Addressing earlier comments regarding “air tight” tubes: What few people realize is that below about -50°C, most of the containers we use are not “air tight”. We evaluated four types of cryogenic vials, three types of conical tubes, two types of glass vials and one type of microtube. At -70°C, none of these products reliably retained seal integrity. Within a few hours, significant exchange of gasses was observed between the sample head space and the surrounding environment.

Interestingly, it seems that the most reliable seals at +20°C are the least reliable at -70°C. Expensive cryogenic storage vials with o-rings or gaskets actually leaked more than cheap snap cap microtubes.

http://www.nature.com/nmeth/journal/v10/n4/full/nmeth.2409.html

I am not a chemist but I think yes if you are talking in absolute numbers. Your eppendorf is not a perfect closed system from a thermodynamic perspective. But for practical purposes one may not live enough to see that volume measurably drops below 300 ul,

Do you add a cryoprotectant like glycerol to your stocks? That would help with denaturation and issues of sublimation.

Just to add that in the short term you may even have an increase in the volume because at the time of closing the tube the water vapor that is present inside the tube will condense and freeze on the wall. If you thaw your sample and centrifuge, these droplets will add to the volume.

If the freezer is regularly/frequently accessed, I would say yes. Especially in the Kansas summers when the Relative Humidity can be rather low. A simple method to prevent loss of solution in longterm storage at -80 is to wrap the top of the each tube several times with parafilm. This will minimize any loss of sample regardless of how often the freezer is accessed.

Interesting question. I found one paper that looked at the effect of 4 years storage at -70 deg on serum. Basically they concluded that the reduced volumes they saw (up to 17%) were due to air displacement pipetting error of the original sample. They used Na+ conc. as an indicator of volume reduction since Na+ conc. should be inversely proportional to the volume loss. Perhaps you can do the same if your observations seem to indicate lower volumes.

Evaluation of serum volume losses during long-term storage.

http://archive.org/download/jresv98n3p355/jresv98n3p355_A1b.pdf

Water does sublimate at -80°C. Eppendorf tubes (and for that matter cryogenic storage vials) are not gas tight at -80°C and parafilm seems to do little or nothing to prevent to exchange of gases in the head space of Eppendorf tubes. (Murphy at al, Nature Methods 10, 278–279 (2013). At least in my circles, people "know" (or think they know) that small samples stored for long periods (>5 years) lose volume. I am sure this happens, but I don't know at what rate. The paper Charles sites (Neal at al) suggests it is rather slow. - Maybe negligible if you don't mind having the top bit if your sample quasi lyophilized.

It would be good if someone with expertise relating sample archiving would chime in. This is a subject that matters in my work, and I don't know of any good studies.

I would like to add, that in my experience sample sublimation/evaporation very much depends on where your samples are actually located in the freezer and how constant the temperature in the freezer is. Samples in the front are more prone to warming when the door is open, than such in the back. The wall of the tubes will warm/cool faster than the frozen liquid and the ensueing gradient will lead to condensation at the walls. That effect is really strong when using so called "no-frost" -20°C freezers in which one can lyophilize small volume samples almost perfectly -after a few months almost all of the water will have crystallised evenly troughout the tube. Actually I could never observe significant loss of overall volume after thawing and centrifugation of the tubes - the water remained in the tube but not actually in the sample. Since I store also frequently samples of 5-10µl at -80°C I checked for that effect there also. It was generally much less pronounced but after a few month samples in the front developed a condensate ring directly above the sample, while those in the back did not. Overall sample volume reduction was very low in the -80 when compared to a sample of similar size stored in a "no-frost" -20

Samples can definitely dry out, regardless of temperature; I've seen it often, even in properly closed tubes. Something I do, tho' I've never actually tested it to make sure that it works, is to store my tubes in a box along with several OPEN tubes to which I add water; my theory (hope?) is that the extra samples will tend to sublime first, increasing the water partial pressure and decreasing the tendency for my samples to dry out. (Or I may just add liquid water to the bottom of the box, if the tubes don't go all the way to the bottom; or a tissue that I've moistened.)

On a similar note: For samples that may denature with freezing/thawing, like proteins, I usually store them in a small Styrofoam box in the freezer, in an attempt to insulate them from defrost cycles.

Many decades ago, when equipment and methods were much less sophisticated, people used to place their alliquots inside carefully sealed thermo flasks and then placed these flasks in their freezers. I have continued with this practice and find that all our important samples such as long-term QCs and Standards are very stable. Not only does the thermos prevent sublimation, but it also buffers for any inevitable temperature fluctuation. Might be worth a try?

We have plant samples more than ten years old at -80°C. When I isolated genomic DNA, samples were practically lyophilized. So if the tube was not sealed, these things may happen.

ALso, I regularly observed sublimation at -20°C.

Great answers to this question. I would like to give +1 to James' comment about humidity and Daniel's comment about location of samples in the freezer. I've spent lots of time in labs where average humidity was low (Albuquerque, Lubbock) and high (Nashville). Freezers ice up much more rapidly in humid climates. Ice in the freezer will contribute to the vapor pressure of water inside the cabinet, which in turn will oppose sublimation. Also, remember that frost-free freezers work by going through warming cycles that hasten the sublimation of the ice. So as Daniel noted, samples in a tube rack near the door will warm slightly and go through a mini defrost cycle every time the door is opened.

