GC/GC5 Troubleshooting

High background gas levels

An overview of the background gas levels can be achieved by running a ‘background scan’ method (Background gas levels); during the method the magnet current changes and the ion source typically has trap current of 200uA to allow direct comparison of target background levels (Ion Beam Background Levels). If background levels are too high it can adversely affect your analysis. Possible reasons could be:

Leaks

  • A leak in system sees the ingress of atmospheric air into the system and can be identified by elevated N2 (m/z28) and Ar (m/z 40) levels, typically these backgrounds should be <1e-10 A and <1e-11 A, respectively. The most efficient way to locate leaks is to tune the IRMS to m/z 40 and direct a stream of argon gas around various fittings.

After routine maintenance

  • After changing a component, e.g. GC column, injector liner, etc. the background may be temporarily high due to an ingress of atmospheric air into the system (high N2, H2O and Ar). Assuming there are no leaks the levels should return to normal.
  • When using new Vespel or 85% Vespel ferrules within the GC oven they may initially provide a leak tight seal, however after the oven temperature has been cycled (i.e. heated up and cooled) a leak may form at these fittings. This is simply degassing from the ferrules and the leaks can be corrected by tightening the fittings after the first oven temperature cycle.
  • After a furnace tube has been changed and is heated to its operating temperature the background gas levels may increase, again this will be due to degassing. Background levels should return to normal without intervention within 1 hour.

High water backgrounds

  • High water backgrounds can potentially cause issues, particularly with d13C measurements, due to protonation in the ion source. Elevated water levels are often evident after routine maintenance, particularly if the IRMS has been vented for extended periods, but should reach normal levels without intervention.
  • If water levels remain high, check that the ‘Nafion flush flow is on’. This is indicated in the software within the System Status Panel.

Gas supply

  • High background gas levels could also be due to the purity of then gases. We recommend the helium carrier gas should have a purity of > N5.0 (99.999%). Please refer to the site requirements documentation.
  • Although unlikely, there is the potential that the actual gas supply cylinder may be contaminated. This is can be identified by swapping the cylinder.

No samples peaks at IRMS or FID

If no sample peaks are detected at the IRMS or FID, first check that the IRMS can see a monitoring gas pulse and that the FID is at temperature and lit. If the IRMS does not detect a monitoring gas pulse, please refer to Error Reporting.

If the IRMS and FID are both operational and no sample peaks are observed (including a solvent peak at the FID) it is likely the problem will be in the VSOS splitter union or somewhere upstream, including potential sample issues.

GC carrier flow and temperatures

  • Make sure that the GC method is set with appropriate gas flows, timings and temperatures. Please refer to Agilents GC user manual for more details.

Sample and injection issues

  • Has the sample evaporated? If so, re-dissolve in appropriate volume of solvent
  • Is the syringe blocked? If so, clean and unblock with solvent or replace
  • If using an autosampler, ensure that the syringe is correctly installed and that it samples and injects the correct sample vial
  • Is there a large leak in the injector? If so, cool down and re-assemble and possibly replace parts. Please refer to Agilents GC user manual for more details.

GC column

  • Is the GC column appropriate for the sample / application
  • Is the column installed correctly? e.g. When using a split / splitless injector the ‘injector end’ of the column should protrude between 4 and 6 mm above the graphite ferrule (see Agilents GC user manual for more details). The other end of the GC column is fed through the VSOS union so it is level with the union or pulled back by approximately (~1mm) (Gas Chromatograph (GC) and GC5 interface)
  • Is the GC column blocked? If so, trim either end (~15 cm) with a ceramic wafer. See Maintenance for more details. Alternatively, replace the column
  • Is the GC column broken? Depending on the location of the break, either replace or re-install. If the break is near the front or end of the column, it may be appropriate to trim and reinstall the column provided the stationary phase has not been damaged.

VSOS splitter union and capillaries

No samples peaks at IRMS, but evident at FID

If sample peaks are detected at the FID, the sample, injector, GC column and capillary to HS valve should be okay. As no peaks are detected at the IRMS, the first step is to check if a monitoring gas pulse can be detected. If not, please refer to Error Reporting.

