RSD failure

[Murshad]

I am in a strange problem during Headspace following USP method IV. During a system suitability first three injection are consistent but after that there is a sudden decrease in area counts which causes system suitability to fail.
Chromatographic conditions are

GC HP5890 II
Column DB-624 30m x 0.53mm x 3.0um
Injector 140°C
Detector 260°C
Oven 40°C for 20min then increase at 70°C/min to 240°C and hold at 20min
Injector is combipal

My colleague and I have spent couple of days. First we used ZB 624 column which previously worked well. I have considered all options , changing the syringe for combipal, changing line, o ring for liner ,column but not able to trace the problem. I have also checked the flow-rate manually for any leakage which seems to be not there. Gold seal have bee checked about a week ago by a certified technician and this injector has not been much used. Being a contract lab employee I cant spend much time on it . Would somebody help me out? It is very rush issue for me.
[/b]

[Schmitty]

Is this the USP method for OVIs? I tried setting that up on our combi, and it didn't work either

We are still doing them by hand injection

Maybe if I get some time I will try this again. BTW, our (2) combipals work excellently for all other assays.

[Murshad]

Yes it is USP method IV. Fortunately it workef couple of times in past.

The way I am making standard by adding

9.1ul of dichloromethane
0.8ul of chloroform
1.0ul trichloroethylene
7.5ul 1,4-dioxane

into water and diluting to 1000ml with water, shaking well and sonicating for at least 30 mints.

Couple of injections go well and then change in area count which causes RSD failure.

Unfortunately I cant deviate from USP method

Thanks a lot for reply

[Schmitty]

Are you heating your samples consistently before injection?

[chromatographer1]

I have found that most headspace errors are generally found in the procedure used in preparing samples and standards rather than in hardware failures, if a reproducibility of the hardware is initially demonstrated.

Look at the exposure time of your standard solution to the air. Do any of the vials remain uncapped for a longer period of time than the others?

I found it convenient to place my OVI's into methanol, DMF, or propylene glycol before adding a small amount to the water. Also, allow for zero headspace in the std solution vial or you will lose some of the OVIs from the solution. I spike samples and stds using a syringe.

Sonication is not usually a good idea in my experience in preparing volatile stds. Also be aware that vial lips are not always consistent and that you can have losses from inconsistent or inappropriate crimping of the septum caps.

I have prepared standards ranging from 1mg/mL to 1µg/mL with R-squared = 0.999+ accuracy for all the USP OVIs as well as 18 other solvents many times.

Look at your preparation procedure and see if the error lies there. As others have noted heating issues can also cause errors.

best wishes,

Rod

[Murshad]

First I want to say thanks to all folks for their replies. It is USP method IV which requires heating of sample at 80°C for one hour. I have also considered the issue of vials which I have, in my opinion almost eradicated.

Now I have deviated from the method a bit. I have prepared first dilution of standard in DMSO and then 0.1ml of this to 100ml with water.

I injected 8 injections for standard but RSDs for methylene chloride, chloroform, tri-chloroethylene were 17%, 15%, and 14% resp. while for 1,4-dioxane it was 38%. When I omitted first three vials it became well within 3% for all analytes. Before sequence I did two trial runs. These trial and first three of sequence run were not consistent. After that it became consistent.

This problem is from the first day. Couple of injections go up and down either in beginning or in middle of run.

Please don’t consider this inconsistency in sample preparation. This is an USP method, being in a quality control lab we cant deviate from USP.

Would someone have done this method?
Thanks

[Peter Apps]

Hi Murshad

Please post ALL the operating settings for the combi pal - syringe temperature, purge time and temperature, injection speed etc etc.

Your problem seems to have reversed itself - first you had three consistent runs and then a drop in area, now you have consistency only after the first three runs. Please post the actual peak areas.

Ar you doing replicate injections from a single vial, or a series of different vials ?

Peter

[Murshad]

Dear Peter

Following were the parameters for Combipal

Sample vol 1ml
Incubat temp 80°C
Incubat time 1hr
Syringe temp 95°C
Fill speed 300ul/s
Inj spedd 300ul/s
Agi speed 300rpm
Syringe flush time 4min

Thanks

[chromatographer1]

What kind of seals are you using?

You should be using non-aluminized septa, preferably teflon lined silicone.

best wishes,

Rod

[Murshad]

Seals(caps) I used are 20mm magnetic crimp cap, LO w/PTFE/Sil Liner

supplied by Canadian Life Science.

[Peter Apps]

Hi Murshad

Using the Gerstel clone of the the Combipal I found that injection speed influenced peak area repeatability. 300 microlitres per second is 20 ml/min. You do not say whether you are operating split or splitless, or what your carrier flow rate is, but you may be getting a pressure pulse on injection that disrupts lows through the split and/or the septum purge. It might help to lower the injection speed to match the total carrier gas flow rate.

I also found that the sample temperature control was none too good. If you run consecutive series of six vials, and plot peak area against run number, do the plots look the same ?

Do all the peaks on the chromatogram change in the same way ?

We still need to see some hard data.

Peter

[Murshad]

Dear Peter

Thanks for reply. 300ul/sec I am using just because at this speed it has been done in past. Carrier flow is 34.5cm/sec and it is splitless injection. The only reason I am using 80°C incubat temp because USP method tells me to do this. I have already did one deviation ie method says to keep ur injector at 70°C while I am keeping it at 140°C. I ran eigth standard vials two blanks, one sample and one standard at the end of sequence. First three vials were inconsistent. While plot for remaining five std vials and last std vial looked same.
All the peaks in chromatogram change in way that if one goes up others go up too or vice versa.

[Peter Apps]

Hi Murshad

At 35 cm/s linear flow the volume flow into the column is about 6 ml/min. You very probably have a pressure pulse during an injection at 20 ml/min, and an inconsistent portion of your sample is escaping out of the septum purge.

Unless you can provide the other information that I asked for there is nothing else that I can suggest.

Peter

[Murshad]

Hi Peter

Thanks for all your help in resolving this issue. I am off from my work for some time. This is the only information I have.

Can u give me some suggestion on injection speed. Should I change this in future because this is the only parameter I can play.

All other temperatures and parameter I have to follow because of USP procedure.

Thanks a lot again

[Peter Apps]

HI Murshad

Set the injection speed to match the carrier gas flow rate.

There could be other problems with your hardware. When you get back to work post a message and we can work through some trouble shooting.

Peter