Journal of Chromatographic Science, Vol.

44, July 2006

HS–SPME Determination of Volatile Carbonyl and

Carboxylic Compounds in Different Matrices

Elena E. Stashenko1,*, Amanda L. Mora2, Martha Cervantes1, and Jairo R. Martínez1

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

1Laboratorio
de Cromatografía-CIBIMOL, Escuela de Química, Facultad de Ciencias, Universidad Industrial de Santander, A.A. 678,
Bucaramanga, Colombia and 2Escuela de Química, Universidad Nacional de Colombia, Sede Medellín, Colombia

Abstract place safety studies (2–4). The achievement of trace-level detec-

tion limits for these polar compounds normally requires sample
Specific chromatographic methodologies are developed for the preconcentration (cryogenic trapping, adsorption on sorbent car-
analysis of carboxylic acids (C2–C6, benzoic) and aldehydes (C2–C10) tridges, or chemical derivatization) (5). Sampling with sorbents
of low molecular weight in diverse matrices, such as air, automotive followed by thermal desorption has been employed successfully in
exhaust gases, human breath, and aqueous matrices. For carboxylic the determination of nonpolar contaminants in air. However, the
acids, the method is based on their reaction with pentafluorobenzyl application of this method to the determination of polar com-
bromide in aqueous solution, followed by the separation and pounds is hindered by their strong adsorption and the partial loss
identification of the resultant pentafluorobenzyl esters by means of of these compounds on the surfaces of injectors and chromato-
headspace (HS)–solid-phase microextraction (SPME) combined with
graphic columns (5,6). A solution to this drawback has consisted
gas chromatography (GC) and electron capture detection (ECD).
Detection limits in the µg/m3 range are reached, with relative
of the combination of chemical derivatization with the use of
standard deviation (RSD) less than 10% and linear response (R2 > solid sorbents (3,7). Thus, the most common analytical procedure
0.99) over two orders of magnitude. The analytical methodology for for the determination of aldehydes and ketones in air involves the
aldehydes is based on SPME with simultaneous derivatization of the circulation of air over a solid sorbent, which has been previously
analytes on the fiber, by reaction with pentafluorophenylhydrazine. impregnated with 2,4-dinitrophenylhydrazine (8–10). The deriva-
The derivatization reagent is previously deposited on the SPME tives are then removed with solvent and quantitated by means of
fiber, which is then exposed to the gaseous matrix or the HS of the high-performance liquid chromatography (HPLC)–UV, which
sample solution. The pentafluorophenyl hydrazones formed on the achieves detection limits in the µg/m3 range. A chlorinated
fiber are analyzed selectively by means of GC–ECD, with detection derivatization analogue, 2,4,6-trichlorophenylhydrazine, has
limits in the ng/m3 range, RSD less than 10%, and linear response been employed with gas chromatography (GC)–electron capture
(R2 > 0.99) over two orders of magnitude.
detection (ECD) to achieve detection limits in the sub-µg/m3
range (11), in a procedure that involves sealing the cartridge after
sampling and heating it to 100°C for 6 min to perform the deriva-
Introduction tization. However, this method requires the addition of an ozone
trap to prevent oxidation and deactivation of the derivatization
agent. Similar detection limits have been obtained in the
Carboxylic and carbonylic substances are part of the great GC–MSD quantitation of the oximes formed by the reaction of
variety of organic volatile compounds that may be present in air carbonyls with O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine
as contaminants. These substances may be dispersed in the atmo- (PFBHA) over a polymeric sorbent (Tenax) (3). Cecinato et al. (12)
sphere from anthropogenic and biogenic primary sources, or they reached parts per trillion by volume detection limits in the deter-
may originate in the atmosphere itself as a result of the different mination of carbonyls in air when they utilized silica cartridges
photochemical reactions that all the hydrocarbon-polluting impregnated with 2,3,4,5,6-pentafluorophenyl hydrazine (PFPH)
agents can undergo (1). The main sources of carboxylic acids are to form hydrazones that were extracted in dichloromethane–ace-
the oxidation of aldehydes and the reactions of alkenes with tonitrile (4:1) and quantitated with GC–MSD in selected ion mon-
ozone. Because of their ample participation in the photochemical itoring mode.
reactions of the troposphere and their injurious effects on the The most common analytical procedures for carboxylic acids in
atmosphere and human health, the determination of these com- air and car exhaust include a sampling step with either liquid
pounds in air has become important for environmental and work- (potassium hydroxide or calcium hydroxide aqueous solutions) or
solid traps (2,13–17). Quartz, glass wool, or cellulose are typical
* Author to whom correspondence should be addressed: email elena@tucan.uis.edu.co. supports used in the latter type of traps to embed potassium

