Chromatography Forum

this is very confusing! i ran std3 right afterwards, and am starting to get peaks again, but its inconsistent:
the y-axis is about 40000 in the following chromatographs

std3 - first run:

def seeing some peaks, but the solvent tail is still making it difficult.



=========================================
Area Percent Report
=========================================
Data File Name : C:\HPCHEM\1\DATA\NV-F0137.D
Operator : Page Number : 1
Instrument : ANALYZER Vial Number :
Sample Name : Injection Number :
Run Time Bar Code: Sequence Line :
Acquired on : 16 Aug 07 06:44 PM Instrument Method: 6584-100.MTH
Report Created on: 16 Aug 07 07:16 PM Analysis Method : 6584-100.MTH

Sig. 1 in C:\HPCHEM\1\DATA\NV-F0137.D
Pk# Ret Time Area Height Type Width Area %
|---|----------|--------------|--------------|----|---------|----------|
1 1.345 2606 165 BV 0.198 0.0003
2 1.567 86 17 VB 0.063 0.0000
3 1.746 7818 5836 BV 0.022 0.0008
4 1.949 8.99766E+008 1.31258E+008 VV 0.085 94.5063
5 2.004 5.0963E+007 5806131 VV 0.106 5.3529
6 4.094 962430 22965 VV 0.496 0.1011
7 5.645 230977 14254 VV 0.199 0.0243
8 9.458 41810 4701 VV 0.148 0.0044
9 9.525 19902 5259 VV 0.063 0.0021
10 9.541 59716 5182 VV 0.192 0.0063
11 17.090 1754 235 BV 0.093 0.0002
12 17.151 1039 265 VV 0.055 0.0001
13 17.214 621 238 VV 0.034 0.0001
14 17.328 1642 244 VV 0.087 0.0002
15 17.358 198 202 VV 0.016 0.0000
16 17.458 1061 170 VV 0.103 0.0001
17 17.704 1371 94 VV 0.184 0.0001
18 17.765 429 139 VV 0.040 0.0000
19 17.805 328 181 VV 0.030 0.0000
20 17.980 2333 350 VV 0.083 0.0002
21 18.043 1287 406 VV 0.053 0.0001
22 18.076 764 440 VV 0.029 0.0001
23 18.117 1115 402 VV 0.043 0.0001
24 18.156 831 374 VV 0.031 0.0001
25 18.199 273 269 VV 0.017 0.0000
26 18.220 562 240 VV 0.033 0.0001

Total area = 9.5207E+008

------------------------------------------------------------------------------

std3 - second run:

where did the peaks go?



======================================
Area Percent Report
======================================

Data File Name : C:\HPCHEM\1\DATA\NV-F0138.D
Operator : Page Number : 1
Instrument : ANALYZER Vial Number :
Sample Name : Injection Number :
Run Time Bar Code: Sequence Line :
Acquired on : 16 Aug 07 07:27 PM Instrument Method: 6584-100.MTH
Report Created on: 16 Aug 07 07:59 PM Analysis Method : 6584-100.MTH

Sig. 1 in C:\HPCHEM\1\DATA\NV-F0138.D
Pk# Ret Time Area Height Type Width Area %
|---|----------|--------------|--------------|----|---------|----------|
1 1.992 20196 504 BV 0.474 0.0042
2 2.266 4.62873E+008 8.8122E+007 VV 0.066 95.3902
3 2.339 2.22797E+007 2979198 VV 0.118 4.5915
4 8.584 8414 737 BV 0.190 0.0017
5 8.631 1578 752 VV 0.028 0.0003
6 8.694 3491 1025 VV 0.057 0.0007
7 8.811 8327 1235 VV 0.085 0.0017
8 8.858 3065 1340 VV 0.032 0.0006
9 8.924 5311 1423 VV 0.053 0.0011
10 8.966 2753 1438 VV 0.027 0.0006
11 8.988 2177 1586 VV 0.023 0.0004
12 9.006 3182 1566 VV 0.028 0.0007
13 9.074 7852 1662 VV 0.059 0.0016
14 9.173 8008 1610 VV 0.071 0.0017
15 9.247 2831 1421 VV 0.033 0.0006
16 9.266 6916 1366 VV 0.065 0.0014
17 9.380 3846 991 VV 0.065 0.0008
18 9.529 814 393 VV 0.029 0.0002

