Evaluation of P2O5 Distribution Inside the Main Clinker Minerals

Cement and Concrete Research 59 (2014) 147–154

Contents lists available at ScienceDirect

Cement and Concrete Research

j ourna l homepage: ht tp : / /ees .e lsev ie r .com/CEMCON/defau l t .asp

Evaluation of P2O5 distribution inside the main clinker minerals by theapplication of EPMA method

Tomáš Ifka a,⁎, Martin Palou a, Jan Baraček b, František Šoukal b, Martin Boháč b

a Institute of Construction and Architecture, Slovak Academy of Sciences, Dúbravská cesta, 9845 03 Bratislava 45, Slovak Republicb Faculty of Chemistry, Brno University of Technology, Purkyňova 464/118, Brno 612 00, Czech Republic

⁎ Corresponding author. Tel.: +421 2 5930 9260.E-mail address: [email protected] (T. Ifka).

http://dx.doi.org/10.1016/j.cemconres.2014.02.0100008-8846/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Article history:Received 17 December 2013Accepted 28 February 2014Available online 6 April 2014

Keywords:Reaction (A)Microstructure (B)Ca2SiO4 (D)Ca3SiO5 (D)Waste management (E)

The formation of Portland clinker phases has taken place in thermodynamically non-equilibrium state betweenmacro-oxides CaO, SiO2, Al2O3, Fe2O3 andMgO from rawmeal and P2O5 frombonemeal. The paper dealswith thestudy of clinker minerals as solid solutions with P2O5 during the clinkerization of raw mixture containing bonemeal (BM). The ash of BM has contributed as a rawmaterial to the formation of different clinker phases. Electronprobe microanalysis (EPMA) method was used to determine the preferential distribution of P2O5 inside calciumsilicate phases and its influence upon C2S/C3S ratio. Basing on these results, composition of solid solution of C2Sand C3S was established.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The use of waste products as partial replacement of fossil fuels in ce-ment manufacturing is well known. These alternative fuels are usedmainly for their calorific values and also because of the similarity be-tween chemical composition of their ashes and raw materials or clin-kers. Some alternative fuels can also be considered as alternative rawmaterials. The ash as remained inorganic part of wastes after burningprocess is incorporated into clinker or reacts with raw material toform clinker minerals or their solid solutions. One of the alternativefuels that could be considered as an alternative raw material is themeat bone meal (MBM).

The ash of theMBM is composedmainly of P2O5 and CaO. P2O5 entersthe clinker minerals forming solid solutions. In solid solutions, the chem-ical composition varieswhile the crystal structure remains the same. Solidsolution is formed by the substitution of ions in crystal lattice by otherions with similar ionic radius. During ionic substitution, the vacanciescan form in normally unoccupied sites. The composition of solid solutionscan be expressed by oxide or elemental analysis. According to [1], thetypical composition of alite (C3S solid solution) with a wide variety of“foreign” atoms in its structure can be expressed as follows:

Na0:01Ca2:90Mg0:06Fe0:03Al0:04Si0:95P0:01O5:00: ð1Þ

The typical composition of belite (C2S solid solution) is expressed byformula [1]:

K0:03Na0:01Ca1:94Mg0:02Fe0:02Al0:07Si0:90P0:01O3:93: ð2Þ

The amount of each “foreign” ion in the structure of calcium silicatesdepends upon the amount of the element in question in the bulk com-position, but aluminum and iron are always present.

The aluminate phase accommodates alkali metal ions and also alarge amount of another “foreign” ions. Taylor quotes a “typical” compo-sition for C3A solid solution [1]:

K0:04Na0:09Ca2:73Mg0:09Ti0:01Fe0:17Al1:66Si0:17O6:00: ð3Þ

Again, the amount of each “foreign” ion depends upon the amount ofthe element in question in the bulk composition, but silicon and iron arealways present.

Of all phases, tetracalciumaluminoferrite has themost variable com-position. In addition to its variable aluminium/iron ratio, it can take uplarge amounts of “foreign” elements. Taylor quotes a “typical” composi-tion for ferrite [1] in clinkers with higher alumina contents:

K0:01Na0:01Ca1:98Mg0:17Ti0:05Mn0:02Fe0:62Al1:00Si0:14O5:00 ð4Þ

and for ferrite in clinkers with lower alumina contents:

K0:01Ca1:98Mg0:17Ti0:05Mn0:02Fe0:90Al0:72Si0:14O5:00: ð5Þ

Table 1Chemical composition of used raw materials.

Composition(wt.%)

Marl Limestone Blast furnace slag Ash of bone meal

CaO 18.88 53.67 8.84 54.14SiO2 41.80 3.65 3.20 0.82Al2O3 11.90 – 0.69 –

Fe2O3 3.82 – 80.91 0.01P2O5 – – 0.20 38.03

Table 2Adjusted basic chemical parameters.

