Subject: Re: Great Raith Quesion: Any Suggestions for SEM Imaging at High EHT?
From: "Lindsay Moore" <lsmoore@stanford.edu>
Date: Wed, 17 Jan 2007 11:39:53 -0800
Wed, 17 Jan 2007 11:39:53 -0800
James, Bing, et al,
When I do writes at 30kV and find that I am having some imaging problems, I
do 2 rounds of alignment and focusing using the same aperture: one at 10kV
and then again at 30kV.  I initially ramp the beam to 10kV and drive around
to locate the stage positions of the latex spheres or other objects that I
want to focus on.  I do the initial focus and alignment, writing down the
stage positions of the features.  I then ramp up to 30kV, and since I am in
almost perfect focus already, and I know the stage positions of the
features, it is much easier to find them and do the penultimate focus and
alignments.  My final step is to burn a very small contamination dot and
focus and align on that.  As James mentioned, I use the SE detector for
this.

Alternatively, it might be worthwhile to add a lithography step where you
write a pattern on your chip corner that you metallize and use for focus and
alignment.  It is much easier to image metal at 30kV.  (keep in mind if you
do this that you are focusing on the chip surface and not the resist
surface, so you will need to make a dose array on a chip where you focus on
a metallized chip surface if you want to get the finest features.)

Lindsay

On 1/17/07, James Conway <jwc@snf.stanford.edu> wrote:
>
>  Bing Dai wrote:
>
>  Hi James,
>
> I tried to do exposure under 25keV, Aperture 30 tonight but the noise was so
> high that I couldn't even find the latex spheres, which were very easy to
> locate under lower EHTs like 10keV. I also noticed that the aperture was very
> dark under emission mode/scan speed 9, although it was centered in Gun
> Alignment. I also tried 25keV, aperture 20 but the gun alignment was even
> worse, the center aperture nearly totally dark.
>
> So do you have any suggestions for adjusting brightness/contrast and doing
> focus/alignment under high EHT? Thanks a lot.
>
> Bing Dai
>
>
> Greetings Bing Dai,
> This is an excellent question and will share my response with the greater
> RAITH community too!
>
> The RAITH's In-Lens Ultra High Efficiency detector does not have very good
> performance above 20 keV acceleration voltage! More likely you will need to
> use the normal Secondary Electron detector called SE in the detector
> selection panel.  This is simply a scintillation phosphor disk and a
> photo-multiplier tube hooked to the back of it. It has much lower point to
> point resolution than the other In-lens SE detector but will perform better
> at higher acceleration voltages. (20 keV and Up...)
>
> Another issue to keep in mind is that at higher acceleration voltages you
> have a greater penetration depth of the Electron beam into, and in this case
> through the latex spheres and into the substrate below.  This gives
> information on the material (Z contrast) but at a loss of some of the
> surface or topographic contrast signal in the image.
>
> Finally note that the latex spheres are simply hydrocarbons, made mostly
> of polystyrene, and as such they have relatively little electron scattering
> cross section resulting in very low secondary electron yield to the
> detectors and reducing the signal to noise levels in the SE detectors.
>
>
> HINT for Users:
> My normal practice is to burn a long contamination shot dot then carefully
> focus and stigmate on this feature.  I can then burn smaller contamination
> dots easily; often by reducing the SEM magnification, which increases the
> field of view, assisting in out-gassing a suitable volume of hydrocarbons
> from the resist to aid in contamination dot formation.
>
> Hoping that this information is useful to you. Maybe other Users will wish
> to expand and add on this thread?
>
> Thank you,
>
> James Conway
>
>
>
>


James, Bing, et al,
When I do writes at 30kV and find that I am having some imaging problems, I do 2 rounds of alignment and focusing using the same aperture: one at 10kV and then again at 30kV.  I initially ramp the beam to 10kV and drive around to locate the stage positions of the latex spheres or other objects that I want to focus on.  I do the initial focus and alignment, writing down the stage positions of the features.  I then ramp up to 30kV, and since I am in almost perfect focus already, and I know the stage positions of the features, it is much easier to find them and do the penultimate focus and alignments.  My final step is to burn a very small contamination dot and focus and align on that.  As James mentioned, I use the SE detector for this. 

Alternatively, it might be worthwhile to add a lithography step where you write a pattern on your chip corner that you metallize and use for focus and alignment.  It is much easier to image metal at 30kV.  (keep in mind if you do this that you are focusing on the chip surface and not the resist surface, so you will need to make a dose array on a chip where you focus on a metallized chip surface if you want to get the finest features.)

Lindsay

On 1/17/07, James Conway <jwc@snf.stanford.edu> wrote:
Bing Dai wrote:

Hi James,

I tried to do exposure under 25keV, Aperture 30 tonight but the noise was so
high that I couldn't even find the latex spheres, which were very easy to
locate under lower EHTs like 10keV. I also noticed that the aperture was very
dark under emission mode/scan speed 9, although it was centered in Gun
Alignment. I also tried 25keV, aperture 20 but the gun alignment was even
worse, the center aperture nearly totally dark.

So do you have any suggestions for adjusting brightness/contrast and doing
focus/alignment under high EHT? Thanks a lot.

Bing Dai

Greetings Bing Dai,
This is an excellent question and will share my response with the greater RAITH community too!

The RAITH's In-Lens Ultra High Efficiency detector does not have very good performance above 20 keV acceleration voltage! More likely you will need to use the normal Secondary Electron detector called SE in the detector selection panel.  This is simply a scintillation phosphor disk and a photo-multiplier tube hooked to the back of it. It has much lower point to point resolution than the other In-lens SE detector but will perform better at higher acceleration voltages. (20 keV and Up...)

Another issue to keep in mind is that at higher acceleration voltages you have a greater penetration depth of the Electron beam into, and in this case through the latex spheres and into the substrate below.  This gives information on the material (Z contrast) but at a loss of some of the surface or topographic contrast signal in the image.

Finally note that the latex spheres are simply hydrocarbons, made mostly of polystyrene, and as such they have relatively little electron scattering cross section resulting in very low secondary electron yield to the detectors and reducing the signal to noise levels in the SE detectors.


HINT for Users:
My normal practice is to burn a long contamination
shot dot then carefully focus and stigmate on this feature.  I can then burn smaller contamination dots easily; often by reducing the SEM magnification, which increases the field of view, assisting in out-gassing a suitable volume of hydrocarbons from the resist to aid in contamination dot formation.

Hoping that this information is useful to you.
Maybe other Users will wish to expand and add on this thread?

Thank you,

James Conway