9-8-98

years I had an instrument that used a LaB6, but I've changed companies.

I have 2 questions.

1. Both the people here and the instrument's service man tell me

that the tungsten filament is more stable than the LaB6. I ask them to

explain further and they really don't get into it. What is the stable thing

about the tungsten filament? My LaB6 seemed fine to me, perfect in fact,

so what is the unstableness of it?

2. I'm also told here that, with the tungsten filament, as the beam sits in

one area on the sample, Carbon will develop in that area. Is this true? If

so how long does it take for the carbon to contaminate the area, and also

does this take any confidence in a carbon analysis and throw it out the

window?

Mark Darus

DARUSM@cle.lg.bfg.com

can't say anything about instabilities. I have heard about field emission

scopes sometimes being unstable, but even that has remedies.

Last I heard, there was no difference between electrons once they left the

gun, be it W, LaB6, or FE. The only difference would be in the number of

them per time in a given space. The lesser vacuum requirements for a W

filament might lead to more rapid C buildup on the sample. But our 840 with

W normally runs 10-6 torr and has only moderate problems with buildup.

I am leary of doing much with C analyses. The absorption factors are

significant sources of uncertainty. I would stick to qualitative comparisons

between points as much as possible. And make sure spectra are collected

under similar conditions of time and current.

Warren S.

Warren Straszheim

wesaia@iastate.edu

stability for periods longer than the response time of the cage+electronics.

High-frequency instabilities are harder to measure, but would show up as

light or dark bands in the image for appropriate scan rates. I have no

experience with SEM, but I don't think there is much difference in stabil-

ity between LaB6 and W. Both are thermionic sources which should be heated

by DC current. In this case the temperature--therefore the emission--should

be constant.

This will happen with any source of electrons, but the contamination

rate will vary with such parameters as vacuum in the specimen area and the

nature of the specimen. The appearance of these contaminant peaks (they look

like little mountains) has been used to measure local specimen thickness and

to identify the position where EDS was done (for single-point analysis).

The HVEM does not leave such peaks in spite of the so-so column vacuum and

its use for plastic sections of biological specimens. Certainly, if you see

carbon peaks on the specimen, you will also see carbon peaks in the EDS spec-

tra, so I wouldn't trust a carbon analysis on such an instrument either for

a single-point spectrum or for a carbon element map.

Yours,

Bill Tivol

tivol@wadsworth.org

I can speak from experience here since I have been using LaB6 & Tungsten for

a number of years.

Tungsten in general will give a short lifetime ( typically 100 hrs+ ) whilst

a LaB6 ( in theory ) will last for months. My experience with LaB6, which

probably depends on the LaB6 supplier ( in this case Denka ), has produced

approx. 6 weeks of more electrons before the filament starts to produce less

and the alignment wander.

This may be due to the type of source of the LaB6, but I have talked

recently to someone running a different supplier's LaB6 and he was suffering

the same problems.

In general operation the LaB6 will not give noticable differences in

stability. Any measurements using a Faraday cage & monitor will give

different results depending on the age of the filament. However if you

have demanding customers, such as Geologists ( apologies to any reading this

! ), when you are carrying out quant analyses and producing totals

approaching 100% the analyses are very dependant on beam stability. I have

not been able to carry out these analyses without using a tungsten source.

Even with a tungsten filament it will be unstable in the early & late part

of its life and serious analyses can only be carried out during the middle

period. I produce longer life and greater stability by cleaning the gun

properly after a blown filament and using the Ion pump to produce a higher

vacuum during tungsten operation. Tungsten filaments also have faster ramp

times and are more tolerant of poor vacuums.

As regards carbon build up I think that it is both vacuum & sample

dependant. With more hydrocarbons in the vacuum you get a greater

build-up. I have two turbo-molecular pumped systems and do not suffer from

much carbon build-up unless certain samples are scanned.

Best wishes,

Colin

Colin Reid,

Electron Microscope Unit,

Trinity College Dublin,

Dublin 2,

Ireland.

Tel: 353-1-6081820

Fax: 353-1-6770438

email: creid@tcd.ie

Although I am a "vendor", allow me to comment on this thread, from the

perspective of someone who has manufactured tungsten filaments and

distributed Denka LaB6 cathodes for more than 10 years.

Colin Reid wrote:

>Tungsten in general will give a short lifetime ( typically 100 hrs+ ) whilst

>a LaB6 ( in theory ) will last for months. My experience with LaB6, which

>probably depends on the LaB6 supplier ( in this case Denka ), has produced

>approx. 6 weeks of more electrons before the filament starts to produce less

>and the alignment wander. This may be due to the type of source of the LaB6,

but I have

>talked recently to someone running a different supplier's LaB6 and he was

suffering

>the same problems.

For any filament, by far the most important parameter affecting material

loss, and, therefore, lifetime is vacuum. This holds true for both

tungsten filaments and LaB6. Filament life, therefore, is highly

dependent on the vacuum conditions of a given microscope, and filament

life will vary substantially from microscope to microscope.

Tungsten filaments are cold-formed from wire, and the stress from the

forming will cause instability if the filaments are not properly annealed

under vacuum. If the tungsten filament has not been properly annealed by

the manufacturer, the user will, in effect, anneal it in the microscope.

The filament will be unstable during this process, and will often have to

be realigned after the annealing.

In the case of LaB6, the crystal is in <100> orientation, and formed into

a point, a round tip or a microflat tip. As the cathode experiences

material loss, the tip flattens. After some 150 hours, the brightness

will fall off, as the emission surface of the tip forms into a larger

flat. In effect, the sharp or round tips become flat tips. Although the

cathode has hundreds of hours of "life" left, it will never again be as

bright as it was during the first ~150 hours. This is true of any

manufacturer's LaB6.

For quantitative analysis, mechanical stability becomes an issue. For

this application, the LaB6 mount should be as robust as possible.

I would be happy to provide additional details to anyone interested. In

particular, we have several Denka technical reports which explain the

relationship between lifetime, brightness, vacuum and material loss for

LaB6 cathodes.

Best regards,

Steven E. Slap, Vice-President

********************************

Energy Beam Sciences, Inc.

The Laboratory Microwave Company

http://www.ebsciences.com

********************************