Quantitative Analysis of Clinker Phases Using XRF

The cement industry faces a very interesting challenge of quantitative analysis of clinker phases. Other than traditional chemical analysis with XRF techniques, the role of XRD has been highly solicited for analyzing phases or minerals.

Analyzing phases such as free lime in clinkers and limestone additions in cement by XRD has been extremely useful in controlling the kiln process and the quality of the end product, respectively. XRD has substituted traditional wet chemical or other methods in many cement laboratories over the past ten years. A study was performed to introduce reliable, cost effective and rapid analytical methods to address clinker phase analysis. This study is focused on analyzing clinker phases in real time kiln conditions and examining the potential use of XRD as a substitute for microscopy or other indirect methods of calculation.

Instrument and Sample Preparation

A Thermo Scientific ARL 9900 Series was used for the study. The instrument has a series of XRF fixed channels (monochromators), an XRF goniometer, and an XRD goniometer integrated such that both XRD and XRF measurements can be done on the same sample with the same tube conditions and under a vacuum.

Over the production time, a series of 30 clinker samples were collected. Microscopy measurements were done in parallel on the 30 samples to obtain quantitative data for C₃S, C₂S, C₃A and C₄AF phases. A set of samples were chosen to obtain a reasonable dynamic range (working range) for concentrations of the four clinker phases. These clinker granules were ground and pressed into pellets before being measured by the ARL 9900 Series Total Cement Analyzer.

While normally observing clinkers microscopically, the biggest limitation lies in the presence of a vitreous interstitial phase which makes it difficult to differentiate the ferrite and the aluminium oxide. This does not happen with the study samples since the clinker samples are cooled slowly allowing C₃A crystallisation. The microscopically determined interstitial phase (alumina and ferrite) on the other hand is always smaller than that calculated by Bogue formulæ as they assume that all the Al₂O₃ and Fe₂O₃ contributes to the formation of C₃A and C4AF without considering the possibility that part of those oxides can form solid solutions with silicates.

The total analysis time by the compact integrated XRD system for free lime and the four clinker phases is less than 4 minutes. Peak intensities are used directly without corrections.

Results and Discussion

The results of chemical analysis (total oxide concentration) obtained by the XRF part of the instrument is shown in Table 1. These concentrations are used to calculate the equivalent concentrations of clinker phases using Bogue formulæ:

%C₃S = (%CaO - %CaO free) x 4.07 - (%SiO₂ x 7.6 + %Fe₂O₃ x 1.43 + %Al₂O₃ x 6.72)

%C₂S = %SiO₂ x 2.87 - C₃S x 0.75

%C₄AF = %Fe₂O₃ x 3.04

%C₃A = %Al₂O₃ x 2.65 - %Fe₂O₃ x 1.69

Table 1. XRF analysis of clinkers.

Figure 1. Calibration curve for alite (C₃S) in a series of industrial clinker samples using the Integrated XRD system

Figure 2. Calibration curve for belite (C₂S) in a series of industrial clinker samples using the integrated XRD system.

Figure 3. Calibration curve for aluminate (C₃A) in a series of industrial clinker samples using the Integrated XRD system.

The results of the clinker phases obtained with the compact XRD system combined in the ARL 9900 in comparison with the values obtained from microscopy and the Bogue calculations are shown in Table 1.

Figure 4 shows that microscopy values and XRD results obtained with the ARL 9900 TCA correlate well while the Bogue data can be seriously in error compared to microscopy. Figure 5 shows the very poor correlation between microscopy as a reference and Bogue calculation for C₃S obtained with the data of Table 2.

Figure 4. Microscopy vs Compact XRD for C₃S: good correlation.

Figure 5. Microscopy vs Bogue XRD for C₃S: no correlation.

Table 2. Quantitative XRD analysis of free lime and clinker phases in comparison with microscopy and Bogue calculations.

Conclusion

Quantitative results were compared in this case study obtained in a series of clinker samples under three methods including Bogue formulæ, microscopy and compact Integrated XRD using an ARL 9900 Series Total Cement Analyzer.

The results from Bogue are not as expected. The comparison between microscopy and quantitative XRD data shows that it is possible to exploit the unique compact integrated XRD system of ARL 9900 Series in real process conditions for clinker phases analysis. As C₃S phase provides resistance to cement such XRD analysis can help predict the 3-day resistance to compression of the future cement obtained from these clinkers while 7-day resistance and 28-day resistance can be influenced by other factors such as fineness and particle size distribution for example.

This information has been sourced, reviewed and adapted from materials provided by Thermo Fisher Scientific - Elemental Analyzers.

For more information on this source, please visit Thermo Fisher Scientific - Elemental Analyzers.