Total area = 4.85241E+008

=======================================

std3 - third run:



===============================================================================
Area Percent Report
===============================================================================

Data File Name : C:\HPCHEM\1\DATA\NV-F0139.D
Operator : Page Number : 1
Instrument : ANALYZER Vial Number :
Sample Name : Injection Number :
Run Time Bar Code: Sequence Line :
Acquired on : 16 Aug 07 08:09 PM Instrument Method: 6584-100.MTH
Report Created on: 16 Aug 07 08:41 PM Analysis Method : 6584-100.MTH

Sig. 1 in C:\HPCHEM\1\DATA\NV-F0139.D
Pk# Ret Time Area Height Type Width Area %
|---|----------|--------------|--------------|----|---------|----------|
1 1.243 9879 495 BV 0.269 0.0011
2 1.734 9.32837E+008 1.44431E+008 VV 0.081 99.8959
3 4.543 425222 13308 VV 0.383 0.0455
4 5.264 216745 6668 VV 0.389 0.0232
5 6.099 137964 6176 VV 0.290 0.0148
6 6.865 27234 1132 VV 0.301 0.0029
7 13.038 8469 338 PV 0.327 0.0009
8 13.128 752 252 VV 0.049 0.0001
9 13.181 1227 215 VB 0.095 0.0001
10 15.982 8772 386 BV 0.279 0.0009
11 16.159 259 171 VV 0.025 0.0000
12 16.215 568 123 VB 0.062 0.0001
13 17.413 53900 2480 BV 0.259 0.0058
14 17.797 6859 905 VV 0.101 0.0007
15 17.872 5444 1018 VV 0.082 0.0006
16 17.954 3941 958 VV 0.054 0.0004
17 18.022 3132 878 VV 0.048 0.0003
18 18.349 42734 1814 VV 0.287 0.0046
19 18.755 9241 956 VV 0.121 0.0010
20 18.848 7974 725 VV 0.183 0.0009
21 19.176 977 277 VV 0.059 0.0001
22 19.345 456 223 VV 0.033 0.0000

Total area = 9.33808E+008

==================================


I'm not sure what causes the lack of repeatability, all of our injections from day 1 last week showed pretty sharp peaks. I wonder if injection technique is to blame.

for all but the last run i drew 1 uL into the syringe and injected it. for the last run above run, I drew in 5 uL, and injected it from 5 down to 4 uL. in all cases i washed the syringe with at least three full 5uL shots of pure heptane immediately after injecting.

I guess my major questions are:
1) any clue as to why am i seeing such lack of repeatability?

2) looking at the 'good' chromatograph above, would you say the column seems to be functional at least?

3) how can i reduce the solvent tailing? replace the guard column and/or clip a meter off the front of the actual column?

4) last night i 'baked' my column at 350C for an hour (sorry bruce!, i started right before your post, and i figured 350C isnt too hot). over the baking time, the high temperature baseline decreased pretty steadily. This morning i replaced the injector septa, and then checked my carrier gas flowrate. The flowrate had substantially decreased. Before we were using 6-8 psi on the EPC to get 3 mL/min carrier gas flow. I had to bump it up to over 14 psi to get 3 mL/ minute this morning! Does that tell us anything about the condition of the column?

5) what is the odd peak in the last chromatagraph above. This is well after the run is over when the GC is cooling down. Again, does it tell us anything about the condition of the column?

thanks everybody for all your advice!

this is very confusing!

I guess my major questions are:
1) any clue as to why am i seeing such lack of repeatability?

2) looking at the 'good' chromatograph above, would you say the column seems to be functional at least?

3) how can i reduce the solvent tailing? replace the guard column and/or clip a meter off the front of the actual column?

