8/6/2019 Sample Preparation for X-Ray Analysis

http://slidepdf.com/reader/full/sample-preparation-for-x-ray-analysis 1/2

SAMPLING & ANALYSIS

he use o X-ray analysis in the cement industry 

has increased considerably in recent years. While

X-ray fuorescence (XRF) is the universally accepted

standard method or raw material control, X-ray di-

raction (XRD) is now also proving to be an essential

technology in this rapidly evolving industry. Te in-

creased use o XRD has been helped by the development

o advanced instrumentation, enabling the method to

replace traditional calibration, wet chemical or Bogue

methods or quantitative clinker and cement crystalline

phase analysis.

X-ray analysis enables optimised use o raw materi-

als, reduces the energy costs o production and controls

product quality, while also providing the necessary level

o analysis or alternative raw uels, such as hazardous

waste uels. Tese attributes are vital in today’s climate

o stringent environmental legislation and rising

energy costs.

Simple to learn, easy to use

Te key requirements o sample preparation or

both XRF and XRD are that samples are homog-

enous and representative o the source material.

Additionally, the composition o the sample should

not be altered beore or during the preparation

process.

Fortunately, X-ray analysis usually requires

only very simple, inexpensive, sample preparation

work. Methods are both easy to learn and easy to

use. Tereore, even when the need or good sample

preparation is taken into account, X-ray analysis is

a much easier and quicker process than almost all

other chemical analysis techniques.

Sample preparation or XRF

For high quality XRF analysis in ce-

ment production, reproducible samplepreparation and a good calibration o 

the XRF spectrometer are essential.

However, this can be di cult as there are ew suitable

reerence materials available. Furthermore, available

standards may not have the same mineralogy as local

samples, which can lead to inaccurate data.

Preparing samples as pressed pellets is widely re-

garded as the most eective and economical way o 

preparing samples or routine control in cement pro-

duction. Good sample preparation can minimise two o 

the major eects that can lead to bad data—particle size

and mineralogical eects:

Particle size efects

Tese arise due to the penetration depth o X-rays. I 

large, unrepresentative particles are analyzed in the top

layer o a sample, X-ray data will provide inormation

about the composition o these large particles, not the

overall sample composition.

Mineralogical efects

Tese happen when a single element is present in di-

erent phases, a sample that has silicon present as both

silica and silicate, or example. o prevent problems,

it is important that when pressed powder samples are

measured, the standards and samples have the same

mineralogical composition.

Sample preparation or XRD

XRD is a powerul tool or the analysis o the phase prop-

erties o samples. As with XRF, good sample preparationhelps minimise errors and ensures good statistical repre-

sentation. A measurement on a single sample must give

Sample preparation or X-ray analysist h e c r i t i c a l i r s t s t e p

One o the many benefts o X-ray uorescence (XRF) spectrometry and X-ray diraction (XRD) is

simple sample preparation. However, even though it is straightorward, it is still vitally important

to get it right. In this article PANalytical demonstrates that good sample preparation is the critical

frst step to consistent and reliable X-ray analysis.

globalglobalcementcementMAGAZINEMAGAZINE May 2008  17

Figure 1, right: Particle

statistics efects.

(a) Good statistics: all

difraction peaks within the

difractogram produced by a

similar number o crystallites

(true pattern).

(b) Bad statistics: difraction

peaks produced by diferent

numbers o crystallites (alse

pattern).

Figure 2, bottom: Mica

thin layer sample: dusted vs.

pressed.

Figure 3, below: Efect o 

sample spinning.

Remy Denker, Nienke Oosten-Nienhuis, Roger Meier PANalytical BV

(a)

(b)

(a)

(b)

8/6/2019 Sample Preparation for X-Ray Analysis

http://slidepdf.com/reader/full/sample-preparation-for-x-ray-analysis 2/2

the same result as an average measure-

ment on a number o the same samples.

Inadequate sample preparation can

lead to signicant variation, which is

di cult to correct ollowing analysis.

Particle size

Particle size is an important aspect insample preparation or XRD. Te par-

ticle should not be too big, because o 

particle statistics (also known as crys-

tal statistics) eects. Crystal statistics

cause incorrectly measured intensi-

ties. In severe cases, these eects also

infuence the shape o the peaks. In

this instance, peak shapes are not

smooth but show a jagged outline,

which aects peak intensities (see Figure 1).

Equally, the particle should not be too small.

Tis can cause damage to the crystal structure or

alteration o phases, making it di cult to identiy orquantiy the actual phases present in the sample.