Heart split valve operation

  • Does the HS valve operate? Establish this by toggling the ‘flow to IRMS’ checkbox in the GC valves section of lyticOS. The needle of the HS valve should move up when the box is unchecked (i.e. flow to the FID, HS valve open) and fall level with the top of the valve when ticked. If this does not happen, check that compressed air or helium is connected to the valve.
  • Are sample peaks detected at FID when HS valve is closed, i.e. ‘flow to IRMS’ checkbox ticked? If so, there is likely to be a leak in the HS valve. This could be due to the capillary from the VSOS not being connected properly into the bottom of the valve or that the ‘back-drilled’ ferrule is leaking.

VSOS splitter union and capillary

  • Is the fused silica capillary from the VSOS splitter union to the furnace installed correctly in the VSOS and furnace tube fittings? See Gas Chromatograph (GC) and GC5 interface.
  • Is the fused silica capillary blocked, damaged or crushed? If so, trim the ends and re-install or replace

Furnace tube

  • Ensure that the correct furnace tube is installed and the furnace is set to the correct operating temperature and the sample line helium is set to between 1.5 and 2.5 psi. See Furnace tube configurations for the EnvirovisION System
  • Check that the furnace tube is not blocked. This can be done by measuring the flow out of the open split; this should be approximately 1 ml/min. If no flow is evident, the furnace tube may be blocked. It is recommended to cool down the furnace and replace the furnace tube.
  • Does the furnace tube require oxidation? For example, the GC-Carbon (‘Natural Gas’) and GC-Nitrogen quartz furnace tubes contain braided metal wires that initially require oxidation before first use. See Furnace tube oxidation strategies

Furnace tube exit to centrION / isoprime visION Mass Spectrometer

  • Is the fused silica capillary from exit of the furnace tube to the Nafion dryer membrane or the Nafion membrane itself blocked?

source inlet valve 

  • Is the source inlet valve set to GC in the software or method? If not, any eluent from the GC will not enter the ion source.

Samples peaks evident at IRMS, but not at FID

If sample peaks are detected at the IRMS, the sample, injector, GC column, capillary to the furnace tube, furnace tube and flow path to the IRMS should be okay. Although not strictly required for compound-specific stable isotope measurements, we recommend the use of a FID as it  is very useful for diagnostic purposes and can establish chromatographic timings when creating GC methods (e.g. when does the solvent elute, etc.)

FID

  • As no peaks are detected at the FID, the first step is to check that the FID is at temperature, with gas connected and is lit, producing a signal. If not, please switch on or refer to the Agilent GC manual.
  • If the FID doesn't ignite then refer to the Agilent GC manual.

Solvent breakthrough to IRMS

  • Has a solvent peak been detected at the IRMS, e.g. ‘Solvent Breakthrough’? If so, check that the HS valve is operating? Establish this by toggling the ‘flow to IRMS’ checkbox in the GC valves section of lyticOS. If the valve does not operate, check that compressed air or helium is connected to the valve.
  • Ensure that the sample line helium pressure is set to between 1.5 and 2.5 psi and the FID helium to approximately 0.3 psi; these pressures are important as they play a key role in the heart split mechanism. Ordinarily with the HS valve open, all the carrier flow will be directed towards the FID, however if the FID helium is set too high this can force the carrier flow through the furnace tube instead. This can also happen when the sample helium is set with too low a pressure.
  • A blockage in the HS valve can also cause ‘solvent breakthrough’. A quick test is to remove the fused silica that connects the VSOS to the HS valve at the VSOS end only and place this end in some solvent. With the HS valve open (i.e. flow to FID), you should see bubbles in the solvent from the FID helium supply. If bubbles are not observed then there will be a blockage, which most likely will be the capillary or the back-drilled ferrule at the base of the HS valve. Please refer to the maintenance section for more details
  • Another possible reason for ‘Solvent Breakthrough’, is a solvent delay has not been specified within the GC method or the sample task list.

Capillary from Heart split valve to FID

  • With no solvent breakthrough and the IRMS detecting peaks, no peaks being observed at the FID could be due to the fused silica capillary connecting the HS valve to the FID being incorrectly fitted, damaged or broken. If so, replace with a new capillary.