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission. 347
 Journal of Chromatographic Science, Vol. 44, July 2006

hydroxide, sodium carbonate, or calcium hydroxide. Potassium- ethyl acetate, and sulfuric acid were purchased from Mallinckrodt
impregnated silica gel-C18 cartridges have been employed as well (Mexico D.F., Mexico). Propionic acid, sodium chloride, and
(18,19). These sampling approaches have been combined with potassium carbonate were obtained from Merck (Schuchardt,
various instrumental techniques, such as ion chromatography, Germany). Zero air (20% O2 in N2) and high-purity gases
liquid chromatography, capillary electrophoresis, and for chromatography were obtained from Aga Fano S.A.
GC (13–19). However, low selectivity, low sensitivity, and long (Bucaramanga, Colombia). Fused-silica fibers coated with PDMS
analysis times are common to these procedures. Ion chromatog- (100 µm), poly(acrylate) (85 µm), or PDMS–divinylbenzene
raphy and GC are the preferred quantitation methods for (DVB) (65 µm) for use in SPME were purchased from Supelco
these analytes, but a sensitive analysis with few interferences (Bellefonte, PA). The performance of all three fiber types in the
requires the introduction of a derivatization step during sample determination of carboxylic acids and carbonyl compounds was
preparation (2,20,21). Sylilation and alkylation are the compared. Because of their superior retention capacity, the
most common approaches. Ester formation has been achieved poly(acrylate) and PDMS–DVB coatings were chosen for the car-

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

by means of reagents such as boron trifluoride–butanol (22–24), boxylic acid and carbonyl compound determinations, respec-
boron trifluoride–methanol (13), diazomethane (25,26), tively.
pentafluorobenzylbromide (PFBBr) (1), and dibromoacetophe-
none (13,17). Air sampling
Solid-phase microextraction (SPME) constitutes an interesting Air samples were collected with an M18 VacBag sampler system
and efficient variation of the trapping–derivatization methods (SK-M180), together with a gas sampling pump (TBP-102) and
needed for the determination of carbonyls and carboxylic acids in 3-L gas sampling Tedlar bags, which were obtained from Apex Inc.
air because it permits the combination of concentration, derivati- Instruments (Holly Springs, NC). A sampling flow of 0.02 L/min
zation, and thermal desorption in a single device. Martos and during 12.5–15 min was employed in all cases. For carboxylic
Pawliszyn employed a polydimethylsiloxane (PDMS)-coated acid determination, the contents of the Tedlar bag were then bub-
SPME fiber, previously impregnated with PFBHA, for the bled at 0.2 mL/min through two impingers connected in series,
GC–MSD determination of formaldehyde in air (27). This method each one containing potassium carbonate solution (50 mg/L
showed good agreement with the results of the National Institute and 30 mL).
of Occupational Safety and Health (NIOSH) method when it was
applied in the field and indoors (28). Pan and Pawliszyn (29) and Carboxylic acid determination
Lee et al. (30) employed diazomethane to derivatize long-chain Stock solutions of the carboxylic acids (10mM) in acetone were
carboxylic acids after their absorption on PDMS-coated SPME used to prepare individual aqueous solutions of various concen-
fibers. For low-molecular-weight carboxylic acids, 1-pyrenyldia- trations (1.27, 2.70, 5.20, 7.20, 10.8, 21.6, 32.0, 43.2, 86.4, and
zomethane was used. It was loaded onto the SPME fiber before its 172.8µM). The derivatization reaction conditions for aqueous
exposition to the sample that contained the carboxylic acids (29). solutions of carboxylic acids were adapted from those used by
Detection limits in the ng/L to µg/L range were obtained with Kawahara (32) with nonaqueous media. Instead of using acetone,
these methods for the determination of carboxylic acids in as in Kawahara’s method, the derivatization reaction was per-
aqueous solutions. formed directly in aqueous solution. The esters thus formed were
This work reports on the development of analytical protocols extracted by means of SPME rather than with solvent extraction
for the determination of low-molecular-weight carboxylic acids (procedure used in Kawahara’s method). Reagent concentration,
and aldehydes in air, using derivatization to pentafluorophenyl or solution pH, temperature, and reaction time were changed, one
pentafluorobenzyl analogues, which were quantitated with high at a time, within intervals determined by means of preliminary
selectivity and sensitivity by means of headspace (HS)–SPME and experiments, in order to determine the conditions that afforded
GC–ECD. An on-fiber derivatization SPME–GC–ECD method- an improved overall esterification yield without increasing its
ology, previously developed in our laboratory for the determina- standard deviation. For the standard derivatization conditions
tion of carbonyls generated during lipid peroxidation (31), was selected, the acid solution (9 mL) was transferred to a 12-mL vial
applied to air analysis. Examples of the application of the analyt- where it was mixed with potassium carbonate (1 mL and 0.5 g/L)
ical protocols to the determination of these analytes in diverse and a nine-fold excess of PFBBr (430mM in acetone). The vials
matrices (breath, foot sweat, car exhaust, and indoor air at a shoe were sealed, stored at 60°C for 4 h, and, finally, transferred to an
factory) are included. ice-water bath for 10 min. This procedure was employed with all
of the samples and with the stock solutions to prepare the
calibration curves for the HS–SPME–GC–ECD analysis.
Quadruplicate injections of an equimolar mixture of the car-
Experimental boxylic acid derivatives obtained by this procedure showed rela-
tive standard deviations (RSDs) below 0.5% for their retention
Reagents and materials times.
Pentafluorobenzyl bromide, pentafluorophenylhydrazine, Poly(acrylate)-coated (85 µm) SPME fibers were used to sample
butanoic, pentanoic, hexanoic, and benzoic acids, as well as the pentafluorobenzyl esters of the carboxylic acids formed in
ethanal, propanal, butanal, pentanal, hexanal, octanal, nonanal, aqueous solution by reaction with PFBBr. The sampling condi-
and decanal were obtained at 99% or higher purity from Aldrich tions were established after a comparative study of the amount of
(Milwaukee, WI). Analytical-grade glacial acetic acid, acetone, PFBBr derivatives extracted when the solution pH (3,7,9), ionic