Raw materials Raw materialcomposition(wt.%)

Characteristicsof cementraw meal

Mineralogicalcomposition(wt.%)

Marl 28.50 LSF 93.84 C3S 62.69Limestone 70.25 SM 2.64 C2S 18.58BFS 1.25 AM 1.62 C3A 8.63Ash of BM – HM 2.16 C4AF 10.11 Fig. 1.Microstructure of the sample with 1.81 wt.%. P2O5 (1st series).

148 T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

The ash of MBM contains P2O5 in the form of calcium phosphates.The influence of P2O5 or calcium phosphate on the formation, stabilityand properties of calcium silicates in Portland clinker has already beeninvestigated by many authors [2–6]. It was demonstrated that phos-phate takes an active part in the reactions during clinker burning pro-cess and shifts the stability limits of individual phases with formationof solid solutions [7]. During the reaction, P2O5 is firstly fixed by calciumin the form of apatite, which gets unstable at higher temperatures,reacting with silicon to form an isomorphous mixture made up ofdicalcium silicate and tricalcium phosphate (C2S–C3P) first, which canbe mixed with β-C2S continuously. If the CaO supply is sufficient, thisisomorphous mixture containing phosphate can react further to formalite containing phosphate. The viscosity of the clinker melt is dimin-ished as a result of phosphate input, which is conducive to alite growth.As phosphate input increases, C2S–C3P isomorphous mixtures having astructure of α′-C2S and α-C2S are formed. As a consequence, alite for-mation is increasingly impeded [7–10].

In [11], it is stated that P2O5 enters the C2S structure and PV ion fillsSiIV position. During this substitution the vacancy is built up as a resultof need to maintain electroneutrality.

CaII þ 2SiIV ¼ NV0 þ 2PV ð6Þ

The substitution of two PV atoms for two SiIV atoms causes the CaII

atom to leave its position and so the vacancy is created.General formula of C2S solid solutions with stabilizing P2O5 follows

[12]:

ðCa2−x=2V0x=2Þ Si1−xPxð ÞO4;where0:02≤x≤0:10: ð7Þ

Table 3The free lime content in the clinker.

Calcining temperature:1450 °C

1st series

Ash of BM(wt.%)

P2O5

(wt.%)f-CaO(wt.%)

0 0 0.050.99 0.38 0.221.96 0.75 0.972.91 1.11 2.13.85 1.46 3.034.76 1.81 4.08

It follows that during the synthesis of stabilized C2S from pure C2Sand P2O5, certain amount of CaO must be released.

Staněk and Sulovský [13] found out that P enters the structure ofboth calcium silicates at least partially through the so called berlinitesubstitution: AlIII + PV = N2 SiIV, where AlPO4 (the berlinite compo-nent), isostructural with quartz, substitutes SiO4 tetrahedrons.

Benarchid et al. [14] studied the influence of chromium and phos-phorus additions on the thermal decomposition of CaCO3 and the for-mation of the tricalcium aluminate. They showed that simultaneousadditions of chromium and phosphorus oxides up to 1.41 and1.31 wt.%, respectively, lower the temperature decomposition ofCaCO3 and improve the burnability of the synthesized doped tricalciumaluminate. The corresponding reduction of temperature decarbonationwas about 34 °C. The same collective of authors [15] also examined the si-multaneous influence of iron andphosphorus inclusions on the CaCO3 de-carbonation and the C2S formation in CaCO3–SiO2 mixtures around themolar ratio CaCO3/SiO2=2. In these reactingmixtures, the onset temper-ature of decarbonation decreased with Fe2O3 and P2O5 additions of,respectively, less than 9.50 wt.% and 8.45 wt.%. The mineralogicalanalysis of the synthesized solid solutions (Ca2 − 2xFexV0x)(Si1 − xPx)O4

(V0: cationic vacancy of Ca2+) with 0 ≤ x ≤ 0.20 showed that at roomtemperature the iron and phosphorus doped C2S is stabilized in theβ, α′ and α forms for additions less than 9.50 wt.% Fe2O3 and8.45 wt.% P2O5.

Another authors [16] studied the effect of combination of manga-nese and phosphorus elements on the formation of solid solutions, byheterovalent substitution of corresponding ions (5/2 CaII + SiIV = N2MnII + PV + 1/2 V0) in C3S lattice, at air atmosphere. These solid so-lutions were described by the following cationic vacancy formula

Calcining temperature:1450 °C

2nd series

Ash of BM(wt.%)

P2O5

(wt.%)f-CaO(wt.%)

0 0 0.051 0.38 0.052 0.76 0.683 1.14 1.54 1.52 2.175 1.9 3.08

Table 4EPMA line analysis of sample shown in Fig. 1.

