Background

X-ray Fluorescence Spectrometry is used extensively for the analysis of cements and associated raw materials. Simple sample preparation, high accuracy and precision are the main reasons for this choice. Accurate cement analysis is important for many reasons, for example, superior process control with associated cost savings; minimizing and controlling contamination and environmental pollution during cement manufacture; quality control of products; and pure research. Accurate quantification requires accurate, spectrally clean net peak intensities, accurate corrections for inter-element matrix effects and good standards.
CEMOXI Synthetic Standards

CEMOXI is a set of synthetic standards together with an application set-up that will handle fused bead major element analysis of cement final product and cement raw materials. CEMOXI can be used as a primary fused bead calibration or to verify customer in-house standards for pressed powder applications.
The Axios-Cement Spectrometer

The Axios-Cement spectrometer with CEMOXI standards and the SuperQ software FP algorithm makes up a unique system that consistently produces very high quality major and minor element analyses in a wide variety of cements and the raw materials used in its manufacture, including limestones, iron ore, clay/shales and gypsum.
Preparation of Standards and Samples

The CEMOXI standards set comprises 9 multi-element synthetic "cement" standards. The standards are delivered as powders to be prepared as fused beads by the customer, using the customer’s flux, dilution ratios and methodology. They are therefore suited and applicable to every cement industry laboratory.

The standards have been made from high-purity chemicals, pre-conditioned, weighed, ground and then ignited at 1050ºC before packing under nitrogen. Prior to fusion the standards are preignited at 800ºC, except for gypsum, which is dried at 500ºC. The use of synthetic standards, made from traceable compounds, brings this method significantly closer to being a primary analytical solution, rather than a strictly comparative method using CRMs as standards.
Example Spectra from Standard Materials

The standards and certified reference materials illustrated in the figures and tables below were prepared as fusion beads using 0.9g ignited sample fused in 9.0g flux (66% lithium tetraborate, 34% lithium metaborate).

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for CaO CEMOXI standards and 26 cement CRMs

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for SiO2 CEMOXI standards and 26 cement CRMs

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for Al2O3 CEMOXI standards and 26 cement CRMs

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for Na2O CEMOXI standards and 26 cement CRMs

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for P2O5 CEMOXI standards and 26 cement CRMs

AZoM - Metals, Ceramics, Polymer and Composites - Accuracy verification: comparison of certified and measured values for SO3 CEMOXI standards and 26 cement CRMs

Figures 1. Accuracy verification: comparison of certified and measured values for 9 CEMOXI standards and 26 cement CRMs.
Accuracy

The accuracy of the Axios-Cement using CEMOXI and the SuperQ FP algorithm for major and minor element analyses in cements and related materials is phenomenal. This is illustrated in the plots for CaO, SiO2, Al2O3, Na2O, P2O5 and SO3 (Figures 1-6) and in Tables 1 and 2 where certified and measured values for 25 Certified Reference Materials (CRMs) are compared. In Figures 1-6 the CEMOXI standards are shown as (▲) and cement CRMs as (●). The regression lines are based on CEMOXI standards only. The combination of CEMOXI standards and the PANalytical FP calibration model enables accurate extrapolation of calibrations outside the range in the standards. For example, high alumina cements (up to 70 wt% Al2O3) have been analysed successfully with CEMOXI (Figure 3, Table 1).

Table 1. Analytical accuracy - Comparison of certified and measured values for six major and minor oxides in thirteen cement CRMs. Values were calculated from the CEMOXI calibration lines shown in Figures 1-6. Average absolute differences and average relative errors are also given.

CRM


Description


CaO


SiO2


Al2O3


MgO


Fe2O3


SO3







Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.