The Parafilm trick is commonly done. It makes sense that a Parafilm seal wouldn't necessarily be any more airtight than the tube cap, especially a screw cap with an O-ring. I generally use Parafilm just to assure that the cap on a 1.5 ml tube doesn't inadvertently pop when the plastic gets cold.

I would say that, for very precious samples, one way to assure that you don't get ANY water loss would be to overlay the solution with clean, light mineral oil. When PCR was first developed we used a mineral oil overlay to prevent loss of reaction volume caused by evaporation and subsequent condensation on the walls of the tubes. Now this phenomenon is prevented by heating the walls of the tube; you still get a little evaporation, but the water vapor doesn't condense on the walls of the tube, so the loss of volume in your reaction is minimal. The same trick should work to block sublimation in a freezer.

Enclosure in air tight tube secured by wrapping with parafilm will minimize sublimation to the point it will not become an issue in your lifetime.

False. In my lifetime I have encountered the sublimation problem, specifically with aliquots of antisera made 30+ years ago. Houston is the 9th most humid big city in the US.

I do notice sublimation in stored protein and nucleic acid samples at -80 or -20 degree C and the sublimation could vary vial to vial even though in tight seal tubes. Whenever I am using samples after long term storage (more than a month), I measured the concentration again and adjust the volume if required. Though it sounds over doing but it avoids concentration variation among samples.

Addressing earlier comments regarding “air tight” tubes: What few people realize is that below about -50°C, most of the containers we use are not “air tight”. We evaluated four types of cryogenic vials, three types of conical tubes, two types of glass vials and one type of microtube. At -70°C, none of these products reliably retained seal integrity. Within a few hours, significant exchange of gasses was observed between the sample head space and the surrounding environment.

Interestingly, it seems that the most reliable seals at +20°C are the least reliable at -70°C. Expensive cryogenic storage vials with o-rings or gaskets actually leaked more than cheap snap cap microtubes.

http://www.nature.com/nmeth/journal/v10/n4/full/nmeth.2409.html

You have to realize that at low temperature the soft plastics of rubber gaskets become rigid and stop sealing the vial. This is what you observe with the expensive vials. Snap cap tubes on the other hand shrink evenly in the cold and thus maintain a better seal. I'd suggest trying a screw-cap system without o-ring (that's what we usually use for the storage of frozen electron microscopy specimen). If you want to try the parafilm method, you should use PTFE tape, the one you use for sealing plumbing. PTFE stays more flexible at low temperatures than parafilm. But screw-cap vials are probably still the best choice.

>tubes for long term storage, can the ice sublimate away,

I agree with James Stoops' suggestion. Furthermore, I would advise the use of screw cap vials containing a rubber O-ring. Using these tubes, I haven't seen volume loss when kept at -80°C for a long period.

Density at 0.01ºC is 0.999. Bellow 0ºC (D: 0.9162) ice contracts slightly while at -80ºC , ICE density increases further more (D: 0.925) . So it may look like you have less solution. At -100ºC (D: 0.926) ice is even more compacted. In case you are using at a graduate vesel.

How about freezing the samples and then storing it inverted? This way, the ice will be trapped in the tube and can not evaporate?

Interesting discussion. :-)

We are about to buy a new -20C freezer in our lab and we eventuated a no-frost technology.  Since sublimation could be a problem in non-air-tight tubes, I find the paper cited by Mark Fitchmun (Murphy et al 2013 Nature) quite helpful.

However, as I understand it, the paper shows the contrary of what Mark says here: there was no measurable gas exchange in tubes exposed to temperatures as low as -70C, meaning that they can be safely considered 'air-tight' at such temperatures.

Does anyone have some qualified comments on the no-frost technology for storage of lab samples at -20C?  

Thanks in adavance.

For Tian, inverting the samples is not going to change the laws of physics if the seal is not air-tight.

For Geziel, no-defrost freezer just means accumulated ice. This is fine if you have regular clean-ups to manually defrost the freezer (and spare capacity for transfers), otherwise it will be Siberia sooner or later.

Hi Mark Jimenez,

Thanks for the comment. My concern was more related to possible negative effects of 'no-frost' technology in concentration and quality of stored samples. 

Hi Geziel,

sublimiation is not only a problem if the lids are not air-tight. The water also re-sublimates on the tube walls, reducing the amount of water present in the original sample. Depending on the sample volume and tube volume in some cases more or less the complete sample water is evenly distributed on the tube walls after some time in no-frost freezers, leaving the sample efficiently lyophilised. We had that effect with no-frost freezers in our labs very reproducibly, which in some cases lead to very unpleasant effects (i.e. loss of activity of enzymes).

One of the problems is that you take such a tube out of the freezer, thaw it, spin it down and the volume is effectively unchanged - after all, the lid was airtight. So everything looks perfect. Whether your sample was damaged or changed by the lyophilisation surely depends on what it contains, but many proteins will not easily tolerate such a process.

I was always told that "no frost" freezers warmed occasionally to melt accumulated ice (Siberia!) and that this was definitely inadvisable for biological samples stored frozen...