A matrix that uses C/60/10 is very tightly packed. Some non-reactive hydrogarnets were observed.

Fig. 4. Microstructure C/60/10, mag.2000x

• In binders with C/90/10 the CSH phase is compact and, according to the chemical analysis, low in calcium and with trace quantities of primary ettringite, see Fig. 5.

Fig. 5. Microstructure C/90/10, mag.2000x

Based on an analysis of all scanning images of the samples after 22 weeks of exposure to corrosion it was found that: • The damage to C/0 binders was high and their CSH phase was highly porous. The EDS chemical analysis shows a reduced calcium ion concentration and numerous products of sulphate corrosion, i.e. gypsum and ettringite. The main compound present in the contact zone and in the pores was gypsum, see Fig. 6. The cement binder became much looser.

Fig. 6. Microstructure C/0, mag.2000x

Clear zonality of corrosion developed in C/0/10 binders. In the external zones gypsum was the corrosion product, while ettringite was found closer to the centre of the samples. Across the volume ettringite and gypsum were found within the CSH phase, see Fig. 7.

Fig. 7. Microstructure C/0/10, mag.2000x

In C/60 binders porosity was much increased, gypsum was observed throughout the samples, but no secondary ettringite. Grains of gypsum were observed in the external zones of C/60/10 binders, as well as in the contact zone where isolated grains had crystallised, see Fig. 8. No corrosion products were found closer to the centre of the samples, which only displayed minor loosening of their structures.

Fig. 8. Microstructure C/60/10, mag.2000x

In C/90 binders corrosion processes caused a loosening of the CSH phase structure, as a result of leaching of calcium ions, see Fig. 9. In SF-modified C/90 binders no corrosion products, such as gypsum and secondary ettringite, were observed.

Fig. 9. Microstructure C/90/10, mag.2000x

All the samples exposed to ammonium sulphate solution were found to be corroded. The effects of acid corrosion were observed including the dissolution products of portlandite and the effects of sulphate corrosion. This primarily included a degraded CSH phase, which became more porous through leaching and overgrown with the products of corrosion. Such products included secondary ettringite and gypsum, which mainly crystallised in pores and in the contact zone. Microscopic observations revealed that the structure of the C/60/10 binder was the most compact and best sealed and its CSH phase was a compact gel of negligible porosity. This tightly sealed material had a slowing effect on corrosion, which also displayed a clear zonality. Its products were particularly clearly visible in the external zones and mostly involved gypsum, while towards the centre of the sample only trace quantities of secondary ettringite were observed. The least change was observed in the microstructure of cement binders using C/90. It only involved increased porosity and loosening of the material's structure due to the leaching of easily soluble components, but no products of sulphate corrosion were identified.

3.2. Macroscopic observations

The destruction of mortars exposed to ammonium sulphate is caused by acid, sulphate and ammonium corrosion. The degree of degradation reflects processes linked with dissolution and expansion. Cracks develop on the surface of samples as a result of the development of sparingly soluble compounds, such as gypsum and ettringite. On the other hand, the easily soluble compounds, which also develop are washed out leaving empty spaces. This considerably loosens the cement structure leading to a deterioration of its original parameters. Examples of images of samples after 22 weeks ofexposureto corrosion are shown in Figures 10-15.

Fig. 13.C/0/10


The greatest degree of deformation was observed in binders made with pure Portland cement (C/0) and with a 60% addition of slag (C/60). Binders with high slag concentrations (C/90) remained unchanged. Samples with silica fume (C/0 and C/60)showed a significant reduction in deformation, especially in a binder with 60% slag and 10% silica fume. On the other hand, C/90 binders with the addition of SF showed visible cracking that was not observed in pure C/90 binders.

3.3. Mass change and deformation

This study was based on the assumption that the resistance of the materials tested to corrosive environments would be measured by their change in mass and distortion over time. The results of these measurements are shown in Figures 16-17.

