Growing Crystals on LEDs and ETextiles




Introduction: Growing Crystals on LEDs and ETextiles

About: designer, leather ninja, tech explorer, ruiner of manicures

Want to grow glowing crystal LEDs? This instructable shows our testing process for growing crystals around LEDs on an eTextile circuit.

Crystals are wonderful as you never quite know how they will turn out. ETextiles are also somewhat unpredictable. For our base material we are using suede leather which is also a material with unusual and intuitive properties.

The process shown here is from the sampling process for a larger wearable piece.

We created etextile circuits on leather with different kinds of LEDs. Both were hacked for etextile sewing:
-small white surface mount LEDs (2 connection points per LED)
-programmable white APA102 LEDs (8 connection points per LED)

This instructable is taken from a project called eTextile Crystallography, a collaboration between Rachel Freire and Melissa Coleman. It explores the idea of growing a luxurious futuristic garment and the aesthetics and practicalities involved. Red carpet dresses and gowns are created at huge expense and often worn only once as they are so unique and recognisable.

Rather than make a single use, incredibly expensive crystal gown with synthesised glass crystals, we pose the idea that you can create a gown which can be grown and re-grown in different forms to be worn on multiple occasions. Technology is expensive and often not sustainably created, so we wanted to make the foundation of a dress - including reactive and interactive electronic elements - which can be re-used and re-created as a different garment for each wearer. The eTextile electronics in the garment are washable and will stay the same. But each time the dress is worn it's crystals can be re-grown and each resulting gown is completely unique.

Work in progress photos on Flickr, which will be updated as the project progresses:

You can follow Melissa here on Instructables.

Step 1: Materials and Tools


- aluminium potassium sulphate (alum)
- conductive fabric
- conductive thread
- fabric base (we used leather)
- surface mount device (SMD) LEDs and/or addressable LEDs
- gel pen or chalk (for drawing the circuit)
- jewellery wire
- insulated wire
- copper tape
- bonding material such as Bondaweb or Heat 'N' Bond (to glue the circuit)
- superglue
- clear nail varnish


- domestic iron
- soldering iron and solder
- hand sewing needles
- wire cutters
- scissors

- weighing scales
- a large bucket/tub (preferably clear plastic so you can watch the crystals grow underneath!)
- measuring jug
- a pan and stove

Step 2: Preparing the Elecrtonics and Making the Circuit


You will need to start by hacking your LEDs so you can sew them to the circuit. We soldered small loops of wire on to the solder pads so they can be stitched to the fabric circuit using conductive thread.

The smaller surface mount LEDs are fiddly, so we left the legs of the loops long on one side and cut them off as each one was stitched. This is a useful way to indicate the plus and minus, so you are less likely to make mistakes when sewing them on to the circuit.

Dab on a bit of flux to make soldering SMD LEDs easier. Don't overheat the LEDs with your iron. Test your LEDs with a coin cell battery once you're done.

Making the circuit: the tech and the sewing

We made two different circuits to figure out the process and the aesthetics of using different types of LED technology in combination with the crystals: SMD LEDs and programmable LEDs. The SMD LEDs are not programmable so require only two connections for power. The addressable LEDs need in total eight connections for each LED: clock (in/out), power (GND and 5V on both sides) and data (in/out).

Our SMD LED circuit contained 7 LEDs (four with soldered loops and three soldered to copper tape) with in total 14 connections to sew or solder.

Our addressable LED circuit had 2 LEDs with in total 12 connections to sew. It's 12 rather than 16 because you don't have to connect the end of the last addressable LED in the circuit.

Draw the circuits on your fabric using a gel pen and iron on conductive fabric or stitch with conductive thread. Stitch on the LEDs using the soldered loops and make sure all connections are strong and secure.

We also soldered loops to small pieces of wire to connect to power. We made the SDM LED circuit connect to a coin cell battery, and the addressable LED circuit connect to a microcontroller. In principle both types can easily be controlled by a microcontroller, but with the addressable circuit you won't have any lights unless you drive it with signals. With the SMD LED circuit the microcontroller is optional.

Step 3: Testing the Circuit

Check everything is working and the connections are stable.. next we seal the electronics and it's hard to go back.

Step 4: Prepare the Circuit for a Bath! Waterproofing and Starter Crystals

To grow the crystal the whole sample will need to be submerged in water.

Use clear nail varnish to paint all the exposed metal to seal it and protect it from the corrosive salt. Paint the wire loops, the LEDs themselves and any exposed metal. On an earlier sample with copper-based connections the connection between the LED and wires started corroding over a matter of days once they were crystallised. On the silver-based circuits it's probably not as urgent but these materials definitely degrade over time so sealing them does help preserve them better regardless. As there are so many factors that can make an e-textile circuit unstable you really want to use all options you have to make your circuit more reliable.

To have more control over where your crystals grow, and to make sure they grow on your LEDs, you can use starter crystals. These are small crystals you make yourself using the same process we'll use to crystallise these samples, but without any material to grow it on. Use superglue to stick single crystals or clusters to each place you want a larger crystal to grow.

Step 5: Growing the Crystals

Measure out the aluminum sulphate and water. We used 70gr / 100ml but this is by no means an exact measure. The main goal is to oversaturate the water with the salt. Many instructions for making crystals will tell you that when your solution is oversaturated the salt will stop dissolving but you might never reach a super clear moment where you're sure of that. Embrace the uncertainty.