Point MgO(wt.%)

Al2O3

(wt.%)SiO2

(wt.%)P2O5

(wt.%)CaO(wt.%)

Fe2O3

(wt.%)CaO/SiO2 Solid solution

1 1.01 1.26 24.8 2.01 70.9 0.00911 2.85 C3S2 1.94 6.35 18.3 1.95 66.4 5.06 3.63 C3S + C6AxFy3 0.803 1.10 23.8 2.06 72.2 0.00933 3.03 C3S4 0.781 1.46 23.9 1.81 70.7 1.37 2.96 C3S5 1.14 2.12 23.8 1.87 71.1 0.00911 2.99 C3S6 3.00 6.12 16.7 1.78 68.7 3.78 4.12 Grain boundary7 2.98 16.8 4.29 0.612 60.9 14.4 14.2 C6AxFy8 2.97 19.8 7.09 0.922 57.7 11.5 8.14 C6AxFy9 0.596 1.29 26.8 3.30 68.0 0.0105 2.54 C2S10 0.00274 1.60 29.1 3.30 64.1 1.89 2.21 C2S11 0.00269 1.85 27.5 4.38 66.3 0.00986 2.41 C2S12 2.58 14.0 13.1 0.781 55.4 14.2 4.24 Grain boundary13 0.772 2.32 28.0 3.88 63.3 1.72 2.27 C2S14 1.40 6.23 20.0 2.62 62.5 7.27 3.13 Grain boundary15 0.00277 1.96 28.0 3.76 66.3 0.0102 2.36 C2S16 1.12 3.57 26.6 4.42 61.6 2.72 2.31 Grain boundary17 0.00258 2.05 28.0 4.00 65.9 0.0106 2.35 C2S18 0.00218 1.63 25.3 3.04 68.7 1.35 2.72 C2S + C19 0.502 1.93 27.4 4.46 64.7 1.02 2.36 C2S20 2.02 9.04 19.2 2.25 58.9 8.52 3.06 Grain boundary21 1.04 4.52 18.8 2.61 68.1 4.98 3.62 Grain boundary22 0.749 1.63 28.5 3.76 65.3 0.0108 2.29 C2S23 1.00 1.58 28.4 3.69 63.8 1.55 2.25 C2S24 1.02 4.68 21.9 2.86 65.7 3.87 3.00 Grain boundary25 0.00218 2.15 15.4 1.09 81.4 0.0113 5.30 Grain boundary

149T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

(Ca3 − 5xMn4xV0x)(Si1 − 2xP2x)O5 with x ≤ 0.005 (for inclusion lessthan 0.69 wt.% Mn2O3). The X-ray analysis showed that the M3 poly-morph of C3S was undoubtedly formed at Mn and P inclusions lessthan 0.69 wt.%, Mn2O3 and 0.62 wt.% P2O5. At higher amounts ofthis order, the alite phase was not formed and solid solutions of C2Stake place with 2 CaO·MnO2 and CaO compounds.

Recent work of Ifka et al. [17] has considered inorganic part of MBMas raw materials. Calcium oxide (CaO) from apatite in bone meal hasbeen accounted in material balance as a potential reactant susceptibleto produce clinker phases. Indeed, by considering bone meal as rawmaterial, the content of limestonewas adequately reduced and the con-tent of P2O5 could climb from 1.11 wt.% without balancing to 1.52 wt.%.

Table 5EPMA line analysis of sample shown in Fig. 2.

Point MgO(wt.%)

Al2O3

(wt.%)SiO2

(wt.%)P2O5

(wt.%)

1 0.562 1.76 24.1 1.812 1.42 4.49 20.2 1.203 1.39 5.41 20.8 1.214 0.878 1.25 22.2 2.005 1.32 0.922 24.3 2.206 1.38 1.92 20.5 1.497 0.755 1.34 21.4 2.268 0.735 1.56 23.1 1.389 0.610 1.23 23.5 1.0410 0.912 1.12 21.6 1.2011 3.17 12.8 14.2 1.4112 4.02 19.1 4.57 0.0030213 0.00129 0.750 0.978 0.0019314 0.00205 1.87 4.07 0.0028715 0.00164 0.680 8.23 1.2616 0.00241 1.90 25.6 4.5117 0.00260 1.60 27.0 4.2218 0.00285 2.20 29.9 2.8819 0.00293 2.21 28.3 3.9420 0.927 2.13 29.6 3.4321 0.00250 2.09 25.8 4.3322 0.00246 1.67 26.5 3.8923 0.356 1.06 22.9 3.4424 0.00277 1.76 27.0 4.50

The aim of the present investigation is to identify the composition ofclinker minerals as solid solutions with P2O5 after the addition of bonemeal (BM) ash to the basic raw mixture.

2. Experimental

2.1. The preparation of clinker samples

The ashes of BM, lime, marl and blast furnace slag with determinedchemical composition were used for preparation of the raw mixtures(Table 1). Chemical analysis of starting raw materials was realized by

CaO(wt.%)

Fe2O3

(wt.%)CaO/SiO2 Solid solution

71.7 0.00822 2.97 C3S67.7 4.93 3.35 C3S + C6AxFy68.3 2.90 3.29 C3S + C6AxFy73.7 0.00867 3.32 C3S71.3 0.00976 2.94 C3S74.7 0.0101 3.64 C3S74.3 0.00888 3.47 C3S73.2 0.00927 3.17 C3S73.6 0.00899 3.13 C3S74.0 1.18 3.43 C3S58.6 9.80 4.12 Grain boundary55.7 16.5 12.2 C6AxFy69.2 29.1 70.8 C2F + C90.9 3.16 22.3 C87.9 1.91 10.7 C68.0 0.0112 2.65 C2S + C65.0 2.10 2.41 C2S63.8 1.19 2.14 C2S65.6 0.0104 2.32 C2S63.9 0.0111 2.16 C2S65.3 2.49 2.53 C2S66.8 1.20 2.52 C2S72.2 0.0116 3.15 C2S + C65.3 1.52 2.42 C2S

Fig. 2.Microstructure of the sample with 1.81 wt.%. P2O5 (1st series).

