gc troubleshooting

SUPELCO

Capillary GC Troubleshooting:A Practical Approach

Gas Separations R&DKatherine Stenerson

Supelco, Supelco Park, Bellefonte, PA 16823

SUPELCO©1999 Sigma-Aldrich Co.

Outline

❚ Basic Troubleshooting Strategy❚ Preventing Problems❚ Identifying Common Problems❚ Recommended Reading❚ Discussion


Troubleshooting Strategy

❚ Have appropriate equipment andsupplies on hand.

❚ Establish a systematic approach.❚ Know what to look for first.❚ Record what you did to correct the

problem.


Suggested Equipment to Have onHand for Troubleshooting:

Electronic Leak DetectorFlow Meter“Test” ColumnReplacement Accessories (Syringes, Ferrules, Septa)Replacement Purifiers


Five Major Sources ofChromatographic Problems:

❶ Operator Error❷ The Sample❸ The Column❹ The GC Electrical System❺ The Gas Flow System (both internal and

external to the GC)


Approaching the Problem

❚ Check first to see if a “fix” for theproblem is already known.

❚ Check the Supelco Capillary GCTroubleshooting Guide (Bulletin 853.)

❚ Check back in the instrumentmaintenance record.

❚ Talk to others in your lab.


Check operating parametersRun Ref. Std.

OK Not OK

OKNot OK

Problem was sample related

Install Test

Column

OK Not OK

Problem was

Column related

Cap off Detector

OK

Not OK Problem is in the Detector

Problem is in the Inlet or with the carrier gas

Isolate the Source of the Problem

Correct the parameter


When reviewing operating methodparameters consider the following:

? Is my starting temp. low enough to allowsufficient sample focusing?

? For splitless injections, is my splitter openingat the appropriate time?

? Is my column flow set to give me maximumefficiency at the most critical point?

? Are heated zones (injectors, detectors,interfaces) set appropriately?


The Best Way to Solve Problemsis to Prevent Them!

❚ Install and maintain proper purificationfor all gases in the GC system.

❚ Maintain the injector by periodicallyinspecting and changing the liner,septa, and seal (H/P™.)

❚ Use the proper injection technique-thisincludes using the right liner for the job.

❚ When necessary, use a guard column toprotect the analytical column.


Gas Purification

❚ Carrier Gas❙ At minimum, remove hydrocarbons, water, and oxygen.

❚ Hydrogen (FID)❙ At minimum, remove hydrocarbons.

❚ Air (FID)❙ At minimum, remove water and hydrocarbons.

❚ Nitrogen make-up (FID, ECD)❙ At minimum, remove hydrocarbons.

❚ P-5 make-up (ECD)❙ At minimum remove hydrocarbons (especially halogen-

containing), oxygen.


Acceptable Purity Levels forChromatography Grade Gases

Impurity / Maximum Concentration

Gas O2 H2O CO2 COTotal

Hydrocarbons

Helium <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppmNitrogen <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppmAir 20-22% <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppmHydrogen <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppmArgon/ Methane <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppm <1.0 ppm


Suggested purifiers

Hydrocarbons Water Oxygen

Carrier Supelcarb HCSupelpure HC

Mole Sieve 5A OMI -2

H2

Air Mole Sieve 5A

N2 makeup

P-5 OMI -2 OMI -2

P-5


What are some signs that mypurifiers need to be changed?

Hydrocarbon TrapsNoise in the baseline (FID)

Increase in backgroundpeaks on tune (MSD)

Higher than normalbaseline reading on FID

Extra peaks visible in run

Molecular Sieve 5AIncrease in column bleed

Water visible in MSbackground

Poor peak shapes forgaseous VOCs (purge andtrap)

Extra peaks visible in run

OMI™-2 color change


Injector Maintenance

❚ Change (as needed):1. Liner and O-ring*

2. Seal and washer **H/P™ GCs

❚ Inspect the inlet periodically-Look for contamination in the liner-Look for residue on the seal


Using the right liner and injectiontechnique can also prevent problems:

❚ Split Injection❙ used for concentrated samples❙ high flow of carrier gas through liner

during injection❙ should use liner designed for split injection

❚ Splitless Injection❙ used for trace analysis❙ low flow of carrier gas through liner during

injection❙ inertness and internal volume of liner used

are critical


Some liners used for splitinjection:

Cup (unpacked)

Cup (wool packed)

Split/splitless- wool packed


Some liners used for splitlessinjection

2 mm ID, straight

Dual-tapered

Single-tapered


If you must clean a liner….