10 8 6 4

□ C/0

□ C/60 ■ C/90

0% SF 3% SF

10% SF

Fig. 16. Change of mass of samples immersion in ammonium sulfate solution, after 22 weeks

„ 70

K 4) 20

□ C/0

□ C/60 ■ C/90

0% SF 3% SF

5% SF 10% SF

Fig. 17.Expansion of samples immersion in ammonium sulfate solution, after 22 weeks

The tests clearly showed that the mineral additives increased the resistance of cement binders to the corrosive effects of ammonium sulphate. Pulverised blast-furnace slag was found to be particularly effective in raising the chemical resistance of cement binders. The best results were observed in samples made of C/90 without an addition of SF, where linear deformation was reduced by 90% and the change of mass by 75% in comparison to samples made of pure Portland cement. Despite its superior resistance to corrosive environments, however, this binder cannot be used as a construction material due to its very low compressive strength of just 10 MPa after 90 days maturation. An introduction of SF to C/90 binders (at 3, 5 and 10%) reduced their resistance, as linear deformations and mass growth increased proportionally to SF concentration.

An addition of 60% slag (C/60) reduced deformation by ca. 25% and mass change by ca. 35%, when compared with samples made with C/0.

The optimal composition of the binder matrix, which ensured the highest resistance to chemical corrosion while maintaining the required compressive strength, was a combination of Portland clinker with pulverised blast-furnace slag at the ratio of 40:60 and 5 or 10% SF. Samples made with this binder had 75% lower deformations and 50% lower mass increase when compared to samples made of pure C/0.

4. Discussion of results

The results shown above indicate that three-component binders are generally more resistant to the corrosive effects of ammonium sulphate. The beneficial influence of blast furnace slag on binder resistance to corrosion specifically involves a reduction in the concentration of crystallised portlandite in the cement matrix. Crystallised portlandite is highly susceptible to acidic and sulphate corrosion resulting in the production of compounds which are both easily and sparingly soluble, such as gypsum and ettringite. In addition, the introduction of blast furnace slag in a cement matrix helps the development of more neutral and chemically resistant phases, such as hydrogarnets and gehlenite hydrate. Generally, the more slag in a matrix the greater its resistance.

An addition of silica fume to C/0 and C/60 binders improves their resistance to ammonium sulphate. This additive by itself is not as effective as slag and its effect is virtually negligible at 3%, only beginning to be discernible at 5% concentration.

The study has shown that the best results in improving the durability of cement binders are achieved with a combined addition of both blast-furnace slag and silica fume. A high concentration of slag makes the binder resistant to corrosive environments, while silica fume closes the structure physically and acts as pozzolana. In reaction with portlandite SF produces additional quantities of the CSH phase, which directly improves the durability and resistance. The study has also shown that an increase in the SF concentration from 5 to 10% produced no significant additional benefits in either durability or resistance. It follows, therefore, that the optimum SF concentration for the mechanical parameters, durability and economy is 5%.

Binders with high slag concentration (C/90) behaved differently. When no SF was added, they showed no change in the parameters measured, but an addition of silica fume caused a drop in resistance. It was also noted that the change of mass and deformation, and thus resistance to corrosive environments, was proportional to the SF concentration. Silica fume in cement matrixes plays a role of both a micro filler and pozzolana, but in C/90 binders its effect is basically reduced to a physical sealing of the structure. Therefore, despite the low reactivity of the C/90 binder, its microstructure density gain from a greater SF concentration reduces its capacity to buffer the emerging products of corrosion leading to increased deformations and an increase in mass. In C/90 binders the potential of silica fume to induce the pozzolan reaction is radically decreased due to a negligible concentration of portlandite in the matrix. This explains the very low gain in resistance of C/90 binder modified by silica fume.

5. Conclusion

The application of three-component binders in cement composites results in the development of a compact gel CSH phase (produced in the presence of Portland cement, blast furnace slag and silica fume) with a reduced ratio of CaO to SiO2. This lower quantity of reactive phase (i.e. calcium hydroxide caused by the lower alite concentration and the pozzolan reaction with SF), and the sealing of the structure with the micro filler, significantly reduce porosity and improve the composite's resistance, thus improving the material's performance in corrosive environments.


[1] Hooton, R. D., 1993. Influence of silica fume replacement of cement on physical properties and resistance to sulphate attack, freezing and thawing and

alkali reactivity, ACI Materials Journal 90, pp. 143-151.

