To add comments or start new threads please go to the full version of: EBL for dielectric substrate
PhysForum Science, Physics and Technology Discussion Forums > Technology > Technology

stefano.park
1st June 2006 - 05:01 PM
I am almost new to EBL system especially with dielectirc substrate.
I am going to pattern on dielectric substrate on oxide substrate and glass substrate.
Do I need to deposit metal layer on both of oxide layer and glass in addition to PR layer. If I have to do so, how much should the metal layer be?
I heard that I don't need to deposit metal on oxide layer on the PR layer if the oxide thickness is small. Do you know how much is the maximum thickness for without metal layer?

Thank you.
Steveo
1st June 2006 - 05:30 PM
I don't know the answers to all of your questions, although my project involves doing EBL on a glass substrate for liftoff. I just deposited about a 5-10nm layer of gold ontop of my photoresist and I have had no problems at all. goodluck.
guiding_light
2nd June 2006 - 12:52 AM
When the photoresist is thin enough it emits more electrons than are incident upon it (secondary yield >1). This allows the surface potential to stabilize better, since if the sample starts to charge more negatively it can pull more secondaries back. A conductive layer speeds up the process.
guiding_light
2nd June 2006 - 12:57 AM
QUOTE
if the sample starts to charge more negatively it can pull more secondaries back


Should be "positively".
guiding_light
2nd June 2006 - 02:33 AM
There is also EBIC (Electron-beam induced conductivity) which could help you out, as long as your oxide is on Si or metal.

What is your beam energy? It should take some thought. A larger beam energy is less deflected by charging but at the same time the substrate charges more negatively since most electrons are trapped under the resist.

guiding_light
2nd June 2006 - 03:49 AM
QUOTE
I am going to pattern on dielectric substrate on oxide substrate and glass substrate.


To complete the circuit, you need to have your conductive film bleed the excess charge away. Since this film has a pretty large resistance itself, the excess charge will take some time to come off.

QUOTE (->
QUOTE
I am going to pattern on dielectric substrate on oxide substrate and glass substrate.


To complete the circuit, you need to have your conductive film bleed the excess charge away. Since this film has a pretty large resistance itself, the excess charge will take some time to come off.

Do I need to deposit metal layer on both of oxide layer and glass in addition to PR layer.
Sounds unnecessary

QUOTE
I heard that I don't need to deposit metal on oxide layer on the PR layer if the oxide thickness is small.


If your substrate is completely insulating (no silicon or metal whatsoever) you will need to deposit metal. Sometimes the thin gold overlayer is enough. Other times, you need 100 nm of Al or Cr under the resist.
guiding_light
2nd June 2006 - 05:27 AM
QUOTE
Sometimes the thin gold overlayer is enough. Other times, you need 100 nm of Al or Cr under the resist.


I can only vouch for beam energy <20 keV. Above this energy, I think the charging is much harder to deal with, since most of the electrons are stopped or generated deep beneath the resist and the resist supporting layer, far away from the possible locations of the conductive layer.

See the other thread post by SteveO, I think he has a process that could work for you:

http://forum.physorg.com/index.php?showtopic=5159
Guest
2nd June 2006 - 08:18 AM
QUOTE
When the photoresist is thin enough it emits more electrons than are incident upon it (secondary yield >1). This allows the surface potential to stabilize better, since if the sample starts to charge more negatively it can pull more secondaries back. A conductive layer speeds up the process.


There should be the same issue as for UPS and XPS, where secondary electrons leave the surface of the sample. How do they come back? Or do they come back?
Steveo
2nd June 2006 - 03:54 PM
stefano.park, whats the plan after your exposure? Doing lift off? Or etching your dielectric away? This might help with giving you advice, if we new what you were doing after the patterning.
Guest_stefano.park
2nd June 2006 - 11:23 PM
First of all, coat the PR on 1) SiO2/Si substrate and 2) Glass substrate.
next , pattern with EBL system.
next , lift off process

What I am wondering is if I need to deposit metal layer(Al or Au) on the PR.

I am using 30keV so far as accelerating voltage.



According to my understanding, I don't need the metal deposition on PR in case of SiO2/Si substrate, but I need the metal deposition on PR in case of glass substrate.


[/QUOTE]If your substrate is completely insulating (no silicon or metal whatsoever) you will need to deposit metal.

Does it mean glass substrate is completely insulating material and SiO2/Si is not completely insulating material?
Steveo
3rd June 2006 - 01:55 AM
For the glass you will need to deposit metal on the resist. I personally am using Au (around 5-10nm) ontop of my resist.
For glass you might need to change your accelerating voltage. I am getting great results using 5KeV. 10KeV gave me to much proximity effects, and 2KeV didn't expose all the way through the resist.

I am doing lift off also, and it works pretty good. You might have to do proximity effect corrections depending on how small your features are.
stefano.park
3rd June 2006 - 04:51 AM
Thanks everyone

Steveo, what kind of PR are you using? and how much is it thick?