4) last night i 'baked' my column at 350C for an hour (sorry bruce!, i started right before your post, and i figured 350C isnt too hot). over the baking time, the high temperature baseline decreased pretty steadily. This morning i replaced the injector septa, and then checked my carrier gas flowrate. The flowrate had substantially decreased. Before we were using 6-8 psi on the EPC to get 3 mL/min carrier gas flow. I had to bump it up to over 14 psi to get 3 mL/ minute this morning! Does that tell us anything about the condition of the column?

5) what is the odd peak in the last chromatagraph above. This is well after the run is over when the GC is cooling down. Again, does it tell us anything about the condition of the column?

thanks everybody for all your advice!

This is what I was expecting. Failure to identify whether the instrument parameters are set correctly will cause serious heartache, and bring in a whole range of "maybes".

My answers, others will differ.
1. You're confronting multiple issues, but not being systematic is solving them.

2. No. The peaks should be really sharp. The best way to establish whether the column is still OK is to perform the CoA test. Frankly, it looks like a 1970s packed column on a bad day. I suspect your column is OK, and doesn't need continued reconditioning - the MSTFA can help move some junk off the columns on each run. The danger with conditioning is that you move the rubbish from the guard to the main column.

3. The tailing can be caused by overload, activity on guard column, activity on main column, poor injection technique, leaking syringe, Incorrect syringe needle, incorrect carrier gas flows, incorrect fitment of the column into the injector or detector, incorrect joining of the two columns, incorrect detector gas flows, incorrect temperatures, and probably a lot more reasons.

4. No. I've never worked with EPC, but I suspect it tells us that you may have either a partial blockage or, more likely, a leak on you injector or column connection. Perhaps your needle dislodged some junk. Pressure tells us nothing about the condition of the stationary phase unless you have injected so much solvent onto the column that the phase swells and blocks the column, very, very unlikely with a 0.32mm diameter column. If there are no leaks, I would separate the columns and measure the back pressure of each on the injector.

5. No. The extra peak could be some of the sample eluting because you have leak and the flow through the column is really slow. Alternatively, as the system cools, a fresh slug of injected material could be pushed through the column as the flow changes ( perhaps leak seals fora while ). There are many other possible reasons....

************************

My suggestion would be to take a step back, and check and verify all instrument parameters, especially carrier and detector gas flows ( measure them at the detector or column effluent using a soap film meter ), and also check for leaks, particularly around the injector.
This should be a systematic process, ensuring that all flows go where expected, and that no leaks occur.

Next, confirm that your injection technique is not at fault.
The quickest way may be to perform replicate injections of a simple mixture of n-alkane hydrocarbon ( any one between nC15 - nC30 should be OK ) in toluene or iso-octane solvent solvent. The peak should be really sharp. You should follow the recommendations for the pace of the on-column injection.

Start with a relatively-high concentration ( 0.1 mg/ml ), and reduce concentration, and then perform replicates. That will identify your injection precision. If time's an issue use simple isothermal conditions with 10 minutes run time.

Once you know your injection precision, add some of your n-alkane to the MSTFA/pyridine/toluene ( but no sample or standards ) and repeat the test. The peaks should have similar sharpness and precision. If not you may have to look at the guard column, injection syringe, injector cleanliness.

Then perform your derivatisation of the internal standards, but use the IS at 5x nominal concentration, injecct multiple times, as see if the peak areas increases. Once they are stable, prepare and inject at nominal concentration. Repeat, adding sample...

You really need to be systematic, because there usually are many reasons for most problems in chromatography. You can randomly try various things, and you might get lucky....

Please keep having fun,

Bruce Hamilton

Golden rule number three of trouble-shooting is never test one doubtful thing with another doubtful thing. You are not sure; 1 that the column is working, 2 that the standards are OK, 3 that the reagents are OK. Derivatising doubtful standards with doubtful reagents and injecting onto a doubtful column is a waste of time.

Make up some alkanes or other hydrocarbons in hexane. Inject some of that. Do you see sharp peaks ? If so the guard column and the analytical column are not grossly contaminated, and then it is worthwhile injecting derivatized standards, getting new suppplies of standards and reagents and trying them etc.

What is the reason for not changing the guard column ? - it is both quicker and cheaper than a new analytical column, or new chemicals.