[2] Melchers, R.E., Pape, T., 2011. Aspects of long-term durability of reinforced concrete structures in marine environments, European Journal of

Environmental and Civil Engineering 15(7), pp. 969-980.

[3] Fiertak, M., Stryszewska, T., 2005. Freeze-thaw resistance of cement concrete with silica fume addition, Cement Lime Concrete6, pp. 307-313.

[4] ACI 201. 2R-1992.Guide to durable cement, ACI Manual of concrete Practice, Part 1: Materials and General Properties of Concrete, Detroit Michigan

[5] Peukert, S., 2000. General and special cements. Polish Cement, Cracow (in Polish).

[6] Moon, H.Y., Lee, S.T., Kim, H.S., Kims, S.S., 2012. Experimental Study on the Sulfate Resistance of Concrete Blended Ground Granulated Blast-

furnace Slag for Recycling, Geosystem Engineering 5(3), pp. 67-73.

[7] Behim, M., Cyr, M., Clastres, P., 2011. Physical and chemical effects of El Hadjar slag used as an additive in cement-based materials, European Journal

of Environmental and Civil Engineering 15(10), pp.1413-1432.

[8] Scislewski, Z., 1995. Durability of structures.Kielce University of Technology. Kielce (in Polish).

[9] Nocuñ-Wczelik, W., 2005. Silica fume- properties and applications in concrete,Polish Cement, Cracow (in Polish).

[10] Nagataki, S., Wu, C., 1995. "A study of properties of portland cement incorporating silica fume and blast furnace slag", Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, proceedings 5, International Conference, ACI SP 153, edited by Malhorta, Milwaukee.

[11] Boudchicha, A., Zouaoui, M., Gallias, J.L., Mezghiche, B., 2007.Analysis of the effects of mineral admixtures on the strength of mortars: Application of the predictive model of Feret,Journal of Civil Engineering and Management13(2), pp. 87-96.

[12] Bubshait, A., Bassam, M., Tahir, B.M., Jannadi, M.O., 1996. Use of microsilica in concrete construction. Reviews state of the art silica fume concrete and discusses the influence silica fume has on the various properties of concrete and the effect on the bond between parent concrete and new concrete, Building Research & Information24(1), pp. 41-49.

[13] Zhang, C., Wang, A., Tang, M., 1996. The filling role of pozzolanic material, Cement and Concrete Research26(6), pp. 943-947.

[14] Chatterji, S., Thaulow, N., Christensen, P., 1982. Puzzolanic Activity of byproduct Silica Fume From Ferro-Silicon Production, Cement Concrete Research 12, pp. 781-784.

[15] Bleszynski, R., Doug Hooton, R., Thomas, M., Rogers, C.A., 2002. Durability of Ternary Blend Concrete with Silica Fume and Blast-Furnace Slag: Laboratory and Outdoor Exposure Site Studies,Materials Journal 99(5), pp. 499-508.

[16] Ahmed, S.F.U., Demura, K., Obama,Y., 1999. Comparison of mechanical properties and durability of mortar modified by silica fume and finely ground blast furnace slag, Journal of Civil Engineering27(2), pp. 143-154.

[17] Stryszewska, T., 2005. Effect of silica fume on the properties of concrete and mortar based on blast furnace slag. PhD dissertation. Cracow University of Technology, Cracow (in Polish).

[18] Ramezanianpour, A.A., Pourkhorshidi, A. R., 2004. Durability of concretes containing supplementary cementing materials under hot and aggressive environment, ACI SP221, pp. 633-646.

[19] PN-EN 196-1:2006 method of testing cement- Part 1 Determination of strenght.

[20] Rouseková, I., Bajza, A., Zivica, V., 1997. Silica fume-basic blast furnace slag systems activated by an alkali silica fume activator, Cement and Concrete Research 27(12), pp. 1825-1828.

[21] Skripkiünas, G., Rudzionis, Z., Vaitkevicius, V., 2002. Complex admixtures for high strength concrete,Journal of Civil Engineering and Management 8(4), pp. 276-280.

[22] Langan, B.W., and Ward, M.A., 1996. The strength and mirostructure of high strength paste containing silica fume, Cement, Concrete, and Aggregates 18, pp. 112-117.