Skip to main content
Chemistry LibreTexts

Optimisation

  • Page ID
    61195
  • \( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \) \( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)\(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\) \(\newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\) \( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\) \( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\) \( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\) \( \newcommand{\Span}{\mathrm{span}}\)\(\newcommand{\AA}{\unicode[.8,0]{x212B}}\)



    Abstract

    Although the principle of the rapid injection method for large volume sampling is extremely simple, a number of parameters have to be optimized before a large volume injection can be performed successfully.

    Leve: Advanced

    The parameters that require some consideration are:

    • Choice of the packing material,
    • Optimization of the splitless transfer and determination of the inertness of the packing material,
    • Determination of the maximum sample volume that can be accommodated by the packed bed,
    • Determination of the solvent-elimination time.

    In the following paragraphs each of these parameters will be addressed in more detail.

    Choice of the Packing Material

    An important parameter in Rapid large volume injection is the material that is used for packing the liner. This material should meet a number of different criteria. Firstly, the material should be capable of retaining a sufficiently large volume of solvent. This means that good wetting of the material by the solvent should occur. In general, maximum loadability occurs when the polarities of the packing material and the solvent match. This means that if a non-polar solvent is used, a non-polar packing material will give the largest capacity. A second important requirement for the packing material is that it is sufficiently inert. If this is not the case, then the components of interest can undergo thermal degradation during splitless transfer to the column. Moreover, the components should not exhibit too strong an adsorption on the packing material as this might hinder their desorption from the material and subsequent transfer to the column. Some interaction between the components and the packing material, however, can be advantageous, in particular for volatile analytes that otherwise might easily be lost together with the solvent during solvent evaporation. Clearly, the packing material should not dissolve in the solvent or swell under influence of the solvent (for example, the structure of the packing material Tenax is irreversibly affected by chlorinated solvents such as dichloromethane). Finally, the packing material should have a sufficient thermal stability and should give good blanks. In table 3.1 an overview of a number of possible packing materials is given. Apart from the materials given in this table there are numerous other materials that can be used as liner packings for Rapid large volume injection. An important parameter to take into account finally is the particle size of the packing material. Good results can be obtained with particle sizes between 60 and 100 mesh. Larger particles such as glass beads, however, can also be used.

    Packing materials for the Rapid injection method

    Packing material

    Vmax* (µl)

    Tmax (oC)

    Glass wool

    115

    -

    Quartz wool

    125

    -

    Glass beads

    40

    -

    Teflon wool

    80

    275

    Tenax TA

    125

    310

    Dexsil coated support

    80

    340

    Supelcoport

    120

    330

    *Based on a liner with an i.d. of 3.4 mm and a packed bed with a height of 25 mm.

    Once the packing material is selected, the liner can be packed with this material. To keep the material in place, it is advisable to use a liner with a frit in its bottom section. The required amount of packing can be packed onto this frit and finally a plug of glass wool can be put on top of it to prevent the packing from being blown out. Prepacked liners with glass frits at either end are nowadays commercially available. When packing the liners, one should realize that in all PTV injectors significant temperature gradients can occur along their length. Due to heat loss at the top and the bottom side of the injector, the temperature in both the top section and the bottom section of the liner will be significantly below the set temperature of the PTV injector. When packing the liners care should be taken that the packed bed does not extend into the cooler regions of the injector. If required, the needle length of the syringe should be adjusted so that the needle does not protrude into the packed bed in the liner. Figure 7. shows some temperature gradients measured along a PTV liner. Information on which part of the injector is properly heated can be obtained from the manufacturer. Alternatively, the temperature profile inside the injector can be measured using a small thermocouple.

    7. Temperature gradient along a PTV liner.

    ZwjivxoHO.jpg

    7b. Temperature gradient along a PTV liner.

