Indian Journal of Engineering & Materials Sciences Vol. 23, February 2016, pp. 88-100

Durability of composite cements containing granulated blast-furnace slag and silica nano-particles

Mohamed Heikala,b*, S Abd El Aleemc & W M Morsid aChemistry Department, College of Science, Al Imam Mohammad Ibn Saud, Islamic University (IMSIU),

P O Box 90950, Riyadh 11623, Saudi Arabia bChemistry Department, Faculty of Science, Benha University, Benha, Egypt

cChemistry Department, Faculty of Science, Fayoum University, Fayoum, Egypt dBuilding Physics Institute BPI, Housing and Building National Research Center, HBRC, Dokki, Giza 11511, Egypt

Received 22 December 2014; accepted 26 October 2015

The durability of concrete has been a major concern of civil engineering. Chemical attack caused by aggressive waters is one of the factors causing damage to concrete. The effect of nano-silica (NS) on the durability of composite cement pastes and mortars cements containing granulated blast-furnace slag (GBFS) subjected to seawater attack has been studied. Chemically combined water, free lime, total chloride and sulphate contents, as well as bulk density, compressive and flexural strengths have been determined. The durability of cement pastes has been monitored using SEM and XRD techniques. The obtained results indicate that NS improves the compressive and flexural strengths subjected to seawater up to 12 months. NS decreases the accessibility of SO4

2- and Cl- to penetrate the pore system, hence the total sulfate and total chloride contents decrease. The accumulation of additional hydration products within the pore system enhances the densification of cement paste matrix to form closed and compact structure with narrow pores.

Keywords: Granulated blast-furnace slag, Nano-silica, Composite cements, Durability, Microstructure

With the expected increase in cement production to meet the need of steadily growing world population and with the urgent need to reduce the amount of energy consumed and CO2 released in the air, there will be increase of the pressure to reduce cement consumption. The problem of producing blended cements has been of considerable scientific and technological interest, because they need less energy for production1-4.

GBFS used as a supplementary cementitious material (SCM) in concrete, either as a mineral admixture or a component of blended cements5,6. GBFS is a by-product formed in the iron manufacture from the fusion of limestone with ash from coke and the siliceous as well as aluminous residue remaining from the iron ore after the reduction to iron metal. In this process, a molten slag forms as nonmetallic liquid that floats on the top of the molten iron, then sudden cooled. Depending on the cooling mode, three types of slag are produced, i.e., air-cooled slag (ACS), expanded or foamed slag and granulated slag7-9. The rapid cooling of molten slag by water prevents the formation of large crystals, and the resulting slag

normally contains more than 95% of glass (amorphous calcium aluminosilicates).

Nanomaterials are gaining widespread attention to be used in construction to exhibit enhanced performance in terms of smart functions and sustainable features. Nanotechnology has changed our vision, expectations and abilities to control the material world10-14. The engineering characteristics of building materials containing nano-materials will generate durable and stronger cement based-materials with smart properties15. Nano-particles have become a focus in the development of new accelerators for cement hydration, which improve the hydration characteristics, physico-mechanical properties, microstructure, sulfate resistance, the pozzolanic as well as filler effects, thermal expansion, reduction of capillary pores, corrosion resistance of steel bars embedded, chloride penetration resistance, and to a refinement of the pore structure of composite cement pastes16-30.

Nanomaterials can change the concrete world, due to their unique physico-chemical properties, which are different from the conventional materials29,30. The concrete properties are affected by the performance of the nano-material. The main hydration product of cement-based materials is the C–S–H gel, which is a

————— *Corresponding author (E-mail: ayaheikal@hotmail.com)

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

89

nano-structured material31,32. The mechanical properties, firing resistance and the durability of concrete depend mainly on the refinement of the microstructure of the hardened cement paste and the improvement of the paste-aggregate interfacial transition zone16-18,33.

Effect of nano-silica (NS) particles on the mechanical and microstructural properties of ultra-high performance cementitious composites was discussed21,22. The hydration process was accelerated by the addition of NS. The porosity and the average pore diameter decreased with the increase of NS content and aging. The effect of NS on the hydration characteristics, thermal expansion and microstructure of cement pastes and mortars was also studied21. NS is highly pozzolanic activity forms additional calcium silicate hydrate (C–S–H gel) improving the microstructure34,35-39. NS enhances the compressive strength, reduces the permeability, sulfate resistance and consequently improves the overall durability of hardened concrete15,20,40. The addition of NS reduces the corrosion resistance of steel bars embedded in ultra-high performance concrete effectively23.

The superior mechanical properties and higher durability of ultra-high performance concrete have been discussed24. The addition of NS particles leads to a reduction of capillary pores and to a refinement of the pore structure of ultra-high performance. It has also been proved that the inclusion of NS improves the durability of concrete25,26. Silica nano-particles addition improves the chloride penetration resistance, enhances the chloride and electrical resistance of concrete as well as the general ionic permeability of Portland cement based mortars27-29. The reduction in pore volume, pore-size distribution has become finer and the improvement in physico-mechanical properties of the mortars after the addition of nano-SiO2 (NS), nano-Al2O3 (NA) and nano-Fe2O3 (NF) powders could be explained by the filler effect or amount of hydration products of cement19.

