5890 rehab and startup

[southwest]

Hi all! I am new to this forum and have some burning questions regarding an old 5890 system.

As stated it is an HP 5890 series one w/ split splitless injector and FID (also a packed injector and photometric flame detector, but I'm not interested in them for my determinations.) The instrument was purchased at auction and appears to be functional for all electronic systems. I am installing a 15meter 0.25x0.25 capillary column using hydrogen as the carrier. The gas is UHP 4.5 or better with a whatman zero air generator and the appropriate traps.

My question is how important a makeup gas flow for the FID is. What sort of sensitivity gain can be expected? I suspect a large number of people have tried this first hand and can tell me if it will broaden a little without makeup, or if the dead volume in the jet is large enough to occlude a baseline resolution. (obviously dipends, but some general idea would be great)

I am replacing all consumable parts and have not ran this thing for the first time yet. I would like to keep startup costs to a minimum and re-invest as needed, and another tank with regs and traps may be another $500. Should I let it rip without makeup and see if it is good enough or do it right/ the best way?


Thanks,
Soutwest

[osp001]

I have a Series II 5890 FID/NPD, and run hydrogen as both carrier gas and makeup gas. I probably lose a little sensitivity, but the baseline is fine and the machine operates normally. There's a sweet spot for the flow rates, so it may require a little tinkering to tune it in.

[southwest]

Awesome osp001, I was kinda hoping a good mixture was all that is needed. I suppose the internals of an FID do not mind a variable change in flame intensity too much. A grounded jet and anode should be stable over a very wide range. . . I will be sure to investigate the dynamic calibration range of this technique and (hopefully) linearity thereof.

Thank you for the timely reply.

I hope other people can learn something from this question.

It seems simple enough to test your method with this gas.

I may even go crazy and direct the vent line from the split injection to be the makeup gas assuming the total volume is purged by the time my detector is reading important values. (but not bloody likely, the H2 is only $120 for 200 cubic feet so lets "KISS")

I should be at ASMS for an update on the next step. . . real data

Cheers

[Peter Apps]

Hi Southwest

If you use hydrogen for makeup all you do is increase the total hydrogen flow to the flame - which makes it much noisier. With hydrogen makeup you need to decrease the hydrogen fuel flow to keep the total hydrogen flow at the optimum, and since both hydrogen and makeup run through the jet nothing in the flow has changed except that you are now supplying hydrogen through two lines instead of one.

As long as the tip of the column is inside the tube below the flame jet, the fuel hydrogen flushes the column effluent up into the flame very effectively, so you will not see peak broadening without makeup. The sensitivity gain with makeup is a maximum of 30%, which is not likely to make a difference. The sensitivity gain with makeup depends on diffusion in the flame, so only nitrogen provides a sensitivity gain.

Peter

[Bigbear]

The fid for 5890's were designed for packed columns ( 0.25" dia.) so you would be better served by using make up. Get a bottle of N2 and set the make up to at least 30cc/min.

[Consumer Products Guy]

On our non-EPC 5890, a co-worker often forgets to turn on the make-up gas flow, and his chromatography results always seems OK. We typically have to have the make-up gas turned off to ignite the FID, then turn it on manually, he forgets.

Typically we use a 30m x 0.32mm capillary there.

[southwest]

Very cool.

I understand improving ionization efficiency chemically by altering the chemistry of the flame, and am not as concerned with that as the dead volume around a capillary into the jet.

From the chromatography classes I have taken: dead volume seems to be the greatest setback in an analytical system. There is no point in achieving a great differential interaction leading to differences in mass transport just to mix it all together before it is detected. Sensitivity is not a very big concern right now, but having sharp peaks is. The analysis I'm preparing to do has good enough resolving power for the species that a little slop is acceptable to begin with. The suggestion of using hydrogen makeup gas makes a lot of sense.

hydrogen is fairly inert (before mixing with some oxygen not much will happen)

hydrogen is already needed for the flame (and carrier gas)

I'm going to give it a shot and i'll report back

Thanks for your time.
Cheers,
Southwest

[Peter Apps]

Using hydrogen as make up gas makes no sense at all as long as you have installed the column correctly. Once they enter the body of the detector both makeup gas and fuel hydrogen flow along exactly the same paths. Adding hydrogen as makeup means that you will have to reduce the flow of fuel hydrogen. For the FID to work properly the total flow of hydrogen needs to be 30 - 40 ml/min and that has the same flushing effect no matter whether you introduce it as fuel or as makeup gas.