Peter

JF,

I agree with Bruce and Peter about isolating the problem. I really agree with running an alkane mix as a good starting point. I would also suggest simply removing the guard column for the purposes of running the alkane mix, then if that works out, return to guard column. I would also add that you are running your injection port right at the boiling point of your solvent which will make a mess of the flash in the injector and not help you solvent peak shape at all. Colder or hotter, either one, but right at the boiling point is worst case. Finally, you know you have an inlet problem, with the EPP errors and flow control errors, so you really need to give that a good look over first. Again, start with a simple analysis (alkanes) on a simple system (no guard) and work from there.

Best regards.

thanks guys, based on the advice above I will be proceeding in the following order

1) procure new column and a few guard columns, to have on hand

2) make up some solution of hydrocarbon in heptane (can anyone suggest a simple, commonly available hydrocarbon that doesnt require derivitization and will give me a fairly narrow peak? I'm thinking diesel fuel, but as a distillate fuel, it may have a broad range of peaks. Dont want to spend money on reagent grade alkane mixtures unless i have another use for them. I see that sigma aldrich has a handful of standard alkane solutions for GC, would this be worth having on-hand?)

3) disassemble/inspect the inlet, check for leaks according to the manual. Recheck column flows after repairing leak

4) inject some blank heptane runs at 80C to see if the inlet work has affected the peak shape. if not, try removing guard column and injecting blank heptane runs. this should tell me something about contamination of the guard column

5) try injection of hydrocarbon in heptane

6) when comfortable with #5, try injection of derivatized internal standards, at 5X concentration

general questions:

1) how does solvent weight and polarity play into this?
our standards are in pyridine, and our samples are dissolved in heptane prior to injection. I have seen discussion of hexane, toluene, and iso-octane solvents here. were these selected for polarity? For example i imagine hexane/heptane/toluene and pyridine/isopropanol/ethanol would be interchangeable respectively. Is there any other reason one solvent is better than another for this purpose (we dont want to order/store extra solvents unless for a good reason). We also happen to have non-reagent grades of isopropanol, ethanol, acetone and xylene around here.

2) i assume the purpose of the hydrocarbon injection is to eliminate derivitization from the variables. in simple terms, what does derivitization do, and why wouldn't i be able to simply see a peak of non-derivatized esters and glycerides?

3) thoughts on the question about common hydrocarbons vs. alkane standards above?

thanks everybody!

2) make up some solution of hydrocarbon in heptane (can anyone suggest a simple, commonly available hydrocarbon that doesnt require derivitization and will give me a fairly narrow peak? I'm thinking diesel fuel, but as a distillate fuel, it may have a broad range of peaks. Dont want to spend money on reagent grade alkane mixtures unless i have another use for them. I see that sigma aldrich has a handful of standard alkane solutions for GC, would this be worth having on-hand?)

3) disassemble/inspect the inlet, check for leaks according to the manual. Recheck column flows after repairing leak

4) inject some blank heptane runs at 80C to see if the inlet work has affected the peak shape. if not, try removing guard column and injecting blank heptane runs. this should tell me something about contamination of the guard column

5) try injection of hydrocarbon in heptane

6) when comfortable with #5, try injection of derivatized internal standards, at 5X concentration

general questions:

1) how does solvent weight and polarity play into this?
our standards are in pyridine, and our samples are dissolved in heptane prior to injection. I have seen discussion of hexane, toluene, and iso-octane solvents here. were these selected for polarity? For example i imagine hexane/heptane/toluene and pyridine/isopropanol/ethanol would be interchangeable respectively. Is there any other reason one solvent is better than another for this purpose (we dont want to order/store extra solvents unless for a good reason). We also happen to have non-reagent grades of isopropanol, ethanol, acetone and xylene around here.

2) i assume the purpose of the hydrocarbon injection is to eliminate derivitization from the variables. in simple terms, what does derivitization do, and why wouldn't i be able to simply see a peak of non-derivatized esters and glycerides?

3) thoughts on the question about common hydrocarbons vs. alkane standards above?

thanks everybody!