Heat the water up in a pan and slowly add the aluminum sulphate so it dissolves. Stir until all the crystals have dissolved completely. The warmer the solution the more salt it can hold so the easier the salt dissolves in it. This is a good way of over saturating the water as the water will not be able to hold the same amount of salt as it cools down and will start attaching itself to any surface it can find.

Fill a tub with your solution while filtering out any remaining crystals and gently place your fabric face down in the water. Our material floated, but a more porous fabric with a large weave might sink, so you may want to prepare some scaffolding in advance! A sample on the bottom will just stick to the bottom and have a flat surface rather than the nice crystal shaped one you're aiming for.

When the fabric is suspended in the water the crystals will grow all around, but since the crystals can form anywhere they will also start in other places and sink to the bottom and then start colonising that space as crystals like to attach to others of their kind. In our case, we wanted as little growth as possible on the back of the fabric so kept it suspended on the surface of the water.

Here is where it becomes interesting.

We found that the more you move the fabric or tub, the smaller the crystals become. This is because each time the liquid is disturbed by vibrations, new crystals effectively form in the liquid. This means if the fabric stays completely still you will get large clear crystalline structures with flat triangular surfaces. If you disrupt the water even slightly during the process the effect will be more fragmented and powdery.

This can only be controlled to a certain extent, but with some practice the crystals can be manipulated and grown in layers to achieve different effects.

Step 6: Finished Samples

Our tiny SMD LED crystals look beautiful and the LEDs completely disappeared under the crystals. The two colours you see are from warm white and cold white LEDs as we decided to test both kinds in the sample to see how they compared.

The addressable LEDs were not so magical. They were just the ones we happened to have available for testing, and it was hard to get enough crystals to grow to cover the large board. If we had left them in the water long enough, they would have been covered, but the whole fabric would have become thick with crystals as we were also still figuring out the crystallising process. Definitely need to try this with smaller PCBs.

Playing with how the crystals grow and how the starter crystals effect it is part of the fun. Combine this with different concentrations of solution, time in the water and stillness/vibrations and you can engineer a crystal fabric!

Lights Contest 2017

Participated in the
Lights Contest 2017

Be the First to Share


    • Barbecue Speed Challenge

      Barbecue Speed Challenge
    • Furniture Contest

      Furniture Contest
    • Explore Science Challenge

      Explore Science Challenge



    Question 6 months ago on Introduction

    Does the type of fabric impact crystal formation? Do you think they’re more likely to stick to say cotton than a polyester?

    Gary Viveiros
    Gary Viveiros

    4 years ago

    Aqueous-based crystals come with a big set of headaches. In Hawaii we have too much experience with highly ionic aqueous-grown crystals (called salt) from the ocean mists around the coastal areas - going inland about a quarter to half-mile inland. They ruin electronics and rust-out appliances and auto bodies. In industrial settings, any conductive traces have to be sprayed with a silicone lacquer, or coated with a conformal coating to prevent etching the traces by the presence of a corrosive salt crystal which absorbs water. Epsom 'salts' are no better. I'd try casting acrylic 'crystals' and experiment with positioning them for best lighting effect, or crack a piece of 'safety glass' bought from an auto junkyard and using clear epoxy or resin to join small pieces into small clumps as the basis of crystal-looking groups (then coating them to prevent sharp edges).


    Reply 4 years ago

    i really like the impermanence of this actually. I am definitely curious to see how they fare over time, in terms of corrosion..

    Mark 42
    Mark 42

    4 years ago

    I think I'll try growing rock candy frrom an LED.


    4 years ago

    I love art, science, and fashion, and I LOVE this! And since punning unfortunately runs in my genes, I have to say the inevitable, "This ROCKS!"


    4 years ago

    I can't get over that cover photo. Gorgeous!

    spark master
    spark master

    4 years ago

    I can see this as an art installation , but clothing? don't walk in the rain as they will all dissolve.

    This brings me back to the 1970's and chem lab. As brother Paul would say :

    "You can Feeell the heat" followed by

    "Poof no more freshman!"

    Ah the good old days


    4 years ago

    wow, it looks amazing. Great work


    4 years ago

    Interesting idea. At the beginning, you talked about growing these crystals for fashion design. I wonder how fragile these crystals are? Is the crystal matrix flexible at all? Fabric needs to be at least slightly flexible to be worn comfortably.


    Reply 4 years ago

    They are surprisingly sturdy, with a bit of experimentation. I think it would be different for every garment, depending on construction, so would involve quite a lot of testing. They definitely shed over time.. for us, that was the point, their impermanence. You can always grow another layer of crystals on top of existing ones.

    In the album linked below there are some images of a collar we began to construct using illusion net (commonly used for swarovski crystal gowns). Because it is sheer the crystals can actually grow through the material rather than sitting only on the surface. You can see some samples in grey. They have some flexibility, but not a lot as we grew the crystals quite thick and used plastic boning:

    In this case, we made a stiff frame to support the crystals so they are suspended on a stable base away from the body. I also lined the bolero part in clear plastic so the crystals can't grow on the inside. I will upload more images and our conclusion when it has been crystalled!