I am going to use PMMA 2%, 100nm thick.

thank you.

guiding_light
3rd June 2006 - 01:01 PM
QUOTE
There should be the same issue as for UPS and XPS, where secondary electrons leave the surface of the sample. How do they come back? Or do they come back?


The lower energy electrons have a higher chance of coming back to the sample, while the higher energy electrons travel further.

This can be thought of as the motion of a free electron in the field of a sheet of (positive) charge representing the surface after ionization has just occurred. The field is given by surface charge density divided by the permittivity of free space.

For an ionization density of 0.001 e/nm^2, a 10 eV electron can travel vertically 553 nm.
guiding_light
3rd June 2006 - 01:20 PM
QUOTE
First of all, coat the PR on 1) SiO2/Si substrate and 2) Glass substrate.
next , pattern with EBL system.
next , lift off process

What I am wondering is if I need to deposit metal layer(Al or Au) on the PR.

I am using 30keV so far as accelerating voltage.


For liftoff, I (and Steveo too) have used a bi-layer resist. The idea is to have the lower layer open wider space than the top layer. This is the ideal resist profile for liftoff.

As Steveo has said, you can put 100 A Au (evaporated not sputtered) on the PR. I don't do this if I plan to etch the 100 nm Al or Cr directly after lithography. The reason is the wet etchant to remove the Au sometimes attacked the metal under the resist. When the metal is on glass, I have been able to avoid charging effects by minimizing dose and exposure area coverage, at 20 keV. It is an even bigger risk when you use larger beam energy.

I am not satisfied with any current proximity effect correction scheme in the machines. So I have generally tried to minimize the density variation. In other words, keep all the features isolated, or all the features at the same density.

I think dense features will never pattern as well as isolated ones in electron beam lithography, because of electrons from neighboring areas.

QUOTE (->
QUOTE
First of all, coat the PR on 1) SiO2/Si substrate and 2) Glass substrate.
next , pattern with EBL system.
next , lift off process

What I am wondering is if I need to deposit metal layer(Al or Au) on the PR.

I am using 30keV so far as accelerating voltage.


For liftoff, I (and Steveo too) have used a bi-layer resist. The idea is to have the lower layer open wider space than the top layer. This is the ideal resist profile for liftoff.

As Steveo has said, you can put 100 A Au (evaporated not sputtered) on the PR. I don't do this if I plan to etch the 100 nm Al or Cr directly after lithography. The reason is the wet etchant to remove the Au sometimes attacked the metal under the resist. When the metal is on glass, I have been able to avoid charging effects by minimizing dose and exposure area coverage, at 20 keV. It is an even bigger risk when you use larger beam energy.

I am not satisfied with any current proximity effect correction scheme in the machines. So I have generally tried to minimize the density variation. In other words, keep all the features isolated, or all the features at the same density.

I think dense features will never pattern as well as isolated ones in electron beam lithography, because of electrons from neighboring areas.

According to my understanding, I don't need the metal deposition on PR in case of SiO2/Si substrate, but I need the metal deposition on PR in case of glass substrate....Does it mean glass substrate is completely insulating material and SiO2/Si is not completely insulating material?


This is where the EBIC comes in. Electron-beam induced conductivity might help if the electron beam penetrates the insulating layers and goes all the way through to the conducting layer or Si substrate. 30 keV probably might do it. Anyway you can always try it on the SiO2/Si substrate.

30 keV into glass. Since the electron range is more than a micron, I think you will trap a lot of negative charge under the PR. Steveo had good results with 5 keV.
plasma_guy
3rd June 2006 - 04:13 PM
Liftoff is not recommended by me personally, for two reasons.

First the space between two deposited features cannot be too small, otherwise the resist in between will collapse.

Second the deposited film should be thinner than the resist itself, and this tends to result in more roughness and impact from pinholes.

So I recommend etching the features if possible.
Steveo
3rd June 2006 - 08:17 PM
QUOTE
  Thanks everyone

Steveo, what kind of PR are you using? and how much is it thick?

I am going to use PMMA 2%, 100nm thick.

thank you.


As Guiding Light said, go with a bilayer, especially if you are doing lift off. I have my bottom layer at about 65nm, and my top layer at about 45nm. I am using PMMA 495, and PMMA 950....I think 495 is the bottom layer, but I would have to look through my notes (which are at work, not at home) to be sure. Anyways, its pretty common knowledge which one should be the bottom, and I am sure any internet search (which I am to lazy to do) would tell you which one is bottom and top.


QUOTE (->
QUOTE
  Thanks everyone

Steveo, what kind of PR are you using? and how much is it thick?

I am going to use PMMA 2%, 100nm thick.

thank you.