Do not use diesel - ever!. It seldom is dominated by n-alkanes these days, and there are copious quantities of additives. If you can find a special boiling point alkane solvent in the 200C plus region, a light paraffin oil, or a high flashpoint, low aromatics, lighting kerosine, you could use those. The best idea is to contact the lab at a local oil refinery and ask if you can purchase a small quantity of one of their standards, also most chemical departments of univeristies should have at least one n-alkane in the C15 - nC30 region.

If you want to quickly buy a mix, you can, but I'd just buy a small quantity of the liquids (<c16) and/or solids (>c16) from a supplier.
You need a mix of n-alkanes to check to see if there is discrimination ( more volatile compounds give higher % than actual ), but as you are using cool on-column, that's probably not an issue we need to worry about... yet.

I endorse the suggestion to remove the guard for testing the hydrocarbon mix, and also adding it and testing with the mix, but put it back before you use the derivatising agent injections.

In answer to your questions,

1. n-heptane ( and make sure you purchase n-heptane with a very low non-volatile residue, and don't use heptane fraction from petroleum) boilis at 100, so does Iso-octane, and toluene boils at 110C. Boiling point and purity are critical, and ofyten iso-octane ( aka 2,2,4-trimethylpentane ) is easier to obtain at the preferred purity.

Toluene, being an aromatic has greater ability to dissolve junk, but will tail more. I use toluene to ensure materials that may have a trace of water stay in a single phase. I would not use n-hexane, the boiling point is too far away from you n-heptane.

My apologies, I didn't recall that you were using n-C7 diluent, and mentioned toluene, as that's what I used to use in the 1980s when I was doing Biodiesel. Use the solvent specified in the method.

2. The lipid components are thermally-reactive and stick to the column, and polymerise. The derivatisation only affects the glycerol, and MAG and DAGs, it doesn't affect the TAGS or FAMEs, which elute unchanged. The MAGs and DAGs will immediately die in a hot injector or guard column to form polymeric gunk that quickly causes tailing of all peaks - they will not elute. The MSTFA is in excess to help prevent that, and makes the molecules much more volatile, as the best solution is lowest exposure to temperature.

3. refer my response above.

Please keep having fun,

Bruce Hamilton

JF
Restek sells a hydrocarbon retention time mix that would ideally suit your needs. Only a couple of mL's so if you don't re-cap it it won't last very long. Generally, this would also give you a hydrovarbon range to look at.

Best regards.

i took apart the inlet, there was a decent-sized chunk of septa lodged in the top of the column.

we are running heptane blanks right now,first run showed some vaporization of some heavy components similar to what we would expect with our standards. Second run was pretty flat. Initial solvent peak is a lot sharper, I'm hoping were in the clear, and it was a valuable learning experience!

The Sigma-Aldrich product manager indicated that he often sees large, elevated baselines with inexperienced biodiesel producers. This can be attributed to two sources:
1. Incomplete derivatization of the sample with MSTFA. This can be caused by:
a. using the wrong MSTFA reagent;
b. not waiting for the reaction to complete;
c. waiting too long after preparing the samples;
d. excess water or alcohol in the sample that uses up the MSTFA.

Elevated baseline due to incomplete derivatization can be overcome by baking out the column.

2. Poorly made biodiesel that contains large amounts of unreacted vegetable oil. This problem cannot be overcome and the column and precolumn must be replaced.

He also provided the following recommendations:

Calibration standards and samples should all be prepped at the same time, using the same lot of derivatization reagents.
The MSTFA derivatization reagent should be purchased in 1-ml ampoules, and fresh reagent should be used every time you prepare standards and samples. The only MSTFA customers should use is "Sigma-Aldrich Derivatization Gradeâ€

excellent advice giacomo

I will purchase mstfa in 1 ml vials from now on.
I might try to put the 23 mL of mstfa we have left into ampules anyway.
We did end up getting good correlation with our standards (>99% on all but triolein, which is obscured due to column bleed at the end of the run. Our next column will be a high-temp metal column, which should resolve this)

the one question i have is with your last statements

The pre-column will need to be changed at least every 100 injections, and perhaps more frequently than that.
You will need to use new ferrules on the pre-column every time you change the pre-column.
The analytical column will last AT MOST for 100 samples, and maybe for far fewer, depending on the quality of the biodiesel.

i can see changing out the guard column every 100 or so injections, using tailing of initial solvent peak as an indicator.

