SEM Gloat

[Jerry]

If you have samples just send them to me.

Done a bunch of work. I found out that the EDX controller I have also has digital capture capabilities. I found the external Digital Beam Control box on ebay for $110 shipped and ordered a couple fiber optic cables to connect it to it's mainframe. I got it all hooked up today and it works well. This is the first digital capture out of the scope, 30x image of a 1/8" dia screw head on a sample holder and the edge of a dime:

SCREW by macona, on Flickr

[Doug Coulter]

Awesome! And, I bet it even looked pretty clean to the naked eye. Amazing how the right tool shows you what it really is like. I know I spill one cat hair into a vacuum and my mass spec can see it for a long time, even though I can't. One fingerprint stays for a week or so...even with some baking. Suddenly there's all these high-mass lines from complex hydrocarbons.

[Jerry]

Got the EDX running, EDX allows you to do elemental analysis by sampling the x-rays that are created when the electron beam hits the target in the microscope.

I picked up a 50L dewar of liquid nitrogen Monday and filled the dewar on the EDX head. You need to let it sit from 4-10 hours to make sure the sensor is cooled. If the sensor is not cool (cryogenic temps) the sensor will be ruined.

I threw a dime in the chamber and scanned it and did an acquisition with the software and it showed me some peaks, a dime's outer plating is about 75% copper and 25% nickel. The histogram showed some peaks pretty close to where they should be, it needs to be calibrated and the gains adjusted but it is pretty good for just hooking it up. I need to go through the manual some more and do all the adjustments.

I also picked up another sensor head off ebay the other day. It has different windows you can use as well as a windowless operation mode that will allow you to sample light elements like carbon and oxygen. The standard sensor used a beryllium window which makes it hard to see light elements.

dimeedx by macona, on Flickr

[Doug Coulter]

That's VERY nice. I suppose you could check precious metals for adulteration too - but if that gets known you'll be swamped with requests

[Jerry]

I calibrated the EDX tonight, the software uses a sample with aluminum and copper to create a spectrum and then figures out the offset and gain to get the peaks in the right place. With that done I zoomed into the little screw in the picture with the dime edge and did an acquisition. This is what I got and the analysis side says the screw is composed of the following:

Code: Select all

Element	Line	Weight%	K-Ratio	Cnts/s	Atomic%
---------------------------------------------------------
Al	Ka	 6.26	0.0115	90.84	13.78
Fe	Ka	 1.68	0.0206	77.55	 1.79
Ni	Ka	 1.06	0.0130	39.00	 1.08
Cu	Ka	55.95	0.5566	1483.97	52.35
Zn	Ka	31.62	0.3132	728.50	28.76
Sr	La	 1.91	0.0066	19.13	 1.30
Mo	La	 1.51	0.0079	23.45	 0.94

Total	99.99

This works out to be a brass with aluminum added. Adding aluminum increases the strength of brass and makes it more corrosion resistant.

screwspectrum by macona, on Flickr

[Jerry]

I threw a piece if U Ore in the scope to see what it would show, here it is:

Code: Select all

			[A N A L Y S I S  R E P O R T]


GENERAL CONDITIONS
------------------
  Result File       : UORE
  File Version      : 1
  Background Method : Foil
  Decon Method      : Gaussian
  Decon ChiSquared  : 6.13
  Analysis Date     : 28-DEC-2014
  Microscope        : SEM
  Comments          :                                                   

ANALYSIS CONDITIONS
-------------------
  Quant. Method       : XPP/ASAP
  Acquire Time        : 400 secs
  Normalization Factor: 100.00

SAMPLE CONDITIONS
-----------------
  kV                     : 30.0
  Beam Current           : 100.0 picoAmps
  Working Distance       : 25.0 mm
  Tilt Angle             : 30.0 Degrees
  TakeOff Angle          : 35.0 Degrees
  Solid Angle*BeamCurrent: 0.9


Element	Line	Weight%	K-Ratio	Cnts/s	Atomic%	
---------------------------------------------------------
Mg	Ka	  3.16	0.0117	  9.87	  5.37	
Al	Ka	  8.21	0.0365	 33.43	 12.59	
Si	Ka	 40.97	0.2064	194.23	 60.36	
Ca	Ka	  7.39	0.0471	 33.73	  7.63	
Ti	Ka	  0.00	0.0000	  0.00	  0.00	
Fe	Ka	  8.69	0.0743	 32.49	  6.44	
Cu	Ka	  3.47	0.0309	  9.58	  2.26	
Pd	La	  0.00	0.0000	  0.00	  0.00	
Au	La	 12.69	0.1014	  8.26	  2.67	
U	La	 15.42	0.1138	  2.89	  2.68	

Total		100.00

I was surprised by the gold. I guess I should have not been, U was first found in Gold mines so says the internets. But I am scanning a very tiny piece of the ore so who knows if it is homogenous throughout the sample.

UOre by macona, on Flickr

[Jerry]

Tried the Map mode of the EDX for the first time last night. In the center window you can see an image of the sample (U Ore) and the surrounding images are heat maps of different elements that are found in that image. the more towards the red end of the spectrum, the more of that element was found. This was a "quick" 256x256 scan at 15ms per pixel dwell time. It took 15 minutes. Looks like it is a very time consuming process but still pretty neat.

uoremap by macona, on Flickr

I analyzed a piece of goldish metal that came out of a machine that was used for washing silicon wafers. It is 81% gold, 11% rubidium, 3% Iron, and the rest copper. I was not expecting to find this in the flake, this is at 490x.

GOLDFLAE by macona, on Flickr

[Doug Coulter]

This is all just total, utter awesomeness! I want one, but also wish I had room for one. Darnit.