The Cameca WDX Spectrometers

The spectrometer configuration

If you only need to check the Cameca's spectrometers, or if you are just starting with a list of elements in hand, you can now begin considering the configuration for the microprobe.
Because a single x-ray spectrometer may have several diffracting crystals from which to choose, many possible configurations for the Cameca exist.   The TV monitor will indicate the Cameca's current crystal configuration ... but our Cameca's spectrometers also offer the following possibilities ...

spectrometer 1 spectrometer 2 spectrometer 3 spectrometer 4
soft x-ray
(low pressure)
soft x-ray
(low pressure)
hard x-ray
(high pressure)
hard x-ray
(high pressure)
TAP
lead sterate
PC1
PC3 (soon)
TAP
PET
PET
LIF
PET
LIF


...   if you need to change a crystal use the Probe for Windows software or be sure to move to a spectrometer position greater than 81000 first. The SX commands are:
 
SX> AMOV spN "position"       (e.g., AMOV SP2 81000)
SX> XTAL spN "xtal name"       (e.g., XTAL SP2 PET)
 

The two most common configurations for the spectrometers are:

spectrometer SP1 SP2 SP3 SP4
most minerals TAP PET LIF LIF
feldspars TAP TAP PET LIF
The first allows spectrometers 3 & 4 to share several transition metal elements which are typically counted for longer periods of time.   The second is optimized for a mineral like feldspar which involves few metals but allows spectrometers 1 & 2 to share Na, Al, Si and allows Mg and Fe free of charge.   Notice how these two possible configurations ask you to organize your work over two different sessions, but your time on the instrument will be less overall.


crystal PC1 OdPb TAP PET LiF
wavelength
range (A)
51 to 12 85 to 20 22 to 5.2 7.35 to 1.75 3.42 to 0.81
K lines C(6) to F(9) B(5) to O(8) F(9) to P(15) Si(14) to Cr(24) Sc(21) to Br(35)
L lines n.a. Ca(20) to V(23) Mn(25) to Y(39) Sr(38) to Eu(63) Te(52) to Bi(83)
M lines n.a. n.a. La(57) to Ir(77) W(74) to Pu(94) n.a.



The table above and Cameca's sine-theta table imply you have some choices ... for some elements you can either measure an element's specific x-ray line with a choice of crystals or you may have a choice of measuring it K, L or M line.   A general rule of thumb is to pick the line with the best signal-to-noise ratio ... another generality is that best S:N ratios exist for K lines and locations on any crystal for large sine-theta values.   An exception for picking the best S:N crystal is for major elements when the best S:N isn't needed and because spectrometer reproducibility is better for small sine-theta values.   For example, this would imply you would choose TAP for Si in silicates but would choose PET for Si in spinels.
Regarding your choices between K,L or M lines, choose the best S:N ratio when needed for sensitivity, but also concern yourself with spreading the elements around.   For example, instead of measuring Cu K in sulfides with LiF measure the L line with the idle TAP spectrometer.

Pulse Height Analysis

PHA can be thought of as "Pulse Height Acceptance"   After x-rays enter the detector and are amplified the PHA does not allow the x-ray counter to count unwanted pulses.   The PHA (1) typically excludes noise by setting a baseline, and (2) the EPMA analyst may also choose not to count higher energy pulses by setting a window or an upper threshhold.   The Cameca's mode of PHA detection can be set to use only a baseline (integral) or to use a window (differential).

SX> SACQ spN MODE INTE       (count all pulses above the baseline voltage)
SX> SACQ spN MODE DIFF       (count only pulses within the window)

A good example of PHA and its benefits is depicted by the measurement of Al x-rays while chromium is present ... e.g., in Cr-spinel.   That is, the Cr K-beta line's wavelength is the 4th order equivelent of Al K-alpha ... the primary difference being its energy ... four times that of the Al line.   Therefore ... as shown below ... the upper threshhold can exclude Cr x-rays from Al's measurement.

Al in Cr-spinel
The optimum window settings can be set by the Cameca by issuing the command:

SX> ADJP spN      or
SX> ADJP SPEC     (all spectrometers)

... which tells the Cameca to have the window follow the spectrometer's position.   We have found this to be accurate and more advisable than determining what is optimum yourself because it is consistent with what everyone else uses.

The window width for the differential mode is generally a consideration for which crystal is used.   You can set the window width with the following command:

SX> ADJP spN CW X     (where X= 1,2,3,4,5)

The following table indicates general guidelines for the CW parameter:
crystal PC1 OdPb TAP PET LiF
Cw 2 2 3 5 5

The PHA is sensitive to the detector gas and the HV bias applied to the detector.   You don't have much control over the gas pressure, but you do over the bias ... it is the primary parameter for making sure the PHA is adjusted properly.   You first have to test the PHAs:

(1)   Move the stage to a standard which can be a source for all x-rays:

SX> MOVE STAG ANDR     (move stage to andradite garnet)

(2)   Move the spectrometers to the element x-ray positions

SX> MOVE SP1 SI KA     (move TAP crystal to Si x-ray position)
SX> MOVE SP2 CA KA     (move PET crystal to Ca x-ray position)
SX> MOVE SP3 FE KA     (move LiF crystal to Fe x-ray position)
SX> MOVE SP4 FE KA     (move LiF crystal to Fe x-ray position)

(2)   Test the PHAs

SX> TPHA SPn     (test each one at a time ... adjust bias if needed)

Adjusting the bias will center the x-ray counts within the window (... baseline and threshhold ...).   Note the current BIAS setting and you can shift the peak left or right by modifying the voltage by as little as 5 volts (... try 10 volts at first ...).
You can set the BIAS with the command:

SX> SACQ SPn BIAS XXXX     (where XXXX is the bias voltage value)


Verifying the Spectrometers

The Probe for Windows software maintains the spectrometer offsets internally using a polynomial fit to calibrated positions separately for the K, L and M x-ray line families. This is preferred to the Cameca "veri spec" command assumes a constant offset over the entire range of the spectormeter.
 
copyrights 1995-2002,
john donovan: July, 2002
Table of goto home page Contents