Introduction

Analysis of coins, jewellery pieces and other items manufactured from precious metals creates special needs for analysis because the monetary value will be determined by the content of precious metals. Different analytical methods will be used for this problem.

Many methods are destructive which means they consume a small amount of the sample. X-Ray spectroscopy offers the possibility of nondestructive analysis. In addition concentrations of other metals in the alloy can also be determined. This is important for other mechanical properties such as color, plasticity and surface quality.

Due to the high monetary value of precious metals a very high precision for the analysis is needed.

Here we will demonstrate using multiple measurements of a precise reference sample, that a small X-Ray spectroemter can fulfill this demand.

Measurements was made:

Measuring instrument

The here reported measurements was realised with a small energy dispersive spectrometer.

Tube parameters

The exciting X-Ray beam will be collimated on the sample with a diameter of 0.5 mm.

The detector is a proportional counter.

Samples

Over 150 very different samples from the company SAFINA Prague were examined. These samples had previously examined by:

Gravimetric results, fire assay, wet chemical analysis and others

Repeatability

For determination of repeatability repeated measurements on the same sample were made. The following measuring conditions were used:

Measuring conditions

For analysis of more than 10 different samples with very different concentrations the following results were obtained:

Absolute standard deviation (1 x s) depends upon the intensity of measured radiation, which in turn depends upon the concentration of the analysed element. Standard deviation will be better for higher concentrations.

For 1xs the range should be < 0.12 %, that means the statistical error should be in range of + 0.25 %.

Accuracy

Measurements obtained

The comparison of X-Ray results with known concentrations is shown in figure 1. These values are in very good agreement with reference values. The standard deviations for repeated measurement are in correlation with expected statistical errors, which are for Gold, approx. 0.2 %.

Figure 1: Comparison of analysed and certified values

Measurement errors

Measuring distance

Different distances from the sample to the tube and detector influence the measured intensity and therefore the analytical results. This can be neglected by a normalisation of intensities. Figure 2 shows the dependence of concentration on the sample position.

Sample errors

In X-Ray spectroscopy of precious metal alloys only a very thin layer gives a contribution to the signal.
Therefore

These types of layers are not homogenous. Therefore they can influence the reproducibility of repeated measurements with different sample positions. This is a way of identifying such layers.

Figure 2: Dependence of distance