The Influence Of Alkali Activator On The Early Hydration And Performance Of Portland Blast Furnace Slag Cement

THE INFLUENCE OF ALKALI ACTIVATOR ON THE EARLY HYDRATION AND PERFORMANCE OF PORTLAND BLAST FURNACE SLAG CEMENT Arnaldo Forti BATTAGIN, Brazilian Association of Portland Cement, São Paulo, Brazil ABSTRACT This paper shows the best behaviour of Na2SO4 among other activators of portland blast furnace slag cement. Sodium silicate, extensively used in alkali activated slag, has a poorer performance than Na2SO4 in the presence of Ca(OH)2. Depending on the characteristics of cement and activator, Na2SO4 can increase the early strength gain of blast furnace slag cements without significantly affecting long term compressive strength. A use of sodium sulfate - rich waste could be an environmental solution for waste disposal but further studies will be necessary. Key-words: accelerators, activation, alkali, blastfurnace slag. INTRODUCTION Many publications have dealt with the role of chemical activators, mainly sodium silicates (water glass) on the slag hydration (Shao-Dong Wang et al, 1995). There is a general agreement about the good behaviour of these compounds as slag activators on so-called alkali-activated slag cements (AAS) but less work has been done about the study of alkaline activation of slag in portland cement. In order to evaluate the most suitable slag activator we carried out preliminary tests based on compressive strength. Two blast furnace slag cement (BFSC) samples representative of industrial production in Brazil were chosen. Cements selected have different mechanical performance according to the Brazilian method by using 50 x 100mm cylindrical test specimens in proportion 1:3 (cement:sand), water cured at 23oC, w/c = 0.48. Both granulated blast furnace slag are basic, refractive index 1.65 and 95% vitrified. Mixes of 99% cement and 1% of six different chemical agents (Reagent Grade) were used with the aim of improving the early strength gain. Results are shown on Table 1. Table 1 - Effects of different slag activators Cement

A

B

Activator None K2SO4 NaOH Na2CO3 A2(SO4)3 Na2SO4 NaSiO3.nH2O None K2SO4 NaOH Na2CO3 A2(SO4)3 Na2SO4 NaSiO3.nH2O

1 day 5.1 8.1 7.2 6.9 5.5 8.6 5.5 4.3 7.2 6.5 5.6 4.0 7.6 4.7

Compressive Strength (MPa) 3 days 7 days 28 days 17.4 35.3 59.1 21.2 35.2 49.8 15.7 23.0 33.8 17.3 26.7 37.4 18.3 32.3 52.1 21.0 32.7 47.1 16.1 27.9 46.2 13.1 20.0 32.8 15.7 22.4 34.7 14.1 19.1 29.5 14.1 18.7 30.6 13.0 21.3 36.0 16.1 22.5 33.8 13.2 20.1 32.0

Based on the preliminary results it was decided to detail the investigation of the Na2SO4 activator, although Jawed and Skalny (1983) have shown it decrease the cement strength at long term age. Sodium silicate had a poorer and unexpected behaviour compared to those reported in literature for slag alone.

ACTIVATION WITH Na2SO4 (REAGENT GRADE) Mixtures of BFSC and Na2SO4 at proportions in mass of BFSC/Na2SO4 of 100/0, 99.5/0.5, 99/1 and 98/2 were made. Compressive strengths were determined at 1, 3, 7, 28, 60 and 91 days and a conduction calorimeter was used to monitor hydration kinetics. Table 2 shows the results obtained as well as additional characteristics of materials. Table 2 - Mechanical and Rheologycal Results Na2SO4 Cement

(%)

Setting Time Soundness (Vicat Test) Le Chatelier 1 3 7 28 60 91 Initial Final (mm) day days days days days days (h:min) (h:min) Compressive Strength (MPa)

0 A (reference) 5.1 17.4 35.3 7.1 20.5 34.1 (58% slag , 5% 0.5 8.6 21.0 32.7 limestone and 419m2.kg-1 1.0 10.7 21.9 29.5 Blaine surface area) 2.0 0 B (reference) 4.3 13.1 20.0 5.6 14.4 20.7 (44% slag, no 0.5 2 -1 7.6 16.1 22.5 limestone and 342m .kg 1.0 8.5 17.0 22.0 Blaine surface area) 2.0

68.6 57.3 54.8 46.5

3:55 3:45 2:55 2:35

5:35 5:15 4:15 4:15

0 0.5 0 0

32.8 39.9 43.1

3:25

4:55

0

32.1 38.4 42.9 33.8 36.0 38.8 33.1 37.1 39.5

3:05 3:10 2:40

4:05 4:20 4:00

0 0 0.5

59.1 51.5 47.1 39.9

60.7 54.8 50.4 44.3

From Table 2 we can conclude that at early ages (1 and 3 days) the higher the Na 2SO4 content, the higher the compressive strength for both samples.For cement A, at 7 days and beyond, there is a remarkable decrease of strength with the increasing quantities of Na2SO4. It can be seen that strength values can reach only 68% of the reference at 91 days with 2% Na2SO4. However, a different behaviour was found for sample B showing smaller loss of strength. In fact, at 91 days a value of 90% of reference for 1% Na2SO4 was observed. Tables 3 and 4 summarize the results of heat of hydration up to 72 hours, taking into account that the peak due to the first minutes of hydration was undetected due to operational conditions. Table 3 - Heat Peak Evolution Cement Na2SO4 Heat peaks at 23oC (W/g) (%) First Second 0 2.32 A 0.5 2.40 1.0 2.42 2.0 2.43 0 1.58 1.89 B 0.5 1.69 1.78 1.0 1.77 1.77 2.0 1.73 1.74

