WHAT ARE EPA’S APPROACHES TO TRACE VOCs?

Sample Preparation Techniques



181



Method 5031 describes an azeotropic distillation technique for the determination

of nonpurgeable, water-soluble VOCs that are present in aqueous environmental

samples. The sample is distilled in an azeotropic distillation apparatus, followed by

direct aqueous injection of the distillate into a GC or GC-MS system. The method

is not amenable to automation. The distillation is time consuming and is limited to

a small number of samples.

Method 5032 describes a closed-system vacuum distillation technique for the

determination of VOCs that include nonpurgeable, water-soluble, volatile organics

in aqueous samples, solids, and oily waste. The sample is introduced into a sample

flask that is in turn attached to the vacuum distillation apparatus. The sample chamber

pressure is reduced and remains at approximately 10 torr (the vapor pressure of

water) as water is removed from the sample. The vapor is passed over a condenser

coil chilled to a temperature of 10°C or less. This results in the condensation of

water vapor. The uncondensed distillate is cryogenically trapped on a section of

1/8-in. stainless-steel tubing chilled to the temperature of liquid nitrogen (–196°C).

After an appropriate distillation period, the condensate contained in the cryogenic

trap is thermally desorbed and transferred to the GC or GC-MS using helium carrier

gas. The method very efficiently extracts organics from a variety of matrices. The

method requires a vacuum system along with cryogenic cooling and is not readily

automated.

Method 5035 describes a closed-system P&T for the determination of VOCs

that are purgeable from a solid matrix at 40°C. The method is amenable to soil/sediment and any solid waste sample of a consistency similar to soil. It differs from

the original soil method (Method 5030) in that a sample, usually 5 g, is placed into the

sample vial at the time of sampling along with a matrix-modifying solution. The sample

remains hermetically sealed from sampling through analysis as the closed-system

P&T device automatically adds standards and then performs the purge and trap. The

method is more accurate than Method 5030 because the sample container is never

opened. This minimizes the loss of VOCs through sample handling. However, it does

require a special P&T device. Oil wastes can also be examined using this method.

Method 5041 describes a method that is applicable to the analysis of sorbent

cartridges from a volatile organic sampling train. The sorbent cartridges are placed

in a thermal desorber that is in turn connected to a P&T device.



38. WHAT IS THE P&T TECHNIQUE?

The P&T method to isolate, recover, and quantitate VOCs in various environmental

sources of water has and continues to remain the premier technique for this class

of environmental contaminants. The technique was developed at the EPA in the early

1970s49 and remains the method of choice, particularly for environmental testing

labs that are regulatory driven. P&T has had the most success with drinking water

samples when combined with gas chromatography and element-specific detectors.

GC detectors will be discussed in Chapter 4. In this section, we will discuss the

EPA methods that use the P&T technique to achieve the goals of TEQA. EPA Method

502.2 summarizes the method as follows:50



© 2006 by Taylor & Francis Group, LLC



182



Trace Environmental Quantitative Analysis, Second Edition



Highly volatile organic compounds with low water solubility are extracted (purged)

from the sample matrix by bubbling an inert gas through a 5 mL aqueous sample.

Purged sample components are trapped in a tube containing suitable sorbent materials.

When purging is complete, the sorbent tube is heated and back-flushed with helium to

thermally desorb trapped sample components onto a capillary gas chromatography

(GC) column. The column is temperature programmed to separate the method analytes

which are then detected with a photoionization detector (PID) and an electrolytic

conductivity detector (ElCD) placed in series. Analytes are quantitated by procedural

standard calibration.… Identifications are made by comparison of the retention times

of unknown peaks to the retention times of standards analyzed under the same conditions used for samples. Additional confirmatory information can be gained by comparing the relative response from the two detectors. For absolute confirmation, a gas

chromatography/mass spectrometry (GC/MS) determination according to EPA Method

524.2 is recommended.



The classical purge vessel is shown in Figure 3.10. Usually, 5 mL of an environmental aqueous sample is placed in the vessel. The sample inlet utilizes a twoway valve. A liquid-handling syringe that can deliver at least 5 mL of sample is

connected to this sample inlet, and the sample is transferred to the purge vessel in

a way that minimizes the sample’s exposure to the atmosphere. Note that the

incoming inert purge gas is passed through a molecular sieve to remove moisture.

A hydrocarbon trap can also be inserted prior to the purge vessel to remove traces

of organic impurities as well. The vessel contains a fritted gas sparge tube that serves

to finely divide and disperse the incoming purge gas. These inert gas bubbles from

the purging provide numerous opportunities for dissolved organic solutes to escape

to the gas phase. The acceptable dimensions for the purge device are such, according

to EPA Method 502.2, that it must accommodate a 5-mL sample with a water column

at least 5 cm deep. The headspace above the sample must be kept to a minimum of

<15 mL to eliminate dead-volume effects. The glass frit should be installed at the

base of the sample chamber, with dispersed bubbles having a diameter of <3 mm

at the surface of the frit.

Figure 3.10 also depicts a typical trap to be used in conjunction with the purge

vessel. The trap must be at least 25 cm long and have an inside diameter of at least

0.105 in., according to Method 502.2. The trap must contain the following amounts

of adsorbents:

•

•

•



One third of the trap is to be filled with 2,6-diphenylene oxide polymer,

commonly called Tenax.

One third of the trap is to be filled with silica gel.

One third of the trap is to be filled with coconut charcoal.



It is recommended that 1 cm of methyl silicone-coated packing be inserted at

the inlet to extend the life of the trap. Method 5030B from the SW-846 series is

more specific than Method 502.2 and recommends a 3% OV-1 on Chromosorb-W,

60/80 mesh, or equivalent. Analysts who do not need to quantitate dichlorodifluoromethane do not need to use the charcoal and can replace this charcoal with more

Tenax. If only analytes whose boiling points are above 35°C are to be determined,

© 2006 by Taylor & Francis Group, LLC



Sample Preparation Techniques



Optional

foam

trap



183



Exit 1/4 IN.

O.D.

14 mm O.D.

Inlet 1/4 IN.

O.D.

Sample inlet

2-Way syringe valve

17 cm. 20 gauge syringe needle



1/4 IN.

O.D. Exit



6 mm. O.D. rubber septum

10 mm. O.D.

10 mm. 14 mm. O.D.



Inlet

1/4 IN.

O.D.



1/16 IN. O.D.

Stainless steel



13% Molecular

sieve purge

gas filter



Purge gas

flow

control



10 mm. Glass frit

medium porosity



Packing procedure

Glass 5 mm

wool

Activated 7.7 cm

charcoal



7 Ω/foot

Resistance

wire wrapped

solid

(double layer)



Grade 15 7.7 cm

silica gel



Tenax 7.7 cm

3% 0 V–1 1 cm

Glass wool 5 mm

Trap inlet



15 cm

7 Ω/foot

Resistance

wire wrapped

solid

(single layer)

8 cm



Construction

Compression

fitting nut

and ferrules

Thermocouple/

controller

sensor

Electronic

temperature

control and

pyrometer

Tubing 25 cm

0.105 IN. I.D.

0.125 IN. O.D.

Stainless steel



FIGURE 3.10 Schematic of a purge-tube-and-trap configuration.



both the charcoal and the silica gel can be eliminated and replaced with Tenax. The

trap needs to be conditioned at 180°C prior to use and vented to the atmosphere



© 2006 by Taylor & Francis Group, LLC



184



Trace Environmental Quantitative Analysis, Second Edition



Helium



N2 liq .

ECD



Trap

Gas chromatograph



Sample

Sparger

Helium



Vent

(a)



N2 liq .

ECD



Trap

Gas chromatograph



Heat



Helium

Vent

(b)



FIGURE 3.11 Schematics of the purge, trap, and thermal desorption for trace VOCs in water.



instead of the analytical GC column. It is also recommended that the trap be

reconditioned on a daily basis at the same temperature. Tenax is a unique polymer

and offers the advantage that water is not trapped to any great extent.

Commercially available P&T units are fully automated and consist of a bank of

purge vessels with switching valves that enable one vessel to be purged after another.

Figure 3.11 is a schematic diagram that clearly depicts the purge, trap, and thermal

desorption steps involved in this technique. A six-port valve, placed after the trap

and interfaced to the injection port of a GC, provides the needed connection. Referring to Figure 3.11a, the P&T step is shown. Inert gas, usually helium, enters the

© 2006 by Taylor & Francis Group, LLC



Sample Preparation Techniques



185



purge vessel while the trap outlet is vented to the atmosphere. Meanwhile, GC carrier

gas flows through one side of the six-port valve directly to the GC. The purge vessel

and the trap are generally kept at ambient temperature. Some commercially available

units provide for a heated purge vessel. Referring to Figure 3.11b, the direction of

gas flow during the desorb step is illustrated. The valve is turned and inert gas enters

and passes over the trap in a direction opposite to the P&T step. The trap is rapidly

heated to the required final temperature and the trap outlet is directed to the injection

port of the GC. GCs that are equipped with cryogenic cooling can deposit the VOCs

from the trap to the GC column inlet as a plug. The GC can then be temperature

programmed from the cryogenic temperature to the final temperature, and hence

complete the gas chromatographic separation of all VOCs.

Operating conditions drawn from EPA Method 5030B for the P&T system differ

slightly based on which determinative method is used and are presented in the

following:

Analysis Method

Purge gas

Purge gas flow rate (mL/min)

Purge time (min)

Purge temperature (°C)

Desorb temperature (°C)

Back-flush inert gas (mL/min)

Desorb time (min)



8015



8021 or 8260



N2 or He

20

15.0

85

180

20–60

1.5



N2 or He

40

11.0

Ambient

180

20–60

4



39. HOW DOES ONE GO ABOUT CONDUCTING

THE P&T TECHNIQUE?

Let us assume that the decisions of whether to screen the sample prior to conducting

P&T have been made and that we are ready to actually perform the technique. We

have available in our laboratory organic-free water, the determinative method has

been defined, and we have reviewed EPA Method 5000 for guidance with respect

to internal and surrogate standards. Environmental samples such as groundwater,

drinking water, wastewater, and so forth, have already been collected, stored in

capped bottles with minimum headspace, free of solvent fumes, and stored at 4°C.

If the sample was found to be improperly sealed, it should be discarded. All samples

should be analyzed within 14 days of receipt at the laboratory. Samples not analyzed

within this period must be noted and data are considered to represent a minimum

value. This is the essence of the so-called holding times. Holdings times, like

detection limits that were discussed in Chapter 2, are a controversial topic between

the regulatory agency and contracting laboratory. Using Method 5030B for guidance,

the sequence of steps needed to carry out the method are as follows:

•



Initial calibration: The P&T apparatus is conditioned overnight with an

inert gas flow rate of at least 20 mL/min and the Tenax trap held at a



© 2006 by Taylor & Francis Group, LLC



186



Trace Environmental Quantitative Analysis, Second Edition



•

•



•

•



temperature of 180°C. The P&T apparatus is connected to the GC, and

each purge vessel is filled to its total volume with organic-free water.

Methanolic solutions containing VOCs at high enough concentration such

that only microliter aliquots need to be added to the water in each purge

vessel are added to a series of purge vessels. A blank and set of working

calibration standards are prepared right in the purge vessels. Internal

standards and surrogates are also added as their methanolic solutions as

standard references in a manner similar to that of the analytes to be

calibrated. Refer to Chapter 2 for a discussion of the various modes of

instrumental calibration.

Conduct the P&T for the set of blanks and working calibration standards.

Initial calibration verification (ICV): Prepare one or more purge vessels

that contain the ICV. Sometimes, it is advantageous to prepare ICVs in

triplicate to enable a preliminary evaluation of the precision and accuracy

of the calibration to be made. The ICV criteria are usually given in a

determinative method such as in the 8000 series of SW-846. These criteria

must be met before real samples can be run.

Adjust the purge gas flow rate referring to the guidance given in the above

table.

Sample delivery to the purge vessel: Remove the plunger from a 5-mL

liquid-handling syringe and attach a closed syringe valve. Open the sample

or standard bottle, which has been allowed to come to ambient temperature. Carefully pour the sample into the syringe barrel to just short of

overflowing. Replace the syringe plunger and compress the sample. Open

the syringe valve and vent any residual air while adjusting the sample

volume to 5.0 mL. This process of taking an aliquot destroys the validity

of the liquid sample for future analysis. If there is only one sample vial,

the analyst should fill a second syringe at this time to protect against

possible loss of sample integrity. Alternatively, carefully transfer the

remaining sample into a 20-mL vial and seal with zero headspace. The

second sample is maintained only until such time when the analyst has

determined that the first sample has been analyzed properly. In this way,

the VOC content of the environmental sample is preserved during the

transfer to the purge vessel.



A GC-gram from EPA Method 502.2 that depicts the separated peaks from both

the PIDs and ElCDs is shown in Figure 3.12. Both chromatograms are overlayed

and show the different response for each chromatographically resolved peak between

detectors.



40. DO ALL VOCs PURGE WITH THE SAME RATE?

No, they do not, and until recently, little was known about the kinetics of purging

for priority pollutant VOCs using conventional P&T techniques. We describe the

findings of Lin et al.,51 who demonstrated that first-order kinetics are followed for

the removal of VOCs from P&T vessels. The authors make the point that because

© 2006 by Taylor & Francis Group, LLC



10.0



Time, min

30.0



20.0

18



43

+

52

35 40.0 44

47

31

41

33 36

45 49

37 46 48

+

34



24



50.0



55.0



57



PID



26

+

27



20

8



13



12 15 19

16



9

ELCD

10



14



7

2



3



1



5

4



17



11



32



50



Sample Preparation Techniques



0.0



51

53



56

21

22

23



25



30



29

28



55



42



58



40 44

45

38

39



54



FIGURE 3.12 Capillary gas chromatogram for the separation of VOCs using EPA Method 502.2.

187



© 2006 by Taylor & Francis Group, LLC



188



Trace Environmental Quantitative Analysis, Second Edition



EPA methods such as 524.2 and 624 require the use of internal standards to obtain

relative response factors, the percent purge efficiency is never considered. Some 28

priority pollutant VOCs were studied. Each VOC, dissolved in methanol, was spiked

into high-purity water at concentration levels of 1 and 10 ppb, with a 10 ppb internal

standard. Spiked samples were purged for 11 min. A GC-MS was used, and the

absolute peak area for the characteristic primary ion was used to quantitate. Experimentally, samples were purged with helium for 11 min and GC-MS data were

obtained. These purged samples were then purged a second time for the same 11-min

duration under identical conditions. Peak areas were obtained in the same manner

as for the first purge.

Let us begin to construct a mathematical view of the kinetics of P&T by first

defining the purge ratio, Pi, as being the ratio of the mean GC-MS area count for

the first purge to the mean GC-MS area count for the second purge according to

A /5

A (ave) ∑

P=

=

A (ave )

∑ A /5

5



i



1

i

2

i



j

5

j



1

ij



2

ij



Each peak area for the ith analyte was obtained by averaging peak areas over j

replicate purges, with j taken from 1 to 5. A selected list of VOCs, along with each

VOC retention time, purge ratio, and coefficient of variation (refer to Chapter 2)

from their work, is given in the following table:

VOC

Chloromethane

1,2-Dichloroethane

1,1,1,2-Tetrachloroethane

1,2,4-Trichlorobenzene

Acetone

2-Butanone



tR (min)a



Pi



% RSD



1.82

9.99

16.98

26.30

4.30

8.34



339

2.21

6.06

3.64

1.05

1.03



36

6

9

8

2

5



a30 m × 0.53 mm DB-624 (J&W Scientific); refer to EPA

Method 524.2 for GC column temperature program and other

GC conditions.



The low percent RSDs, except for the very volatile chloromethane, indicate that

Pi is constant over the range of concentrations studied. According to these authors,

consistent values of Pi would indicate that the purging VOCs from an aqueous

solution can be mathematically described by first-order chemical kinetics. How fast

any of the 28 VOCs studies are removed from a fixed volume of an aqueous sample

such as groundwater in a conventional purge vessel is expressed as –dC/dt, where

the negative sign indicates that the concentration of a VOC decreases with time.

First-order kinetics suggests that the rate is directly proportional to the concentration

of each VOC.52 Expressed mathematically,

© 2006 by Taylor & Francis Group, LLC



Sample Preparation Techniques



189



dC

= − kC

dt



(3.32)



where C is the analyte concentration at any time t and k is the first-order rate constant.

Differences in the magnitude of k depend on the chemical nature of the particular

VOC and the degree of intermolecular interaction with the sample matrix — in this

case, a relatively clean water sample. Equation (3.32) is first rearranged as

dC

= − k dt

C

Integrating between the initial concentration C1 and the concentration after the

first purge, C2, on the left side, while integrating between time t = t0 and time t =

t0 + ∆t on the right side is shown below. Identical integration is also appropriate

between C2 and C3. ∆t is the time it takes to purge the sample. We then have



∫



C2



C1



dC

= −k

C



∫



t0 + ∆t



dt



t0



Evaluating the definite integrals for both sides of the equation yields

ln(C 2 − C1 ) = − k ∆t

Utilizing a property of logarithms gives



ln



C2

= − k ∆t

C1



Expressed in terms of exponents,

C2

= e − k ∆t

C1

∴C2 = C1e



(3.33)



− k ∆t



Equation (3.33) implies that C2 can be expressed in terms of C1, k, and ∆t.

The experiment starts with each VOC at a concentration C1. After purge 1, each

analyte is at concentration C2. After purge 2, each analyte is at concentration C3.

This is depicted by

C1



Purge 1→ C 2





© 2006 by Taylor & Francis Group, LLC



C2



Purge 2 → C 3





190



Trace Environmental Quantitative Analysis, Second Edition



Because VOCs are continuously removed from a fixed volume of aqueous

sample, it is also true that

C 3 < C 2 < C1

The peak area obtained after the first purge A1 is proportional to the difference

in concentration between C1 and C2; that is,

A1 ∝ C1 − C2

Substituting for C2 using Equation (3.33) gives

A1 ∝ C1 (1 − e − k ∆t )

Likewise, the peak area obtained for each analyte after the second purge is

proportional to the difference in concentration between C2 and C3 according to

A2 ∝ C2 − C3

Substituting for C3 using Equation (3.33) gives

A2 ∝ C2 (1 − e − k ∆t )

The ratio of peak areas between the first and second purges for a given analyte,

P, can be related to these differences in concentration as follows:

P=



C1 − C 2 C1

=

C2 − C3 C2



Substituting for Equation (3.33) yields

P=



C1

C1e − k ∆t



This simplifies to

P = e k ∆t



(3.34)



Equation (3.34) suggests that P is always greater than 1 and is independent of

the initial analyte concentration. The development of Equation (3.34) has assumed

first-order kinetics. The authors observed good precision over the five replicate

experiments for all 28 VOCs studied, with the exception of the very volatile chloromethane.

If P is determined experimentally, by first obtaining the GC-MS peak area, A1,

and then obtaining A2, Equation (3.34) can be solved for the first-order rate constant, k.

© 2006 by Taylor & Francis Group, LLC



Sample Preparation Techniques



191



1

In addition, when C 2 = 2 C1, the time it takes for half of the analyte to be

removed can be found according to



1/ 2

C1 C1

C

= 1 C1 = 2 1 = e kt

C2

C1

2

1/ 2



2 = e kt ]



(3.35)



ln 2 = kt1/ 2

∴ t1/ 2 =



1

ln 2

k



Using Equation (3.35), the t1/2 values given for the three VOCs from the author’s

work are shown in the following table:

VOC

Chloromethane

1,1,1,2-Tetyrachloroethane

Acetone



t1/2 (min)

1.3

4.2

156



A purge time, ∆t = 11 min, is used in indicated EPA methods. For hydrophobic

VOCs, the t1/2 seems to be too low, and for hydrophilic VOCs, the t1/2 seems to be

too high. Other generalizations emerge with respect to the chemical nature of the

VOCs studied. The effect of deuterating 1,2-dichlorobenzene yields values for t1/2

that are quite close to each other, 5.6 and 5.8 (1,2-dichlorobenzene-d4), respectively.

Differences in structural isomers such as 1,1,2,2- vs. 1,1,1,2-tetrachloroethanes are

reflected not only in their volatility in water, but also in their gas-phase stability of

both neutral and ionic states. It would also appear that the choice of internal standard

should reflect the kinetics of purging in addition to GC relative retention time.

Choosing internal standards based on relative retention times has been the hallmark

of most EPA methods. This work would suggest that t1/2 also be taken into consideration when deciding on which internal standard to use when using P&T techniques.

We leave sample prep for VOCs and return to SVOCs. Conventional sample

extract cleanup techniques are introduced followed by SFE and then SPE. We already

discussed how LLE is used to remove unwanted neutral interferences when the target

analyte of interest is a weak acid and can therefore be ionized. Cleanup introduced

here refers to the need to remove neutral polar interferences from neutral targeted

SVOC analytes.



41. WHAT IS CLEANUP?

It is not the use of detergent to remove dirt. It is the removal of chemical interferences

from the extract following any of the extraction techniques discussed earlier. The

© 2006 by Taylor & Francis Group, LLC