❚ Handle liners with gloves or forceps.❚ Use clean compressed gas and/or a fine

brush to remove particles.❚ Rinse liner in an appropriate solvent

and dry with clean compressed gas.❚ Use mineral acid and/or detergent only

if absolutely necessary.


Using a Guard Column

❚ Choose a guard column which has beendeactivated.

❚ Usually, the ID of the guard matchesthe analytical column.

❚ A 5-10 meter length is normally used.❚ Connect with either a GlasSeal™ or butt

connector.


Common Problems

1 Poor Peak Shapes (either tailing,fronting, or just generally ugly.)

2 Nonlinearity3 Baseline Noise and /or Drift4 Ghost Peaks5 Missing Peaks / Poor Response6 Insufficient Resolution


1. Poor Peak Shapes

❚ Fronting can indicatecolumn overload.

❚ Tailing can indicateactivity in the systemor improper columninstallation.

||

||


1. Poor Peak Shapes (cont.)

❚ Generally ugly peaks, suchas a,a-dimethylphenethylamine,can be caused by a varietyof problems.


➨ Column overload➨ Detector overload➨ Standards preparation➨ Poor peak shape resulting in improper

integration

2. Nonlinearity

The most common causes are:


An Example of Overload:

Peaks have broad bases,fronting on last few visible inrun.


Preventing column overload forsplitless injections:

❚ Inject a smaller amount / use a 1 ulsyringe.

❚ Use a thicker film column.❚ Use a column with a wider ID.❚ Decrease upper limit of calibration

range.❚ Use a column of slightly different

polarity.


An example of poor peak shapeaffecting linearity:

❚ Benzoic acid istypically of poorshape when doingsplitless injections.


3. Common causes of baselinenoise / drift.

❚ Column bleed❚ Dirty detector❚ Contaminants in carrier gas / carrier

gas purity


10

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

6000

6500

7000

7500

8000

8500

9000

9500

10000

10500

11000

11500

12000

Effect of carrier gas purity on baselinenoise:

H2 carrier from tank

H2 carrier from a generator


Column bleed results from thenormal degradation of thestationary phase.

❚ All columns bleed to some extent.❚ Bleed increases with temperature.❚ The amount of bleed will increase in the

presence of oxygen.


A Typical Bleed Profile:

4

Bleed measured as thedifference between 1 and 2.

1

2


Column Bleed on an MSD

❚ Visible as baseline rise in the TIC.❚ Check spectra for key bleed ions:

❙ PTE™-5: 207, 281❙ SPB™-1: 73, 207, 281❙ SPB™-624: 207, 269

❚ Make sure interface temp. is < columnmax. temp.


Ion 207 corresponds to a fragment known as D3:

O Si

SiO

Si

O

CH 3 CH 3

CH 3

CH 3

CH 3

+


40 60 80 1001201401601802002202402602803003203403603804000

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

6000

m/z-->

Abundance

Scan 2895 (27.487 min): 1118015.D207

2817313396

191

253

50

327223162 355115 415

MS spectra of bleed from a PTE™-5 Column

207

281


40 60 80 1001201401601802002202402602803003203403603804004200

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

6000

6500

7000

7500

8000

8500

9000

9500

10000

10500

11000

11500

m/z-->

Abundance

Scan 1941 (28.793 min): 0122002.D73

44

207

147

281341

96 119 179 253 429401313

MS spectra of bleed from an SPB™-1 Column

73

207

281


MS spectra of bleed from an SPB™-624 Column

40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 3400

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

22000

24000

m/z-->

Abundance

Scan 8581 (33.704 min): 1022006.D207

269

44253

19113373 96177147

283119 322 343239163 22359298

207

269


40 60 80 1001201401601802002202402602803003203403603804004200

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

22000

24000

26000

28000

30000

32000

34000

36000

38000

40000

42000

44000

m/z-->

Abundance

Scan 604 (7.460 min): 1201001.D73

147

281

32745

207415399

131 191 343 383251 29795 223163115

73

147

281

MS Spectra of Septa Bleed


Prevent column bleed!

❚ Sufficiently purge column with carrier beforeramping it up in temperature.

❚ Make sure carrier gas is filtered for water andoxygen.

❚ Check integrity of all fittings leading to thecolumn.

❚ Do not heat the column above its maximumtemp.

❚ Precondition the column prior to use.


4. Ghost Peaks

❚ Residue in the inlet liner and at thehead of the column

❚ Contaminated syringe / and or washsolutions on an autosampler.

❚ Sample carryover❚ Contaminated carrier gas


0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000

2000000

2200000

Response test mix, before

If pieces of septa get into an inlet liner...


0

500000

1000000

1500000

2000000

2500000

3000000

3500000

Time-->

…even a simple analysis can be ruined.

Response test mix, after


5. Missing Peaks / PoorResponse

❚ Sample decomposition❙ Activity in the inlet or column❙ Injection port temperature too high❙ Sample not stable enough for GC❙ Standards not stable

❚ Coelution❚ Insufficient run time / final temperature❚ Sample not volatile enough for GC❚ Improper column installation


0

50000

100000

150000

200000

250000

300000

350000

400000

450000

2,4-

din

itro

phen

ol

4-n

itro

phen

ol

2-m

ethy

l-3,

5-di

nitr

ophe

nol

pent

ach

loro

phe

nol

Nasty samples can damage a columnby creating active sites:

Before Sample Injection


0

100000

200000

300000

400000

500000

600000

700000

800000

900000

Time-->

Abundance

2,4-

DN

P &

4-N

P s

hou

ld b

e he

re

2-m

ethy

l-3,

5-di

nitr

ophe

nol

Pen

tach

loro

phen

ol s

houl

d b

e he

re

After sample injection

Here, the responses of some acidswere affected:


Response can also be affected by theposition of the column in the inlet:

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000

2000000

2200000

2400000

2600000

2800000

8 mm above top of ferrule


0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000

2000000

2200000

Here, the column was not inserted farenough:



Here, the column was inserted too far:

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000



6. Insufficient Resolution

❚ Wrong column❙ Longer columns increase resolution.❙ Smaller ID columns increase resolution.❙ A different phase altogether may be

needed.

❚ Wrong Conditions❙ Carrier gas flow too fast or slow .❙ Oven ramp rate too fast.


Recommended ReadingSupelco Bulletins

1. 741: The Supelco Guide to Leak-Free Connections2. 783: Cleaning Flame Ionization Detectors3. 853: Capillary Troubleshooting Guide4. 875: Supelco Capillary GC Selection Guide5. 895: Installation and Maintenance Instructions for

.25 mm and .32 mm ID Fused Silica CapillaryColumns

6. 897: Installation and Maintenance Instructions for.53 mm ID Fused Silica Capillary Columns

7. 898: Gas Management Systems for GC8. 899: Capillary GC Inlet Sleeve Selection Guide9. 916: Purge and Trap System Guide10. 918: Selecting Purifiers for Gas Chromatography


Help is just a phone call ormouse click away!

❚❚ Supelco Technical ServiceSupelco Technical Servicephone: 1-800-359-3041email: [email protected]❚❚ Supelco Customer ServiceSupelco Customer Servicephone: 1-800-247-6628email: [email protected]❚❚ Sigma-Sigma-Aldrich WebsiteAldrich Websitewww.sigma-aldrich.com

gc troubleshooting

Documents