Concrete durability has been a major concern of civil engineering professionals, a considerable scientific and technological interest; because sea structures are exposed to simultaneous action of chemical deterioration of the harmful combined effects seawater constituents on cement hydration products and the corrosion of reinforcing steel. The sulphate and chloride ions can enter into deleterious reactions leading to the dissolution of CH and precipitation of gypsum as well as sulfoaluminates hydrates (ettringite and monosulfate) and chloroaluminate

hydrate (Friedel’s salt). These products cause expansion and softening of concrete, respectively41-43. CaCl2 formed from the reaction of MgCl2 with liberated lime increases the leaching and the porosity of concrete, leading to strength loss. Mg(OH)2 dissociates C–S–H which produces CH and silica gel44. The latter may react with Mg(OH)2 to form magnesium silicate hydrate (Mg–S–H). Also, C–S–H in hydrated cement is decomposed by MgSO4 in aggressive solutions to give gypsum, which is known as decalcification hydrated silica and Mg–S–H, but the later, unlike silica gel, has little or no binding properties. MgSO4

has a more aggressive effect than Na2SO4, because MgSO4 decomposes the C–S–H during long-term exposure to produce Mg–S–H and CH. The presence of Mg2+ as MgSO4 or/and MgCl2 leads to the distortion of C–S–H through a dissolution of Ca2+ and/or an exchange with Mg2+ to give also Mg–S–H and CH. All the above mentioned reactions are accompanied by decrease in strength 45. Ammonium, magnesium, calcium and sodium sulphate are dangerous on concrete leading to expansion, cracking, spalling and loss on strength41.

The chemical resistance of blended cements results mainly from the pozzolanic reaction leading to the consumption of free CH and formation of additional amounts of calcium silicate and aluminosilicate hydrates. These products fill up some of open pores forming more homogeneous and compact microstructure. Therefore, the blended cement pastes show good resistivity for sulphate and chloride attack.

Despite the presence of several studies describing the main properties and characteristics of concrete containing NS particles, most of these studies focus on the application of NS and concrete durability46–52. This work aims to study the characteristics and durability of NS-blended cements in seawater. Experimental Procedure Materials

The starting materials used were Type I ordinary Portland cement (OPC), granulated blast-furnace slag (GBFS), nano-silica (NS) and polycarboxylate superplasticizer. OPC Type I was provided from Beni-Suief Portland Cement Company and GBFS was supplied from Iron and Steel Company, Helwan, Egypt. Chemical analyses of OPC, GBFS and NS are given in Table 1. The Blaine surface area of OPC and GBFS were 3050 and 4000 cm2/g, respectively. NS was supplied from nano-technology laboratory, Beni-

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

90

Suief University, Beni-Suief, Egypt. NS was used in powder form; with an average particle size of 15 nm. XRD, TEM and SEM of NS are given in Figs. 1 and 2. The Blaine surface area of NS was 198 m2/g. Conplast SP-610 superplasticizer was obtained from Fosroc Company, 6 October City, Egypt. It is an opaque light yellow liquid with density 1.08 g/mL and chloride content < 0.1 mass%. Experimental techniques

The mix composition of the prepared cement blends is given in Table 2. The cement blends were mixed in a rotary mixer according to the following sequence: (i) NS was stirred with 1.0% superplasticizer by mass of cement powder and 25% of mixing water at high speed of 120 rpm for 2 min,21,35 (ii) the cement and the residual amount of mixing water were added to the mixer at medium speed (80 rpm) for 1 min, (iii) the mixture was allowed to rest for 90 s and then mixed for 1 min at high speed. For the preparation of mortars, the sand was added gradually and mixed at medium speed for additional 30 s after step (ii). The mortars were prepared by mixing 1 part of cement and 2.75 parts of standard sand with water content to obtain a flow of 110±5 with 25 drops of the flowing table45. Freshly prepared cement mortars were molded in 50×50×50 mm cubic moulds, and then vibrated for few minutes for better compaction. The mortars are cured in humidity chamber for 24 h, then remolded and cured under tap water for 28 days (zero time), then immersed in seawater for 1, 3, 6, 9 and 12 months42,45.

The required water of standard consistency was determined according to ASTM specification53. The hydration of cement pastes was stopped using 1:1

methanol–acetone mixture, then mechanically stirred for 1 h. The mixture was filtered through a G4, washed two times with diethyl ether, dried at 70 °C for 1 h, and stored in desiccators for analysis54.

The chemically combined water content is considered as the percent of ignition loss of the dried sample (on the ignited weight basis) up to 1000 °C for 1 h soaking time. The chemically combined water contents were corrected for the water of free portlandite42,45,55,56. The chemically combined water content (Wn) was calculated using the following equation:

1 2

2

[ ]% 1n

w wW

w

−= × … (1)

where W1 is the mass of the dried sample before ignition and W2 is the ignited mass of sample.

Free portlandite contents of the hydrated cement pastes can be thermally determined. 0.5 g sample of the hardened cement was placed in a porcelain crucible introduced into a cold muffle furnace (room temperature). The temperature was increased up to 390 °C then to 550 °C at heating rate of 3 oC/min. The loss of weight occurred between 390 °C and 550 °C with soaking time of 15 min is equal to the weight of water of calcium hydroxide. Therefore, the free portlandite can be calculated54,57.

pH–value of hardened cement paste was estimated by placing about 10 g of pre-dried specimen into 100 mL of distilled water, stirred at room temperature for 30 min using a magnetic stirrer, then measured using pH meter, Schott. GLAS Mainz-Type CG 843- Germany.

