News:

:) To protect your privacy, you must first log in before you can view forum statistics and user profiles

Main Menu

beam damage and voltage

Started by Ben Buse, March 11, 2026, 05:18:01 AM

Previous topic - Next topic

Ben Buse

Hi

A long time ago Goodhew 1975 and Goodhew & Gulley 1974 showed that glass was more stable at higher voltages - due to charge gradient

Have there been papers looking at impact of voltage where damage not through ion migration of alkalis - e.g. in carbonates, quartz etc. Is a higher or lower voltage better.

Ignoring TEM voltages, within EPMA/SEM I'd assume higher voltages are always more stable (larger interaction volume, in particular greater depth penetration outweighing greater power input), is this drawn out by published studies for different geological materials?

Ben Buse

Looks like I've in part answered my question back in 2015 with an EMAS poster - such a long time ago... I attach the poster.

sem-geologist

#2
nice poster... albeit I have a suspicion that thickness units used in graphics are wrong. Are those not supposed to be nm instead of µm? 1µm would be 1000nm and we normally coat 20 nm of carbon by default...

Also, considering F-Apatite - Increasing electron penetration actually increases the rate of F migration (which is well documented by few independent experiments), so increasing beam voltage would increase the migration rate, AND decrease the Pk/bkgd ratio for F - which then would require increasing counting time for better statistics - and in the end the e-dose is then bigger with less accurate result. Instead it is better to reduce the voltage for maximum peak/bkgd ratio (as far I had experimented it seems it is somewhere between 11 and 12kV) which simultaniously also reduce F migration (compared to 15kV) and also allows to use lower dose (due to better peak/bkgd).

Practically we subdivide the apatite measurement into three conditions: low beam energy with "TDI" (equivalent on peaksight) for F, Ca and P (P can be measured on the TAP without issues, if Ca higher order interference is corrected there), then minor elements at higher beam current (i.e. Sr, Cl, Na...) and then traces at highest beam current.

BTW the graph on T vs HV suggest that at 0kV we have 0 degrees of T. In real settings, and in particular with Cameca EPMA the 0kV starts at 40-50 C° due to stage motors being kept in tension continuously. Reducing that temperature down to room (i.e. 22C°) would improve things a lot, maybe at -90 C° the migration is completely halted, but practically any reduction of temperature would slow it down. In case of F-migration the major pain in applying TDI is that migration is not linear. Slowing it down would make us deal only with linear part of migration making TDI much more reliable.

Ben Buse

#3
I think 1um C referred to 1um beam size, and carbon coat.

Yes F-Apatite can be very complicated as Carl Henderson showed. And there may be reasons don't choose the higher voltage: stability vs other competing factors. I guess I just wondering how well we understand the effect of voltage on different material stability

I was looking at how much additional heat from the beam [voltage].

Probeman

Quote from: sem-geologist on March 11, 2026, 07:13:18 AMIn real settings, and in particular with Cameca EPMA the 0kV starts at 40-50 C° due to stage motors being kept in tension continuously. Reducing that temperature down to room (i.e. 22C°) would improve things a lot, maybe at -90 C° the migration is completely halted, but practically any reduction of temperature would slow it down. In case of F-migration the major pain in applying TDI is that migration is not linear. Slowing it down would make us deal only with linear part of migration making TDI much more reliable.

After we added these small fan motors on our SX100, we noticed that the internal temperature of the sample chamber was reduced considerably:

https://smf.probesoftware.com/index.php?topic=639.msg8574#msg8574

It also keeps the motor controller electronics from frying itself so quickly.
The only stupid question is the one not asked!

Probeman

#5
Quote from: Ben Buse on March 11, 2026, 08:08:32 AMYes F-Apatite can be very complicated as Carl Henderson showed. And there may be reasons don't choose the higher voltage: stability vs other competing factors. I guess I just wondering how well we understand the effect of voltage on different material stability

I was looking at how much additional heat from the beam [voltage].

This is just a "rule of thumb" calculation:

https://smf.probesoftware.com/index.php?topic=598.msg5409#msg5409

But it shows that temperature rise correlates strongly with beam voltage:

Beam conditions of 10 keV, 30 nA,  1 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  102.8571 C

Beam conditions of 15 keV, 30 nA,  1 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  154.2857 C

Beam conditions of 20 keV, 30 nA,  1 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  205.7143 C

Beam conditions of 25 keV, 30 nA,  1 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  257.1429 C

The thermal conductivity is for obsidian glass...

The best strategy according to this calculation is to defocus the beam (and use TDI of course!):

Beam conditions of 10 keV, 30 nA,  10 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  10.28572 C

Beam conditions of 15 keV, 30 nA,  10 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  15.42857 C

Beam conditions of 20 keV, 30 nA,  10 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  20.57143 C

Beam conditions of 25 keV, 30 nA,  10 um
Assuming thermal conductivity of 0.014 W/cmK
Temperature rise =  25.71428 C
The only stupid question is the one not asked!