348
Journal of Chromatographic Science, Vol. 44, July 2006

strength (0, 100, 150, 200 g/kg NaCl), extraction temperature detector temperatures were maintained at 250°C and 270°C,
(30°C, 40°C, 60°C), and time (20, 30, 40, 80 min) were varied sys- respectively. The split ratio was 10:1.
tematically. The extraction was performed by exposure of the fiber Mass spectra (electron impact at 70 eV) were obtained with an
to the HS of the magnetically stirred (900 rpm) aqueous solution HP-5890A Series II GC coupled to an HP 5972 MSD, equipped
(2 mL) in 5-mL vials for 40 min at 30°C, after addition of NaCl with a split/splitless injector (split 1:30) and a data system (HP
(200 g/kg) and pH adjustment (pH = 7). ChemStation) with NBS75K and Wiley 138K mass spectra
libraries (John Wiley & Sons, NY, New York). An HP-5MS 50-m ×
Aldehyde determination 0.25-mm (i.d.) capillary column coated with 5% phenyl
The procedure described by Stashenko et al. (33) was employed poly(methylsiloxane) (0.25-µm film thickness) was used. The GC
to prepare hydrazone derivatives of the individual aldehydes. A oven was programmed from 60°C (2 min) to 250°C (2 min) at
PFPH methanolic solution was reacted with a 50% stoichio- 8°C/min. The temperatures of the ionization chamber and of the
metric excess of each one of the carbonyl compounds. The transfer line were 185°C and 285°C, respectively. Mass spectra

