5890 injector maintenance

[sllimbri]

I have always thought it was not good to do injector maintenance without cooling the injector and oven to room temp for a gc/fid for other reasons than you might get burned. Is this correct or does it matter? Will it affect the column?

[Peter Apps]

If the injector is opened, air gets in. If air gets into a hot column the phase can be damaged. Hot in this context means above about 80-100C, no need to wait for everything to cool to room temp.

Peter

[Bruce Hamilton]

Just for a change, I disagree with with Peter.

There are stationary phase, eg some polar ones, that will die if they see air at 100C. I prefer to wait the extra 10-15 minutes for both the injector and column to be below 50C before breaking open a system.
If you allow the gunk in the injector to oxidise, it's often harder to dissolve.

Non-polar stationary phases are more robust, but I like to give columns every chance to have long lives.

Please keep having fun,

Bruce Hamilton

[Peter Apps]

On sober reflection Bruce is probably right. Isn't it odd though that cooling the detector never seems to be inthe instructions ? - with carrier flow turned off (and no pressure at the inlet in any case) the detector end of the column is jsut as vulnerable to air intrusion as the inlet end. I suppose that it is less of a problem becuase it gets blown out of the end of the column, rather than travelling all the way along it.

Peter

[Bruce Hamilton]

Isn't it odd though that cooling the detector never seems to be in the instructions ? - with carrier flow turned off (and no pressure at the inlet in any case) the detector end of the column is jsut as vulnerable to air intrusion as the inlet end. I suppose that it is less of a problem because it gets blown out of the end of the column, rather than travelling all the way along it.

Peter

I suspect back diffusion is very small, and the thermal convection probably also prevents much air intrusion.

In fact, I prefer to leave FID heaters on 24/7, especially on 5890, as the internal O-Rings seal and slowly char giving a good, long life seal. The only time I cool the detector is to clean silica from the electrodes and outlet.

Given that we squirt 300 - 400 cc/min of air into the detector during operation, I suspect cooling would have minimal impact. I have left the FID gases on, and watched bubbles come out of the column inlet due to back pressure from jet nozzle.

TCD detectors are one exception - failure to cool them can result in slow oxidation of the filaments, failure to turn them off can result in rapid oxidation of the filaments

Please keep having fun,

Bruce Hamilton

[aldehyde]

another exception would be the NPD: make sure to heat in 25 C increments with a few minutes to adjust between increases. At least until you hit 150 or so, the bead is made of ceramic and absorbs water from the atmosphere when cool. If you heat it quickly (and even worse, hit it with voltage quickly) it can shatter.

[gpronger]

The other issue is oxidation of the new liner at elevated temperatures. If what you're running is sensitive to active sites on the inlet sleeve, if you drop the new one in while the inlet is hot, you can get a lot of active sites created. If you're running compounds like substituted phenols, amines, etc., you may find better response if you allow the inlet to cool completely, preform your maintenance, put the inlet together, restore flows, and then wait about 15 - 30 minutes to allow any potential oxygen to completely purge the inlet (this is a fairly conservative time allotment, but it keeps you on the safe side of things).

Greg

[aldehyde]

Thats an interesting point Greg, I try and do inlet maintainance at <100C at the very least to avoid getting GC-hands but you're probably right about making active sites at high temp.

[gpronger]

In the early '90's before there were a lot of pre-silanized inlet liners available I spent considerable time trying to understand issues around achieving the SPCC criteria for the EPA's method 8270. Part of the criteria was a minimum response factor for several problematic compounds (N-nitroso-di-n-propylamine; hexachlorocyclopentadiene; 2,4-dinitrophenol; and 4-nitrophenol with the last 3 being the most problematic). What I came away with was using Sylon CT to deactivate the liners and the need to be really careful with the injection port; heat it up too fast and you undid all your work to sylanize the liner.

Another bit of information I came away with is that if your injection port itself is stainless steel, it can also be sylanized. The sylanizing agent can bond with the oxygen atoms that are part of the stainless steel.