BCS353


Portland cements (sulphate-resisting)


64.80


64.54


20.50


20.43


3.77


3.83


0.10


0.10


0.077


0.081


2.25


2.24

BCS354


While Portland cement


70.00


69.02


21.80


21.42


4.85


4.85


0.10


0.08


0.120


0.131


2.25


2.26

BCS372


Portland cement


65.80


65.78


21.30


21.12


5.35


5.38


0.21


0.17


0.190


0.172


2.35


2.39

NIST1880a


Portland cement clinker


63.83


64.28


20.31


20.40


5.18


5.22


0.19


0.19


0.220


0.223


3.25


3.22

NIST1881a


Portland cement clinker (blended with slag)


57.58


58.13


22.26


22.45


7.06


7.18


0.20


0.21


0.146


0.148


3.37


3.38

NIST1882a


Calcium aluminate cement


39.29


38.89


4.01


4.09


39.14


39.28


(0.02)


(-0.01)


(0.070)


(0.070)


2.83


2.83

NIST1883a


Calcium aluminate cement


29.52


29.24


0.24


0.25*


70.04


69.98


0.30


0.28


0.003


0.004


4.61


4.59

NIST1884a


Portland cement clinker


62.26


62.15


20.57


20.56


4.26


4.29


0.22


0.22


0.128


0.131


4.62


4.54

NIST1885a


Portland cement clinker


62.39


62.30


20.91


20.86


4.03


4.01


1.07


1.14


0.122


0.123


3.16


3.10

NIST1886a


White portland cement (low iron)


67.87


67.46


22.38


22.23


3.88


3.85


(0.02)


(0.03)


(0.022)


(0.020)


2.13


2.19

NIST1887a


Portland cement clinker


60.90


60.48


18.64


18.50


6.20


6.17


0.48


0.50


0.306


0.300


2.68


2.71

NIST1888a


Portland cement clinker


63.23


63.17


21.22


21.15


4.27


4.33


0.11


0.14


(0.080)


(0.080)


0.11


0.11*

NIST1889a


Portland cement blended with limestone


65.34


65.07


20.66


20.49


3.89


3.81


0.20


0.20


0.110


0.108


2.94


2.91

Av Abs Diff (wt%)


0.33


0.12


0.05


0.02


0.005


0.03

Av Rel Diff (%)


0.58


0.76 (1.13) *


0.82


8.7


5.0


1.04 (1.47)*

The data in Table 1 illustrate clearly the remarkable accuracy achievable with the Axios-Cement-CEMOXI-FP system. Average absolute differences for all six oxides are extremely low. Average relative differences for CaO, SiO2, Al2O3 and SO3 are better or close to 1% when excluding the very low values marked with an (*). The average relative differences for Na2O and P2O5 (8.7% and 5.0%, respectively) are excellent considering the very low concentrations of these minor elements.
Accuracy in High Concentration Components

The measured Al2O3 values for CRMs NIST 1882a and 1883a demonstrate clearly the ability of the Axios-Cement-CEMOXI-FP system to extrapolate from the highest standard concentration (10% Al2O3) to concentrations as high as 39% and 70% Al2O3 with accuracies of better than 0.4% relative (see also Figure 3). Similar extrapolations are possible for all elements.
Accuracy in Low Concentration Components

Similarly, results of very high quality are obtained also for CaO, SiO2, Al2O3, MgO, Fe2O3 and SO3 in limestone and gypsum CRMs where the average absolute differences are also very low (Table 2).

Table 2. Analytical accuracy - Comparison of certified and measured values for six major and minor oxides in twelve limestone and gypsum CRMs. Values were calculated from the CEMOXI calibration lines shown in Figures 1-6. Average absolute differences and average relative errors are also given.

CRM


Description


CaO


SiO2


Al2O3


MgO


Fe2O3


SO3







Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.


Cert.


Meas.

BCS393


Limestone


55.40


54.30


0.70


0.70


0.12


0.13


0.15


0.17


0.05


0.04


(0.02)


(0.01)

GSR-6


Limestone


35.67


36.07


15.60


15.61


5.03


5.07


5.19


5.33


2.52


2.54


0.09


0.11

GSR-13


Limestone


51.10


51.14


6.65


6.75


0.68


0.64


0.71


0.69


0.21


0.20


(0.01)


(0.01)

NBS88a


Dolomitic Limestone


30.17


30.19


1.20


1.30


0.19


0.21


21.29


21.86


0.28


0.29


(0.01)


(0.01)

