A word of advice in the hope that nobody does what I did. I was developing a tricky procedure on a cation exchange column to analyse 7 amine compounds simultaneously. It was all working swimmingly and then someone suggested I try THF as the organic modifier - just to see how it differed from ACN and MeOH - selectivity wise. I mixed 10% THF with 20mM phosphate using the binary pump. Problem was I used GPC grade THF which contains butylated hydroxy toluene as a stabiliser which was 'dumped' by the buffer. It's taken me 2 days and a new flow cell to put things right. If you do need to use THF use HPLC grade without BHT. Apologies to the wiser of you who know this already, just thought I might prevent someone messing up like I did.

Yup - always use unstabilized THF for HPLC - and blanket it with nitrogen - and test it for peroxides every few months if you don't goo through it quickly.

We use unstabilized HPLC grade THF from Fisher, and buy 1-liter bottles to minimize contact with air. And we have those HPLC systems and waste bottles in fume hoods.

What happens if THF is not stabilized and air enters into the bottle while in storage? Please advise.

Since THF is an ether, exposure to oxygen will generate peroxides when no stabilizer is present. The danger of course is in the decomposition of the organic peroxide formed.

Regards,
Mark

And some of these perodixes have strong UV absorbance. Therefore, we should use stabilized THT unless it causes problems.

I think that the stabilizers absorb in the UV range too. That's why they are generally avoided. Nitrogen blanketing the THF stock bottles (1L or otherwise) is a good way to minimize peroxide formation.
Once the THF is diluted in water, the peroxide formation rate drops a lot, so you need not worry about blanketing with nitrogen etc. once you've made your mobile phase (unless you keep it around for a long time).

Is THF without stabiliser dangerous? Is it likely to explode spontaneously or anything or is it just a problem when heated?

It's a problem when it is neglected. If given the opportunity to evaporate over time with some oxygen exposure, shock sensitive peroxides can form and be detonated easily.

Peroxide test strips are available. We test our THF with them quarterly.

It's the same set of hazards commonly asssociated with diethyl ether but I think the bad reactions happen more slowly with THF.

It may take years from production to consumption of THF. Even sealed and filled with nitrogen, a lot of things can happen in the mean time. So, we can not do much before we open the bottle.

Therefore, I am still skeptical of unstabilized THF. Maybe we should try not to use THF at all.

THF offers some selectivity and separation benefits when good ol' water-methanol and water-ACN mixtures don't separate analytes. We use THF when it helps us.

Good THF vendors do not allow years to lapse between production and distribution. B&J, for example, puts a 1 year expiration date on their HPLC grade THF bottles. If you can't get the separation with MeOH, ACN, water - THF can be helpful. The only real reason I don't like having it around is the smell and the higher pressures it tends to generate.

Besides the safety issue, peroxides in THF may degrade your chemically bound phase and eventually destroy your column.
Keep unstabilized THF under Ar in a tightly closed bottle in refregerator in a dark bottle and filter it through Alumina (Al2O3) every 3-4 month depends how frequently you open the bottle. Alumina destroys peroxides without adding chemicals into the solvent.

With all the panic about THF:
You can get wonderful selectivity differences in your separation. These can not be achieved with the usual solvents, methanol and acetonitrile.

I know... it smells bad ... it explodes ... but it performs like a charm.

To all the HPLC people: you may want to know that THF is a very much used solvent in GPC!

To add another voice to this discussion:
THF is a very interesting solvent, especially when the compound(s) can't be solved and/or seperated in/with ACN or MeOH. Sure, acetic acid may be another possibility to solve a sample, but THF is less dangerous.
Best THF is fresh produced, keeped unter N2 (or another inert gas) and stored in a refregator in a brown bottle (just as Alexander wrote earlier). It has a relative low uv absorbance (compared to THF stored at room temperatur and without inert gas). For reducing back pressure I add ACN when seperation tolerates this.
In my opinion the only real flaw is the ugly baseline when running a gradientat lower wavelenght. If someone knew a good workaround, let me know.