INVESTIGATIONS INTO THE INFLUENCE OF LIMESTONE ADDITIONS TO PORTLAND CEMENT CLINKER PHASES ON THE EARLY PHASE OF HYDRATION J Stark E Freyburg K Lohmer Bauhaus-Universitat Weimar Germany A B S T R A C T . This paper is concerned with the influence of limestone powder on the hydration of cement. Differences in dilation and shrinkage of concretes made from cements with limestone additives prompted an investigation into the effects of the use of limestone in cement on the process and the products of hydration of the four major cement clinker phases. In order to do so, separately produced cement clinker phases were hydrated one by one with a variety of four different limestone powders together with sulfatic additives. By means of Xray diffractometry for phase analysis monocarbonate (MC) was detected, proving the participation of the calcium carbonate added in the chemical reaction with both aluminatic phases C A and C AF. It showed that calcium carbonate has an influence on the heat of reaction and the conversion of the aluminatic phases as well as C S during the early age of hydration. These effects were noticeably increased solely by extending the specific surface of the limestone powder. 3 4 3 Keywords: (3-C S, C A, C AF, C S, Cement clinker, Conduction calorimetry, Dolomite powder, Limestone powder, Monocarbonate, Thaumasite, X-ray diffractometry. 2 3 4 3 Professor J S t a r k is Director of the F.A.Finger-Institute of Building Material Sciences at the Bauhaus-Universitat Weimar. The main fields of research are the use of binders, the durability of concrete, structural light-weight concrete, high-performance concrete, and historic mortars. D r E F r e y b u r g is a Research/Teaching Fellow at the F.A.Finger-Institute of Building Material Sciences at the Bauhaus-Universitat Weimar. His main research interests include the use of binders, the durability of concrete and especially the assessment of damaged concrete structures. K L o h m e r is a research assistant at the F.A.Finger-Institute of Building Material Sciences at the Bauhaus-Universitat Weimar. She specialises on the use of binders, in particular on the improvement of the properties of cement. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 70 Stark, Freyburg, Lohmer INTRODUCTION By German and European standards, except for CEM II/A-L and CEM II/B-L, the addition of limestone powder is permissible as "subsidiary component" in cements and as "inorganic substance" in plaster and mortar binders. In these cases limestone powder does not have to meet any requirements concerning its chemical reactivity. From various investigations [1,4,5,6] it is known that the addition of even relatively small amounts -still ranging as subsidiary component- changes the processing features of cement. It had to be asked whether or not the use of limestone in quantities classified as " subsidiary components" according to ENV 197 part 1, i.e. about 5 %, accounts for any chemical reactions besides the purely physical effect of a filler. Therefore it was investigated if in the presence of very small amounts of limestone detectable chemical reactions do occur and how they might influence the process of hydration (heat and phases of hydration). In order to be able to classify possible effects and to identify converted compounds the complex matter "cement" was abstracted as singular clinker phases. BASIC M A T E R I A L S AND C O M P O S I T I O N O F MIXTURES The cement clinker phases C S, C S, C A and C AF were produced, processed and tested separately. In order to maintain comparable conditions it was defined that each cement clinker comprises of 100 % of the cement clinker phase investigated. So, the actual proportions of clinker and additives (or the stoichiometrically precise ratio) were not considered. Common to all mixtures was the addition of 6 % sulfatic substance, corresponding to an S0 -content of 3.4 %. The sulfatic substance contained 50 % natural anhydrite and 50 % synthetically produced hemihydrate. 3 2 3 4 3 The reference sample contained nothing but one cement clinker phase each and sulfatic substance. To all other samples 3 and 6 % limestone powder were added to respectively. As additives were used: two different types of limestone powder (A) and (B), type (B) of different fineness (Ba) and (Bb), and dolomite powder (C). Table 1 shows the specific surface and the calcite/dolomite content of each powder used. Table 1 Fineness and chemical purity of the stonepowders used TYPE SPECIFIC SURFACE [cm /g] CALCITE CONTENT A Ba Bb C 13,700 6,800 10,500 11,500 98% 85% 85% 98 % (dolomite) 2 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Influence of L i m e s t o n e on H y d r a t i o n 71 The value for the specific surface of the cement clinker phases ranged between 3,300 and 4,200 cm7g. By means of calorimetric testing only reference samples