Table 1—Chemical analysis of OPC, GBFS and NS (mass%)

SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O LO I Total

OPC 21.30 3.58 5.05 63.48 1.39 2.05 0.26 0.22 2.57 99.90 GBFS 43.21 9.97 0.59 35.96 5.43 1.37 0.79 0.67 1.98 99.97

NS 87.59 0.52 0.09 0.53 0.13 0.02 1.44 0.08 9.60 100.00

Table 2—Mix composition of OPC and composite cements, mass%

Mix No. OPC GBFS NS

M0 100 0 0 M1 55 45 0 M2 54 45 1 M3 53 45 2 M4 51 45 4 M5 49 45 6

Fig. 1—XRD of nano-silica

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

91

Bulk density was carried out before the specimens subjected to compressive strength determination through weighing the hardened specimens suspended in water and in air (saturated surface dry). Each measurement was conducted on at least three similar cubes of the same age45

. The following equations were used for calculating the bulk density:

Bulk density (dp) = sample of volume

weightsaturated g/cm3 … (2)

Volume of sample =

(water) liquid ofdensity

weightsuspended - weight saturated … (3)

Bulk density (dp) =

1 weightsuspended - weight saturated

weightsaturated× … (4)

A set of fifth cubes was used for the determination of compressive strength of cement paste as described by ASTM Specifications58. The measurements were done on a compressive strength machine of Seidner, Riedinger, Germany, with maximum capacity of 600 KN force.

The total sulfate content was gravimetrically estimated by using 1 g sample dissolved in 100 mL of distilled water with 5 mL of concentrated HCl were added, then boiled for 5 min, filter and washed several times with distilled water. Add to the filtrate 10 mL of 10% BaCl2, digest, filter and ignite at 1000°C for 30 min. The total sulfate content was calculated as:

34.3ppt.ofweight

%SO 3 ×=

M … (5)

where, M is the weight of sample.

The total chloride was determined by weighting 2 g into a stopper conical flask, dispersed with 25 mL water and then added 10 mL of nitric acid (specific gravity 1.42). 50 mL hot water was added, heated to near boiling and kept warm for 10 to 15 min. If the supernatant liquid is turbid, filter through filter paper (41) and wash with hot water, then cool to room temperature. An excess of standard 0.1N AgNO3 was added and 2-3 mL of nitrobenzene to coagulate the precipitate, stopple the flask and shake vigorously, then add 1 mL of ammonium ferric alum as indicator and titrate against standard 0.1 N ammonium thiocyanate42,45.

The powder method of X-ray diffraction (XRD) using Philips diffractometer PW 1730 with X-ray source of Cu kα radiation (λ=1.5418 Å) was carried out. The diffractometer was scanned from 5° to 65° (2θ) in step size of 0.015° and the counting time per step was 1.8 s. The X-ray tube voltage and current were fixed at 40 KV and 40 mA, respectively. An on-line search of a standard database (JCPDS database) for X-ray powder diffraction pattern enables phase identification for a large variety of crystalline phases in a sample.

The morphology of the hydration products was investigated through ESE-MEDX after specified hydration times were using ESEM ‘’Inspect S’’, FEI Holland.

Results and Discussion

Required water of standard consistency and setting times

The required water of standard consistency, initial and final setting times of the prepared OPC-GBFS-NS composite cement pastes are graphically represented in Fig. 3. The required water of standard consistency of M1 (55% OPC, 45% GBFS, 0% NS) decreases

Fig. 2—(A) TEM and (B) SEM photographs of NS

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

92

(Fig. 3A), whereas the initial and final setting times elongated, this is mainly attributed to the lower hydraulic reactivity of GBFS in comparison with OPC during the very early ages of hydration, due to the formation of acidic oxide film on the outer most layers of slag grains (Fig. 3B, C). The required water of standard consistency and setting times of OPC-GBFS-NS composite cement pastes increases with NS content59. Replacement of NS instead of OPC increases the required water of standard consistency, due to the high surface area of NS60. The initial and final setting times are elongated with NS content in the presence of 1 mass% polycarboxylate superplasticizer10. The retardation of the setting process is due to the presence of 1% of polycarboxylate superplasticizer as well as the excess of mixing water required for standard consistency60. Chemically combined water contents

The chemically combined water contents of hardened composite cement pastes immersed in seawater for 0, 1, 3, 6, 9 and 12 months are graphically

represented in Fig. 4. The chemically combined water contents increase with curing time for all hydrated cement pastes, due to the continuous hydration of the cement clinker phases to form more hydrated products with increased amount of crystalline water (Wn). Generally, the chemically combined water contents are sharply increased up to one month, due to the acceleration effect of MgSO4 and MgCl2 on the cement hydration42. Mix M1 gives lower values of chemically combined water contents than those of mix M0. The chemically combined water contents of composite cement pastes cured in seawater up to 12 months increase with NS content up to 4%. Silica nano-particles act as a nano-fine nucleating agent, which accelerate the hydration of OPC cement phases61. The composite cement paste containing 4 mass% NS (M4) has the higher values of chemically combined water up to 12 months. Silica nano-particles react with the liberated CH during the hydration of cement phases to form additional hydrated products mainly as C-S-H, C-A-H and C-A-S-H hydrated products. The chemically combined content decreases with the increase of NS content up to 6 mass% (M5), but still higher than those neat pastes (M0 and M1). It can be concluded that 4 mass% NS is the optimum mix composition. The decrease of chemically combined water contents at 6 mass% NS due to the agglomeration of silica nano-particles around cement grains which leads to retard cement hydration or due to the defects generated in dispersion of nano-particles that cause weak zones62. Free portlandite contents and pH values