Fig. 3.Microstructure of the sample with 1.81 wt.%. P2O5 (1st series).

150 T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

X-Ray Fluorescence (XRF) Analysis Spectro 2000. The preparation of ashis reported in [17].

Raw mixtures were prepared according to calculated recipe(Table 2) and elected ash of BM additions. The value of lime saturationfactor (LSF) was kept below 100 at maximum ash of BM addition.

Two series of samples with different contents of the ash wereprepared.

1st series: The ash of BM was added to the basic mixtures withoutcomputing its CaO into the total CaO needed to react withSiO2, Al2O3 and Fe2O3. The lime saturation factor is higherthan that of mixture without ash addition.

2nd series: The CaO from the BM ash was considered in the calculationof total CaO in the mixture. The lime saturation remainsconstant, but the content of limestone decreases.

The homogeneous rawmeals were pressed into pellets 3 g inweight(11 MPa, pressure hold 2 min). The pellets were fired at the tempera-ture 1450 °C. The rate of temperature increase was set to 10 °C/minand isothermal hold to 1 h. At this temperature, the samples weretaken out from furnace and cooled rapidly by air flowing. Chemicalanalysis of free lime (f-CaO) in the samples was achieved according toethylene glycol extraction method [18]. In order to identify clinkerminerals and to detect P2O5 content in their solid solutions, ElectronProbe Micro Analyzer (EPMA JEOL JXA-840A, EDS parameters —

15 KV, Takeoff Angle 40.0°) was used. All samples selected for micro-analysis were carbon coated.

3. Results and discussion

3.1. Chemical analysis of free lime

The results of free lime analysis for both series of samples andelected addition of BM ash to raw mixtures are shown in Table 3.

The free lime content increaseswith increasing ash of BM in both se-ries of analyzed samples. The increase in f-CaO content in clinker is dueto the fact that P2O5 enters C2S and stabilizes it. Once P2O5 stabilizes C2S,it becomes hard to react with CaO to form C3S. The increasing contentof P2O5 is the cause that the formation of alite is increasingly retardedor later totally blocked. As it was mentioned by [19], P2O5 can be usedto stabilize β-C2S against its transformation to γ-C2S. The comparisonof the free lime content in the 1st and 2nd series of samples burnt at1450 °C shows that the free lime content is lower in the 2nd series. Itmeans that considering CaO from the BM ash in the calculation oftotal CaO in the mixture allows to burn a higher amount of MBM incement rotary kiln without negative influence to clinker properties.

3.2. Chemical microanalysis results

The line analysis conducted by EPMA with detected oxide composi-tion in each point in the line was done. C2S and C3S were identifiedaccording to SiO2 content and CaO/SiO2 weight ratio.

As presented in Fig. 1, alite forms sharply bordered crystal (points 1–5). Belite crystals havemostly rounded shape (points 9–11, 13, 15, 17–19,22–23). Oxide composition in points 7 and 8 which is located betweencalcium silicate crystals showed the presence of C4AF in this area.However, pure C4AF doesn't occur in the clinker. The compounds froma variety of solid solutions between C2A and C2F are present. Therefore,the more correct designation of aluminoferrite phase is C6AxFy(Table 4).

Because cooling rate of the clinker is rapid, some calcium silicatecrystals have cracks. These cracks form as a result of thermal expansiondifferences between alite or belite and interstitial phase (aluminate andferrite). As shown in Table 5, the content of P2O5 in C2F and C6AxFy isminimal. The alite and belite accommodate nearly all the phosphorusin the system.

The f-CaO phase, present in the clinker with maximum P2O5

content mainly due to the stabilization of C2S, is located in the area ofpoints 15–16 and 20–22 (Fig. 3, Table 6).

The intervals of P2O5 content in alite and belite for both series ofsamples are shown in Figs. 4 and 5. The P2O5 content in both phases in-creases with the content of P2O5 in the clinker. Figs. 4 and 5 present thatP2O5 content in belite is higher than that in alite. The content of P2O5 inalite and belite is higher in the 2nd series. These results and the resultsfrom chemical analysis of free lime (lower content of f-CaO in the 2ndseries of samples) showed that CaO in the form of calcium phosphateis less reactive than CaO as product of limestone decomposition. CaOas product of limestone decomposition reacts preferentially. Becauseof the decreased addition of the limestone, CaO from calciumphosphateis required to react with SiO2. This CaO and also P2O5 from calciumphosphate enter the reaction to form belite with more P2O5 in itsstructure.

