Submitted by Barbara Reisner / James Madison University on Thu, 10/30/2008 - 16:07

As a solid state chemist, I've done suprisingly little crystal growth from solution at temperatures under 100 oC.  (I've also never taken an inorganic chemistry laboratory course; there wasn't one offered at my undergraduate institution.  I've run across a few paper that talks about layering solvents - a 1:1 ethanol:water solution containing the ligand was layered over an aqueous solution containing the metal source.  I'd like to give the reaction a try. Having never done this, I thought that I could do this reaction in a small vial.  A postdoc in the lab that I'm working in thought that you needed to do this in a thin tube (he has no experience with layering reactions either).  

 Could anyone provide some advice on growing crystals via layering? 

Hilary Eppley / DePauw University
My experience is from recrystallizing things via layering rather than doing actual reactions between the layers, but I think a small vial would probably work for what you are trying to do (I've done it in small Erlenmyers even when I am making starting materials).  Generally the thinner the tube, the better for growing large crystals because diffusion takes longer and the powder that forms initially when you layer together is less because the overall interface is smaller.   If you are looking for making bulk compound, obviously doing it in something like an NMR tube doesn't make a lot of sense, though if you want to grow a single crystal for X-ray, that might be a better choice.  The key to good layering is to get a clean interface.  I generally fill "bottom-up" (adding the second, more dense solvent to the bottom of the vial already containing the first one by sticking a syringe to the bottom of the vial) rather than "top-down" (trying to layer the less dense second solvent on top of the first) and try to use solvents with densities as different as possible.   Hope that helps!  Let me know how it goes.  --Hilary
Thu, 10/30/2008 - 18:20 Permalink
Dean Johnston / Otterbein University
I typically use a narrow glass tube, maybe 7-8mm I.D., with a small "pinch" where the two solvents will meet. The pinch helps limit the mixing of the solvents when you first layer them.  My experience is limited to non-aqueous solvents, however, so I don't know if this will work for your situation.
Fri, 11/14/2008 - 02:40 Permalink
Nancy Williams / Scripps College, Pitzer College, Claremont McKenna College

I've done layerings in a number of ways, all of which work, but some of which are more elegant that others. I've used everything from a thin tube (best when you want to get that one phenomenal crystal for an x-ray structure) to a bucket (best when you're trying to purify that 20 g synthesis of Pd2(dba)3 and you just want a high yield of reasonably good stuff from which you have separated (a) the dust bunnies and (b) the Organic Road Tar (O.R.T, pronounced phonetically)). 

Here are a few moderately useful things I've learned along the way:

 1) If you need to do it carefully, use vapor diffusion, not a true layering.

2) If you need to do it fairly carefully, use a gradient. On the bottom, have water, followed by a thin layer of 50 proof, followed by 100 proof, followed by Bacardi 151, followed by a big layer of Everclear. This gives you slower crystal growth, and mimics the effect of vapor diffusion, but works in places where vapor diffusion is too slow (e.g. a mesitylene/ligroin recrystalization in a -35C glovebox freezer).

3. Decaf is your friend when pipetting on the top layer.

4. If you need it done crude, you can put your bottom layer down (which presumably contains your solute) followed by a thin cushion of the same solvent, followed by your co-solvent. When recrystalizations go pear-shaped, one of the biggest reasons is that your layers mixed while you were making up the layering. The cushion gives you a safety margin in case your hands aren't rock-steady (see point 3).

5. Sometimes the best way to do it is to fill your vial with the co-solvent, and pipette or cannula your solute solution onto the bottom of the flask, which raises the co-solvent as it fills in. This works amazingly well with some solvent pairs, and is so easy, it should be illegal.

6. I always start a layered recrystalization at -35, and if that doesn't give me something, I'll move to RT to accelerate mixing.

7. The best way to monitor your crystals is to forget that you're growing crystals. When you clean out the freezer on lab cleanup day, voila! You have xtals.

Sat, 11/15/2008 - 23:07 Permalink
Sibrina Collins / Marburger STEM Center (MSC) at Lawrence Technological University

I actually do layering (crystal engineering) quite a bit. I usually dissolve about 20-25mg of my ligand in dichloromethane or chloroform(5 mL), and place that in a testtube. I then dissolve an equal molar amount of my metal salt in methanol (5 mL) and carefully layer that on top of the ligand solution. I then cover the test tube with parafilm, puncture a hole at the top, and let it sit for a week. Check for crystals every couple of days. We have been very successful growing single crystals like this at room temperature. Hope this helps. One of our papers describes this (Acta Cryst. 2007, C63, m436-m439).

Sibrina Collins, PhD College of Wooster

Tue, 08/17/2010 - 01:21 Permalink
Chip Nataro / Lafayette College
And don't forget the most important step. Learned this one in grad school. You must bow to your crystallization three times out of respect for the chemistry that is happening.
Tue, 08/17/2010 - 08:21 Permalink
Adam Johnson / Harvey Mudd College
.... and leave it in the back of the hood for a year.
Tue, 08/17/2010 - 19:23 Permalink
Kyle Grice / DePaul University

I would second Scott's advice.

 Also, there is a simple guide called "Layman's Guide to Crystal Growing", of which you can get a pdf if you google it (hosted by texas A&M).  I don't know who the original author was, but its a useful little document. 

In my view, crystallization depends significantly on the desired product.  It also involves a lot of luck.  You've got a head start if you are following a reported prep, as you don't have to try all sorts of solvents.  

Tue, 08/17/2010 - 21:18 Permalink
Sibrina Collins / Marburger STEM Center (MSC) at Lawrence Technological University

Yes, you do need to pray to the crystal Gods to get it to work.

Sibrina Collins, PhD College of Wooster

Wed, 08/18/2010 - 15:58 Permalink
gka / university of canterbury

I have experience of H-tube which is a sort of layering, and it works very well.

All you need a good clean practice + some good luck and lot of pray to the chemistry God.

Thu, 02/28/2013 - 16:54 Permalink
Chip Nataro / Lafayette College

I've recently been introduced to this fairly nice diffusion method. Take a disposable glass pipette and flame seal the thin end. Essentially put the flame right about where the pipette begins to taper, heat, pull and round off the closed end. If you are luck (and want them) you can even get some TLC spotters out of it. Dissolve your sample in this sealed pipette. Place it in a culture tube ( that contains the solvent you want to diffuse in. Of course make sure you do have too much solvent in the culture tube, you don’t want it to flow directly in to your pipette when you put it inside. I’ll generally start at room temp and see what happens. Go cold if need be. The fairly narrow opening of the pipette works pretty nice to keep things slow. I’ve also learned it isn’t a bad idea to make a really concentrated solution of the sample you want to grow crystals of, take a little bit of that and dilute it in a second pipette. This way you set up two attempts for the same sample. This technique also works nicely if you want to do a slow evaporation. Dissolve your sample as before, but put something like DMF or DMSO in the culture tube. This will give you fairly slow ‘reverse diffusion’ as your solvent diffuses into the DMF/DMSO.

Tue, 03/05/2013 - 06:58 Permalink
Mohammed Hammam / The British University in Egypt

In reply to by Chip Nataro / Lafayette College

that is actually very helpful, I have a question though, how much do you fill the sealed pipette? I understand you should fill it enough to allow the co-solvent to diffuse, so can we say it should be half filled as a start?

Tue, 10/27/2020 - 04:16 Permalink