Time (h:min) First Second 12:25 12:30 12:15 11:00 8:40 15:50 8:30 17:00 7:10 14:50 8:20 15:20

Table 4 - Total heat liberation up to 72h Sample A

Na2SO4 (%) 0 0.5 1.0 2.0 0

2 1 1 1 2 2

4 4 5 5 6 8

6 12 14 13 17 18

8 22 26 26 32 29

Total heat liberated after (hour) (J/g) 12 16 20 24 30 40 50 76 91 103 118 139 56 88 106 119 134 154 57 90 111 125 140 159 66 96 117 132 147 165 51 76 97 108 121 137

50 156 169 173 178 149

60 169 183 184 187 159

72 182 196 193 194 168

B

0.5 1.0 2.0

2 3 2

9 12 9

19 24 20

31 37 32

55 61 56

80 87 81

104 110 105

120 126 123

135 141 139

152 158 156

164 170 169

174 181 180

184 191 190

From Table 3 we can conclude that first peak of 99/1 and 98/2 mixes is slightly accelerated compared to reference for both sample. The same phenomenon is found for the second peak of heat for sample B. First peak is supposed to be due to clinker fraction hydration and second peak is attributed to slag hydration. However, sample A exhibits only one peak. These atypical results indicate that, in this case, maximum hydration rate of both clinker and slag are simultaneous. Acceleration of both clinker and slag fraction by using Na2SO4 as activator is confirmed by results of total heat liberation listed on Table 4. ACTIVATION WITH SODIUM SULFATE WASTE The use of an activator is usually expensive and may impair its use in practice. Aiming the reduction of the costs and trying to find an environmental solution for waste disposal an attempt was made concerning the use of a waste as activator. So we replaced the Na2SO4 (Reagent Grade) by a glass - industry waste rich in this component (27.1% SO3 and 47.5% Na2O). Similar tests to those of compressive strength shown in Table 2 were carried out. Preliminary results are listed on Table 5. Table 5 - Results of sodium sulfate waste as an activator Cement Waste Compressive Strength (MPa) Content (%) 1 day 3 days 7 days 28 days 0 5.1 17.4 35.3 59.1 0.5 3.0 16.0 30.9 55.5 1 1.0 2.4 16.3 30.6 50.5 2.0 1.1 16.7 28.5 43.8 0 4.3 13.1 20.0 32.8 2 0.5 3.7 10.8 17.3 28.9 1.0 5.1 13.2 18.6 29.5 2.0 1.1 10.0 15.3 26.1 In spite of bad results for cement 1, the use of waste brought up some advantages to cement 2, mainly with 1% Na2SO4. This fact stimulates further studies to a better understanding of the behaviour of waste activated cement as a contribution for waste elimination. DISCUSSION AND CONCLUSION • Sodium silicate is the most used activator for alkali activated slag but its behaviour was not good in presence of Ca(OH)2 liberated from clinker hydration.

• At 1% levels, Na2SO4 seems to be the most suitable activator among those used in this study. • Unique heat peak found in cement 1 is probably not due to influence of activator because it also occurs in reference sample. In this case a higher specific surface area of slag and a poorer reactivity of clinker compared to sample 2 accelerated slag hydration and retarded clinker hydration. • Compared to cement 1, cement 2 presented a higher heat liberation at first 12 hours probably due to its lower slag content in spite of being coarser than sample 1. At 24h and beyond (72 h) heat liberation was higher for sample 1. Results are in agreement with compressive strength results. Setting time is accelerated with the increase of Na2SO4 contents and unsoundness due to its use was not detected.

• At 1 and 3 days the behaviour of both samples activated with Na2SO4 was similar, with a strength gain of aproximately 100% and 25-30% respectively for 2% Na2SO4. However, a different behaviour was found after 7 days, with a harmful influence of Na2SO4 on the cement of better performance (sample 1). On the other hand, long term strength was little affected in cement of worse performance. • Optimum contents of Na2SO4 depends on characteristics of both cement and activator and further studies will be necessary for supplementary consideration about the potential use of sodium sulfate-rich waste as activator. REFERENCES Jawed, I and Skalny, J. “The influence of alkali sulfates on the properties of cement and concrete.” World Cement, 1983, Nov., p325. Wang S.D. et alii, “Alkali-activated slag cement and concrete: a review of properties and problems.” Advances in Cement Research, 1995, 7 no 27, july, 93-102.