Free portlandite contents and pH values of the different hardened composite cement pastes containing

Fig. 3—Water of consistency, initial and final setting times of OPC-GBFS-NS pastes

Fig. 4—Chemically combined water contents of cement pastes immersed in seawater up to 12 months

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

93

NS immersed in seawater up to 12 months are plotted against curing time in Figs 5 and 6. The portlandite content of the OPC paste decreases with curing time up to 3 months then increases up to 1 year. Free portlandite contents and pH values of composite cement pastes decrease with curing time up to 12 months, this is due to higher surface area of and the pozzolanic activity of silica nano-particles, which reacts with the liberated portlandite from the hydration of OPC cement phases (β-C2S and C3S) to form additional hydrated products. As NS content increases, the portlandite contents and pH values decrease, due to the deficiency of Ca2+. Addition of NS in the presence of 1% superplasticizers tends to form dispersed-hydrated products with low C/S and low Ca(OH)2 content, leading to form additional hydrated products having close and compact nano-structure16-18,35. C–S–H gels having a lower Ca/Si ratio and different gel porosities as shown in the pore structure of composite cement pastes with GBFS and NS addition were reported63,64. The microstructure displays a more denser arrangement of nanocrystalline C-S-H as the main hydration products with minor or nil amounts of Ca(OH)2 in the presence of superplasticizer and silica nano-particles16-18. The addition of NS promotes the densification of the cement matrix composites. Despite of this densification, the apparent porosity of the paste increases with NS content, similarly to those results obtained for the permeable porosity of superplasticized pastes16-18. NS tends to achieve more homogeneous structure with capillary porosities. The increase of the gel porosity can be caused by the increase in the amount of C–S–H gel in the paste as well as produced by the pozzolanic reaction the higher hydration degree and the

properties of the gel produced by the pozzolanic reaction of the nano-silica particles16-18,36.

With the increase of NS content, the free portlandite contents and pH values decrease, due to the higher pozzolanic affinity of NS as well as GBFS35. This indicates that cement pastes containing different contents of NS up to 4% were more durable in seawater than OPC pastes, due to the formation of additional amounts of C-S-H to fill the available pores and leading to restrict of the penetration of the aggressive ions (SO4

2- and Cl-) 45. The free portlandite contents of OPC cement paste have the higher values than other cement pastes. The decrease of free portlandite up to 3 months was mainly due to the reaction with MgCl2 and MgSO4 forming CaCl2, CaSO4 and Mg(OH)2. Magnesium ions form a layer of brucite at the exposed surface of cement grains in presence of free porltindite. Due to the low solubility of brucite, the penetration of Mg2+ beneath of the brucite layer into the intererior of the paste is restricted. The brucite formation consumes a high amount of Ca(OH)2 released from hydration and then depleted65. After 3-6 monthes, the brucite forms a protective layer around the cement grains, which decreases the accessibility of Mg2+ ions penetrations, so the portlandite content of the OPC paste increases from 3 months up to 1 year. Furthermore, the presence of sodium chloride increases the solubility of Ca(OH)2

42,45. Mg(OH)2 dissociates C–S–H producing Ca(OH)2 and silica gel. This gel reacts with Mg(OH)2 to form magnesium silicate hydrates66.

Compressive, flexural strengths and Bulk density Compressive, flexural strengths and bulk density of

hardened composite cement mortars containing NS in comparison with OPC mortar immersed in seawater

Fig. 5—Free portlandite contents of cement pastes containing NS immersed in seawater up to 12 months

Fig. 6—pH values of cement pastes containing NS immersed in seawater up to 12 months

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

94

up to 12 months were graphically represented in Figs 7-9. The results show that the compressive, flexural strengths and bulk density of all composite cement mortars containing different contents of NS increase up to 1 year. In contrast, the compressive as well as flexural strength and bulk density of OPC mortar (M0) increase up to 3 months, while mix (M1) increases up to 6 months then decreases up to 12 months. The increase of compressive strength of cement mortars up to 3 months is due to the acceleration effect of sulfate and chloride ions on the hydration of cement phases, to form C–S–H and AFt and/or AFm hydrates. The amount of C–S–H increases greatly during the first 3 months of immersing in seawater. Mix M1 has a higher compressive strength than those of neat OPC mortars. This is due to the pozzolanic reaction of GBFS to form dense structure. The increase of compressive strength is due to the formation of additional amounts of C–S–H. Parts of the Mg2+ ions are adsorbed and

incorporated into the particles of calcium silicate hydrates, which stabilize and enhance its crystallization as well as strength of cement pastes during the first 6 months. On one hand, a part of the Na+ is absorbed on the particles of C–S–H and incorporated into its interlayer space44,45, to give better ordered C–S–H and the strength increases. On the other hand, the decrease of compressive strength after the first 6 months and up to one year is mainly due to the aggressive attack of chloride and sulphate ions in seawater.

The results indicate that the compressive and flexural strengths increase with curing age. The reactivity of NS is controlled by its specific surface area and the amount of Q2-3 (silanol) Si–OH groups, which cover the NS particle. These groups constitute condensation sites for monomeric silica units released from the cement phases; consequently, the hydration of the cement mortars is accelerated46. The strong and instantaneous interaction between NS and portlandite leading to the formation and accumulation of excessive amounts of hydrated products mainly as C-S-H, which fill the open pore systems, and tend to increase the gel/space ratio, causing packing effect16-18.