and type (A) mixtures containing 3 and 6 % limestone powder were analysed. All mixtures were produced with a water-to-binder ratio of 0.5. METHODS OF INVESTIGATION Calorimetric measurements were run by means of a conduction calorimeter (DCA-06, ToniTechnik). Only the rate of heat evolution was taken under consideration. Each measurement lasted over a period of 72 hours. The influence of the stone powder content was determined concerning the time and peak value at which a maximum in heat evolution is reached. Furthermore, these maxima determined served as an indicative to limit the hydration period prior to X-ray analysis. Before being analysed by means of XRD, the samples have been mixed for 90 seconds and were then stored quasi-adiabaticly. The hydration was stopped by grinding the sample in isopropanol and subsequent rinsing with acetone. The samples were dried at 40 °C; to examine them a X-ray diffractometer was used ( D 5000", Siemens). The results were evaluated qualitatively and semi-quantitatively, the latter by measuring and comparing peak values of certain reflections. M The d-values of the reflections are listed in Table 2. Table 2 d-values of certain reflections COMPOUND REFLECTIONS [nm] (d-value) Cement clinker mineral C,S: 0.276; 0.218; p-C S: 0.219; 0.287; Portlandite Anhydrite Calcite Dolomite Gypsum Thaumasite Ettringite / AFt-phase Monosulfate (MS) / AFm-phase Monocarbonate (MC) C AH /C (A,F)H C A H /C (A,F) H C A: 0.191; C AF: 0.181; 0.205; 0.725 0.49 0.35 0.209; 0.228; 0.303 0.289 0.761; 0.761; 0.428 0.966 0.973 / 0.56; 0.973 0.223; 0.893 0.378; 0.756/0.757 0.23; 0.513/0.515 0.287 - 0.289; 0.796 - 0.834 2 3 3 x 6 y 3 z 6 x y z 4 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 72 Stark, Freyburg, Lohmer RESULTS AND I N T E R P R E T A T I O N Conduction Calorimetry With the C S samples an exothermal effect of 5 J/g-h was indicated within minutes after the samples were exposed to water which has to be explained with the moistening of the solid material. A second exothermal effect of about 20 J/g-h was observed after 10 to 12 hours. If limestone powder had been added the exothermal maximum occurred up to one hour earlier, as illustrated in Figure 1. 3 During the period of observation the (3-C S samples did not show any exothermal effect besides the one due to moistening. 2 —~~"~" C S - C S + 3 % LST. A - - 3 - SAMPLE 3 CJS + 4 8 12 16 20 6% LST. A 24 T I M E OF HYDRATION [ H ] Figure 1 Course of heat emission of C S samples 3 All C A samples are of very high reactivity during the first minutes of hydration. The heat release of this effect was found to be reduced if limestone powder had been added. A slight acceleration is suggested. 3 Figure 2 illustrates the course of heat emission of C A samples during the first minutes of hydration. 3 The reduction in heat release is evident. This could be attributed to a "thinning" of the reactive substance i.e. the clinker phase with "inert" filler as well as to a formation of compounds involving calcium carbonate. The latter is supported by the results of the X-ray analysis (see next chapter). Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Influence of L i m e s t o n e on H y d r a t i o n 10 20 30 40 50 73 60 time of hydration [minutes] Figure 2 Course of heat emission of C A samples 3 Regarding C AF samples three exothermal effects were detected. The first one corresponds to the C A effect, but it occurs later, lasts longer and is less intense than the one of the ironfree calcium aluminate. A second and third effect was observed during the following five hours. 4 3 Figure 3 illustrates the course of heat emission of C AF samples containing no limestone and 3% of stonepowder respectively. 4 100 time of hydration [h] Figure 3 Course of heat emission of C AF samples 4 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 74 Stark, Freyburg, Lohmer The addition of limestone powder prompted an acceleration of the reaction, so that all three effects occurred earlier. The first effect then was, as with the C A samples, less intense but in the following two effects the samples containing limestone tended towards a higher rate of heat release. 3 X - r a y Diffractometry The influence of limestone additives on the hydration of C S, as described in other publications, could be explained mineralogically by means of X-ray diffractometry. In the presence of limestone (3 and 6%) an intensified conversion of C S was observed, especially in the beginning. 3 3 A related acceleration in the formation of portlandite could not be established. Figure 4 illustrates the different conversion of C S in samples containing no and 6% limestone powder respectively after two and 24 hours, taking the reflections of 2-0=32.07 and 41.1 (d-value=0.276 and 0.218) as examples. 