Poor sample sensitivity

If there is a loss in sample peak sensitivity, it is advisable to first to check if the IRMS has also lost sensitivity when a monitoring gas is introduced; please refer to Error Reporting. If the monitoring gas sensitivity is as expected the problem will likely be associated with either the actual sample, sample introduction, combustion / pyrolysis efficiency or using the wrong method for analysis.

Leaks

Sample injection issues

  • Has the sample been diluted in too much solvent? Check by analyzing a standard of known concentration.
  • Check if the injection method being used is correct, e.g. the correct temperatures and gas flows, split or splitless, etc. In the case of liquid injections with a split splitless injector, ensure that the temperature of the injector is high enough to vaporize the sample and check that correct split ratio is being used when in split mode. In splitless mode, check the sample ‘purge time’ is long enough to transfer the vaporized sample onto the GC column (typically 0.75 to 1.0 min). For manual injections you also need to ensure that the Prep Run button is pressed on the GC keypad just prior to injection, this shuts the split vent valve for the specified sample ‘purge time’ this procedure is done automatically when an autosampler is attached.

Furnace tube

Poor chromatography on both the IRMS and FID

If poor chromatography is observed at both the IRMS and FID, it is likely the source of the problem will be an issue with the injector, the GC column or the GC method used. Below, we refer to some common issues, but for a more comprehensive guide to chromatography please refer to the vast array of literature available on the subject and/or Agilents GC user manuals; it is also advisable to refer to guidelines provided with each GC column.

Injector

  • Ensure that the injector is regularly maintained. In the case of a split splitless injector, the septum, injector liner, liner O-ring and gold seal all require regular replacement. Instructions on how to carry out this maintenance can be found in Agilents GC manual
  • Are you using the correct liner for your analysis? Split, splitless, PTV, etc. Please refer to the manufacturers for more details.

GC column and methods

  • Ensure that you are using an appropriate GC column for your sample type/ application. Agilents GC column section guide provides useful information on this subject.
  • Is the GC column installed correctly and has it been conditioned ready for sample analysis? Refer to the GC column manufacturer’s guidelines for instructions
  • Use a suitable temperature program. Do not exceed the maximum temperature for the column. See column manufacturer’s specifications.
  • Are you using an appropriate GC carrier flow? Typically for IRMS applications the GC columns have IDs of between 0.20 to 0.32 mm which normally require GC flows of between approximately 0.8 and 1.5 ml/min

Samples

  • Does your sample require derivatisation or has it become partially underivatised? Derivatisation is often used to make samples more amenable to GC analysis, e.g. Fatty acids to fatty acid methyl esters (FAME). Compounds that exhibit poor volatility, thermal stability or contain reactive functional groups are all candidates for derivatisation. The derivatisation process, where compounds are chemically altered, results in samples producing better chromatographic peak shapes relative to their underivatised or partially underivatised equivalents.

 

Example chromatographic issues

Fronting peaks

These can be caused by injecting too much sample and overloading the GC column.

Tailing peaks

Cause by interaction with active sites within, for example, the injector liner or the GC column.

Broad peaks

These can arise from dead volumes in the system or inappropriate GC temperature programs (temperature too low or a slow ramp) or carrier gas flows (too low).

Split peaks

Typically a result of issues with the injector or the injection method

Co-eluting peaks

Caused by an inappropriate GC method (e.g. temperatures, gas flows) or an unsuitable GC column

Ghost peaks

these are carryover peaks from the previous sample run. This is caused when a GC method is not long enough or does not reach a high enough temperature to elute all the sample components during one run.

Poor chromatography on the IRMS, but good on the FID

If poor chromatography is only observed at the IRMS, then the source of the problem is likely to be an issue with the VSOS splitter or anywhere downstream towards the IRMS.

VSOS splitter union and capillary to furnace tube

  • Is the fused silica capillary from the VSOS splitter union to the furnace installed correctly in the VSOS and furnace tube fittings? Ensure that the fused silica has been cut cleanly and there are no dead volumes in the fittings. See Gas Chromatograph (GC) and GC5 interface.
  • Is the fused silica capillary partially blocked or crushed? If so, trim the ends and re-install or replace.
  • Avoid active sites in the silica capillary by only using a deactivated fused silica.