The compressive and flexural strengths of cement mortars containing NS are higher than those of M0 and M1. The compressive and flexural strengths of the cement pastes increase with NS from 2% up to 4% NS. Composite cement mortars containing 4% NS possess higher values of compressive and flexural strengths than those of the other neat and composite cement mortars. The addition of NS tends to improve the durability of OPC cement mortars in seawater aggressive medium. This is attributed to the effect of NS, which behaves not only as a filler to improve microstructure16-18, but also as a reactive pozzolana to

Fig. 7—Compressive strength of hardened cement mortars of cement pastes containing NS immersed in seawater up to 12 months

Fig. 8—Flexural strength of hardened cement mortars of cement pastes containing NS immersed in seawater up to 12 months

Fig. 9—Bulk density of hardened cement mortars of cement pastes containing NS immersed in seawater up to 12 months

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

95

promote pozzolanic reaction as well as acts as nucleating sites to form more accumulation and precipitation of calcium silicate, aluminate and alumino-silicate hydrates in the open pores originally filled with water, leading to the formation of homogeneous, dense and compact microstructure.

OPC mortars give the lower values of compressive strength at all immersing ages, this is mainly due to the high content of portlandite, as a result of this CaCl2 and CaSO4 are formed. The chloro- and sulphoaluminate hydrated products are formed, which cause the softening and expansion hydrated products. On the other hand, MgCl2 also reacts with CH forming CaCl2 and Mg(OH)2. CaCl2 formed from the reaction of MgCl2 with CH produce calcium chloroaluminate hydrates, which deleterious effect that decreases the strength after 3 or 6 months and up to 12 months for M0 and M1, respectively.

NS is ultrafine material with high surface area and improves the packing of particles at the surface of the cement grains, which also favors the pozzolanic reaction, i.e., the formation of denser and high stiffness C–S–H phases38,60,67. The change in the microstructure in the presence of NS is responsible for the increase of compressive strength composite cement mortars. Composite cement mortars containing NS show higher values of compressive and flexural strengths, due to the decreased accessibility of sulfate and chloride ions towards the more dense structure and low capillary pore structure of the hardened pastes during hydration55,56. It can be concluded that mix M4 is the optimum mix composition, which has a high durability and can resist the effect of the sulfate and chloride attack. Total chloride contents

Figure 10 represents the variations of total chloride contents of composite cement pastes immersed in seawater for 0, 1, 3, 6, 9 and 12 months. The total chloride contents increase with curing time up to 12 months for all hardened cement pastes. This is mainly due to the chemical reaction of Cl- ions with free portlandite and C3A to produce CaCl2 and calcium chloroaluminate hydrates. The total chloride contents of OPC pastes (M0) show the higher values C3A and C4AF than those of OPC-GBFS pastes due to the decrease of portlandite in composite cement pastes.

Composite cement pastes containing NS decrease the Ca2+ concentration and total porosity to produce a compact structure with very much finer pores, thereby inhibiting the chloride ion penetration46. The increase in the gel porosity may be resulted from a higher

hydration degree as a result of the pozzolanic reaction of NS particles. NS promotes the densification of the cement matrix. Despite this densification, the total porosity of cement paste decreases with NS content. The decrease of the total porosity of cement paste improves the durability of composite cement pastes; hence the total chloride contents decrease45,55,56. The accumulation of excessive amounts of hydrates within the open pore systems produces a more compact structure with very much finer pores, thereby inhibiting the chloride ion penetration; and decreases the accessibility of chloride after prolonged hydration. It was found that, blending of OPC with NS is beneficial to decrease of chloride ions diffusion.

Total sulphate contents

Figure 11 represents the total sulphate contents of composite cement pastes immersed in seawater for 0, 1, 3, 6, 9 and 12 months. Total sulphate contents increase with curing time up to 12 months for all hardened cement pastes. This is due to the reaction of some hydration products forming calcium sulphate

Fig. 10—Total chloride contents of cement pastes containing NS immersed in seawater up to 12 months

Fig. 11—Total sulphate contents of hardened cement pastes containing NS immersed in seawater up to 12 months

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

96

dihydrate (gypsum) and a tricalcium sulphoaluminate (ettringite) phase42,45. Neat OPC paste gives a higher value of total sulphate content than those of composite cement pastes containing NS up to 12 months; this is due to the higher content of portlandite in OPC paste. The total sulphate content increases sharply at one month, and then gradually increases at later ages of hydration. The high reactivity and fast pozzolanic activity of NS particles produced a more refined microstructure. The accumulation of excessive amounts of hydrates within the open pore systems produces a more compact structure with very much finer pores, inhibiting the sulphate ions penetration, then decreases the accessibility of sulfate ions. Radius of Cl- ions is smaller than sulphate ions, so the attack of Cl- ions is more effective at all curing ages than the sulphate attack. On the other hand, when NS is added, the cement pastes become almost impermeable to the penetration of SO4

2- ions. This can be explained by the presence of homogeneous microstructure, characterized by compact and nano-sized C–S–H gel, which leads to refinement of the microstructure of the paste due to higher amount of pores with ink-bottle shape46. Interpretation of XRD diffraction patterns

Figure 12 illustrates XRD diffraction patterns of M0, M1 and M4 mixes immersed in seawater for 0, 1, 6, and 12 months. X-ray diffractograms of composite cement pastes containing different contents of NS show the presence of diffraction lines corresponding to Ca(OH)2, β-C2S, C3S, CaCO3 and CSH. XRD patterns corresponding to anhydrous phases (β-C2S and C3S) decrease with curing time, due to the continuous hydration. Calcium chloroaluminate hydrate is absent during immersion in tap water (0 month). With the increase of the hydration times, the intensity of the peaks corresponding to the calcium chloroaluminate hydrate increases, whereas the intensity of the peaks corresponding ettringite hydrate decreases. Seawater contains 12543 ppm Cl- and 8680 ppm SO4

2-, therefore, the effect of chloride ions on the cement pastes is higher than sulphate ions, this is due to the small radius of Cl- ions in comparison with SO4

2- ions. The competition exists between sulphate ions and chloride ions. So, sulphate ions decrease the chloride-binding capability greatly. Obviously, the lower of suphate content of composite cement pastes containing NS is one of the reasons that composite cement containing NS has higher performance to resist the chloride-induced corrosion attack, due to the higher radius of SO4

2- ions and their effect on the chloride binding is less68.

