Precautions for using Headspace Samplers

What factors are related to the analysis results of the headspace sampler gas chromatograph? The following is a brief analysis and explanation. 

1  Properties of the sample

The biggest advantage of headspace & GC is that it does not require complex processing of the sample, and the headspace gas is directly taken for analysis. We do not need to consider the impact of non-volatile components in the new sample on GC analysis. However, the properties of the sample still have a direct impact on the analysis results. The sample referred to here refers to the "original sample" placed in the sample vial, rather than the "volatile matter" entering the GC, so it is necessary to consider the properties of the sample in the entire sample vial.

For gas samples or samples that can be completely converted into steam under certain conditions, there is only gas phase and no condensed phase in the sample vial. So, this sample is not much different from regular GC analysis. It should be noted that the sampling temperature and sample storage temperature of gas samples may be different, often with the latter being lower than the former. When storing samples at relatively low temperatures, some components may condense. Therefore, during analysis, it is necessary to place the samples at equilibrium temperature for a certain period of time to achieve a uniform gas phase, in order to eliminate errors caused by condensation of some sample components. If the liquid sample is converted into a gas, the conversion process requires a certain amount of time, unlike the sample vaporization at the sample inlet in ordinary GC. Incomplete vaporization can cause the composition of the headspace sample to be different from the original sample, thereby affecting the accuracy of the analysis results. Therefore, sufficient time should also be balanced at a certain temperature.

Liquid and solid samples are more complex, and there are at least gas-liquid or gas-solid phases in the sample vial, or even gas-liquid-solid three-phase coexistence. The content of each component in the headspace gas is not only related to its own volatility, but also to the sample matrix. Especially for components with high solubility (distribution coefficient) in the sample matrix, the "matrix effect" is more pronounced. This is a major characteristic of headspace sampling, which is that the composition of the headspace gas is different from that of the original sample, which has a particularly serious impact on quantitative analysis. Therefore, the standard sample cannot be prepared solely with the standard sample of the substance to be tested, and must also have the same or similar matrix as the original sample, otherwise, the quantitative error will be significant.

There are some methods to eliminate or reduce matrix effects in practical applications, mainly including:

(l) Using salting out, which involves adding inorganic salts (such as sodium chloride) to an aqueous solution to change the distribution coefficient of volatile components. Experiments have shown that salt concentrations below 5% have almost no effect, so high concentrations of salt are commonly used, and even saturated concentrations are used. It should be pointed out that the effect of salting out on polar components is much greater than that on non polar components.

(2) Add water to an organic solution, of course, to be compatible with the organic solvent used. This can reduce the solubility of organic compounds in organic solvents and increase their content in headspace gas. For example, when determining the residual amount of 2-ethylhexylacrylate in polymers, the sample is dissolved in dimethylacetamide and then water is added, which can increase the analytical sensitivity by hundreds of times.

(3) Adjusting the pH of the solution can change the dissociation degree of alkali and acid by controlling the pH, or increase the volatility of the tested substance, which is beneficial for analysis.

(4) The diffusion coefficient of crushed substances in solid samples is 1 to 2 orders of magnitude smaller than in liquid samples. The diffusion rate of volatile compounds in solid samples is slow and often takes a long time to reach equilibrium. Sampling solid samples with small particles as much as possible is beneficial for shortening the equilibrium time. However, it should be noted that general crushing methods can cause sample loss. For example, if grinding generates heat, volatile components will be lost. Therefore, freeze crushing technology is often used in headspace GC to prepare solid samples. At the same time, wetting the sample with water or organic solvents (three-phase system) can also reduce the adsorption of the analyte on the solid surface.

In addition, diluting the sample is also a commonly used method to reduce matrix effects, but its cost is to reduce sensitivity.

Finally, the moisture in the sample is also an influencing factor. Although the moisture content in static headspace samples is often not as high as in dynamic headspace samples, water vapor can affect the GC separation results when the concentration of aqueous solution samples is high.

2  Sample volume

Sample volume refers to the volume of sample in the headspace sample vial, and sometimes also refers to the amount of sample entering the GC. In fact, the latter should be referred to as the injection volume.

In headspace GC analysis, the injection volume is controlled by the injection time (pressure balance system) or the size of the quantitative loop, as well as the pressure inside the quantitative loop (pressure control quantitative loop system). It is also influenced by factors such as temperature and pressure. In fact, the absolute injection volume in headspace GC analysis does not have much significance. What is important is the reproducibility of the injection volume. As long as the injection conditions can be fully reproduced, it ensures the maximum amount of reproducibility.

When reaching gas-liquid equilibrium at a certain temperature, there are:

K=CL/Cg；

Co·VL=Cg·Vg+CL·VL ；

then

Cg=Co/(K＋Vg/VL)，

Cg=Co/(K＋β)，

Cg: Concentration of the sample in the gas phase 

Co: Original concentration of sample 

VL: Sample volume

Vg: Gas phase volume 

K: Distribution coefficient 

β： Vg/VL

In equilibrium, the composition of the gas phase is proportional to the original composition of the sample. After obtaining Cg through GC analysis, the amount of the original sample can be calculated.

The sample volume in the headspace sample vial has a significant impact on the analysis results, as it directly determines the comparison β. For a given gas-liquid equilibrium system, K and C0 are constants, β It is proportional to the concentration in the headspace gas. It can also be said that when the sample volume VL increases, β Decreasing leads to an increase in Cg, resulting in an increase in sensitivity. But for specific sample systems, it also depends on the size of K. In other words, K>> β The change in sample volume has little impact on analytical sensitivity. And when K<< β The impact is significant.

For example, analyzing dioxane and cyclohexane in aqueous solutions, equilibrate at 60 ℃ using a 20mL sample vial. At this point, the K of dioxane is 642, while that of cyclohexane is 0.04. When the sample size changed from 1mL to 5mL, the analytical sensitivity (peak area) of dioxane only increased by 1.3%, while cyclohexane increased by 452%. Therefore, the sample size should be determined based on the properties of the sample system.

Another issue related to sample size is its reproducibility. Because static headspace GC often only takes samples from one sample vial once. When conducting parallel experiments, it is necessary to prepare several samples and place them in different sample vials. At this point, the reproducibility of the volume of each sample also affects the analysis results. The smaller the distribution coefficient of the component to be tested (the greater the solubility in the condensed phase), i.e. the lower the K value, then β The size of the sample is more important, that is, the larger the result error caused by the fluctuation of sample volume; On the contrary, the larger the distribution coefficient, that is, K>> β， The smaller the impact. However, in practical work, the distribution coefficient of the sample system is often unknown, so we suggest that the volumes of each sample should be kept as consistent as possible at all times.

When conducting specific analysis, the sample volume is also related to the volume of the sample vial. The upper limit of the sample volume is 80% of the full volume of the sample vial, in order to have sufficient headspace volume for sampling. 50% of the sample vial volume is often used as the sample volume. Sometimes only a few microliters of sample are used. The sample nature, analysis purpose, and method are the main factors that determine the sample volume.

3 Equilibrium temperature

The equilibrium temperature of the sample is directly related to the vapor pressure, which affects the distribution coefficient. Generally speaking, the higher the temperature, the higher the vapor pressure,

The higher the concentration of headspace gas, the higher the analytical sensitivity. The lower the boiling point of the component to be tested, the more sensitive it is to temperature. Therefore, headspace GC is particularly suitable for analyzing low boiling point components in samples. From this perspective alone, a higher equilibrium temperature is beneficial for analysis, as it can shorten the equilibrium time.

However, in headspace GC, temperature changes only affect the distribution coefficient K and do not affect the comparison β。 As mentioned earlier, we must consider both parameters simultaneously. For a given sample system, β It is a constant, and the concentration of the headspace gas is inversely proportional to the distribution coefficient K. As mentioned above, when K>> β The influence of temperature is very obvious. When K<< β Time. An increase in temperature causes K to decrease, but K+ β The change in concentration is very small, so the concentration of the headspace gas also changes very little. For example, we analyze methanol, methyl ethyl ketone, toluene, n-hexane, and tetrachloroethylene in an aqueous solution, and Table 6-3 shows the distribution coefficient K values of this system at different temperatures. Using a 6mL sample vial with a sample volume of 1mL, the comparison is 5. The table also lists 1/(K+ β) Value. Assuming that the concentrations of each component in the original sample are the same, the equilibrium temperature of 80 ℃ will increase the concentration of methanol in the headspace gas by 5 Fifteen times, methyl ethyl ketone increased by 2.61 times, toluene only increased by 25%, while n-hexane and tetrachloroethylene increased by 2 6% and 10.4%. It can be seen that the influence of temperature varies depending on the composition. For methanol and methyl ethyl ketone, increasing the equilibrium temperature can greatly improve the analytical sensitivity; For toluene and tetrachloroethylene, the impact is minimal, while for n-hexane, its impact can be completely ignored. Therefore, the equilibrium temperature should be selected based on the analysis object.

In practical work, it is often necessary to choose a lower equilibrium temperature while meeting the sensitivity requirements (and other methods can also be used to improve the analysis sensitivity!). This is because excessively high temperatures may lead to the decomposition and oxidation of certain components (with air in the sample vial), and can also increase the pressure of the headspace gas, especially when using organic solvents (therefore, organic solvents with higher boiling points should be chosen). Excessive headspace gas pressure will put forward higher requirements for the next step of pressurization, which may also cause air leakage in the instrument system.

The influence of temperature fluctuations in the heating furnace can cause significant changes in the distribution coefficient of certain samples, such as ethanol, at different temperatures. If the temperature fluctuates greatly, it is difficult to achieve repeatability. For n-hexane, its K value changes little at different temperatures, which means it is not sensitive to temperature fluctuations and may be easier to achieve repeatability.

In addition to balancing the temperature, the sampling tube, quantitative tube, and the connecting tube to the GC should all strictly control the temperature. These temperatures are often higher than the equilibrium temperature to avoid sample adsorption and condensation.

4 Balance time

The equilibrium time essentially depends on the diffusion rate of the tested component molecules from the sample matrix to the gas phase. The faster the diffusion rate, i.e. the larger the molecular diffusion coefficient, the shorter the required equilibrium time. In addition, the diffusion coefficient is related to molecular size, medium viscosity, and temperature. The higher the temperature, the lower the viscosity, and the greater the diffusion coefficient. So, increasing the temperature can shorten the equilibrium time.

Due to the diverse nature of the samples, it is difficult to predict the equilibrium time. Generally, it needs to be determined through experiments. The method is to fill a series of sample vials (5-10) with the same sample, with each vial using a different equilibrium time, and then perform GC analysis. By plotting the equilibrium time t with the peak area A of the substance to be measured, the required equilibrium time can be determined. When the equilibrium time exceeds, the peak area basically does not increase, indicating that the sample has reached equilibrium.

The equilibrium time required for gas samples or liquid samples that can be completely converted into gas is shorter (the diffusion coefficient of gas molecules is 104-105 times that of liquid molecules), usually around 10 minutes. The situation of liquid samples is more complex, and in addition to being related to sample properties and temperature, the equilibrium time also depends on the sample volume. The larger the volume, the longer the balance time required.

And the sample volume is related to the sensitivity requirements of the analysis. As mentioned earlier, for components with small distribution coefficients, increasing the sample volume can greatly improve the analytical sensitivity and correspondingly increase the required equilibrium time. For components with large distribution coefficients, increasing the sample volume has little effect on improving sensitivity, so a small sample volume can be used to shorten the equilibrium time.

Another effective way to shorten the equilibrium time of liquid samples is to use stirring technology. Modern instruments generally have this function, either mechanical vibration stirring or electromagnetic stirring, and there are several stirring speeds that can be selected based on the sample's gap. Experiments have shown that for samples with small distribution coefficients and low solubility in the condensed phase, the sampling and stirring method can shorten the equilibrium time by more than half. But for samples with large distribution coefficients, the impact is relatively much smaller.

Solid samples require longer equilibrium time. In addition to shortening the equilibrium time by increasing the temperature, reducing the size of solid particles and increasing the specific surface area can effectively shorten the equilibrium time. In addition, dissolving solid samples in appropriate solvents or soaking solid samples with solvents are commonly used methods in practical work.

5  Factors related to sample vials

(1)Sample vial

The requirements for headspace GC sample vials are accurate volume, ability to withstand certain pressure, good sealing performance, and no adsorption effect on the sample.

Although ordinary glass vials have been used in the past, now most headspace sample vials made of borosilicate glass have excellent inertness

Analysis of the vast majority of samples.

When conducting quantitative analysis, it is important to compare β The accurate value requires us to know the exact volume (volume) of the sample vial. High demand applications should not simply use the manufacturer's nominal volume. A simple method is to first weigh the empty vial with a balance, then fill it with water before weighing. The accurate volume of the sample vial can be calculated based on the density of water at the weighing temperature (such as 0.9971g/L at 25 ℃). Experiments have shown that there is an error of about 1% between the nominal volume and the actual volume of the commercially available headspace sample vials. For the same batch of sample vials, the true volume of 5 of them can be accurately measured, and the average volume can be used as the true volume of the batch of sample vials.

It is best to use the headspace sample vial only once. If you want to use it repeatedly, make sure it is clean and clean! The recommended cleaning method is to first use detergent to clean (too dirty vials can be soaked in detergent), then wash with distilled water, rinse with chromatography pure methanol, and finally dry in an oven. For newly purchased sample vials, they can generally be used without cleaning, but attention should be paid to the supplier's reputation. If it is the first time using a new supplier's product, it is best to conduct a blank analysis first to confirm whether the sample vial is clean.

(2)Sealing cap

The sealing cap is composed of a plastic or metal cover with an encrypted gasket. There are two types of screw caps that can be used multiple times and single use pressure caps. Nowadays, instruments often use disposable aluminum glands, which can ensure sealing performance after being compressed with a Capper.

There are three main types of materials for sealing gaskets, namely silicone rubber, butyl rubber, and fluororubber. The price of butyl rubber pads is low, silicone rubber pads have good high-temperature resistance, and fluorine rubber pads have good inertness. In order to prevent the adsorption of sample components by the sealing gasket, PTFE lined sealing gaskets are now commonly used, and the selection depends on the analysis conditions (temperature) and the specific situation of the sample. Low priced butyl rubber pads can be used for routine analysis, while lined silicone rubber pads are preferred for trace analysis. If necessary, confirm through blank analysis that the volatile matter in the sealing gasket does not interfere with the analysis.

After puncturing the sealing gasket once (sampling), it may leak air, and the inner gasket loses its protective effect after puncturing, and the rubber substrate may adsorb sample components. So, when multiple injections are required from a sample vial, it is best to proceed continuously, and do not leave the sample vial tied through the sealing gasket for a period of time before reuse. Correspondingly, when preparing samples, all samples should be added before sealing. For example, when adding internal standards, if sealed before adding them to the vial, the sealing gasket must be pierced, which is unfavorable for analysis.

The above content refers to "Chromatographic Analysis Methods and Applications" edited by Liu Huwei.