Peter

[southwest]

thanks, I was under the impression makeup was added as a sheath flow to focus the outlet stream. What you are telling me is that the fuel gas does this also. Are the fuel gas and makeup mixed prior to introduction of the jet body in all cases?

[Peter Apps]

If you have a look at a diagram of the FID you will see that both fuel and makeup are introduced next to the column where it goes into the flame tip. The then both flow up through the flame tip, around the column. The Air is introduced above the flame tip through a ring of holes or a circular slit.

At the scale of the FID, with the flow rates and gasses involved (diffusion in hydrogen is very rapid) there is no "sheath" effect, or focussing. All that happens is that the fuel and/or makeup flows going upwards past the column tip where it sits in the flame tip stop the column effluent from diffusing backwards into the dead space below the flame tip, which would cause peak tailing.

Peter

[rb6banjo]

As Bigbear said above, the FID's were originally constructed for packed column use. That means very large flows. To use smaller bore columns without redesigning the FID, our chromatographic forefathers had the very good idea to add a makeup gas flow at the column outlet to make the FID think that it still had a packed column feeding it. The makeup gas functions to sweep the column effluent into the flame as rapidly as possible so that you don't get broadening of your chromatographic bands (your "dead volumes"). Makeup gas is very important as you go to smaller and smaller bore columns. You might be able to get by without it on a 0.53 mm i.d. column (I don't, I use makeup gas all of the time and I just use carrier gas, helium or nitrogen) but I would not think to go without it for anything smaller than that.

As Peter stated, it's more critical to keep your fuel/oxidant ratio correct so that your flame burns nicely. If you use H2 as your makeup gas, be sure to adjust it so that makeup+H2 (from the source) is optimal relative to the air.

[Peter Apps]

Hi Southwest

Does your flame tip look similar to this ... http://www.chem.agilent.com/en-US/produ ... -80560.gif ?

and is your column installed with its tip inside the tube below the screw thread ?

If so, you can work without make-up because the flow of fuel hydrogen up through the tube around the column tip and into the flame is enough to flush the column effluent into the flame with no dead volumes or band broadening. The flame will generate ions a little less efficiently, but you say that you can live with that.

Peter

[CE Instruments]

As Bigbear said above, the FID's were originally constructed for packed column use. That means very large flows. To use smaller bore columns without redesigning the FID, our chromatographic forefathers had the very good idea to add a makeup gas flow at the column outlet to make the FID think that it still had a packed column feeding it.

A very good idea? That would be those who were too lazy to make an FID work for capillary
The Carlo Erba designed an FID that does not require make up and works fine with standard columns. There is merit in the use of Nitrogen make up as a Hydrogen flame burns better with a higher ratio of Nitrogen than in standard air so it is an option to add make up but most of the GCs I sold were without.

[Peter Apps]

As Bigbear said above, the FID's were originally constructed for packed column use. That means very large flows. To use smaller bore columns without redesigning the FID, our chromatographic forefathers had the very good idea to add a makeup gas flow at the column outlet to make the FID think that it still had a packed column feeding it.

A very good idea? That would be those who were too lazy to make an FID work for capillary
The Carlo Erba designed an FID that does not require make up and works fine with standard columns. There is merit in the use of Nitrogen make up as a Hydrogen flame burns better with a higher ratio of Nitrogen than in standard air so it is an option to add make up but most of the GCs I sold were without.

Indeed, the first capillary columns (which were made of metal) were connected to FIDS that had 1/8 or 1/4 inch fittings to take packed columns, and huge internal volumes that were flushed with carrier gas coming from the packed columns that they were designed for. Adding auxillary makeup gas was a stopgap measure to save rejigging the FID hardware. Also the flame tips had wide holes in them to accommodate the high gas flow rates - carrier gas in those days was nitrogen or argon so diffusion was much slower. With slower gas flows from a capillary column the peaks were broad and tailed. Some manufacturers produced adapters that converted packed FID plumbing to capillary connections, especially when glass columns came along. Carlo Erba redid its FID plumbing in the Fractovap model to produce a capillary GC that did not need converters or makeup.

Adding nitrogen makeup to a Fractovap increased peak heights by about 30% but increased noise as well, so the net gain was rather small.

For most applications (after all who uses FID for trace analyses these days ?) FID makeup is a historical holdover from packed column days. Other "living fossils" that people have got used to are the 1 ml default loops in automated headspace samplers, and inlets that have evolved remarkably little from their origins as the top 3 inches of a 1/4 inch packed column.

Peter