• If you are looking for a monolayer (1L), then MoS2 is much easier than WSe2. But still, both are okay.
• For 1L flakes, PDMS would be very difficult than SiO2. But thick layers are possible in PDMS as well as SiO2 substrates.
• For 1L flakes, SiO2 is the best substrate so far in our experience. However SiO2 surface state is supercritical, otherwise, it will not work.

According to your description, I think you are making your SiO2 surface bad by doing the chemical treatments. We always use new (company packaged and sealed) wafers and always avoid any chemical treatment. The chemical treatment makes bad surface states for exfoliation.

O2 plasma treatment is very good for exfoliation, it makes two things. 1) Activated surface (after 15-20 min, activated surface deactivated due to atmosphere), 2) Dry clean SiO2 surface for contaminations. If you don't have O2 plasma facility, still exfoliation is possible but a bit more difficult, you will get a smaller flake size (~10 um).

You may follow these steps:

I am quite sure you will get it easily.

We are getting this every day, so don't be stressed up.

Everything will be good.

Good luck.

Chandan

Although I followed your comment, only thick WSe2 (colored yellow or green in OM) is observed rather than thinner one.

I tried more than 5 times on empty scotch tape to make them thinner but it still doesn't work. I am now suspicious about the state of bulk WSe2...[(maybe damaged?) - I bought the bulk crystal from 2D semiconductors]

O2 plasma treatment was done for 50W, 50sccm for 10 minutes.

I'm now really struggling on exfoliation and having a hard time going further steps. Could you please give me any additional comments?

Hi Youngkyu Ko,

We also use 2D semiconductor crystals. It should work.

It seems you are thining down your tape (before exfoliation) too much.

5 times thining down is too much. All crystals will be broken into small fragments and strongly attach to the tape glue. Getting a monolayer would be very difficult in this case. 2-3 times thining is the best. Try this.

Let me try to give you more details.

Monolayer or few-layer flakes come from the shiny regions only.

For ideal case, shiny to non-shiny region ratio should be around 80% : 20% (or 90% : 10%). This is critical for getting monolayer or few-layer flakes.

Try this within 10-15 min after O2 plasma treatment. At the end of the day, the exfoliation yield is not so good. But if you make 3-4 substrates like the above, you should have a few 1L flakes.

You are very close to your target. You will get it. I am very confident. Just be patient. Let me know if it worked for you.

Good luck,

Chandan

Chandan Biswas I had tried the following comments and found that exfoliation improved! Although a few number of few layer WSe2 is observed, there's some progression. Thank you for the detailed explanation. It helped me a lot!!

Dear Youngkyu Ko,

I am glad it worked for you.

Keep moving forward.

Chandan

I would like to ask, which tape you are using? or there is a different tape for different crystals?

Dear Mohammed A Aly Nouh ,

Any standard office tape will work. We normally use, Scotch® Magic™ Tape https://www.scotchbrand.com/3M/en_US/scotch-brand/products/catalog/?N=4335+3294529207+3294603601&rt=rud

Good luck with your experiments.

Chandan

Dear Dr. Chandan Biswas , I was wondering how you remove the tape residue after exfoliation? It is becoming increasingly difficult to identify layers due to tape residue in the substrate under microscope. I tried soaking it in Acetone for 20 mins and then shake it in a bottle of IPA for 30 seconds and then rinse with DI water for a minute. After that I find other colorful residues

Dear Tawsif Ibne Alam,

You don't need to bother about tape-glue residue due to the following reasons:

Let me explain to you in a bit more detail.

Your tape contains roughly three regions:

1) Only glue region (glue directly touch your substrate, no flake in-between)

2) Flake+glue region (flake & glue both touch the substrate side-by-side)

3) Only flake region (flakes directly touch the substrate, no nearby glue region)

You will get monolayer to few-layer TMD flake exfoliation only in the region "3) Only flake region". You may get monolayer TMD in the "2) Flake+gule region", but the size of the flakes will be around 1-2 microns. This would be a too-small size for most of the device applications. Plus, thick glue residues will be nearby and nearly impossible to remove without damaging the flake.

Therefore, when you check OM, you should only find flakes in the region "3) Only flake region". Look at my attached OM-image, I clearly marked all three regions in the image and an example of monolayer WSe2 flake position in the region "3) Only flake region". In our lab, we often get 20-30 micron flakes in this region, a sufficient size for complicated device fabrications.

If your exfoliation does not result in the region "3) Only flake region" at all, then your "thinning-order" is too high. This means, most of the flakes in your tape are too-much thin for a good exfoliation. For a good exfoliation, you need "2nd-order or 3rd-order thining" of your tape crystal. Read my previous post on this thread carefully, where I explain "thinning-order" in detail.

Hope this will provide you with sufficient help and ecouragement.

I am very sure, you will get it very soon. Just be patient and enjoy your work. Let me know if it worked for you.

Good luck,

Chandan

Hi Chandan Biswas , it is good to see another group doing exfoliations sharing details. Our group perform exfoliation at a regular basis and I have noticed some differences in our steps. Since exfoliation a highly empirical process, there is a high chance that we might just be using the inappropriate steps or vice versa. Can we have a discussion of them to see if we can both improve our procedures?

1. The number of thinning. In our group, we will perform the thinning many times (~5 to 8 times). In fact we would fold the tape to do the thinner process to get the material across the entire tape. It is in our understanding that shiny regions tend to yield large pieces but most of them are just bulk materials, non-shiny regions will yield smaller yet thinner materials. Thus, to get monolayers, the ideal shiny-to-non-shiny ratio in our case would be something like 20% to 80% or even 10% to 90%, very different from your procedures. Have you tried more thinning steps and realized the chances of finding monolayers dropped by a significant amount?

2. The speed of peeling tapes. In one of your comments, you described that the tape detachment speed should be high here. It is in our general understanding that too fast tape detachment would break layers into small pieces. Thus, we do the detachment in a very slow pace. Is there any reason behind the high detachment speed here?

3. The wafer exfoliation condition. In our group, the wafer (after making contact with the tape) would be heated on hotplate at ~50 to 70 C for 1 hr while flatting the surface on the back to make sure no bubble is present. This in general improves material coverage at the compensation of increased amount of tape residues and I think this paper Article Facile Exfoliation for High-Quality Molybdenum Disulfide Nan...

Hi, we are also struggling with the exfoliation of TMDC, specifically with PtSe2. We are using Scotch tape for exfoliation purposes. We tried exfoliation on PDMS, chemically treated SiO2, and also on O2 plasma-treated SiO2 substrate. We did a post-heating (After bringing tape and substrate in contact) at 100 C. But in every case, our exfoliation yield is not as per expectation. We are getting flakes in bulk only (>20nm). For the analysis, we inspected our scotch tape (4-5 times peeled off to thin down the flakes on the tape itself) under a microscope (reflected and transmitted mode booth). We observed all the flakes on the tape were almost of the same thickness. And if we are further thinning the flakes on the tape itself, the layers of the material are not separating, but the flake is breaking in small sizes. we don't know whether there is some issue with the crystal itself or the tape; even we tried with new scotch tape and 2D crystals from "HQ Graphene" and "2D semiconductors".

Any suggestion form your side is highly appriciated.