    ZwjivxoHQ.jpg

    Once the liner is filled with packing it should be conditioned. This can be done by temperature cycling the liner several times to a temperature that is roughly 20oC above the temperature that is going to be used for the splitless transfer of the solutes to the column. Care should be taken not to exceed the maximum allowable operating temperature of the material selected. The time required for conditioning strongly depends on the type of material used. Glass wool, quartz wool and Supelcoport can be conditioned by heating it to 325oC for one hour. Conditioning of Tenax, on the other hand, requires many hours. In all cases, the conditioning process should be carried out in an inert atmosphere completely free of oxygen and water. Prior to starting the conditioning of the material, all air should first be removed from the liner at low temperature by purging it with carrier gas for a few minutes.

    Inertness of the packing material

    Once the liner is conditioned a first series of experiments can be conducted to optimize the splitless transfer of the solutes from the injector to the column. This can be done by injecting a small volume (e.g. 1 µl) at different PTV final temperatures and splitless times. It should be emphasized here that the transfer of solutes from a packed liner can be somewhat slower than from an empty liner. This because the packing might show some interaction with the solutes. It is therefore dangerous to transfer a PTV program directly from a method that uses an empty liner to a method in which a packed liner is used. A first guess of the required PTV final temperature can be obtained from the elution temperature of the solutes from a standard GC column. A PTV final temperature some 25oC above the elution temperature in general gives quantitative transfer of the solutes to the column in a short time. Standard splitless times are 30 to 45 seconds, calculated from the moment the PTV injector reaches its final temperature.

    At this stage, the inertness of the packing material has to be determined. This can be done by comparing the optimum chromatogram from the splitless-transfer optimization experiments with a reference chromatogram recorded using on-column injection or, if this is not available, using a PTV injection with an empty liner. In general, thermal degradation during a PTV injection is reduced if lower heating PTV programming rates are applied. For practical reasons, heating rates below 1 to 2oC/sec are only rarely used.

    Max Sample Volume in Packed Bed

    Once the splitless transfer conditions have been optimized and it is clear that no unacceptable thermal degradation or adsorption of the components occurs inside the PTV liner, a start can be made with the optimization of the actual large volume injection. The first step in this optimization is the determination of the maximum volume of solvent that can be retained by the packing material in the liner. As this volume depends (slightly) on the liner temperature, the gas flow through the injector and the solvent used, first these parameters have to be fixed. For the selection of the optimum liner temperature and the gas flow some guidelines have been given in section 3.3.2. In summary: when using hexane as the solvent, a good initial PTV temperature is 30oC. For other solvents the initial PTV temperature should be adapted according to the difference of the boiling point of this solvent and that of hexane. To obtain rapid evaporation of the solvent, it is advantageous to use a split flow that is rather high (>100 ml/min). A split flow of 200 to 250 ml/min is, as stated before, more or less standard in all forms of PTV large volume sampling.

    When the initial PTV temperature and the split flow have been selected, the maximum volume of solvent that can be accommodated by the packed liner has to be established. The easiest way to do this is based on visual observation. To this end the packed liner is installed in the injector and the column is disconnected. In doing this the carrier gas flow is kept on. Now a large volume of the solvent that has been selected is injected (e.g. 100 µl). During the rapid injection of this large volume one should carefully observed the outlet of the injector, i.e. the place where normally the column is installed. If droplets of liquid are observed squirting into the oven, the maximum capacity of the liner has been exceeded. By varying the injected volume, the maximum allowable injection volume can be rapidly determined. To be on the safe side, in a true large volume experiment the injected volume should be only some 60 to 70% of the maximum volume determined using this visual method. The method described here is a relatively crude method, although for most of the situations it is sufficiently reliable. A more sophisticated method is based on the injection of increasing sample volumes of a non-volatile component in separate GC runs starting from a very low value to a very high value (e.g. from 1 µl to 200 µl). This should be done with the selected solvent at the preselected initial PTV temperature and flow rate. In the sample, only one component has to be present, preferably a high boiling component (boiling point roughly above that of n-C20). A plot of the peak area of this component versus the injected volume should increase linearly until a plateau value is reached. The volume where the plateau occurs is the maximum allowable injection volume. If in a Rapid large volume injection experiment a volume in excess of the maximum liner capacity is injected, liquid sample will be lost via the split exit. It is also possible that liquid sample can enter the column. This latter phenomenon manifests its self as peak splitting. All peaks in the chromatogram will show a 'pre-peak'. An example of these pre-peaks is shown here:

    Peak splitting, caused by a too large injection volume

    ZwjivxoHS.jpg

    8. Peak splitting caused by too large an injection volume in the Rapid injection technique.

    It is apparent that the maximum volume of liquid that can be retained by a given liner depends on the temperature, the gas flow, liner diameter and type and amount of packing material in the liner. Liners with a larger inner diameter clearly have a larger sample capacity. Therefore, these liners are to be preferred for the Rapid large volume injection method. Unfortunately, however, the use of wide-bore liners also has a few disadvantages. The most important one being the longer time required for splitless transfer. This increases the risk of sample degradation during splitless transfer. Also in this case, bleeding of the packing material can not be completely avoided, the fact that for the Rapid injection technique a large i.d. liner with a larger amount of packing material has to be used is a disadvantage.

    solvent-elimination time.

    In the Rapid or 'at-once' injection method a large volume of sample is injected into a packed liner. After injection the injected sample spreads out as a thin film of liquid over the packing material. Due to the large flow of gas through the injector the solvent film starts to evaporate. The components present in the sample will be trapped in the cold-spot that is created due to the large heat consumption of the evaporation process. The solvent vapour created in the evaporation process is discharged via the split exit. A small fraction of this, approximately 0.5 to 1% enters the GC column. Once the solvent film is evaporated, the split exit can be closed and the components remaining in the liner can be transferred to the column in the splitless mode.

    An important parameter in Rapid large volume injection is the time of solvent evaporation. If this time too short, there will still be a significant volume of solvent left in the liner at the moment the split exit is closed. When the injector is heated, with the split exit closed, a very rapid, uncontrolled evaporation can occur in the liner, creating a pressure pulse. In the worst case this pressure pulse could adversely affect the integrity of the packed bed in the liner and packing material particles, solvent vapour and components could be forced back up in to the carrier gas inlet of the GC, resulting in contamination of the system. The worst case scenario described above only occurs if a very large volume of solvent (some 10 µl or more) is still present in the liner at the moment heating starts. If on the other hand, the time for solvent elimination is excessively long, the liner is evaporated to dryness. When all of the solvent is evaporated, the cold spot, that helps to retain the volatile sample constituents in the liner, disappears and volatile components are lost. From this it will be clear that a careful optimization is required for successful and reliable large volume injections using the at-once method. Under optimized conditions, the split exit is closed at the moment that a few microliters of solvent are left inside the liner. As long as there is some solvent left, volatile components will not be lost.

    After careful optimization, even very volatile components can be analyzed sucessfully using this method. A boiling point difference of only some 30 to 40oC between the solvent and the most volatile components is sufficient for obtaining quantitative recovery. For components with a higher boiling point the optimization of the solvent vent time is less critical. If hexane is used as the solvent and solvent elimination is carried out at a temperature of 35oC, solutes with a boiling point above that of roughly n-C16 will not be lost, even if an already dry liner is purged with a high flow of carrier gas for some time.

    In principle there are four methods for determining the evaporation time of a given volume of solvent.

    Method 1:
    In the split exit line of the GC a monitor detector, as for example a thermal conductivity detector, can be installed. This detector detects the presence of solvent vapour in the carrier gas leaving the split exit. Once solvent elimination has reached completion, the signal of the detector returns back to its original value. At this point the split exit can be closed. Superficialy, this method is very attractive, because it allows automation of the split exit control. Unfortunately, however, it is a relatively expensive method as it requires an additional TCD and control software.

    Method 2:
    A very simple approach for the determination of the solvent elimination time is based on visual observation of the evaporation process by igniting the vapour leaving the split exit. When evaporation is complete the flame extinguishes. Now the split can be closed and the injector can be heated. Alternatively, the evaporation time can be measured a few times with large volume injections of a pure solvent. In a first real experiment an evaporation time some 5% shorter can this value can than be used. As the time allowed for evaporation is slightly shorter than the required time some solvent is still left in the liner. This solvent helps in retaining the volatile sample constituents. This approach works well for hexane and pentane (yellow flame) and ethylacetate (weak blue flame). Other solvents such as dichloromethane or chloroform do not give self-sustaining flames. The vapours of these solvents, however, can be monitored by allow the split gas to flow through, for example, a candel flame.

    Method 3:
    Another method for determining the solvent elimination time uses the GC detector itself as a monitor detector. Solvent entering the column during solvent elimination will reach the GC detector after the column dead-time. When solvent elimination is complete the detector signal returns to the base-line. The width of the solvent peak (at half height) now equals the duration of solvent elimination. An example of this is given below. If this method is used, the oven temperature during solvent elimination should be some 10 oC or more above the initial PTV temperature. This is to ensure rapid transport of the vapours formed through the column. This method can not be used if the solvent is eliminated at zero injector pressure. When a mass spectrometer is used the vacuum level of the system can also be used to determine the width of the solvent plug entering the detector. It is advisable to measure the width of the plug using a few injections of the pure solvent at the desired volume. This method can be used for most GC detectors including the FID, NPD, FPD or SCD. For the ECD it is slightly more complicated. To allow accurate measurement of the peak width at half height it will often be necessary to perform the experiments at a reduced detector sensitivity setting. In a first real experiment one could again use a solvent elimination time that is some 5% shorter than the time required for complete evaporation.

    Determination of the solvent elimination time.

    Determination of the solvent elimination time.

    9. Determination of the solvent elimination time (method 3) in a Rapid large volume injection. Further details: see text above.

    Method 4:
    The last method for the determination of the optimum solvent elimination time is based on trial and error optimization. In this approach a large sample volume of a sample containing various components covering a wide range of boiling points in the solvent of interest is injected. The identity of the components is not important. A homologous series of normal alkanes is therefore very well suited for this purpose. The selection of the solvent is of course important. In the first experiment a very long solvent elimination time is used. At this very low evaporation time the more volatile components present in the sample will be lost. In a series of experiments the solvent elimination time is now reduced in small steps to a value where good recoveries are obtained for all solutes. Figure 3.10 shows the chromatograms obtained for a 100 l 'at-once' injection of a hexane solution of normal alkanes obtained at solvent elimination times that are too short, good and too long, respectively. As a rule of the thumb, the evaporation rate of a solvent increases by a factor of 2 if the PTV temperature is increased by 10oC to 15oC or if the split-flow is doubled.

    If the solvent elimination time is far too short, a large volume of liquid is still present in the liner at the moment the split exit is closed. Upon heating the injector this volume of solvent rapidly evaporates ('explodes') and creates a huge volume of gas inside the liner of the injector. Eventually the packed bed of packing material in the liner might be disrupted by the rapid evaporation of the solvent. Moreover, solvent vapours might back-up into the carrier gas lines resulting in severe contamination of the inlet. This situation should always be avoided. During optimization of the solvent elimination time one should therefore always start the optimization procedure with elimination times that are too long and successively reduce the time for solvent elimination in small steps. A very helpful figure is figure 3.10. The top figure shows the chromatogram of a large volume of an alkane standard at too long a solvent elimination time. Losses of volatile components show that the liner is vented to dryness. The middle chromatogram shows the analysis at an optimized solvent vent time while the bottom chromatogram shows a chromatogram obtained at too short a solvent elimination time. In the latter case a very characteristic peak distortion is observed in the middle part of the chromatogram. The early eluting peaks have good peak shapes and so have the peaks more towards the end of the chromatogram. The middle part is distorted.

    Peak distortion and sample loss during 'at-once' LVI.

    ZwjivxoHW.jpg

    10. Peak distortion and sample loss during 'at-once' large volume injections of an alkane standard in hexane at a long (top), optimum (middle) and short (bottom) solvent elimination time.


    Optimisation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by LibreTexts.

    • Was this article helpful?