NBS1a


Argillaceous Limestone


41.32


40.95


14.11


14.02


4.16


4.09


2.19


2.14


1.63


1.58


0.67


0.63

NBS1b


Argillaceous Limestone


50.90


50.95


4.92


4.92


1.12


1.14


0.36


0.34


0.75


0.77


#


0.04

FGD-1


Calcium Sulphate Dihydrate


32.70


32.48


0.13


0.14


0.02


0.03


0.007


-0.002


(0.01)


(0.01)


46.40


46.52

GYP-A


Gypsum


32.90


32.87


0.45


0.54


0.10


0.08


0.18


0.17


0.05


0.04


46.20


46.59

GYP-B


Gypsum


32.80


32.43


1.05


1.07


0.17


0.16


1.80


1.74


0.07


0.06


41.00


40.77

GYP-C


Gypsum


30.40


30.27


3.50


3.51


0.79


0.78


5.35


5.34


0.40


0.40


33.00


33.48

GYP-D


Gypsum


28.20


28.25


8.70


8.98


2.03


2.00


1.73


1.70


1.08


1.07


36.70


37.36

TIG-1


Calcium Sulphate Dihydrate


32.30


31.97


0.11


0.099


0.57


0.59


0.12


0.11


0.26


0.26


43.40


43.75

Av Abs Diff (wt%)


0.26


0.06


0.02


0.08


0.01


0.29

Av Rel Diff (%)


0.65


4.5


6.1


4.3


5.0


4.0
Precision and Instrument Stability

The precision, repeatability and reproducibility of the Axios-Cement system are excellent, not only for short-term measurements (20 consecutive measurements, Table 3), but also for longer term measurements (measurements carried out over a period of twelve days). For comparison, the counting statistical error (CSE in wt%) is also shown in Table 3.

Table 3. Analytical precision for CEMOXI fusions of NIST-1889a

Element


CaO


SiO2


Al2O3


Na2O


P2O5


SO3

Certified values


65.34


20.66


3.89


0.195


0.110


2.69

Repeatability (20 consecutive measurements)

Mean (wt.%)


65.11


20.35


3.89


0.175


0.142


2.67

RMS (wt.%)


0.05


0.03


0.01


0.004


0.002


0.004

RMS (ref%)


0.08


0.16


0.24


2.07


1.28


0.16

Reproducibility (measurements over 12 days)

Mean (wt.%)


64.97


20.30


3.88


0.174


0.142


2.66

RMS (wt.%)


0.15


0.05


0.01


0.003


0.002


0.008

RMS (ref%)


0.23


0.25


0.27


1.95


1.11


0.031

Counting statistical error

CSE (wt.%)


0.22


0.33


0.22


0.033


0.022


0.11

CSE (rel%)


0.04


0.16


0.27


0.68


0.67


0.19

Twenty consecutive measurements of a single sample (fusion disc of NIST- 1889a) demonstrate relative standard deviations better than 0.3-1% at typical concentration levels for many of the elements commonly analysed in cements, e.g. CaO, SiO2, Al2O3, Na2O, P2O5 and SO3. More importantly, this level of precision is maintained for measurements carried out over a period of 12 days, illustrating the long-term stability of the system.
Summary and Conclusions

Elemental analysis with XRF is already the key to quality and control of production processes in the cement industry.

The Axios-Cement-CEMOXI-FP system further extends the advantages of XRFS as the best analytical method for cement analysis, and is capable of measuring all elements required to control the manufacture of cements. It has been demonstrated that analyses are accurate and precise and the method benefits from simple fusion sample preparation. The reproducibility described in the ASTM C114 norm is easily met.

Furthermore, the stability of the system is such that individual calibrations can be used for months. Time-consuming re-standardizations are unnecessary and the resultant data are highly consistent over time.

Although the Axios-Cement is configured to meet industry requirements (see System Configuration box), it can be upgraded easily with features such as increased power; continuous loading for extra speed of analysis; Pro-Trace for trace element analysis down to sub-ppm level (especially important for the analysis of alternative fuels); and IQ+ for complete standardless analysis.