I really like these things, I think we could use them as wallets too, and remembered me a lot of the https://bt.irlbtc.com/view/5111005 bitcointalk collectible, could be waaay cheaper to implement than any metal wallet.

I don’t like to give out my address on this forum for obvious reasons but these are fucking awesome yogg.

Kudos.

I find the prospect of mass-adopted physical wallets very exciting for a number of reasons and was fortunate to get 2 Coldkeys from Yogg for testing. Here's a description and results of my first test.

The photos below are low-res to save bandwidth. Here's a high-res album: https://imgur.com/a/a0qqmO6

Inspired by the S1 Casascius solvent hack, I decided to give a variation of that method a shot.

For this test, I decided to test 2 different solvents. Chloroform (pictured) and another non-polar solvent. Unfortunately I was out of glass syringes so I used plastic. And the smallest needle I could find.


I attempted to pierce the non-adhesive part of the hologram and inject in some chloroform, however the needle was blocked by a layer of paper/cardboard between the hologram and the private key. It also left a pretty obvious track mark (See photo below). The next thing I tried was trying to slice the edge of the hologram with a scalpel, and trying to inject some chloroform between the plastic card and the hologram. This worked to some extent, but soon, the word 'void' was clearly visible. I won't get into the details of the hologram now, but there's two layers. If the two layers are no longer adhered to one another, the word 'VOID' becomes visible. (see photo below).



Next, I decided to attack the hologram from the LEFT side using my other solvent and pretty much the same technique as above.
The second solvent was a lot better at dissolving the adhesive and I was able to peel back most of the hologram without the 'VOID' becoming visible. However, it's still visible at the edges from the original incision/prying open the adhesive to inject solvent in. I then sliced the top and bottom off with the scalpel.
Yogg promised me a "surprise" and he delivered. The private key is printed on a polymer that is different from the card plastic and almost the entire key melted (probably too fuzzy to be recoverable), as a result of the chloroform. However, as you can see in the photo below, the LEFT side, where I used my other solvent, is not melted. So I'm pretty confident that I've found a solvent that can remove the adhesive without dissolving the private key, but will test this more in the coming days.


As you can see below, the hologram is pretty mangled, however, I think if I practice my technique and stick with the second solvent, I might be able to do a much better job. I'm going to spend 4 or 5 days thinking about how to attack my second Coldkey and will a few more ideas on this one.

 I'm also going to try a few tricks to reattach the hologram without leaving too many traces (as it's pretty obvious this one has been tampered with) and will see if I can find a way to get rid of the 'VOID' text.
I have access to a lot of equipment and chemicals and am definitely open to suggestions if anyone has some. I'll also be testing some optical attack vectors and will see if I can alter the public key although it looks tough.


Below is a photo of the tampered coldkey (left) and the untampered one (right). I think after about 1 second of inspection, it's pretty obvious this Coldkey has been tampered with. Make sure to zoom in on the high res photos: https://imgur.com/a/a0qqmO6

To be honest, congratulations for your achievements with the above pic.
This got me thinking to find a solution that would not allow to re-use the hologram, if we imagine the case where the VOID pattern doesn't appear.

So far, this is what I found as a solution :



This way, the holo cannot be re-used if we emboss something on it.
It doesn't have to be a word. Maybe simple "BTC" or the ticker of the coin. I don't know yet. It can be solely some character.

By the way, test generic hologs were 20x20mm, but the definitive ones will be 22x22, so more space to crush something on them.

Thanks so much for your review Sa0shisGhost.
Yes, I will try to munch a little on the holos in the future, but this has to be thought of during the card designing stage.

I am curious to know how easily can the back of the card be laminated again. The embossing should make this difficult to reproduce.

---

/!\ Don't do this at home
I have been experimenting a little further with Coldkeys, particularly with the private key QR code.

First, I started by carving the privkey QR code with a pair of scissors and made some strip.



Then, why not, I completely cut off the QR code in two.



And the results are surprising !
This card that's been so badly damaged is still functioning !

This morning I had an apointment with a company that does quality control on PCBs.
They really have some high-end hardware, it was a very interesting tour.

Anyway, the engineer who was doing the tests on Coldkeys said he can't think of a method that can reveal the private key without damaging the card.

There is nothing to be seen, that's all.



I brought in two Loaded Coldkeys.
One that is done, loaded, and funded. It's serial "BTC-19.5M.14".
The other one was a test Coldkey with the privkey exposed, to compare tests.



The best they could come up with is this :



We can see the shape of the square paper that is underneath the hologram.

I made a short video when they were fiddling with settings to check what can be extracted from the Coldkey with x-rays.

Video : https://streamable.com/z532h
There's nothing to see here.

They will try with a few other methods, one called "laminography" which is "3d xray" or something like that.

As the Xrays are sensible to the change ("contrast") in matter's atomic density, they told me that overall, they don't expect to see much from these tests.

---

Hello guigui371,

Thanks a lot for your proposal. It is very interesting !

I'd love to see how a Coldkey looks under an MRI scan device.

Alright, I can do it and mark it in a way it's good only for such use.
I'll reach out to you to arrange it all and talk about it in more details.

Just bear in mind, that Coldkey, the operating company and myself cannot be held liable for any kind of damage that would happen with the MRI scan device.

Thank you.

---

Any idea on how to crack these Coldkeys ?

We have pushed the engineering tests a little bit further, this time testing a non-invasive method called "laminography".
Laminography, basically, generates a 3D sample of the inside of an item, putting in evidence cavities and/or irregularities.

Here is the best image of the laminography of a Coldkey :



There are other pictures (different "depths"), but this is the sharpest one.

Here is a video of the shot, and going through the hologram, to see what's further underneath it :

https://streamable.com/dynhc

No QR code appears.

---

As the non-invasive methods that would allow to reveal the private key QR code are defeated one after the other;
we are headed towards delivering very elegant and secure solutions for everybody to hold onto their cryptos.