The intensity of CH peaks decrease with NS content which is due to the pozzolanic reaction of NS with CH to form additional calcium silicates hydrates. XRD diffraction patterns of M4 show that the peaks of CH and anhydrous silicate phases (β-C2S and C3S) decrease with immersion time up to 12 months. These additional hydrated products precipitated in the available open pores to form more compact and closed microstructure with higher stiffness C-S-H. The more compact and close structure of cementituous hydrated products, inhibiting the SO4

2- and Cl- ions penetration; leads to decrease the accessibility of SO4

2- and Cl- ions towards the more compact structure.

The suphate ions influence the chloride-binding capability, because of the preferential reaction between suphates and C3A. More C3A reacts with suphate ions; less C3A can bind the chloride to form calcium chloroaluminate hydrates. But HO− ion can be exchanged with chloride ion, because the HO− ion is attracted by the electrovalent bond to form calcium

Fig. 12—XRD patterns of hardened M0, M1 and M4 immersed in seawater for 0, 6 and 12 months

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

97

chloroaluminate hydrate as well as structure of C4AH13 can be denoted as 2[Ca2Al(OH)6OH.H2O]69. The aggressiveness of chloride is lower than that of sulphate ions, because the chemical bounded compounds are not expansive. Chloride chemically reacts with the hydrated aluminate phases to form Friedel’s salt: 3CaO.Al2O3.CaCl2.10H2O. Another oxychloride compounds (xCa(OH)2.yCaCl2.zH2O) can

be also formed in the presence of concentrated chloride salts. The presence of chloride limited the sulphate attack, due to the increase of solubility of portlandite (Ca(OH)2) and the formation of hydrated CaCl2 compounds70.

Microstructure

Figures 13-15 show SEM micrographs of hardened cement pastes of M0, M1 and M4 mixes immersed in

Fig. 13—Microstructure of hardened M0 immersed in seawater for 0, 6 and 12 months

Fig. 14—Microstructure of hardened M1 immersed in seawater for 0, 6 and 12 months

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

98

seawater for 0, 6 and 12 months. Figure 13 shows SEM micrograph of hardened pastes of M0 immersed in seawater for 0, 6 and 12 months. Figure 13A shows a flocculent and porous structure of ill crystalline C-S-H and sheets of massive Ca(OH)2 embedded through regions of C-S-H gel with a wider pores are available for crystallization of the formed hydrates. Figure 13B represented the presence of ettringite needles, hexagonal plates and Friedel’s salt

(3CaO.Al2O3.CaCl2.10H2O) within the pore system. Increase the immersion time up to 12 months, hexagonal plates of Friedel’s salt and plates of portlandite growing in pores as shown in Fig. 13C.

Figure 14 shows SEM micrographs of hardened pastes of M1 immersed in seawater for 0, 6 and 12 months. Figure 14a shows the porous structure of ill crystalline C-S-H and sheets of massive Ca(OH)2 within the pores. Increase of the immersion time NaCl activates the pozzolanic reactivity of GBFS (mix M1 contains 45% GBFS), leading to the formation of calcium aluminate layered structure (Fig. 14b), which are subsequently transformed into Friedel’s salt via HO-/Cl- ionic exchange inside the interlayer space (Fig. 14c)71. At 12 months hexagonal plates of Friedel’s salt and plates of portlandite growing in pores are as shown in Fig. 14c.

SEM micrograph depocits the formation of compact and small-sized C–S–H gel within the structure of hardened composite cement pastes containing NS. NS caused a refinement of the microstructure and induced the precipitation of nano-sized C–S–H gel, probably having a higher stiffness and lower Ca/Si ratio50. Figure 15 shows SEM micrograph of hardened pastes of M4 immersed in seawater for 0, 6 and 12 months. Figure 15A shows that the micrograph is composed of dense structure of platelet-like hydrates of C-S-H and crystalline hydrates having close structure with interlocking arrangements. NS has high pozzolanic activity reacted with CH to form additional C-S-H. The occurrence of C-S-H in large quantity is responsible for bridging cement particles and producing a rigid system. SEM micrograph represents the formation of closed structure arrangement (Fig. 15B). The micrograph also shows that the paste has a high degree of hydration and lower porosity. After 12 months immersion in seawater somewhat hexagonal plates of Friedel’s salt and ettringite needles on the surface of wider pore system and absence of sheets of hexagonal plates of portlandite (Fig. 15C).

Conclusions From the above results it can be concluded that:

(i) Initial and final setting times elongate with NS content.

(ii) The composite cement paste containing 4 mass% NS (M4) has the higher chemically combined water contents up to 12 months, then decrease at 6% NS (M5), but still higher than M0 and M1 pastes.

