Aztec XRF Sample Analysis

The AZTEC employs the X-ray fluorescence technique whereby gold K X-rays are excited by high energy X-rays. The novel detector is able to efficiently measure the K X-rays to give the necessary sensitivity. This non-destructive technique enables repeat measurements to be made at any time for the sample can be stored in the sample container.

Download a detailed Product Brochure of the Aztec Analyzer here.

For ease of servicing the circuits are modular using the Euro card format. Computerised fault diagnosis cuts down-time to a minimum and self calibration to international and mine standards ensures that the instrument is correctly calibrated at all times.

Design

A plan schematic of the of the AZTEC irradiation and detection geometry is shown in Figure 1 below. An X-ray generator provides X-rays at a mean energy of 120 keV, to excite the sample. X-rays emitted or scattered by the sample are detected by an array of cryogenically cooled pure germanium detectors, coupled to high count rate signal processing electronics.

Figure 1

 

A sample changer collects the sample from a magazine, and places it into the irradiation chamber. It also presents instrument calibration and reference samples when required. Sample changer capacities of up to 792 samples are available. A separate diagnostic system warns of faults. The whole instrument is computer controlled and analysis is automatically carried out. Remote reporting and diagnostic capabilities are provided.

A feature of the instrument is that the sample is loaded into a thin plastic walled tube approximately 2 cm diameter and 30 cm long. The X-ray beam is about 7 cm wide when it intersects the tube, and excites about 30g of sample at any one time. The sample can be scanned vertically so that the X-ray beam intersects different zones in the sample.

Performance of the AZTEC 3 analyser

Accuracy

The K x-ray technique requires a 2 point calibration, (a) sensitivity (slope) and (b) zero (offset), which also compensates for variations in sample density and gross sample matrix effects .

The sensitivity calibration is derived from accurate synthetic standards, and systematic error in the sensitivity calibration will be less than 1%, for Au or Pt. The zero calibration is derived from background standards containing none of the elements to be measured.

Systematic errors can also be caused by x-ray line overlaps; these can be minimised by various techniques. The most difficult X-ray line overlap to compensate for is from Hg; essentially the gold measurement has to be carried out by the Au Ka 2 peak, which is only half the intensity of the normally used Au Ka 2 peak, so longer counting times are required.

Precision of measurement

Random errors due to radiation counting statistics are the main contributor to imprecision. However, they are Gaussian in nature, and are accurately pre-definable. An inverse-square law applies. Counting for 4 times as long, or at 4 times the count rate, will improve the precision by a factor of two, i.e. precision is inversely proportional to the square root of the counting time.

With current technology, using an array of 10 detectors, counting in parallel, a precision of 0.5 g/t (at 68% confidence level) is obtained in 45 seconds. Precision, sample count times, and sample throughput are summarised in the table below.

Precision (g/t)	Sample Count time (s)	Samples per day
0.5	45	1,130
0.35	100	650
0.2	300	250
0.1	1,200	65

 

Sample averaging

The random error is well-defined, being governed by known physical phenomena, and is relatively constant over the concentration range of most gold and platinum ore samples.

This results in the analytical accuracy of an average measurement being significantly improved as the number of samples from an ore body increases. For example, one sample might have a measurement precision of 0.3 g/t, but the average measurement of 25 samples will have a precision of 0.3/(25)0.5 = 0.06 g/t.

Generally, sampling and sample preparation errors contribute more to the overall error in a measurement than the analytical precision.