# How to solidify sand

I'm doing a research project where we analyzing the structure of sand after a certain process. To get a better idea of the internal structure of the sand, we use different colors to visualize the movement of the particles. However we can't actually observe that structure since sand is not a solid substance. Therefore we need a way to solidify the sand without changing the structure.

We figured we would need some sort of liquid which we can pour into the sand container (the container is 2x1x0.2 meters) so that it will solidify. In other words, how can we solidify sand?

Using a transparent container is not really an option since we are curious about the structure in the center which we cannot see... unless we use transparent sand ;)

Note: we are using 'regular' beach sand.

EDIT: I have read through all the answers and they are great! I can't accept any answers just yet since we need to figure out which method we will use, but it looks promising! We started with 0 and now we have some great ideas going with your help!

• You mentioned colours to visualize the movement of the particles (which I presume consist of quartz, since you are using beach sand). How could you still measure movement if you solidify the material, e.g. by embedding in matrix? – Klaus-Dieter Warzecha Sep 22 '14 at 14:47
• 2 x 1 x 0.5 is 400 L or almost 100 gal. of volume. At those volumes (lets say 20% void space), the cost of any binder will be significant. – Nick T Sep 22 '14 at 18:39
• Which of your dimensions is the height? if it's the 0.2m its a much easier job than if it's 1m. – Level River St Sep 23 '14 at 9:19
• @steveverrill sorry for my sloppy dimensioning. It's dimension as lxhxw. – Ortix92 Sep 23 '14 at 15:57
• @NickT I forgot to mention that we won't be filling the entire container with sand. That 1 meter will vary during each experiment – Ortix92 Sep 23 '14 at 15:59

Based on your comments, it sounds like you have some sort of colored sand and you want to immobilize it. My suggestion would be some sort of polymer matrix.

There are multiple possibilities, but one of the easiest is polydimethylsiloxane PDMS which is optically clear and pretty easy to work with (i.e., safe). There's the "base" and a curing agent which creates cross linking.

Probably the easiest how-to comes from U. Wisconsin.

If done right (i.e., no air bubbles) you should be able to pour in the PDMS, cure, and have a polymerized matrix around the sand. Then you could cut it up or do whatever needed to see different regions.

Geologists prepare soil profiles for a living. The task is easy in principle - dig a pit, apply glue to the wall, then transfer the glued layer to a sheet of cloth. The practice is really difficult, the glue recipe and transfer technique are trade secrets. See here (warning, German language): http://www.sand-abc.de/sandkorn/lackprofil/lackprofil.htm

Why not remove layer by layer and then take photos of the cross-sections? That would be easier than messing about with casting resin and vacuum chambers. Alternatively, enquire in your friendly Department of Geology.

I upvoted the water idea, after all it is possible to build sandcastles over 0.2 m high with just water as a binder. I also think coring as mentioned in that answer is a good idea. I also upvoted the comment about the freezer, although expansion could be an issue.

I’m going to suggest a hot gelatine solution (or even dessert jelly just to try it out.) There are various other food gelling agents (pectin, etc.) If it’s strong enough (and you don’t want it too strong, because you want to cut it open) it would be a good option. Gelatine is cheap, and you ought to be able to recover it by heating your used sand and filtering if cost is really an issue.

I’m thinking you may have a serious problem with air trapped at the bottom, if you are planning to distribute the sand using air then add the liquid afterwards. Chromatography columns have a porous frit at the bottom, which initially lets the air out (you see the colour darken as you add the first solvent) and then obviously the liquid, too. A wire or plastic mesh over a platform grating would probably work in your case. Pressure is usually applied above a chromatography column to push the solvent through, but in your case you may find it easier to suck from below, if a pressure difference turns out to be required. A bottom connection could also be useful for preheating the bed with steam.

You also need to think about how to distribute the liquid without disturbing the sand. A matrix of pipes with sprinkler holes, and/or a plastic mesh on top of the sand is probably the best way.