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

resulting derivatives were used to prepare hidrazone solutions of and reconstructed chromatograms were obtained by automatic
1.0, 2.5, 5.0, 10.0, 20.0, 30.0, 40.0, and 60.0µM, which were used scanning in the mass range m/z 40–350 at 3.5 scan/s.
for the external calibration and the identification of the analytes
of interest by comparison of their retention times with those of
the standards.
The gaseous aldehyde standards were prepared using volu- Results and Discussion
metric injection. A measured amount of an aldehyde standard
(C2–C10) was deposited in a Tedlar bag, which contained 3 L of Analytical methods
zero air, through the septum that these bags have using a The analytical methods employed in this work to determine
microsyringe (0.1–1.0 µL). After the sample was prepared, the carboxylic acids and carbonyl compounds at low concentrations
needle of the SPME syringe was inserted through the septum, were designed to obtain high sensitivity and selectivity and to
immediately exposing the PDMS–DVB coating to the analytes. include as few steps as possible, in order to make them swift and
After 30 min, the fiber was withdrawn and inserted in the injec- to decrease the likelihood of cross contamination and analyte
tion port of the GC. The air samples were collected in 3-L Tedlar loss. The analytes were converted into their pentafluorophenyl or
bags and analyzed following the same procedure. pentafluorobenzyl derivatives in order to obtain high sensitivity
with the ECD, which in turn reduces the amount of sample
Chromatographic analysis required for the analysis. The use of SPME for on-fiber derivatiza-
Sample analysis was carried out in an HP-5890A Series GC tion of aldehydes or for HS–SPME sampling of the PFBBr car-
(Hewlett-Packard, Palo Alto, CA) equipped with split/splitless boxylic acid derivatives formed in slightly acidic aqueous solution
injection and ECD (63Ni-ECD). The chromatographic data were eliminated the steps of solvent extraction and concentration
processed with HP ChemStation A.06.03 software (Hewlett- under nitrogen, common to many existing analytical methods.
Packard). The chromatographic column was a DB-5 (J&W For the determination of carboxylic acids in air, the sample may
Scientific, Folsom, CA) fused-silica capillary column coated with be collected in the field and brought to the laboratory in Tedlar
5% phenyl poly(dimethylsiloxane) (30 m × 0.25-mm i.d., with bags, or it could be bubbled directly into 2- × 30-mL potassium
0.25-µm df ). Helium (99.995%) was used as the carrier gas (1 carbonate (50 mg/L) solutions. The partial ionization of the car-
mL/min) with a column inlet pressure of 90 kPa. An boxylic acids under these conditions diminishes losses because of
argon–methane (9:1 v/v) mixture was used as the auxiliary gas in volatilization. The recovery calculated after derivatization and
the detection system at 60 mL/min. The oven temperature was HS–SPME was near 80% for acetic and hexanoic acids and higher
programmed from 60°C (2 min) to 250°C (2 min) at 8°C/min for than 99% for acids of 3-, 4-, and 5-carbon atoms. However, it was
the analysis of the acid derivatives and from 120°C to 250°C (4 very low (6%) for benzoic acid, but the derivatization and
min) at 8°C/min for the hydrazone derivatives. The injector and HS–SPME extraction showed relatively similar yields for all car-
boxylic acids examined, and the trapping in aqueous solution
showed lower recoveries for benzoic acid caused by condensation
Table I. Calibration Data for the Determination of losses because of its lower volatility (boiling point = 249°C), rela-
Carboxylic Acids in Air tive to the other carboxylic acids examined (for hexanoic acid,
boiling point = 205°C). Thus, the recovery calculated for the
Recovery (%)
method, excluding the trapping step (bubbling through potas-
Carboxylic LOD LOQ From sium carbonate solution), shows high values for all carboxylic
acid (µg/m3) (µg/m3) R2 Overall aqueous trap acids studied (Table I). The calibration curves and response fac-
tors for each analyte were calculated from the chromatograms
Ethanoic 20.16 40.32 0.9994 86 97.2
obtained by means of HS–SPME and GC–ECD for esters formed
Propanoic 11.56 23.11 0.9961 97 94.8
from standard acid solutions of different concentrations. Table I
Butanoic 25.38 50.75 0.9986 100.2 98
Pentanoic 36.77 73.53 0.9968 100.0 98 contains the calibration curve results for the determination of
Hexanoic 105.94 211.88 0.9965 88 95.9 carboxylic acids in air. Sub-µg/m3 quantitation limits were
Benzoic 271.11 542.21 0.9965 6.5 92.4 reached for all carboxylic acids examined, and the linearity of the
calibration curves presented R2 values above 0.99 in all cases.

349
 Journal of Chromatographic Science, Vol. 44, July 2006

The identity of the PFBBr carboxylic acid derivatives was the combination of SPME sampling with derivatization and chro-
confirmed by MS. Their mass spectra contained molecular ions matographic electron capture detection.
with relatively low abundances (1–32%), which is in agreement
with the observations made by Brill (34), who found abundances Applications
between 1% and 20%. The base peak in all of the mass spectra of Aldehydes at a shoe factory
the PFBBr derivatives was found at m/z 181, assigned to the The air in the main floor of a local footwear factory
[C6F5CH2]+ ion. It was accompanied by a signal at m/z 161, corre- (Bucaramanga, Colombia) was sampled in two areas that corre-
sponding to a loss of HF from the pentafluorobenzyl fragment. sponded to two distinct manufacturing operations. In area 1, the
Aldehyde determination by on-fiber derivatization with PFPH leather pieces are mounted and glued around the shoe frame-
has been showed to permit femtogram-level detection limits works, and edges are burned off. In area 2, outsoles are lacquered,
(31,33). The method previously developed in our laboratory was and cleaning agents are applied to remove impurities and provide
applied to the air samples contained in Tedlar bags. Synthetic air luster to the leather. All eight aldehydes were detected at concen-

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

samples containing known amounts of aldehydes were used to trations that could cause allergies and discomfort to the workers.
determine the fiber exposure time and PFPH load amounts most In contrast, a duplicate determination of aldehydes in air from a
appropriate to the analyte levels examined. These samples were pedestrian sidewalk at our university campus did not detect any of
also employed to construct the calibration curves (Table II). Very these aldehydes at concentrations above the method’s detection
good linearity (R2 > 0.99) was achieved for all the aldehydes limit. Air sampling (two Tedlar bags per site) was performed on a
studied, and their quantitation levels were in the sub-ng/m3 Friday afternoon, at the end of the working day (5 p.m.) and week.
range. Table III contains the aldehyde concentrations found in the air at
The novelty of the methods applied in this work to the determi- the two sampled factory sites. However, the NIOSH-recom-
nation of polar contaminants in diverse sample matrices resides mended time-weighted average exposure limits for acetaldehyde
in the particular conjugation of rather simple setups, a smaller (324 mg/m3) and pentanal (175 mg/m3) were above the concen-
number of steps, shorter experimental time, and lower environ- trations determined at the footwear factory. Currently, there are
mental impact (negligible solvent consumption) to achieve high no NIOSH recommended limits for workplace exposure to the
selectivity and sensitivity. These advantages result primarily from other aldehydes studied (35).