Fig. 15—Microstructure of hardened mix M5 immersed in seawater for 0, 6 and 12 months

HEIKAL et al.: DURABILITY OF COMPOSITE CEMENTS

99

(iii) The free portlandite contents of the OPC pastes decrease with curing time up to 3 months and then increase up to 1 year. On the other hand the free portlandite contents and pH values of composite cement pastes decrease with curing time for all composite cement pastes up to 12 months.

(iv) The compressive, flexural strengths and bulk density of OPC cement mortar (M0) increase up to 3 months, whereas the values of mix (M1) increase up to 6 months then the two value decrease up to 12 months. Mortars containing 4% NS possess higher values of compressive and flexural strengths than those of the other mortars containing NS.

(v) SEM micrograph depocits the formation of compact and nano-sized C–S–H gel within the structure of hardened composite cement pastes containing NS. NS improves the microstructure and induced the precipitation of nano-sized C–S–H gel, probably having a higher stiffness and lower Ca/Si ratio. The occurrence of C-S-H in large quantity is responsible for bridging cement particles and producing a rigid system. The micrograph also shows that the paste has a high degree of hydration and lower porosity.

(vi) The increased in the gel porosity may be resulted from a higher hydration degree as a result of the pozzolanic reaction of NS particles. NS promotes the densification of the cement matrix. The increased in the gel porosity may be resulted from a higher hydration degree as a result of the pozzolanic reaction of NS particles. NS promotes the densification of the cement matrix.

References

1 Heikal M, El Didamony H & Moustafa M A, Sil Ind, 74 (5-6) (2009)155-16.

2 Singh N B & Middendorf B, Cem Inter, 6 (4) (2008) 76-91. 3 Singh N B & Middendorf B, Cem Inter, 6 (2009) 78-93. 4 Mukesh K, Singh S K, Singh N P & Singh N B, Constr Build

Mater, 36 (2012) 68-686. 5 Shi C & Qian J, Resour Conserv Recycl, 29 (2000) 195-207. 6 Nazari A & Riahi S, Energy Build, 43 (2011) 864-872. 7 Li C, Sun H & Li L, Cem Concr Res, 40 (2010) 1341-1349. 8 Chen W, Hydration of slag cement: Theory, Modeling and

Application, Ph. D. Thesis, University of Twenty, 2007. 9 Siddique R, Waste materials and by-products in concrete,

(Springer-Verlag, Berlin, Heidelberg), 2008. 10 Yu R, Spiesz P & Brouwers H J H, Constr Build Mater, 65

(2014) 140-150. 11 Du H, Du S & Liu X, Constr Build Mater, 73 (2014) 705-712. 12 Oltulu M & Sahin R, Constr Build Mater, 53 (2014)

658-664.

13 Du H & Pang S D, Key Eng Mater, 629 (2015) 443-448. 14 Rong Z, Sun W, Xiao H & Jiang G, Cem Concr Compos, 56

(2015) 25-31. 15 Sobolev K, Flores I & Hermosillo R, Nanomaterials and

nanotechnology for high performance cement composites,

Proc ACI Session on Nanotechnology of Concrete: Recent Developments and Future Perspectives, , Denver, USA, Nov 7, 2006, 91-118.

16 Heikal M, Ali A I, Ismail M N, Awad S & Ibrahim N S, Constr Build Mater, 70 (2014) 339-350.

17 Heikal M, Al-Duaij O K & Ibrahim N S, Constr Build Mater, 93 (2015) 1067-1077.

18 Heikal M, Ismail M N & Ibrahim N S, Constr Build

Mater, 91(2015) 232-242. 19 Oltulu M & Sahin R, Constr Build Mater 53 (2014) 658-664. 20 Tobón J I, Payá J & Restrepo O J, Constr Build Mater, 80

(2015) 92-97. 21 Abd El Aleem S, Heikal M & Morsi W M, Constr Build

Mater, 37 (2014) 151-160. 22 Rong Z, Sun W, Xiao H & Jiang G, Cem Concr Compos, 56

(2015) 25-31. 23 Ghafari E, Arezoumandi M, Costa H & Júlio E, Constr Build

Mater, 94 (2015) 181-188. 24 3rd Int Symp on Ultra high performance concrete and

nanotechnology for high performance construction

materials, Kassel, Germany, 2012. 25 Beigi M H, Berenjian J, Omran O L, Nik A S & Nikbin I M,

Mater Des, 50 (2013) 1019-1029 26 Jalal M, Mansouri E, Sharifipour M & Pouladkhan A, Mater

Des,34 (2012) 389-400. 27 Madani H, Bagheri A, Parhizkar T & Raisghasemi A, Cem

Concr Compos, 53 (2014) 18-24. 28 Zhang M & Li H, Constr Build Mater, 25 (2) (2011)

608-616. 29 Said A M, Zeidan M S, Bassuoni M T & Tian Y, Constr

Build Mater, 36 (2012) 838-844. 30 Sanchez F & Sobolev K, Constr Build Mater, 24 (2010)

2060-2071. 31 Skinner L B, Chae S R, Benmore C J, Wenk H R & Monteiro

P J M, Phys Rev Lett (PRL), 104:195502 (2010) 1-4. 32 Constantinides G & Ulm F, J Mech Phys Solids, 55 (2007)

64-90. 33 Nili M, Ehsani A & Shabani K, Proc 2nd Int Conf on

sustainable construction materials and technologies, Universita Politecnica delle Marche, Ancona, Italy, June 28-30, 2010.