Orientation efects

Additionally, preerred orientation eects,

which can be induced merely by touching a

sample, play an important role or the analysis

o, or example, gypsum and calcite samples.

Figure 2 shows comparison between dusted

(random orientation) and pressed (preerred

orientation) samples. Refections o type hk0

disappear in the pressed sample.

Instrumental parameters and samplepreparation solutions

For good crystal statistics, a large number o 

crystallites should contribute to the diraction

process. Tis can be achieved by grinding sam-

ples to reduce the average particle size. A range

o grinders is available or this process, many o 

which are optimised or specic powder types.

It is important to identiy the optimum level o 

sample grinding, or broadening o peaks can

still occur i particles are too small or i lattice

damage occurs. Tereore, thorough investiga-

tion o dierent grinding methods is essential.

Tere are many sample preparation tech-

niques that address the problem o preerred

orientation eects, including back- or side-

loading, surace roughening, use o amorphous

ller or isostatic pressing. Additionally, modern

X-ray analysis instruments are equipped with

spinning sample holders, greatly reducing the

infuence o sample orientation (see Figure 3).

Inuences o sample preparation

parameters

A very important aspect o sample prepara-

tion, particularly or cementitious materials, isensuring that a sample’s physical properties are

not altered during preparation. Tis is achieved

by monitoring the eects o certain parameters,

such as applied pressure and milling time, prior

to actual sample measurement.

Figures 4-7 show data rom studies to de-

termine the infuence o various parameters

on sample preparation. Te data underline

the importance o monitoring and controlling

infuential parameters during the sample prepa-ration process. Tese results cannot, however,

be used to extrapolate a trend to other samples

because cements dier widely in composition.

Inuence o pressure

Te Portland cement analysed in the diracto-

gram in Figure 4 was used to study the eects

o applied pressure when samples are pressed

as part o the preparation method. Te results

on peak intensity are clearly visible in Figure

5. Te higher the applied pressure, the lower

the intensity o the peak. Tis eect is caused

by the loss o crystal water, which alters thephysical property o the cement.

Inuence o milling time

Various milling times were applied to samples

o the Portland cement. Te results are shown in

Figure 6. It can be concluded that the longer the

milling time, the lower intensity o the peak.

Measurements under vacuum

Once milling time and pressure parameters

have been optimised, the selection o the right

measurement sequence must also be considered.

With XRD technology becoming increasingly 

commonplace in the cement industry, a manu-

acturer may have the option to perorm both

XRD and XRF analysis. However, with XRF

analysis, cement samples are usually measured

under a vacuum, which has an eect on subse-

quent XRD measurement.

Figure 7 shows the infuence o a vacuum on

cement samples. Here, an XRF measurement

has been conducted under vacuum, ollowed by 

XRD analysis. Tis results in reduced peak inten-

sity. In this example, the reduced intensity o the

gypsum peak occurs because the gypsum phase

is partly transormed into a semi-hydrate phase.

Conclusion

For many years, XRF has been at the heart o pro-

duction control in every modern cement works.

Now, to keep pace with industry needs, advanced

XRD technologies are also becoming increas-

ingly commonplace. Both are used routinely—as

the principal means o controlling composition

o raw materials, raw eed, clinker and cement.

As reliance on X-ray analysis increases,

so too does the importance o good sample

preparation. Tis, and the correct measurementsequence, ensures the highest possible accuracy 

and precision.

SAMPLING & ANALYSIS

18  globalglobalcementcement MAGAZINEMAGAZINE May 2008

Figure 4: Difractogram o an

ordinary Portland cement - theblue bars indicate the position o 

gypsum related peaks.

Figure 5: The efect o 

applied pressure in sample

preparation.

Figure 6: The efect o milling

time during sample preparation.

Figure 7: The inuence o 

vacuum on XRD analysis.

As world leaders in X-ray analysis

systems, with over 50 years o 

experience, PANalytical is at the

oreront o X-ray analysis sample

preparation methods. The com-

pany’s comprehensive range o 

manual and automatic, advanced

sample preparation systems

helps users create perectly pre-

pared samples.

  The range includes the compact,

bench-top MiniMill 2, which grinds

samples at low cost with a high

degree o reproducibility; and the

MiniFuse 2 and Perl’X 3 instruments

which oer, respectively, semi- and

ully-automatic high quality used

bead preparation.

Expertise rom sample preparation to analysis result

From left: Bench-top MiniMill 2, Perl’X and MiniFuse 2.