Aldehydes and carboxylic acids in car exhaust

Table II. Calibration Parameters for the Determination of The exhaust fumes from two compact cars with 4-cycle com-
Aldehydes in Air bustion engines, with and without catalytic converter, were sam-
pled by means of a 1-m stainless steel tube (0.8-mm i.d.) inserted
LOD LOQ 10 cm into their exhaust pipes. The stainless steel tube was con-
Aldehyde (ng/m3) (ng/m3) R2 Recovery (%)
nected to the air sampling Tedlar bag by means of Teflon tubing
Ethanal 0.111 0.222 0.9989 98
Propanal 0.129 0.257 0.9983 98 Table IV. Concentration of Carboxylic Acids and
Butanal 0.128 0.256 0.9983 98 Aldehydes Found in Car Exhaust Gases
Pentanal 0.151 0.302 0.9980 98
Hexanal 0.160 0.320 0.9989 98 Concentration (mg/m3)*
Octanal 0.036 0.073 0.9979 98
Nonanal 0.035 0.069 0.9980 99 Without With
Decanal 0.298 0.596 0.9968 99 Acid catalytic converter catalytic converter

Acetic 43.6 ± 0.90 5 ± 3.1

Propanoic 6.2 ± 0.17 0.0 ± 0.1
Table III. Concentration of Aldehydes in Air at Two Butanoic 8.8 ± 0.25 0.053 ± 0.0063
Sections of a Footwear Factory Pentanoic 0.217 ± 0.0070 0.18 ± 0.017
Hexanoic 0.20 ± 0.011 0.045 ± 0.0045
Concentration (ng/m3)* Benzoic 2.10 ± 0.070 0.274 ± 0.0022
Aldehyde Site 1 Site 2 Aldehyde Concentration (ng/m3)*

Ethanal 54 ± 4.3 53 ± 1.2 Ethanal 800 ± 140 430 ± 15

Propanal 58 ± 1.7 103.3 ± 0.93 Propanal 170 ± 20 160 ± 19
Butanal 9.8 ± 0.37 7.2 ± 0.21 Butanal 100 ± 14 54.1 ± 0.81
Pentanal 15 ± 1.1 15.4 ± 0.66 Pentanal 250 ± 25 100 ± 1.5
Hexanal 12 ± 1.1 13.2 ± 0.37 Hexanal 33 ± 3.4 73 ± 2.5
Octanal 10 ± 1.1 8.0 ± 0.26 Octanal 15.2 ± 0.85 47 ± 5.0
Nonanal 11.0 ± 0.84 13.4 ± 0.20 Nonanal 110 ± 16 12 ± 1.8
Decanal 34 ± 2.4 25.8 ± 0.50 Decanal 64 ± 5.0 19 ± 3.1

* Average concentration ± s (n = 2). * Average concentration ± s (n = 2).

350
Journal of Chromatographic Science, Vol. 44, July 2006

sufficient in length (6 m) to cool down the air before its admission in the concentrations of acetic and propionic acids, which varied
into the sampling bag. The Teflon tubing also served to trap most by factors of up to three-fold. Blanks of the shoes, talc, cotton
of the particulate matter released by the car without a catalytic insole, and aqueous solution employed in the extraction of
converter, which formed a dark deposit on the tubing walls. Very the insole revealed the presence of small amounts of esters
few particulates reached the bag, but none reached the impingers of acetic and propionic acids as contaminants of the derivatization
used to trap the carboxylic acids. The car with a catalytic con- reagent. Benzoic acid (40 mg/kg) was found in the aqueous
verter did not release particulate matter. Table IV shows the con- extract (100 mL, 37°C) of the sport shoe employed in the
centrations of aldehydes and carboxylic acids found in duplicate study. This indicates that the shoe materials contributed to the
samples taken from the car exhausts. The NIOSH-permissible benzoic acid determined in the volunteer’s foot sweat. However,
exposure limits (8-h time-weighted averages) for acetic, GC–MS studies have documented the presence of benzoic
propanoic, and pentanoic acids were 25, 30, and 176 mg/m3, acid among 346 different compounds collected from human skin
respectively (36). The C3 and C4 carboxylic acids were the species emanations (37–39). Indeed, samples of foot sweat from volun-

