At ALS, we have a red epoxy-on-concrete floor, which is "lN2 resistant" 
but still cracks if abused.  However, we have not repainted it for many 
years now.  We found that one very good way to extend the life of the 
floor is to keep users from dumping their cryogens onto it.  In 
Berkeley, it is easiest to appeal to their "green" sense and offer lN2 
"recycling".  Simply place a coffee filter into a funnel (foam funnels 
are best) and pour your old, icy lN2 through the filter and voila! 
ice-free "recycled" nitrogen!  It actually works very well (no snowball 
crystals), and you can see the pile of white fluffy snow left behind in 
the filter.  Coffee filters are good, but a few layers of Kimwipe(TM) 
will also do in a pinch.  Recycling also has the advantage of reducing 
the other factor that limits the lifetime of cryo-room floors, which is 
the crushing force of heavy supply dewar wheels (which you will be 
replacing less often).

I have spent some time reading about this problem in the past, so 
although I do not have contact info for a cryogen-proof floor supplier, 
I can speak to the general principles.  lN2 is the world's best paint 
remover, and this is because very few paints have exactly the same 
thermal expansion coefficient as the surface you painted.  When exposed 
to cryogen, both substances become stiff, but shrink at different rates, 
and so tend to shear along the surface (peel).  There are three ways to 
combat this:

1) find a paint that matches the expansion coefficient of the surface.  
Epoxy and concrete are a fairly good match, (and by some lucky miracle 
of civil engineering, so are steel and concrete), but for changes in 
temperature of 200 degrees, even slight differences in expansion 
coefficient can be important.  This is probably why epoxy flooring is 
"lN2 resistant" and not "lN2 proof".  You might be able to find a better 
epoxy, but will probably have to do experiments to be sure.  I doubt 
your paint supplier will have CTE data on their paint, and there is no 
way for them to know what it is for your floor.

2) find a "paint" that does not become brittle at cryogenic 
temperatures.  Most steels are good for this, and also have the 
advantage of remarkably low thermal conductivity (for a metal).  Invar 
steel has no expansion coefficient at all, but has the disadvantage of 
being the only "stainless" steel that rusts.  Another one is teflon, 
which is still flexible under lN2, but (like most plastics) suffers from 
a high thermal expansion coefficient, meaning that a teflon floor would 
be under tremendous stress when cooled.  The bigger the piece, the more 
"peeling force" generated at its ends.  This brings me to the third point.

3) A trick I call "macroscopic bonding".  Expansion coefficient mismatch 
is only important if the stress induced by the temperature change 
exceeds the strength of the bond holding the floor together.  If this 
"bond" is simply the Van der Waals between two smooth surfaces (like 
paint) it is relatively easy to make this "slip" along the plane of the 
surface (peel the paint).  For example, if you try to epoxy together two 
smooth plates of aluminum (which has a VERY different expansion 
coefficient from epoxy) and then dunk the result into lN2, you will see 
everything readily crack apart.  However, if you first drill a lot of 
little holes at criss-crossing angles into the aluminum, allowing the 
epoxy to fill the holes, then you will find the resulting bond 
remarkably resistant to lN2.  This is because the covalent bonds in the 
bulk of the epoxy (occupying the holes) must be broken before the two 
materials can separate. In fact, this is what I did to get my crystal 
mounting robot's aluminum tong paddles to stick to the wooden pencils 
that hold them.  This bond has shown no signs of loosening after 
thousands of dunkings into lN2.

I think this is probably what happened in David Roberts's painted foam 
floor room.  Cushioning probably plays a role too, but since foam has 
lots of nooks and crannies, the paint must break the covalent bonds 
holding its bulk together before it can separate from the foam.  The 
foam probably shrinks tremendously more than the paint, but because the 
painted surface is not smooth, the forces of expansion-coefficient 
mismatch do not build up, but instead are transferred between the 
surfaces by the "rivets" of paint at each nook and cranny.  No doubt, 
the "nooks and crannies" on the underside of high quality tile form a 
"macroscopic bond" to the mortar/grout beneath.

So, perhaps a good BSL2 floor would be to find some stainless steel that 
is smooth on one side and "holey" on the other, then epoxy the "holey" 
side onto a (roughened) concrete floor?  That is, if you don't mind 
working in a room that looks like a prison shower.

-James Holton
MAD Scientist

On 2/16/2011 4:59 AM, David Roberts wrote:
> We went through a building renovation and this has been a problem for 
> us too.  In the old days, our building was simply sealed concrete - 
> and you could do anything with it with no issues.  Now, it's an epoxy 
> floor, but really what happens is the liquid nitrogen cools it down 
> and breaks the seal between the epoxy and the concrete - causing 
> cracks and ugliness.  We use a series of throw rugs and large 
> containment pans to hold the nitrogen.  It doesn't really work - so 
> every now and then we simply repaint the floor.  It's only in a few 
> places that we use this, so it's not too bad.
>
> I do have 1 room that they put the wrong floor down first.  It's a 
> spongy floor - designed to not carry a static charge.  Instead of 
> removing the covering (which they can't do), they just went over that 
> floor with an epoxy coating.  Oddly enough - that room doesn't crack 
> when I pour liquid nitrogen on the floor.  So maybe that's the trick - 
> put a cushion between the concrete and the epoxy.
>
> They wouldn't leave sealed concrete because it looked bad (we did a 
> renovation, not a new building).  Too bad
>
> Dave
>
>
> On 2/16/2011 7:32 AM, Richard Edward Gillilan wrote:
>> This has been a problem for us too. Sorry, I don't have a solution to 
>> offer except, recently, we have provided metal buckets filled with a 
>> layer of aquarium gravel at each station and encouraged users to dump 
>> their excess nitrogen there instead of on the floor.
>>
>>
>> Richard
>>
>> On Feb 16, 2011, at 7:20 AM, Nicholas Keep wrote:
>>
>>> Can anyone recommend a floor coating that passes category 2 
>>> containment (ie not wood) that is resistant to liquid
>>> nitrogen.  Ie you can fill dewars on without cracking.  Various 
>>> solutions our estates people have fitted have all proved
>>> unsatisfactory.
>>> Bets wishes
>>> Nick
>>> -- 
>>>