EDIT 1: some more ideas, which while more expensive, will give you a sharp, predictable freezing point. “Glauber’s salt” (sodium sulphate decahydrate) melts at 32 °C, and sodium thiosulphate pentahydrate melts at 48 °C. Sodium acetate trihydrate melts at 58 °C. These compounds can all be supercooled below their melting point, then made to suddenly freeze by adding a single crystal for nucleation (though I expect with sand about they will nucleate immediately.) Videos of this can be found on Youtube. The compounds have been tried for medium term storage of low grade solar heat for space heating, and/or are used in personal heat packs, as all the heat is released at the same temperature (the melting/freezing point) when they cool.

Sodium carbonate decahydrate “washing soda” apparently melts at 32 °C. I’ve never seen any fun experiments with supercooling melted washing soda, so maybe it does not supercool. The situation is confused by the existence of a heptahydrate too. However it is certainly cheap and readily available. It’s strongly alkaline, so would tend to pick up carbon dioxide from the air to become the bicarbonate, but in the volumes required I doubt that would occur to the extent required to produce any change in characteristics. Alkalies are harder on the skin than acids, because they turn your natural oils to soap.

I also thought about candle wax, but with the quantities involved extreme caution would be required to keep it between its melting point and flash point.

Any contact between molten salts and metals should be checked carefully for corrosion. In particular alkali + $\ce{H2O}$ dissolves the protective oxide layer on aluminium, leading it to corrode extremely rapidly with evolution of hydrogen gas.

EDIT 2: even though the OP has now indicated that the height is 1 m not 0.2 m, I still like the gelatine idea, though it’s probably not strong enough to support a metre column. It would probably mean tipping the tank on its side and removing the front (which will have become the top.) It’s a question of balancing the strength of the matrix against the ease of cutting (which will be extremely arduous for some of the higher strength suggestions mentioned here.)

• BTW, I am a chemist turned chemical engineer. Hence my answer is far more about chemical engineering than chemistry. Your research would be of interest to chemical engineers, and they may be able to help you out with the problems in performing it. The relevant maths is applicable in both chemical engineering and earth science / civil engineering. I wonder if there is anything relevant on Earth Science stack exchange. – Level River St Sep 23 '14 at 12:44

I don't know what kind of resolution you're looking for here. If you want a true 3d structure you might be able to create one by taking slices off of the sand bed using adhesive paper. Adding a glue or epoxy to the sand bed will also disturb the sand while holding it in place.

Do whatever process you've got to do to the sand to get it to its final state, then take a picture of the surface. Apply an adhesive sheet to the top layer and remove. Take another picture. Apply a sheet, remove, picture, repeat. It's going to take a lot of sheets to empty the bed, but if you want a really fine structure of the sand, it ought to work.

Another option would be a container with walls and bottom that can move independently. Image, slide the sides down by 1mm, use a wiper to remove the protruding sand, repeat. It might prove difficult to avoid disturbing the lower layers as the upper sand is wiped, but the consumable cost is virtually nothing with this method. (This idea is basically the reverse of powder bed 3d printers, if it's not clear.)

ETA: A cheaper version of my first idea - Prepare your sand bed. Image, spray a heavy layer of spray glue on top. Apply a sheet of printer paper. Wait a minute, remove. Repeat. I think you would pick up more sand with each pass this way and it shouldn't cost too much.

the container is 2x1x0.2 meters

I guess you want something cheap in this case.

option 1: water. Wet fine sand keeps shape. Not strongly, mind you, but still. option 2: drop it, push glass tube into sand (reasonably wide with thin walls) option 3: use diluted silicate glue. Would work, but you will have troubles drying it without oven.

Anything else is too costly for fun personal project.

• option 1 + an industrial freezer – Pete Kirkham Sep 22 '14 at 22:36

Sand particles are held together by covalent bonds. The only way to get 2 separate sand particles to join is to heat them to a temperature at which covalent bonds form between particles. This is the process used in making glass. So, to do what you are proposing requires you to make glass from the sand.

• Well we can't really put the container into an oven... – Ortix92 Sep 22 '14 at 14:40
• Still sounds like the right direction. You may also want to consider using a low-melting plastic for your color tracer. That avoids the need to melt the sand outright. You might be able to get away with grilling the sand. – MSalters Sep 22 '14 at 22:02