But surely, with a good guard column in place, we would expect to get more than 100 samples out of a gc column? I'm guessing 90 of the samples we inject will be astm-quality (<0.25% free and total glycerin). I find it hard to beleive that the labs with an autosampler that run their columns 24/7 are changing out their columns every week.

Why 100?

Determine the need to remove a foot of guard column by the performance of the system.

I have seen systems used for hundreds of injections

My experience is somewhat different, admittedly I haven't performed more than a few hundred injections of Biodiesel FAMES with MSTFA, preferring acetylation, and it's well over a decade since I was processing thousands of biodiessel FAME samples, however...

MSTFA stores OK in larger volume containers, provided you have good processes when removing reagent. Typically, the bulk containers are about 50% of the price per ml in ampule, once you get to 20 mls. Work out your usage per month, and decide whether potential $ savings justify the effort of having good reagent protocols. Just ask any Organic synthesis chemist for details, or look at the information provided by suppliers like Pierce.

As you are analysing products from a defined process, you should get excellent life from both the precolumn and analytical column. I'd expect at least 300 samples from a precolumn, and 1000 sample from the main column.

Anything less than those numbers, then I'd be looking very closely at the sample preparation, and especially the presence of moisture or junk in the samples.

I'd also be very, very wary of trying to bake out the column to reduce baseline caused by incomplete derivatisation. If the derivatisation is imcomplete, you will have active MAGS and DAGs that will polymerise. The solution is to change the precolumn, and fix the derivatisation problem.

Once again, the standard method is known to work reliably and consistently, and if you are going to report results to that method, then you must follow the protocols carefully - they work.

Please keep having fun,

Bruce Hamilton

Bruce's comments are on the mark.

Reagents:

When in constant use I would use 100mL bottles of silylating reagents and consume the entire bottle without problems.

Columns:

Biodiesel column customers report column life of high temperature metal capillary columns ranging from 3 to 6 months performing 24/7 final product biodiesel analyses.

Excepts occur when a leak occurs at the union of the guard and analytical columns. The column will degrade VERY quickly if this happens, it takes only a few hours.

There are several quality columns on the market. I have noted a few of these previously (Varian, Restek) as well as SGE's high temperature polyimide low bleed FSOT column which I have seen endure the high temperatures required much better than other FSOT columns. And there are others from Japan and elsewhere.

best wishes,

Rod

we havent seen any problems with our MSTFA yet, I'm always careful to pipette out just enough for the days needs, and reseal the main bottle with a helium purge.

Is there an easy way to tell if the column/guard column has a leak? since it is a mechanical connection, i imagine its not trivial to seal it 100% every time. Is there a preferred type of connector that works better than any other?

Bruce's comments are on the mark.

Reagents:

When in constant use I would use 100mL bottles of silylating reagents and consume the entire bottle without problems.

Columns:

Biodiesel column customers report column life of high temperature metal capillary columns ranging from 3 to 6 months performing 24/7 final product biodiesel analyses.

Excepts occur when a leak occurs at the union of the guard and analytical columns. The column will degrade VERY quickly if this happens, it takes only a few hours.

There are several quality columns on the market. I have noted a few of these previously (Varian, Restek) as well as SGE's high temperature polyimide low bleed FSOT column which I have seen endure the high temperatures required much better than other FSOT columns. And there are others from Japan and elsewhere.

best wishes,

Rod

All butt connectors can fail under the strenuous thermal stress the connection undergoes during the oven temperature programming involved in the biodiesel analysis.

Some ideas:

Certainly any inert connector that uses graphite ferrules can be additionally tightened.

Using quartz connectors with polyimide glue might be an adequate solution.

I have no definitive solution. Many connecting system should work, can work, but certainly all CAN fail at one time or another.

Metal and FSOT columns may require different solutions for a leak-free connection.

Please share with the Forum and with me any special solutions you find work for you.

best wishes,

Rod