As Guiding Light said, go with a bilayer, especially if you are doing lift off. I have my bottom layer at about 65nm, and my top layer at about 45nm. I am using PMMA 495, and PMMA 950....I think 495 is the bottom layer, but I would have to look through my notes (which are at work, not at home) to be sure. Anyways, its pretty common knowledge which one should be the bottom, and I am sure any internet search (which I am to lazy to do) would tell you which one is bottom and top.


I am not satisfied with any current proximity effect correction scheme in the machines. So I have generally tried to minimize the density variation. In other words, keep all the features isolated, or all the features at the same density.


Proximity corrections help, but it always takes a lot of fine tuning, and trial and error. Patterning isn't an exact science, because your looking for the best structures, not the closest match between design feature size, and actual size. As Plasma guy said, lift off has its draw backs. With my project I have been fighting those draw backs. I am trying to make long, wide electrodes with a small space between them. The hope originally was under 30nm gap (so far I have gotten 60nm once, and consistently been able to get 70nm). The funny thing is, in my design the electrodes are spaced 500nm apart, and I vary the dose to vary the electrode gap.

I posted on another EBL thread a link to the Casino program. Its an Electron Monte Carlo simulation program. It is pretty helpful for deciding on an electron energy to use. You can set up your substrate with resist, and metal layers on top (if you want). It gives you a rough idea of penetration depth, and the spreading of the beam due to scattering. Search for it, or look in the other thread where I posted a link. I think it was in the thread that guiding light linked to here.
Question
13th June 2006 - 12:32 AM
Shouldn't photomask substrates be charging negatively with current electron-beam machines? The electrons can't move very much after being implanted into quartz!
Steveo
13th June 2006 - 04:38 AM
QUOTE
Shouldn't photomask substrates be charging negatively with current electron-beam machines? The electrons can't move very much after being implanted into quartz!


You don't use a photomask for Electron Beam Lithography. The electron beam is deflected with electric and/or magnetic fields to write a pattern. It is direct writing, so its slow. Good for prototyping and research, not for mass production at this time.
guiding_light
13th June 2006 - 05:45 AM
QUOTE
Shouldn't photomask substrates be charging negatively with current electron-beam machines? The electrons can't move very much after being implanted into quartz!


For photomasks, laser writer is far more popular because of its speed. Also, charging draws particles.
guiding_light
13th June 2006 - 07:36 AM
I have made an error in my earlier post:

QUOTE
The field is given by surface charge density divided by the permittivity of free space.
It should be half of this (gauss's law for a thin sheet of positive charge).


QUOTE (->
QUOTE
The field is given by surface charge density divided by the permittivity of free space.
It should be half of this (gauss's law for a thin sheet of positive charge).


For an ionization density of 0.001 e/nm^2, a 10 eV electron can travel vertically 553 nm.
This result should consequently be doubled.

Sorry...

Guest
22nd June 2006 - 02:51 PM
QUOTE
You don't use a photomask for Electron Beam Lithography.


Absolutely right. Just ran the CASINO program, had electrons lodged more than 20 microns deep!
Guest_kira
22nd May 2009 - 08:00 AM
QUOTE (Steveo+Jun 1 2006, 05:30 PM)
my project involves doing EBL on a glass substrate for liftoff. I just deposited about a 5-10nm layer of gold ontop of my photoresist and I have had no problems at all.

Hi,

I'm also trying Liftoff on a glass substrate. currently I've a about 50nm Copolymer layer on top of which is PMMA. When you have the gold on top of the resist, do you remove it after EBL and before developing?
Guest_kira
22nd May 2009 - 08:07 AM
QUOTE (Steveo+Jun 1 2006, 05:30 PM)
my project involves doing EBL on a glass substrate for liftoff.  I just deposited about a 5-10nm layer of gold ontop of my photoresist and I have had no problems at all.

Hi,

I'm also trying Liftoff on a glass substrate. currently I've a about 50nm Copolymer layer on top of which is PMMA. When you have the gold on top of the resist, do you remove it after EBL and before developing?
Guest
28th May 2009 - 02:14 AM
yes

have you calculated the resistance of the gold film on the way to the grounding clip?
ebl
29th May 2009 - 01:55 AM
The problem with overcoat is loss of resolution. The coat has some graininess, plus there is electron emission from the coat.
Malter
26th August 2009 - 09:31 PM
QUOTE (Question+Jun 13 2006, 12:32 AM)
Shouldn't photomask substrates be charging negatively with current electron-beam machines? The electrons can't move very much after being implanted into quartz!

The electrons in the insulator substrate should be attracted toward the surface. It could cause defects and variability in the exposure.
PhysOrg scientific forums are totally dedicated to science, physics, and technology. Besides topical forums such as nanotechnology, quantum physics, silicon and III-V technology, applied physics, materials, space and others, you can also join our news and publications discussions. We also provide an off-topic forum category. If you need specific help on a scientific problem or have a question related to physics or technology, visit the PhysOrg Forums. Here you’ll find experts from various fields online every day.
To quit out of "lo-fi" mode and return to the regular forums, please click here.