GC / Furnace interface

  • Avoid cold spots at the GC / furnace interface by ensuring that the Cal3300 controller is set to the correct temperature, normally 350°C.

Furnace tube

Poor stable isotope ratio measurements (precision, linearity and accuracy)

If system is producing poor stable isotopic data, it is important to try and diagnose the problem by first determining if the IRMS is functioning as expected. This can be done by introducing monitoring gas pulses into the IRMS and measuring their stability and linearity; if the data from the monitoring gas pulses are also poor then please refer to the Operating the EnvirovisION System.

If the IRMS is performing well with the monitoring gas pulses, then the next step is to investigate the rest of the system including potential issues with the samples / standards being analyzed.

Chromatography

  • A prerequisite for producing the best possible stable isotope ratio measurements is good chromatography with minimal co-elution. See section 6 and 7.

Incomplete combustion / pyrolysis

  • Ensure the correct furnace tube, gas flows and temperatures are being used. See Furnace tube configurations for the EnvirovisION System.
  • Does the combustion tube require oxidation? Stable isotope values can often be seen to be more negative and chromatography can deteriorate when a combustion tube requires oxidation. To oxidize the furnace tube or information on how to analyze your samples with a small bleed of O2, please refer to Furnace tube oxidation strategies.
  • Does the pyrolysis tube require conditioning? It can be observed that instead of a normal Gaussian peak shape the tail can be distorted with a step when the furnace tube requires conditioning. To condition the GC-H high temperature furnace tube, please refer to Conditioning GC-H (high temperature).

Nafion membrane

  • High water backgrounds can cause issues with stable isotope measurements, particularly d13C analysis where elevated water backgrounds can cause protonation in the ion source. It is important to ensure that the ‘Nafion flush flow is on’, i.e. a counter flow of helium flows outside the Nafion membrane aiding the removal of water from the gas stream.

Sample concentration and peak heights

  • Are the sample peak heights too small to achieve reliable data? If so, it is advisable to inject more (be careful not to exceed the volume of the injector liner), inject with a reduced split ratio (or in splitless mode), increase the concentration of the sample or optimize the IRMS ion source tuning (i.e. higher trap current, e.g. 600uA).
  • Are the sample peak signal heights too high and are overloading the amplifier? If so, inject less, inject in split mode, dilute the sample in more solvent or optimize the IRMS ion source tuning (i.e. lower trap current, e.g. 50 to 100uA.

Sample integration and calibration

  • A possible reason for poor stable isotope sample data could be due to inappropriate sample integration. These settings can be optimized in lyticOS..
  • If sample precision is good, but the accuracy is not, this could be due to the calibration methodology. We recommend to use multi-point calibration techniques employing a principle of identical treatment, i.e. samples are calibrated against standards that are analyzed in the same way as the samples.

Communication Problems

Verify that the PC can “see” the GC

  1. Start command prompt (Start -> Run -> “cmd”)

  2. Type “ping 192.168.0.2” and press [Enter].

    (substitute the correct GC IP address)

  3. Validate that the response is OK and no time-out errors

Check PC network connections

  1. Start command prompt (Start -> Run -> “cmd”)
  2. Type “ipconfig” and press [Enter].
  3. Validate that the expected network devices are displayed and that they display as connected

Verify that the PC and the GC are using the same Subnet Mask

Devices can only communicate if they have the same Subnet Mask configured.

Firewall

If you have already verified that the PC can “see” the GC, then the reason for the communication issue is probably Firewall related.

You can open the Firewall settings, to check which type a network is (Domain, Private, Public). Firewall rules are usually configured for 1 or more different network types.

If multiple networks are present;

e.g. 1. Domain network (Internet)

2. Private / Public network (GC)

lyticOS may only have been allowed network communication on one (or possibly none) of the networks. You should check the PC firewall settings, to ensure that inbound rules exist to allow lyticOS to communicate on the network that the GC is connected to:

  1. UDP (Any Ports)
  2. TCP (Any Ports)

This usually requires 2 separate rules to be configured.

lyticOS Debug Messages

lyticOS writes debug messages to the Log Messages window about how it is communicating to the GC (e.g. recording relevant IP addresses).