3.3. The formulas of C2S and C3S solid solutions

The oxide composition obtained by the line analysis showed that inaddition to CaO and SiO2, alite and belite contain minor quantities ofMgO, Al2O3, P2O5 and Fe2O3 (Table 7). According to the quantities of

Table 6EPMA line analysis of sample shown in Fig. 3.

Point MgO(wt.%)

Al2O3

(wt.%)SiO2

(wt.%)P2O5

(wt.%)CaO(wt.%)

Fe2O3

(wt.%)CaO/SiO2 Solid solution

1 0.465 1.90 26.9 3.62 64.8 2.38 2.41 C2S2 0.406 1.91 26.8 3.90 65.2 1.80 2.43 C2S3 2.45 12.8 4.98 0.00249 64.9 15.0 13.0 C6AxFy4 1.08 4.23 18.3 2.03 70.8 3.55 3.88 C2S + C6AxFy5 0.711 1.71 26.7 4.42 66.5 0.00996 2.49 C2S6 0.00159 1.59 15.5 1.28 79.8 1.86 5.15 C2S + C7 0.00260 2.05 28.0 4.71 65.3 0.00931 2.33 C2S8 0.00275 2.50 29.2 4.56 62.6 1.17 2.14 C2S9 0.537 3.73 26.6 3.92 62.6 2.57 2.35 C2S10 0.542 2.29 27.8 3.55 64.0 1.84 2.30 C2S11 0.449 2.19 26.2 3.23 65.7 2.16 2.51 C2S12 0.350 1.98 14.8 1.54 77.8 3.55 5.27 Grain boundary13 0.782 2.40 20.6 2.41 72.1 1.64 3.50 C3S14 0.748 5.45 15.7 2.26 68.4 7.37 4.35 Grain boundary15 1.22 0.00200 0.00233 0.00247 98.8 0.00757 42389.7 C16 1.23 0.00251 0.959 0.00288 97.8 0.00953 102.0 C17 0.00229 1.34 25.2 3.47 70.0 0.0108 2.78 C2S + C18 2.24 9.47 10.7 0.643 67.0 9.88 6.23 CxS + C6AxFy19 0.706 6.37 11.9 1.21 70.5 9.26 5.93 CxS + C6AxFy20 0.750 0.00203 0.00256 0.00264 99.2 0.00809 38799.3 C21 1.12 0.00187 0.00219 0.00223 98.9 0.00754 45119.0 C22 1.36 0.695 2.50 0.00275 95.4 0.00835 38.2 C23 0.538 3.25 21.6 2.19 69.4 2.98 3.22 C3S24 0.00251 2.16 25.0 3.51 69.3 0.0104 2.78 C2S + C

151T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

these oxides, the formulas of C2S and C3S solid solutions were calcula-ted. For example, C3S solid solution (bolded in Table 7) formula fromline analysis shown in Fig. 6 is:

C2:951M0:049S0:936A0:028 F0:007P0:029 ð8Þ

where C = CaO, M = MgO, S = SiO2, A = Al2O3, F = Fe2O3, and P =P2O5.

Tables 8–11 contain final formulas of C2S and C3S solid solutionsfor all samples from both series. Each final formula of solid solutionwas calculated from 5 line analysis of the sample.

As shown in previous tables, the value of SiO2 stoichiometric coeffi-cient decreases with increasing value of P2O5 stoichiometric coefficientin belite and alite formulas. Decreasing value of SiO2 stoichiometriccoefficient is in accordance with ionic substitution stated in [11]. Theresults also showed that PV enters the structure of C2S and C3S solid so-lutions with AlIII and FeIII. The values of Al2O3 and Fe2O3 stoichiometriccoefficients increase with increasing P2O5 stoichiometric coefficient inthe formulas of calcium silicate solid solutions. The comparison of the

0.0 0.4 0.8 1.2 1.6 2.00

2

4

6

Con

tent

of P

2O5

in th

e ph

ase

(wt.

%)

P2O5 added from the ash of BM (wt. %)

C2S

C3S

Fig. 4. Content of P2O5 in belite and alite for the 1st series of samples.

values of Al2O3 and Fe2O3 stoichiometric coefficients in belite and aliteformulas showed that these calcium silicates incorporate more AlIII

than FeIII into their structures.According to the dependences shown in Fig. 7, increasing content of

P2O5 added to the rawmeal from ash of BM lowers the value of stoichio-metric coefficient of SiO2 in both calcium silicate solid solutions.

The dependences in Figs. 8 and 9 present that the values of Al2O3

and Fe2O3 stoichiometric coefficients in solid solution formulas of bothcalcium silicates in burnt clinker samples increase on the increasingP2O5 addition from ash of BM to the raw meal.

Further increasing the addition of P2O5 to the raw meal will lead toinability of belite and alite to incorporate more AlIII or FeIII into theirstructures. The content of Al2O3 and Fe2O3 in calcium silicate solidsolutions should approach to some limit value.