3 Occasionally, some thaumasite was found; its presence was proven through analysis by means of SEM of samples hydrated for five days. This suggests at least the possibility of a chemical reaction of siliceous and sulfatic phases with calcium carbonate, as discussed by some. Yet, it can not be excluded that atmospheric carbon dioxide penetrated into the fresh samples (small volume, large surface) and participated in a chemical reaction. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Influence of L i m e s t o n e on H y d r a t i o n 75 Because of the low reactivity of P-C S no influence of the limestone additive could be observed. 2 Besides the slight increase in the conversion of C A when limestone was added, the reaction of calcium carbonate with both aluminatic phases was established. Calcium aluminate monocarbonate hydrate as well as calcium ferrite monocarbonate hydrate was detected, its amount increasing in relation with the supply of more calcite. 3 Figure 5 is an X-ray diffraction pattern depicting the different formation of monocarbonate in varying C A mixtures after four days of hydration. 3 --Theta - Scale Roentgenlabor F .A.Finger-I nst itut f .Baustoffkunde W e i m a r -UHI - 16-Sep-1998 14:08 MC = M o n o c a r b o n a t e 1 = C A - S a m p l e without Lst. 3 2 = C A + 3 % Limestone A 3 3 = C A + 6 % Limestone A 3 MC MC 3 MC aJLa H 1 6 Figure 5 X-ray diffraction pattern of monocarbonate in C A samples after four days of hydration 3 There were no apparent differences between the two types (A) and (B) of limestone used. When dolomite powder (C) or the more coarse type (Ba) was used less monocarbonate was formed, especially in C AF samples, due to a lower reactivity of the additives. 4 Figure 6 illustrates that fact with the help of the monocarbonate reflections at 2-0 = 11.68 and 23.5 (d-value = 0.757 and 0.378). The detection of monocarbonate in samples with dolomite stonepowder added was possible using the same d-value. In that, the peak changed into the shape of a "shoulder" at d = 0.757nm, as depicted in Figure 7. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 76 Stark, Freyburg, Lohmer 2-Theta £3 I - . . Scale . . . Roentgen 1abor . F .A . F i n g e r - l n s t i t u t lleimar -UNI- 1&-Sep-1998 14-09 , : 10 f .Bans tof fkunde 12 14 16 18 20 22 24 Figure 6 X-ray diffraction pattern of C AF samples with 6% stonepowder each after four days of hydration 4 2-Theta - Scale R o e n t e n 1abor F .A . F i n g e r - l n s t i t u t 9 f .Baustoffkunde Weimar -UHI - lb-Sep-1998 14 09 O O Q L_, r _, , ( , 10.9 , , , ! , 11.0 , , , ( , 11.1 , , , ! , , 11.2 , , ! 11.3 , , ! , 11.4 , , .., | , 11.5 , , , ! , 11.6 , , , ( , , r _, 11.7 ] ,—, 11.8 , , f 11.9 Figure 7 X-ray diffraction pattern depicting monocarbonate in C AF samples with limestone (A) and dolomite (C) powder 4 The above mentioned influence of limestone powder on the second and third exothermal effect during the hydration of C AF samples could not be investigated by means of phase analysis. Detection was possibly hindered by the formation of minor amounts of calcium aluminate ferrite sulfate hydrate phases as well as by not yet appropriately gentle enough ways of preparation. 4 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Influence of Limestone on Hydration 77 CONCLUSION It was established that the addition of 3 and 6% limestone powder indeed influences the early phase of hydration of C S, C A and C AF. 3 3 4 The finding that thaumasite is formed in an early stage by reaction of C S with the sulfatic substance and possibly with limestone powder (at least, the C S reaction is accelerated in the presence of limestone powder) may be exploited by purposefully producing this "rising" mineral to control the shrinkage of concrete from early on comparable to ettringite. 3 3 The conjecture that the thaumasite found originates from the reaction of the source materials and is not formed during carbonation is currently being reviewed. If existing unspecified concrete has to be identified with respect to its original binding agent, these results indicate that monocarbonate may have been formed primarily, not only by carbonation of ettringite. REFERENCES 1. FREYBURG, E, STARK, J, SPRINGENSCHMID, R AND MANGOLD, M. EinfluB der Zementeigenschaften auf die Hydratationswarmeentwicklung und die ReiBneigung des Beton. Wissenschaftliche Zeitschrift der HAB Weimar, 1993, Vol. 39, No. 3, pp 149-155. 2. INGRAM, K, D, AND DAUGHERTY, K, E. Limestone additions to Portland Cement : Uptake, Chemistry and Effects. 9. International Congress on the Chemistry of Cement (ICCC), 1992. 3. PERLT-SCHUSCHNIG, E. EinfluB von Kalksteinmehl auf die ReiBneigung von jungem Beton. HAB Weimar, Master Thesis 1993. 4. SCHIEL, T. Eigenschafiten von jungem Beton bei Einsatz von Kalksteinmehl, Silicastaub und unterschiedlichen Zuschlagen. HAB Weimar, Master Thesis 1994. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved.