Recent Development on GC

7/29/2019 Recent Development on GC

1/5

Current Analytical Chemistry, 2005, 1, 79-83 79

Some Recent Developments in Headspace Gas Chromatography

J. Y. Zhu*,1 and X.-S. Chai2

1USDA Forest Service, Forest Products Laboratory, One Gifford Pinchot Drive, Madison, WI 53726-2398, USA

2Institute of Paper Science and Technology, Georgia Institute of Technology, 10th Street, N. W, Atlanta, GA30332, USA

Abstract: In this study, recent developments in headspace gas chromatography (HSGC) are briefly reviewed.Several novel HSGC techniques developed recently are presented in detail. These techniques were developedusing the unique characteristics of the headspace sampling process implemented in commercial HSGC systemsand therefore can be easily applied in laboratory and industrial practices. These techniques greatly expand thecapability of traditional HSGC from simply analysis of volatile species to the measurements of nonvolatilespecies in solid systems and their properties in aqueous systems. The basic principles of the techniques alongwith specific applications are presented. These HSGC techniques are practical, accurate, and simple for bothlaboratory and industrial applications.

Keywords: Headspace, chromatography, vapor-liquid equilibrium, solubility, kinetics, reaction conversion, phase, fullevaporation, multiple headspace extraction.

INTRODUCTION

Headspace gas chromatography combines sampling in aheadspace of a closed container, for example, the spaceoccupied by gases or vapor in a sample vial, with analysis ofthe sampled gases or vapor by gas chromatography. Theadvantage of the headspace sampling is that direct liquid orsolid probing is avoided and complex sample matrix in aliquid or solid sample can be simplified or even eliminatedin its vapor phase. The basic principle of headspace gaschromatography (HSGC) and many useful techniques can befound in textbooks [1-3] and review articles [4-6]. The keyin HSGC is the sampling and transfer of the samples inheadspace to GC. Several techniques have been developedand some are commercialized, such as "purge and trap", vial

pressurization for static headspace sample transfer, andmultiple headspace extraction (MHE) in static headspace.Recently, solid phase microextraction (SPME) attracted greatattention for its capability in analyzing at the part per billion(ppb) level [7]. With the advances in gas chromatographictechnology, the application of HSGC has been broadenedsignificantly. Recent trends in HSGC application haveshifted to agricultural and food science [8-10], pharmace-utical and medical analysis [11-13], criminal justice and law[ 14, 15], and environmental analysis [16, 17]. A recentreview of different sample transfer methods and theirapplications can be found in the article by Snow and Slack[18]. A recent book also reviewed the application of HSGCwith mass spectrometry [19].

Traditional static HSGC relies on the vapor-liquidequilibrium (VLE) of a volatile species in a nonvolatilematrix within a closed container for determining theconcentration of the volatile species in the sample. It isbelieved that the concept of VLE, which HSGC was initially

*Address correspondence to this author at the USDA Forest Service,Forest Products Laboratory, One Gifford Pinchot Drive, Madison, WI53726-2398, USA; Tel: (608) 231-9520; Fax: (509) 277-0420; E-mail:

[email protected]

1573-4110/05$50.00+.00

developed from, should not limit further development ofHSGC technology. For example, the purge and traptechnique and the multiple headspace extraction techniquesslightly deviate from the traditional application by using theidea of "achieving complete extraction" of the species ofinterest from the sample. In commercial headspace systems,the headspace is the open space of a sample vial that isplaced in an isothermal environment. This unique isolatedisothermal environment along with the versatility andautomation of commercial systems in sampling and sampletransfer offer many opportunities for new developments inHSGC, which can provide unavailable capabilities forapplications in laboratory and industry. The present authorshave devoted much of their recent research effort todeveloping simple, accurate, and practical analytical

procedures using advanced commercial HSGC instruments.Our research focus is on the full utilization of the well-controlled headspace environment in commercial headspacesystems to explore their capabilities. This article presentspart of our own and some other research efforts on thedevelopment and applications of several HSGC techniques.

The objective is to inspire innovations and developments ofHSGC technology without being limited by the initialintentions of its invention and to fully explore its hiddenpotentials.

FULL EVAPORATION HEADSPACE TECHNIQUE

The full evaporation (FE) technique was one of the earlyHSGC techniques. It utilizes the headspace sampler as anevaporator rather than an enclosed static VLE space toachieve quantitative analysis of volatile species in samples.The technique was initially developed by Markelov andGuzowski [20] in 1993. A very small sample size was usedin a sample vial to achieve a near-complete evaporation ortransferof analytes from a condensed matrix into a vaporphase in the headspace of the sample vial in a very shortperiod of time; therefore, VLE and sample pretreatments arenot required. Detailed mathematical derivation of the FE

2005 Bentham Science Publishers Ltd.


2/5

80 Current Analytical Chemistry, 2005, Vol 1, No. 1

HSGC method can be found in the work of Markelov andGuzowski [20]. To achieve the benefit of the intendedpurpose of FE HSGC (that is, sample matrix independence)a near-complete (or constant percentage of) transfer of theanalyte into the vapor phase is required, but completeevaporation of the analyte is not necessary, which makes thetechnique practical and versatile.

We recently applied the FE HSGC technique using apulsed flame photometric detector (PFPD) [21] to solve avery difficult analytical problem in the pulp and paperindustry - measuring organic sulfur compounds, forexample, methyl mercaptan (MM), dimethyl sulphide(DMS), and dimethyl disulphide (DMDS), in kraft pulpingspent (black) liquor. The tedious and time-consumingsolvent extraction technique is most commonly used inindustry practice. Because no liquid form of MM can beobtained in an ambient environment, it is difficult toquantify MM in liquid samples by some GC methods, forexample, the standard addition method. The near-completeevaporation of the species of interest in FE HSGC makes itpossible to use an external gas standard for calibration. Itwas found that near-complete evaporation of DMS can beachieved with sample sizes as large as 400 L at headspacetemperature of 80C as demonstrated by the linearrelationship between the square root of PFPD signal countswith the sample size of the liquor [21], a uniquecharacteristic of PFPDs. The accuracy of the measurementswas demonstrated by the standard deviation of the PFPDcounts of only 1.6% when the same amount of 1-g DMSwas spiked into a set of water solutions of volume rangingfrom 6 to 21 L. The matrix-independence of the measure-ments was demonstrated by the very small variance of 4.1%of the measured PFPD signal counts of DMS from severalsamples with different composition. The samples wereobtained by spiking a fixed amount of DMS into severalnonsulfur-containing pulping liquors obtained from various

pulping stages.

Another recent application of FE HSGC was carried outby Chai and others [22] to determine the residual monomerin polymer latex. The determination of monomer in methylmethacrylate (MMA) polymer latex is very important inproviding process conversion efficiency for processoptimization. In a high conversion (>95%) polymer latexsample. the monomer species can be assumed to bedistributed in both the aqueous and solid polymer particlephase. A material balance of an analyte dispensed into asampling vial can be described by the following equation:

Zhu and Chai

denominator decreases with increase in temperature. At acertain temperature, the second term is much smaller thanVG and can be ignored. Therefore, the monomer in thepolymer latex can be accurately quantified simply byanalyzing the vapor phase in the headspace. Chai and others[22] plotted the ratio of the measured peak areas of themonomer obtained from two consecutive headspaceextractions under an equilibrium time of 5 min to indirectlyverify whether or not near-complete evaporation wasachieved. As shown in Fig. 1, a close to zero ratio wasobtained at about 110C. A headspace extraction efficiencyof 94% (estimated from the first two consecutive extractions)was achieved for a sample of 30 mg. Furthermore,measurements from a set of six synthetic samples containingthe same amount of MMA and different amount of poly-MMA ranging from 0% to 25% resulted in a standarddeviation of only 3.8%, indicating full-evaporatingheadspace analysis can be accurately applied to monomeranalysis in polymer latex at 1 10C.

(1)

where VG and VR are volume of headspace and sampleresidual volume, respectively, and CG and Caq are concentra-tion of monomer in headspace and aqueous phase,respectively. S is amount of monomer partitioned (coeffic-ient Kl) to unit mass of solid X is volumaotric content ofsolid. Therefore, the monomer concentration in headspacecan be

(2)

Because K is proportional to temperature while K1 isinversely proportional to temperature, the second term in the

Fig. (1). Effect of temperature on the ratio of the gaschromatography (GC) signals of the first two consecutiveheadspace extractions (A zero ratio indicating completeevaporation of the analyte). (Reprinted from [22] withpermission from Elsevier).

PHASE REACTION CONVERSION HEADSPACETECHNIQUE

The phase reaction and conversion (PRC) HSGCtechnique was first developed by the present authors for thedetermination of a nonvolatile species (carbonate) in kraftpulping liquor [23]. The technique completely drifts awayfrom the VLE concept, which development of HSGC wasbased on. The PRC HSGC technique is based on theconversion of a fixed percentage, including completeconversion, of an unknown nonvolatile analyte in a liquid orsolid state into a gas state through chemical reactions. Theanalyte is then analyzed through the measurements of thegasproducts in HSGC. By maintaining a constant or completeconversion efficiency in the headspace, quantitative analysisof the nonvolatile analyte in a sample can be achievedthrough calibration. With a one-step conversion reaction, thefollowing relationship can be obtained between the unknown

in the headspace gases, which can be analyzed by HSGC.species massmB in the sample and its molar concentration


3/5


(3)

where CB is the molar concentration of the analyte B in theoriginal sample to be determined, a is the reactionconversion efficiency of analyte B, and b and q are thestoichiometric coefficients of the sample B and the productgas Q to be analyzed by HSGC in the one-step reaction. VT

and VL are the volumes of the reactor (sample vial) and thepost-reaction residual liquid, respectively, and VT - VL is thepost-reaction headspace volume. Vs is the initial volume ofthe unknown sample used in measurement. CQ is the molarconcentration of the product gas Q in the headspace at thecompletion of the reaction and can be measured by GC.

We used sulfuric acid as reactantR to convert carbonatein kraft pulping liquor to carbon dioxide (CO2) to determinecarbonate using (Eq. 3) [23]. The CO2 was measured by athermal conductivity detector (TCD). The sample vial waspurged with nitrogen before adding reactants to eliminate theeffect of CO2 in ambient air on measurements. It was foundthat the standard deviation of measured TCD signal wasonly 1.3% when the same amount of 1-g sodium carbonate

was spiked into a set of aqueous solutions with sample sizeranging from 100 to 350 L. Repeatability tests indicatedthat the measurement precision was 0.62% and 3.74% forcarbonate measurements in aqueous solutions and pulpingspent liquor, respectively. Simultaneous measurements ofmultiple species, that is, carbonate and sulfide, in threetypes of kraft liquors were also demonstrated through themeasurements of CO2 and H2S. Comparisons with meas-urements obtained from coulometry and titration methodsshowed the differences are within 5%.

We later applied the PRC HSGC technique to determinecarboxyl groups in wood fibers, another difficult industrialproblem [24]. It was found that carboxyl groups in woodfibers can be converted to CO2 using bicarbonate after the

fiber is pretreated by hydrochloric acid. It was found that aconstant fraction of conversion of treated fiber to CO2 wasachieved within 10 min of reaction at 60C when 0.05 g ofair-dried fiber was used in a 4-mL 0.001 mol/L sodiumbicarbonate solution. We found that the sodium chloridepresent in the reaction solution can affect the measured CO2signal counts using TCD, which could be caused by thereduced solubility of CO2 in the solution because ofincreased ion concentration and the increased release rate ofhydrogen ions from carboxyl groups. However, when thesolution is doped with more than 0.1 mol/L of sodiumchloride, the effect of sodium ion effect on TCD signalstrength can be eliminated.

Most recently, Guzowski and others [25] applied thephase reaction conversion technique to determinehydroxylamine in pharmaceutical preparations using HSGC.Hydroxylamine was converted into nitrous oxide (N2O) in asingle-step reaction through the reaction with FeC13 in abuffer solution of sodium acetate. The N2O product gas wasmeasured by an electron capture detector (ECD). The effectof matrix composition (Fe3+ ion, base, buffer, solvent) wasfound to be significant on an ECD-measured signal of N2O.It was found that the ECD response was greatest when equalmolar amounts of iron to base were used. A hydroxylamine

recovery efficiency of 98% was obtained from a samplematrix that only contains drug substance.

Cuuent Analytical Chemistry, 2005, Vol. 1, No. 1 81

HEADSPACE GAS CHROMATOGRAPHY FORSOLUBIL ITY DETERMINATION

Solubility determination using HSGC was firstdemonstrated by the present authors [26] through themeasurements of the VLE behavior of a volatile speciesdoped in a salt solution. The technique is based on the well-known fact that the VLE partitioning behavior of a volatile

species is affected by the presence of salt in the solution dueto molecular interaction (the so-called"salting in"or"saltingout"effects). The effect of inorganic salt on VLE behavior ofmethanol aqueous solution was well documented in one ofour early studies [27]. It was found that the presence ofinorganic salt reduced the solubility of methanol in water,that is, the VLE partitioning coefficient of methanol isproportional to the salt concentration. Further studyindicates that the logarithmic VLE partitioning coefficient iscorrelated to the ionic strength in the solution regardless ofthe type of salt used. It is expected that when the salt issaturated in the solution, a constant VLE partitioningcoefficient can be obtained. Therefore, the solubility of thesalt can be obtained. Because the headspace sampler is an

excellent isothermal chamber, the solubility of a salt at agiven temperature can be determined. Fig. 2 shows thedetermination of the solubility of sodium carbonate andsodium sulfate in aqueous solutions at 60C. Goodagreement between the measured solubility data shown in(Fig. 2) and those reported in literature were obtained. Thedifference was about 7%. We also found that the solubilityof sodium carbonate in a two-salt aqueous solution and in aspent pulping liquor (complex matrix with multiple saltsand other organic materials ofvarious concentrations) can bedetermined using the HSGC technique [26].

Fig. (2). Effect of sodium sulfate on the vapor-liquidequilibrium behavior of methanol in sodium sulfate aqueoussolutions (GC, gas chromatography). (Reprinted from [26] withpermission from Elsevier).

Some colloidal in suspension can interact with volatilespecies, which can be used to quantify suspended colloidalin the system. Our laboratory study indicated that bymeasuring the effect of colloidal on the VLE behavior of the

volatile species, such as toluene (Fig. 3), the amount of


4/5

82 Current Analytical Chemistry, 2005, Vol. 1, No. 1

colloidal (pressure sensitive adhesive) in suspension can bequantified, which is very important in paper recyclingpractice because adhesives on wastepaper can disintegrate andform stickies that affect paper machines.

Zhu and Chai

where a and b can be calculated [35] and only depend on theheadspace operating conditions and dimensional Henry'sconstant of the solute in the system. We applied Eq. 6 tostudy the kinetics of slow methanol formation process inspent pulping liquor at temperature of 70C. Because ofmethanol formation, the logarithmic linear relationshipbetween the GC signal peak area and extraction number [33]is no longer valid as shown in Fig. 4. The logarithmic linear

relationship was also found in dynamic headspace and purgeand trap techniques by Careri and others [36], a characteristicof the diffusion process of a solute from a solution to aheadspace. The authors studied the kinetics of the dynamicheadspace and purge and trap techniques to determine theeffectiveness of the two techniques for quantifying traceamounts (ppb level) of nitrous oxide in seawater. Thedeviation from the logarithmic linear relationship is moreevident when a longer time delay between two consecutiveheadspace extractions was used. The same amount ofmethanol formation was obtained when different headspaceextraction delay times were used (Fig. 5), indicating thevalidity of the measurements and technique developed.

Fig. (3). Effect of colloidal substance on the vapor-liquidequilibrium behavior of toluene in aqueous solution (GC, gaschromatography).

MULTIPLE HEADSPACE EXTRACTION FORPROCESS KINETICS STUDY

Multiple headspace extraction (MHE) is similar to

dynamic headspace gas extraction but is carried out in steps.Early development of MHE GC was carried out byMcAuliffe [28, 29] and Suzuki and others [30]. The methodwas further developed by Kolb [31], Ettre and others [32],and Kolb and Ettre [33]. However, MHE GC was not apractical tool until the present authors developed andexperimentally demonstrated an MHE HSGC technique thatcan simultaneously determine solute concentration andHenry's constant [34]. The technique was based on VLE andmaterial balance of solute to obtain a key relationship amongthe MHE-measured GC detector signal peak areas,Ai [34]:

Later, we incorporated a sink or source term in thematerial balance equation (Eq. 4) to account for massconservation due to chemical reaction or adsorption ordesorption by a solid phase of the solute in the system understudy [35]. When the rate of change of the solute due to thesource or sink in the system is much slower than the rate ofVLE of the solute and the increase of the solute did notaffect the solute infinite dilution assumption, Henry's lawstill holds approximately. Furthermore, the ratio between thechanges of the concentrations of the solute in the vapor andin the liquid phase can be approximated by thedimensionless Henry's constant. Then, Eq. 4 can bemodified as follows:

(4)

where DAi is the change of GC signal peak area due to thechange of solute concentration in the vapor phase(headspace). Therefore, the kinetics of the solute can beexpressed as:

Fig. (4). A typical relationship between the gas chromatography(GC) signal peak area and the headspace extraction number inmultiple headspace measurements for the system with solutemass formation. (Reprinted from [35] with permission fromElsevier).

SUMMARY

Headspace gas chromatography is a mature and reliable

technology. Many HSGC techniques can be implemented incommercial systems. Because of the unique isothermalenvironment of the headspace and the headspace samplingprocess along with chromatographic separation of species,HSGC has great potential for many laboratory and industrialapplications other than quantitative analysis of volatilespecies, which was the initial intended purpose of itsdevelopment. This review article presents some recentdevelopments in exploration of HSGC potential. Theobjective of this article is to encourage further innovations

in HSGC.

(5)


5/5


Fig. (5).The measured time-dependent methanol formation in akraft black liquor as percentages of the methanol in the originalsample at temperature 70C. (Reprinted from [35] with

permission from Elsevier).

REFERENCES

Cuuent Analytical Chemistry, 2005, Vol. 1, No. 1 83

Received: 16 September, 2004 Accepted: 30 September, 2004

Recent Development on GC

Documents

Transcript of Recent Development on GC