As seen on the attached image, we intended to produce Au nanobricks by EBL. The resist was PMMA 120 nm, metallization parameters are: Cr 0.5 nm, Au 40 nm. The SEM image of the produced nanobricks reveals some rim or shoulder at the edges of the bricks, which appear as brigth lines around them on the SEM image. AFM measurement approved the height difference at the edges. Does anyone have an idea how to avoid this effect and create flat surface bricks?
Hy,
Edges are often brighter on secondary electron detector images because the emitted ration of secondary electron is higher (there is more surface for emission inside the interaction volume). Normally, the use of the backscattered electron detector will avoid that problem. You can also try a lower accelerating voltage to get a better description of the surface of the nanobricks. Some remaining PMMA can also induce charging effect and BE image can also help in that case.
Hi, thanks for the answer. Yes, we know about this effect, but it is not that. Backscattered images and AFM proved that these are real protrusions. The question is, how to get rid of them. There should be some tricks in the metallization or in the liftoff.
I understand better know. I don't think metallization is faulty. Did you try low voltage image ? something like 2KV ?
No, but we are still convinced that this is not an imaging effect. Atomic force microscopy has proved that.
It could be metallisation happening on the walls of PMMA as some metal atoms are arriving at the surface of the chip at small angles (away from the normal to the surface). To reduce the "hat" effect, one would increase the ratio between the PMMA and metal layer thicknesses (120 : 40 = 1 : 3 may be not sufficient for a clean lift-off). There are two options: (i) to try to increase the PMMA thickness to let's say 1:5 (200 nm in your case) or higher, or (ii) to use a double layer PMMA/MMA to have a pronounced undercut to not have a polymer wall next to the Au structures.
What PMMA concentration, exposure voltage and electron dose were used during EBL? Those may be artefacts of underdevelopment.
Assuming exposure parameters were OK; increasing the thickness of the PMMA resist should help to create an undercutting effect in the exposed areas which will remove the walls of the PMMA with greater ease and prevent crowding of the Au on the edges of the patterns.
See section 2.3 in the attached:
Dear Sergey Gorelick and Jakub Jadwiszczak , thank you both for your suggestions. I will definitely try increasing the PMMA thickness, and will also try decreasing the beam energy in order to increase undercutting. In the meantime, Raith suggested the same.
Hi Zsuzsanna Marton ,
It is a common issue in metal patterning by using POSITIVE resist + lift-off for Au without a strong ultrasonic step. I can say that It is not imaging effect, it is a metal protrusion ( or " ear" forming ) at the edges of the patterns. You can see our very old work on this ear forming at the edges of metals. It happened both at micro and nano scale patterning.
https://research.utwente.nl/en/publications/platinum-patterning-by-a-modified-lift-off-technique-and-its-appl
( our work done when negative resist or LOR resist were not common yet).
Generally, ear or protrusion metal caused by a combination of lithography ( resist side wall, resist hardening by E-beam and hot arrived Au during deposition ) + Metalization ( which metal ? Au is really ductile, other metal like Pt met the same issue but less proven, and deposition procedures- slow/or fast deposition) + lift off step ( with/without using (strong) ultrasonic during lift-off).
So, using a little thicker resist with a negative resist profile ( if you can tailoring the resist profiles with E-beam ???) will solve the issue .
Second approach is a creation an under cut below the PMMA by using th sacrificial layers. For your pattern ( as it is not a high density, ), a simple way is to utilize a LOR resist from MicroChem (LOA 0.5A, spinning at maximum speed you can to have a ca, 30 nm LOR).
Or using 30 nm Ti as the sacrificial one. Ti can be etched/removed in a very diluting HF without effect to both PMMA (for under cut) and Au/Cr.
Good luck with your experiment and email us directly if you need further advice (with sacrificial tricks).
Hien Duy Tong
- Badges
- Science method