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

most effectively reduced by the catalytic converter, and the acetic teers who used talc showed a 60% decrease in benzoic acid
acid emissions were practically the same in both types of cars concentration. Thus, the benzoic acid determined in foot sweat
sampled. Figure 1 presents chromatographic profiles for the samples resulted from metabolism and contributions from the
PFBBr ester derivates of carboxylic acids in an aqueous solution shoe materials.
and in the exhaust fumes from the car without a catalytic con-
verter. On the other hand, all aldehydes examined were found at a Breath analysis
lower concentration in the exhaust from the car with a catalytic The on-fiber derivatization method was used for the
converter. HS–SPME–GC–ECD determination of aldehydes in human
breath. SPME breath sampling was performed with a device in
Carboxylic acids in foot sweat which the SPME holder was attached to the end of a glass tube
During a separate forensic research project conducted in our equipped with a side arm, similar to that employed by Grote and
laboratory, five male volunteers were asked to wear sport shoes for Pawliszyn (40). After placing the other end of the glass tube in
8 h and a cotton-gauze insole within their socks, maintained in their lips, volunteers held their breath for 10 s and then slowly
close contact with their feet by means of plastic wrap. After the released it through the glass tube and over the PFPH-loaded
insole was collected from the volunteers, it was cut into 2-cm SPME fiber for 50 s. The procedure was repeated five times, for a
strips. A weighed strip was transferred to a 25-mL flask and total period of 10 min.
extracted with HPLC-grade water (25 mL) for 1 h at room tem- The PFPH-loaded PDMS–DVB (65 µm) SPME fiber was exposed
perature. An aliquot (9 mL) of this extract was subjected to the (10 min) to the breath of individuals before and after smoking or
PFBBr derivatization and HS–SPME sampling procedure
described in the Carboxylic acid determination section. Table V
Table V. Concentration of Low-Molecular-Weight
shows that acetic, propionic, and benzoic acids were the most
Carboxylic Acids Found in Foot Sweat from Five Male
abundant carboxylic acids in the volunteers’ foot sweat. Volunteers
Volunteers presented concentration profiles that differed mainly
Carboxylic acid concentration (mg/kg)
Volunteers C2 C3 C4 C5 C6 Benzoic

A 432.1 46.4 1.6 2.6 2.6 7.4

B 585.8 41.7 1.2 2.0 1.3 15.3
C 233.1 96.4 1.1 1.3 0.9 24.4
D 460.9 28.3 1.1 2.0 1.2 19.7
E 256.7 87.3 0.1 0.0 0.3 22.1

Figure 1. Gas chromatograms obtained for esters after the acid derivatization Figure 2. Chromatographic profile obtained with ECD of the PFPH-hydrazone
with PFBBr for: a standard aqueous solution (42µM) (A) and an exhaust gas derivatives of aldehydes present in the breath of a volunteer before and after
sample obtained from a car without a catalytic converter (B). the ingestion of 200 mL of red wine.

351
 Journal of Chromatographic Science, Vol. 44, July 2006

Acknowledgments

Financial support from Colciencias (Grant 1102-05012401) is

gratefully acknowledged.

References
1. Ch.-J. Chien, M.J., Charles, K.G. Sexton, and H.E. Jeffries. Analysis of
airborne carboxylic acids and phenols as their pentafluorobenzyl
derivatives: gas chromatography/ion trap mass spectrometry with a
novel chemical ionization reagent, PFBOH. Environ. Sci. Technol.