Fig. 10 shows the dependences of P2O5 stoichiometric coefficient incalcium silicate solid solution formulas on the content of P2O5 addedto the raw meal from ash of BM. The value of P2O5 stoichiometriccoefficient in both calcium silicate solid solutions increases with theaddition of P2O5 up to 1.9 wt.% to the raw meal. The comparison of the

0.0 0.4 0.8 1.2 1.6 2.00

2

4

6

Con

tent

of P

2O5

in th

e ph

ase

(wt.

%)

P2O5 added from the ash of BM (wt. %)

C2S

C3S

Fig. 5. Content of P2O5 in belite and alite for the 2nd series of samples.

Table 7EPMA line analysis of sample shown in Fig. 6. The values bolded were used to calculate the formula of C3S solid solutions.

Point MgO(wt.%)

Al2O3

(wt.%)SiO2

(wt.%)P2O5

(wt.%)CaO(wt.%)

Fe2O3

(wt.%)CaO/SiO2 Solid solution

1 0.000569 0.000794 0.00176 0.00186 81.4 18.6 46256.0 C2F + C2 0.000861 0.404 1.18 0.00217 88.8 9.67 75.5 C2F + C3 0.982 1.07 22.2 1.61 74.2 0.00936 3.34 C3S4 0.613 0.881 18.4 0.00314 80.1 0.0101 4.36 C3S + C5 0.00199 0.721 14.6 0.561 84.1 0.00961 5.75 C3S + C6 0.886 0.995 20.9 1.06 76.1 0.00949 3.64 C3S7 0.745 1.40 22.8 0.902 73.2 1.01 3.21 C3S8 4.27 17.7 3.91 0.00322 56.3 17.9 14.4 C6AxFy9 0.00130 3.69 1.90 0.00263 74.1 20.3 39.0 C2F + C10 0.00214 0.00258 19.7 0.862 79.5 0.0101 4.05 C3S + C11 1.09 0.00288 21.9 1.59 75.4 0.00934 3.44 C3S12 0.801 1.16 21.5 1.93 74.6 0.00861 3.47 C3S13 1.07 1.20 22.3 1.50 72.5 1.46 3.26 C3S14 0.965 0.803 22.6 1.84 72.8 1.03 3.22 C3S15 1.00 0.919 23.7 2.36 70.2 1.82 2.96 C3S16 0.793 2.00 24.1 1.87 71.3 0.00902 2.96 C3S17 4.07 14.8 10.6 0.824 58.2 11.5 5.49 Grain boundary18 0.655 3.06 25.5 3.21 65.5 2.06 2.57 C2S19 1.63 8.76 8.69 0.00307 67.2 13.7 7.73 C6AxFy20 0.648 1.56 22.4 2.08 73.3 0.00818 3.27 C3S21 1.73 6.01 1.19 0.639 86.5 3.92 72.8 C6AxFy22 1.91 8.83 14.3 0.00323 70.3 4.59 4.90 Grain boundary23 0.924 1.08 22.5 1.91 73.6 0.00881 3.28 C3S24 0.643 1.28 22.6 1.12 74.4 0.00858 3.29 C3S25 0.502 0.630 18.8 1.67 78.4 0.00943 4.16 C3S + C26 2.01 3.45 17.0 1.27 74.2 2.07 4.35 Grain boundary27 0.501 4.78 3.10 0.00268 77.5 14.1 25.0 C6AxFy

Fig. 6.Microstructure of the sample with 1.81 wt.%. P2O5 (1st series).

Table 8The formula of C2S solid solution in the 1st series of samples.

Addition of P2O5

(wt.%)

00.380.751.111.461.81

152 T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

values of P2O5 stoichiometric coefficients in belite and alite in both se-ries of samples shows that more PV is incorporated in the structure ofcalcium silicates in the 2nd series.

4. Conclusions

The electron microanalysis results showed that belite and aliteaccommodate nearly all the phosphorus in the clinker. It was found outthat belite and alite contained more P2O5 in the samples where CaOfrom ash of BM was considered in the calculation of total CaO in theraw mixture. The formulas of C2S and C3S solid solutions werecomputed from measured oxide composition. The increasing addition ofP2O5 to the basic raw mixture leads to decrease in SiO2 and increase inP2O5, Al2O3 and Fe2O3 in both calcium silicate solid solutions. PV entersthe structure of calcium silicates with AlIII and FeIII. The content of Al2O3

in calcium silicates is higher than that of Fe2O3. It can be expected thatthe quantity of SiO2, P2O5, Al2O3 and Fe2O3 in C2S and C3S solid solutionswill approach to some limit value (degree of saturation) with increasingaddition of P2O5 from the ash of BM. Searching for these limit valuesdoesn't have significance in practical terms because the clinker will con-tain high content of free lime at higher addition of the BM ash.