34 Singh L P, Karade S R, Bhattacharyya S K, Yousuf M M & Ahalawat S, Constr Build Mater, 47 (2013) 1069-1077.

35 Heikal M, Abd El-Aleem S & Morsi W M, HBRC J, 9 (2013) 243-255.

36 Bjornstrom J, Martinelli A, Matic A, Borjesson L & Panas I, Chem Phys Lett, 392 (2004) 242-248.

37 Ji T, Cem Concr Res, 35 (2005) 1943-1947. 38 Gaitero J J, Campillo I & Guerrero A, Cem Concr Res, 38

(2008) 1112-1118. 39 Belkowitz J S & Armentrout D, Proc Concrete Sustainability

Conf, National Ready Mixed Concrete Association, USA, 2010, 1-15.

40 Nazari A & Riahi S, Compos Part B Eng, 42 (2011) 570-578. 41 Wee T H, Suryavanshi A K, Wong S F & Rahman A K, ACI

Material J, 97 (5) (2000) 536-549.

INDIAN J. ENG. MATER. SCI., FEBRUARY 2016

100

42 Abd El Aziz M, Abd El Aleem S, Heikal M & El Didamony H, Cem Concr Res, 35 (2005) 1592-1600.

43 Ghosh S N, Sarkar S L & Harsh S, in Durability of silica

fume concrete, edited by Moukwa M, vol. 4, (ABI Books, New Delhi), 1993, 467-493.

44 Amin A M, Ali A H & El-Didamony H, Zem-Kalk-Gips, 50(3) (1997) 172-177.

45 Abd-El-Eziz M A & Heikal M, Adv Cem Res, 21(3) (2009) 91-99.

46 Quercia G, Spiesz P, Hüsken G & Brouwers H J H, Cem

Concr Compos, 45 (2014) 69-81. 47 Quercia G & Brouwers H J H. Application of nano-silica

(nS) in concrete mixtures, 8th FIB International Ph.D. Symposium in Civil Engineering, Lyngby, Denmark, June 20–23, 2010, 431-436.

48 Collepardi M, Ogoumah J J, Skarp U & Troli R, Influence of

amorphous colloidal silica on the properties of self-

compacting concretes, Proc Int Conf Challenges in Concrete Construction-Innovations and Developments in Concrete Materials and Construction, Dundee, Scotland, UK, 9-11 September, 2002, 473-483.

49 Maghsoudi AA & Arabpour-Dahooei F, Effect of nanoscale

materials in engineering properties of performance self

compacting concrete, Proc 7th Intl Congress on Civil Engineering, Iran, 2007, 1-11.

50 Khanzadi M, Tadayon M, Sepehri H & Sepehri M, Influence of

nano- silica particles on mechanical properties and permeability

of concrete, Proc 2nd Int Conf on Sustainable Construction Materials and Technologies, Universita Ploitecnica delle Marche, Ancona, Italy, June 28–30, 2010, 1-7.

51 Baomin W, Lijiu W & Lai F C, J Wuhan Univ Technol,

Mater Sci, 23(1) (2008) 85-88. 52 Wei X & Zhang P, Mod Appl Sci, 5(4) (2011) 68-73. 53 ASTM Designation: C191, Standard method for normal

consistency and setting of hydraulic cement, (ASTM Annual Book of ASTM Standards), 2008.

54 Abd-El Aziz M A, Abd El Aleem S & Heikal M, Constr

Build Mater, 26 (2012) 310-316. 55 El-Didamony H, Aiad I, Heikal M & Al-Masry S, Constr

Build Mater, 50 (2014), 281-290. 56 Aiad I, El-Didamony H, Heikal M & Al-Masry S, Constr

Build Mater, 48 (2013) 1092-1103. 57 El-Didamony H, Heikal M & Abd El Aleem S, Constr Build

Mater, 37 (2012) 269-276. 58 ASTM C109, Strength test method for compressive strength

of hydraulic cement mortars, 2007. 59 Heikal M, Aiad I & Helmy I M, Cem Concr Res, 32 (2002)

1809-1812. 60 Qing Y, Zenan Z, Deyu K & Rongshen C, Constr Build

Mater, 21 (2007) 539-545. 61 Land G & Stephan D, J Mater Sci, 47 (2012) 1011-1017. 62 Liao K, Chang P, Peng Y & Yang C A, Cem Concr Res, 34,

(2004), 977. 63 Jalal M, Mansouri E, Sharifipour M & Pouladkhan A R,

Mater Des, 34 (2012) 389-400. 64 Lotenbach B, Scrivener K & Hooton R D, Cem Concr Res,

41 (2001) 1244-1256. 65 Bonen D & Cohen M D, Cem Concr Res, 22 (1992) 707-718. 66 Ali A H, El-Didamony H & Mostafa K A, Sil Ind, 62

(11–12) (1997) 199-203. 67 Mondal P, Shah S P, Marks L D & Gaitero J J, J Transport

Res Board, No. 2141. Washington, DC: Transportation Research Board of the National Academies, 2010, 6-9.

68 Luo R, Cai Y, Wang C & Huang X, Cem Concr Res, 33 (2003) 1-7.

69 Lea F M, The chemistry of cement concrete, (China Construction Industry Publishing House, Beijing), 1980, 256-290.

70 Glasser F P, Marchand J & Samson E, Cem Concr Res, 38 (2008) 226-46.

71 Goñi S, Lorenzo M P & Sagrera J L, Cem Concr Res, 24(8) (1994) 1403-1412.