Granulated ground blast furnace slag (GGBFS) is a key constituent of CEM II/A-S, CEM II/B-S and CEM III cements. Blast furnace slag is mostly an amorphous product, comprising little amounts of crystalline phases such as Akermanite (Ca2MgSi2O7), Merwinite (Ca3MgSi2O8), or Quartz (SiO2). It offers hydraulic properties same as those of Portland Cement and thus may partly substitute the clinker component in cement blends.
Clinker burning is an important source of CO2 emissions. Thus, blast furnace slag cements are increasingly produced so as to reduce cement which makes CO2 emissions. Present regulations permit several methods to determine the slag content in cements, such as gravity separation, selective dissolution, microscopic analysis, or the estimation based on the chemical composition. These methods are either laborious or requires knowledge about the sample. This report illustrates how amorphous phase analysis can be integrated into the TOPAS Rietveld calculation without the need for a standard or further calibration.
Rietveld Quantification of Amorphous Phases
Conventional Rietveld analysis takes only crystalline phases into consideration. The relative weight fractions are normalized to 100 wt.%. To calculate amorphous amounts indirectly, a known weight fraction of an internal standard can be added to the sample. This is known as the “spiking method”. This method cannot be realized in an automated laboratory and systematic errors can come up due to microabsorption effects caused by differences in mass absorption of sample and standard. Precise analysis using the spiking method needs similar mass absorption and grain size distribution of sample and standard. Unlike traditional Rietveld, TOPAS takes into account phases with partially or no crystal structure in the calculation (PONKCS) with the help of hkl_Phases (Pawley or Le Bail fitting).
Quantitative Rietveld Analysis Using Structures
The weight fraction wi of the i-th phase in Rietveld analysis is defined by the following equation:
where si is the scaling parameter, Vi is the volume of the unit cell, and Mi Zi is the weight of the atoms (M = mass of one formula unit, Z = number of formula units in cell) inside the unit cell.
Rietveld Quantification Using hkl_Phases Instead of Structures
The intensity values are obtained from peak intensity values while using hkl_Phases.
When an hkl_Phase is used in quantitative Rietveld analysis, only the volume of the unit cell, V, is known. This method requires the “calibration” of the mass (MZ) of the hkl_Phase, because of the lack of structural information.
The measurements were performed using a D4 ENDEAVOR diffractometer in Bragg-Brentano Geometry having the one-dimensional LynxEye compound silicon strip detector (Figure. 1). The settings are provided in Table 1. The quantitative phase analysis was carried out using the DIFFRACplus TOPAS (Version 4) software.
Table 1. D4 ENDEAVOR configuration with the LynxEye Detector
||D4 ENDEAVOR Theta/2Theta
||2.2 kW Cu long fine focus
||35 kV / 50 mA
||Divergence slit fixed to 0.5° 4° Soller slit
||Rotating sample holder
||Nickel Kâ Filter 4° Soller slit
||LynxEye (opening 3.9°)
|time per step
|angular range (2Theta)
||10° to 65°
|Total Measuring time
||9 min 50 s
A round robin was conducted on quantitative phase analysis of blast furnace slag cements with the help of XRD methods. A group of samples of three Slag Cements of different compositions was distributed to the participants for analysis. The preliminary results published provided the reference values of the slag amounts in each cement sample
Table 2. Composition of the Slag Cement Samples used in the VDZ Round Robin in wt.%
||Slag Content in wt.%
||CEM III/B 32,5 N-NW/HS/NA
||CEM III/B 42,5 N-NW/HS/NA
Figure 1. The one-dimensional LynxEye compound silicon strip detector
About 10 g of each sample were ground in the POLAB APM automatic preparation unit, using Polysius tablets as binder and the samples were pressed in steel rings.
Setup of the hkl_Phase Model
The modeling of the amorphous diffraction data is done using the steps listed below:
- Determination of the pure blast furnace slag
- Whole powder pattern decomposition of the amorphous intensities by Pawley fitting using an arbitrary start model
- Empirical calibration of the mass (MZ) of this model to satisfy the results of the reference samples
This method results in a complete description of the amorphous diffraction characteristics shown in Figure 2.
Figure 2. Structureless modelling of amorphous blast furnace slag. The blue curve represents the measured data. The calculated model is represented by the red curve. The difference is plotted in grey.
All the three round robin samples were measured using the same hkl_Phase model for determining the blast furnace slag. The repeatability of measurement was studied by analyzing each sample five times. For each run, the sample was unloaded and reloaded to the diffractometer. Figure 3 shows measurement data of sample 1 and the result of TOPAS quantification.
Figure 3. Measurement result (blue) and TOPAS calculation (red) of Slag Cement sample 1. The difference of both is given in grey. The marks indicate the peak positions of each phase with a known structure. The blue curve above the difference curve indicates the intensity contribution of the amorphous blast furnace slag.
The advantages of the TOPAS PONKCS method are given below:
- This method provided precise results (Tables 2 and 3) across a wide range of slag concentration.
- The repeatability was much better than that of established methods.
- The absolute standard deviation of the determined slag concentrations was 0.2wt.%.
Table 3. TOPAS quantitative phase analysis of the VDZ round robin slag cement samples (values given in wt.%)
||Slag Content in wt.%
TOPAS proved to be the next generation of Rietveld analysis through the fundamental parameters approach. Its excellent performance in quantitative phase analysis is based on mathematical stability, powerful minimization algorithm, and unrivalled speed of calculation. The quantitative analysis capabilities of TOPAS are significantly extended with the addition of PONKCS. With the help of modern detector technology, such as the ultra fast Compound Silicon Strip Detector LynxEye, results can be obtained in a matter of minutes.
The benefits of the PONKCS method are listed below:
- Easy to implement since no calibration is needed at the plant level
- Independent of tube aging
- Fast and operator-independent analysis
- Rapid measurement allows implementation of this analysis for process control
About Bruker AXS
Bruker AXS, an operating company of Bruker Corporation (NASDAQ:BRKR) is a global market and technology leader in materials research and quality control instrumentation for elemental and crystalline structure investigations. We develop and manufacture high-quality analytical X-ray systems and complete solutions for material analysis. For the mining & exploration activities, Bruker offer handheld XRF analyzers and XRD based equipment.
This information has been sourced, reviewed and adapted from materials provided by Bruker AXS.
For more information on this source, please visit Bruker AXS