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

32: 299–309 (1998).
2. E. Dabek-Zlotorzynska and M. McGrath. Determination of low-
molecular-weight carboxylic acids in the ambient air and vehicle
Figure 3. Typical chromatographic profile obtained with ECD of the emissions: a review. Fresenius’ J. Anal. Chem. 367: 507–18 (2000).
PFPH–hydrazone derivatives of aldehydes present in the breath of a volunteer 3. S.S.H. Ho and J.Z. Yu. Feasibility of collection and analysis of air-
after smoking a complete cigarette. borne carbonyls by on-sorbent derivatization and thermal desorp-
tion. Anal. Chem. 74(6): 1232–40 (2002).
4. United States Environmental Protection Agency. Compendium of
wine consumption. Figure 2 shows the variation of the chro- Methods for the Determination of Toxic Organic Compounds in
matographic profile as a function of time after the ingestion Ambient Air, 2nd ed. Ohio EPA, Center for Environmental Research
of 200 mL of red wine. The crest in acetaldehyde released at Information, Cincinnati, OH, June, 1999.
5. K. Dettmer and W. Engewald. Adsorbent materials commonly used
60 min agrees with the known ethanol metabolic process in the
in air analysis for adsorptive enrichment and thermal desorption of
human body. Figure 3 contains a typical chromatographic profile volatile organic compounds. Anal. Bioanal. Chem. 373: 490–500
of the breath of a volunteer after smoking a full cigarette. (2002).
Methanal, ethanal, propanal, and traces of pentanal were 6. N. Masqué, M. Galià, R.M. Marcé, and F. Borrull. Functionalized
detected. Because the derivatives are obtained as the E and Z iso- polymeric sorbents for solid-phase extraction of polar pollutants.
J. High Resolt. Chromatogr. 22(10): 547–52 (1999).
mers, the most intense peak of each pair was used for quantitation
7. J. Hollender, F. Sandner, M. Möller, and W. Dott. Sensitive indoor air
purposes. monitoring of monoterpenes using different adsorbents and thermal
desorption gas chromatography with mass-selective detection.
Carboxylic acids in rain J. Chromatogr. A 962: 175–81 (2002).
Two combined samples were formed with 50-mL rain water 8. F. Sandner, W. Dott, and J. Hollender. Sensitive indoor air monitoring
of formaldehyde and other carbonyl compounds using the 2,4-dini-
samples collected at five spots that corresponded with the
trophenylhydrazine method. Int. J. Hyg. Environ. Health. 203:
corners and the center of a 100-m square. The combined samples 275–79 (2001).
were subjected to the PFBBr derivatization procedure (see 9. A. Sakuragawa, T. Yoneno, and T. Okutani. Trace analysis of carbonyl
Carboxylic acid determination section) and the chromatographic compounds by liquid chromatography-mass spectrometry after col-
analysis of the derivatives thus formed showed that acetic lection as 2,4-dinitrophenylhydrazine derivatives. J. Chromatogr. A
844(1-2): 403–408 (1999).
acid was the single carboxylic acid present, at a concentration of
10. A. Levart and M. Veber. Determination of aldehydes and ketones in
113 ± 8.3µM. air samples using cryotrapping sampling. Chemosphere. 44:
701–708 (2001).
11. D.W. Lehmpuhl and J.W. Birks. New gas chromatographic-electron
capture detection method for the determination of atmospheric alde-
Conclusion hydes and ketones based on cartridge sampling and derivatization
with 2,4,6-trichlorophenylhydrazine. J. Chromatogr A 740: 71–81
The analysis methods employed in this work were based on the (1996).
12. A. Cecinato, V. Di Palo, R. Mabilia, and M. Possanzini.
combination of SPME sampling with the derivatization of low-
Pentafluorophenylhydrazine as a coating reagent for the HRGC-MS
molecular-weight carboxylic acids or carbonyl compounds to determination of semi-volatile carbonyl compounds in air.
pentafluorophenyl or pentafluorobenzyl analogs to form simple Chromatographia. 54: 263–69 (2001).
analytical procedures with few steps and solvent-free extraction 13. C.G. Nolte, M.P. Fraser, and G.R. Cass. Gas phase C2–C10 organic
with no evaporation routines. The minimum levels of detection acids concentrations in the Los Angeles atmosphere. Environ. Sci.
Technol. 33(4): 540–45 (1999).
(signal-to-noise = 5) of the method for carboxylic acids were
14. K. Kawamura, L.-L. Ng, and I.R. Kaplan. Determination of organic
between 0.028 and 0.0185µM for aqueous matrices and between acids (C1 – C10) in the atmosphere, motor exhausts, and engine oils.
20.16 and 271.1 µg/m3 for gaseous matrices. The analysis of alde- Environ. Sci. Technol. 19(11): 1082–86 (1985).
hydes in air reached detection limits in the ng/m3 range. The use 15. S.R. Souza and L.R.F. Carvalho. Seasonality influence in the distribu-
of Tedlar bags in combination with SPME constitutes a very fast, tion of formic and acetic acids in the urban atmosphere of São Paulo,
Brazil. J. Braz. Chem. Soc. 12(6): 755–62 (2001).
simple, and reproducible technique for monitoring VOCs. The
16. A.G. Allen and A.H. Miguel. Biomass burning in the Amazon—
examples presented in this work illustrate the easy and diverse characterization of the ionic component of aerosols generated from
applications of these solvent-free, clean, and highly sensitive flaming and smoldering rain-forest and savanna. Environ. Sci.
methodologies to various fields of research. Technol. 29(2): 486–93 (1995).