Solid solution formula of C2S(C = CaO, M = MgO, S = SiO2, A = Al2O3, F = Fe2O3, P = P2O5)

C1.988 ± 0.007 M0.012 ± 0.007 S0.97518 ± 0.00001 A0.019 ± 0.001 F0.005 ± 0.002P0.00005 ± 0.00001

C1.988 ± 0.002 M0.012 ± 0.002 S0.957 ± 0.004 A0.025 ± 0.003 F0.009 ± 0.004 P0.009 ± 0.001

C1.987 ± 0.007 M0.013 ± 0.007 S0.940 ± 0.007 A0.029 ± 0.005 F0.009 ± 0.004 P0.022 ± 0.003

C1.987 ± 0.005 M0.013 ± 0.005 S0.924 ± 0.004 A0.031 ± 0.001 F0.011 ± 0.002 P0.034 ± 0.002

C1.987 ± 0.002 M0.013 ± 0.002 S0.909 ± 0.004 A0.035 ± 0.003 F0.011 ± 0.003 P0.045 ± 0.002

C1.987 ± 0.006 M0.013 ± 0.006 S0.897 ± 0.005 A0.038 ± 0.003 F0.012 ± 0.003 P0.053 ± 0.001

Table 10The formula of C2S solid solution in the 2nd series of samples.

Addition of P2O5

(wt.%)Solid solution formula of C2S(C = CaO, M = MgO, S = SiO2, A = Al2O3, F = Fe2O3, P = P2O5)

0 C1.988 ± 0.007 M0.012 ± 0.007 S0.97518 ± 0.00001 A0.019 ± 0.001 F0.005 ± 0.002 P0.00005 ± 0.00001

0.38 C1.988 ± 0.002 M0.011 ± 0.002 S0.936 ± 0.005 A0.025 ± 0.003 F0.012 ± 0.003 P0.027 ± 0.005

0.76 C1.982 ± 0.007 M0.018 ± 0.007 S0.926 ± 0.007 A0.033 ± 0.005 F0.013 ± 0.003 P0.029 ± 0.001

1.14 C1.984 ± 0.005 M0.016 ± 0.005 S0.911 ± 0.001 A0.033 ± 0.001 F0.013 ± 0.002 P0.043 ± 0.002

1.52 C1.97 ± 0.02 M0.03 ± 0.02 S0.894 ± 0.006 A0.036 ± 0.003 F0.014 ± 0.004 P0.057 ± 0.008

1.9 C1.987 ± 0.002 M0.013 ± 0.002 S0.886 ± 0.004 A0.038 ± 0.002 F0.016 ± 0.003 P0.060 ± 0.003

Table 11The formula of C3S solid solution in the 2nd series of samples.

Addition of P2O5

(wt.%)Solid solution formula of C3S(C = CaO, M = MgO, S = SiO2, A = Al2O3, F = Fe2O3, P = P2O5)

0 C2.966 ± 0.009 M0.034 ± 0.009 S0.9922 ± 0.0008 A0.0075 ± 0.0008 F0.000170 ± 0.000003 P0.0000595 ± 0.0000009

0.38 C2.96 ± 0.02 M0.04 ± 0.02 S0.95 ± 0.01 A0.022 ± 0.003 F0.010 ± 0.005 P0.015 ± 0.006

0.76 C2.964 ± 0.009 M0.036 ± 0.009 S0.950 ± 0.008 A0.023 ± 0.004 F0.011 ± 0.003 P0.016 ± 0.004

1.14 C2.960 ± 0.008 M0.030 ± 0.008 S0.94 ± 0.01 A0.025 ± 0.003 F0.013 ± 0.003 P0.026 ± 0.009

1.52 C2.945 ± 0.003 M0.055 ± 0.003 S0.928 ± 0.003 A0.031 ± 0.002 F0.014 ± 0.005 P0.028 ± 0.004

1.9 C2.95 ± 0.01 M0.05 ± 0.01 S0.91 ± 0.01 A0.033 ± 0.007 F0.014 ± 0.002 P0.040 ± 0.007

Fig. 7. The dependence of SiO2 stoichiometric coefficient in the formula of belite and alite on the content of P2O5 added to the raw meal from ash of BM.

Table 9The formula of C3S solid solution in the 1st series of samples.

Addition of P2O5

(wt.%)Solid solution formula of C3S(C = CaO, M = MgO, S = SiO2, A = Al2O3, F = Fe2O3, P = P2O5)

0 C2.966 ± 0.009 M0.034 ± 0.009 S0.9922 ± 0.0008 A0.0075 ± 0.0008 F0.000170 ± 0.000003 P0.0000595 ± 0.0000009

0.38 C2.95 ± 0.01 M0.05 ± 0.01 S0.973 ± 0.009 A0.016 ± 0.002 F0.007 ± 0.002 P0.0042 ± 0.0001

0.75 C2.959 ± 0.008 M0.041 ± 0.008 S0.97 ± 0.01 A0.0161 ± 0.0005 F0.008 ± 0.001 P0.011 ± 0.004

1.11 C2.962 ± 0.008 M0.038 ± 0.008 S0.952 ± 0.003 A0.022 ± 0.004 F0.0081 ± 0.0002 P0.018 ± 0.003

1.46 C2.955 ± 0.003 M0.045 ± 0.003 S0.94 ± 0.01 A0.027 ± 0.001 F0.010 ± 0.001 P0.0236 ± 0.0032

1.81 C2.949 ± 0.002 M0.051 ± 0.002 S0.928 ± 0.008 A0.031 ± 0.003 F0.011 ± 0.004 P0.030 ± 0.001

Fig. 8. The dependence of Al2O3 stoichiometric coefficient in the formula of belite and alite on the content of P2O5 added to the raw meal from ash of BM.