352
Journal of Chromatographic Science, Vol. 44, July 2006

17. W.F. Rogge, M.A. Mazurek, L.M. Hildemann, G.R. Cass, and 30. M.-R. Lee, Y. Yao-Chia, W.-S. Hsiang, and H. Bao-Huey. Solid-phase
B.R.T. Simoneit. Quantification of urban organic aerosols at a molec- microextraction and gas chromatography–mass spectrometry for
ular level: identification, abundance and seasonal variation. Atmos. determining chlorophenols from landfill leaches and soil.
Environ. 27A: 1309–30 (1993). J. Chromatogr. A 806: 317–24 (1998).
18. K.S. Docherty and P.J. Ziemann. On-line, inlet-based trimethylsilyl 31. E.E. Stashenko, M.A. Puertas, W. Salgar, W. Delgado, and
derivatization for gas chromatography of mono- and dicarboxylic J.R. Martínez. Solid-phase microextraction with on-fibre derivatisa-
acids. J. Chromatogr. A 921: 265–75 (2001). tion applied to the analysis of volatile carbonyl compounds.
19. D. Grosjean. Atmospheric chemistry of toxic contaminants, 1. J. Chromatogr. A 886: 175–81 (2000).
Reaction rates and atmospheric persistence. J. Air Waste Manage. 32. F.K. Kawahara. Microdetermination of pentafluorobenzyl ester
Assoc. 40: 1397–1402 (1990). derivatives of organic acids by means of electron capture gas chro-
20. K. Blau and J.M. Halket. Handbook of Derivates for matography. Anal. Chem. 40(13): 2073–75 (1968).
Chromatography, 2nd ed. John Wiley & Sons, New York, NY, 1993, 33. E.E. Stashenko, M.C. Ferreira, L.G. Sequeda, J.R. Martínez, and
p. 369. J.W. Wong. Comparison of extraction methods and HRGC-ECD and
21. J. You, W. Zhang, and Y. Zhang. Simple derivatization method for HRGC-MSD-SIM techniques for the determination of volatile car-
sensitive determination of fatty acids with fluorescence detection by bonyl compounds. J. Chromatogr. A 779: 360 (1997).

Downloaded from https://academic.oup.com/chromsci/article-abstract/44/6/347/373686 by guest on 20 May 2020

high-performance liquid chromatography using 9-(2-hydroxyethyl)- 34. J.H. Brill, B.A. Narayanan, and J.P. McCormick. Selective determina-
carbazole as derivatization reagent. Anal. Chim. Acta 436: 163–72 tion of pentaluorobenzyl ester derivatives of carboxylic acids by GC
(2001). using microwave plasma and mass selective detection. Appl.
22. K. Kawamura and K. Usukura. Distributions of low molecular weight Spectrosc. 45(10): 1617–20 (1991).
dicarboxylic acids in the North Pacific aerosol samples. J. Oceanogr. 35. National Institute of Occupational Safety and Health.
49: 271–83 (1993). Carcinogenicity of acetaldehyde and malonaldehyde, and muta-
23. K. Kawamura. Identification of C2–C10 ω-oxocarboxylic acids, genicity of related low-molecular-weight aldehydes. Current
pyruvic acid, and C2–C3 ω-dicarbonyls in wet precipitation and Intelligence Bulletin 55: 166–67 (1991).
aerosol samples by capillary GC and GC/MS. Anal. Chem. 65(23): 36. National Institute of Occupational Safety and Health. NIOSH Pocket
3505–11 (1993). Guide to Chemical Hazards. Publication No. 2005-151. Available at
24. K. Kawamura and K. Ikushima. Seasonal changes in the distribution http://www.cdc.gov/niosh/npg/npgd0652.html, National Institute of
of dicarboxylic acids in the urban atmosphere. Environ. Sci. Technol. Occupational Safety and Health, Washington, DC, (2005).
27(10): 2227–35 (1993). 37. U. Bernier, D. Kline, D.R. Barnard, C.E. Schreck, and R.A. Yost.
25. J.E. Lawrence and P. Koutrakis. Measurement and speciation of gas Analysis of human skin emanations by gas chromatography/mass
and particulate phase organic acidity in an urban environment 1. spectrometry. 2. Identification of volatile compounds that are candi-
Analytical. J. Geophys. Res. 101(D4): 9159–70 (1996). date attractants for the yellow fever mosquito (Aedes aegypti). Anal.
26. J.E. Lawrence and P. Koutrakis. Measurement and speciation of gas Chem. 72: 747–56 (2000).
and particulate phase organic acidity in an urban environment 2. 38. X.-N. Zeng, J.J. Leyden, H.J. Lawley, K. Sawano, I. Nohara, and
Speciation. J. Geophys. Res. 101(D4): 9159–70 (1996). G. Preti. Analysis of characteristic odors from human male axillae.
27. P.A Martos and J. Pawliszyn. Calibration of solid phase microextrac- J. Chem. Ecol. 17: 1469–91 (1991).
tion for air analyses based on physical chemical properties of the 39. X.-N. Zeng, J.J. Leyden, J.G. Brand, A.I. Spielman, K.J. McGinley, and
coating. Anal. Chem. 69: 206–15 (1997). G. Preti. An investigation of human apocrine gland secretion for axil-
28. J.A. Koziel, J. Noah, and J. Pawliszyn. Field sampling and determina- lary odor precursors. J. Chem. Ecol. 18: 1039–55 (1992).
tion of formaldehyde in indoor air with solid-phase microextraction 40. C. Grote and J. Pawliszyn. Microextraction for the analysis of human
and on-fiber derivatization. Environ. Sci. Technol. 35: 1481–86 breath. Anal. Chem. 69: 587–96 (1997).
(2001).
29. L. Pan and J. Pawliszyn. Derivatization/solid-phase microextraction: Manuscript received September 17, 2005;
new approach to polar analytes. Anal. Chem. 69: 196–205 (1997). revision received January 12, 2006.

353