153T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

Fig. 9. The dependence of Fe2O3 stoichiometric coefficient in the formula of belite and alite on the content of P2O5 added to the raw meal from ash of BM.

Fig. 10. The dependence of P2O5 stoichiometric coefficient in the formula of belite and alite on the content of P2O5 added to the raw meal from ash of BM.

154 T. Ifka et al. / Cement and Concrete Research 59 (2014) 147–154

Acknowledgment

This work was supported by courtesy of the Slovak Grant Agency(VEGA 1/0064/12) and by Project Centres for Materials Research atFCH BUT, reg. no. CZ.1.05/2.1.00/01.0012.

References

[1] H.F.W. Taylor, Cement Chemistry, 2nd ed. Thomas Telford Publishing, London, 1997.[2] R.W. Nurse, The effect of phosphate on the constitution and hardening of Portland

cement, J. Appl. Chem. 2 (1952) 708–716.[3] B. Matkovich, J.F. Young, Dicalcium Silicates Doped with Phosphates, in: Finep (Ed.),

2, 8th ICCC, Rio de Janeiro, Brazil, 1986, pp. 276–281.[4] W. Gutt, High temperature phase equilibria in the system 2CaO·SiO2–3CaO·P2O5–

CaO, Nature 197 (1963) 142–149.[5] W. Gutt, Manufacture of Portland cement from phosphatic rawmaterials, 15th ICCC,

Tokyo, Japan, 1968. 93–105.[6] L. Halicz, Y. Nathan, L. Ben-Dor, The influence of P2O5 on clinker reactions, Cem.

Concr. Res. 14 (1) (1984) 11–18.[7] S. Puntke, Auswirkungen des Phosphateintrages in Drehofenanlagen der

Zementindustrie auf Klinkermineralogie und Zementeigenschaften, Dissertation,TU Clausthal, 2004.

[8] A. Diouri, A. Boukhari, J. Aride, F. Puertas, T. Vazquez, Formation d'hydroxyapatiteCa5(PO4)3OH en milieu silicaté, Mater. Constr. 45 (236) (1994) 5–13.

[9] A. Diouri, A. Boukhari, J. Aride, F. Puertas, T. Vazquez, Research of the lime richportions of the CaO–SiO2–P2O5 system, Mater. Constr. 45 (237) (1995) 3–13.

[10] A. Diouri, A. Boukhari, J. Aride, F. Puertas, T. Vazquez, Elaboration of aVL-C2S form ofbelite in phosphatic clinker. Study of hydraulic activity, Mater. Constr. 48 (249)(1998) 23–32.

[11] H. Saalfeld, K.H. Klaska, The crystal structure of 6 Ca2SiO4·1 Ca3(PO4)2, Z. Kristallogr.Cryst. Mater. 155 (1–2) (1981) 65–73.

[12] K. Fukuda, H. Taguchi, T. Fukuda, Effect of substituent ions on martensitic transfor-mation temperatures in dicalcium silicate solid solutions, J. Am. Ceram. Soc. 85 (7)(2002) 1804–1806.

[13] T. Staněk, P. Sulovský, The influence of phosphorous pentoxide on the phase compo-sition and formation of Portland clinker, Mater. Charact. 60 (2009) 749–755.

[14] My.Y. Benarchid, A. Diouri, A. Boukhari, J. Aride, R. Castanet, J. Rogez, Thermal studyof chromium–phosphorus-doped tricalcium aluminate, Cem. Concr. Res. 31 (2001)449–454.

[15] My.Y. Benarchid, A. Diouri, A. Boukhari, J. Aride, J. Rogez, R. Castanet, Elaboration andthermal study of iron–phosphorus-substituted dicalcium silicate phase, Cem. Concr.Res. 34 (2004) 1873–1879.

[16] A. Diouri, A. Boukhari, J. Aride, F. Puertas, T. Vazquez, Stable Ca3SiO5 solid solutioncontaining manganese and phosphorus, Cem. Concr. Res. 27 (1997) 1203–1212.

[17] T. Ifka, M.T. Palou, Z. Bazelová, The influence of CaO and P2O5 of bone ash upon thereactivity and the burnability of cement raw mixtures, Ceram. Silik 56 (1) (2012)76–84.

[18] M.P. Javellana, I. Jawed, Extraction of free lime in portland cement and clinker byethylene glycol, Cem. Concr. Res. 12 (3) (1982) 399–403.

[19] M. Zhao, Quantitative control of C2S crystal transformation, Appl. Mech. Mater.121–126 (2012) 311–315