Reprinted from
JOURNAL
of
CHROMATOGRAPHY A
Journal of Chromatography A, 867 (2000) 143-149
Comparative methodology in the determination of
a-oxocarboxylates in aqueous solution
Ion chromatography versus gas Chromatography after oximation,
extraction and esterification
Edward T. Urbansky , W. Jerry Bashe
"United States Environmental Protection Agency (EPA), Office of Research and Development, National Risk Management Research
Laboratory, Water Supply and Water Resources Division, Treatment Technology Evaluation Branch, Cincinnati, OH 45268, USA
b73V & Associates, Inc., Andrew W, Breidenbach Environmental Research Center, Cincinnati, OH 45268, USA
Received 19 My 1999; received in revised form 27 October 1999; accepted 2 November 1999
ELSEVIER
-------
JOURNAL OF CHROMATOGRAPHY A
INCLUDING ELECTROPHORESIS MASS SPECTROMETRY AND OTHER SEPARATION AND DETECTION METHODS
Scope. The Journal of Chromatography publishes papers on all aspects of separation science including chromatography,
electrochromatography, electrophoresis, hyphenated and other multi-dimensional' techniques, sample preparation as well as
detection methods such as mass spectrometry. Contributions consist mainly of research1 papers dealing with chromatographic
and electrophoretic theory, instrumental developments and their analytical and preparative applications. Section A covers all
areas except biomedical sciences and biomedical applications of separation science, which are published in section B-
Blomedlcal Sciences and Applications.
Submission of Papers. The preferred medium of submission is on disk with accompanying manuscript. Manuscripts (in English-
four copies are required) should be submitted to: Editorial Office of 'Jobrhal of Chromatography A, P.O. Box 6811, 1000 AR
Amsterdam, The Netherlands, Telefax (+31-20) 485 2304. Review articles are invited or proposed in writing to the Editors who
welcome suggestions for subjects. Ah outline of the proposed review should first be forwarded to the Editors for preliminary
discussion prior to preparation. Submission, of. an article is understood to imply that;the article is original and unpublished and
is not being considered for publication elsewhere. : / , ' ' . ; .
Publication Information. Journal of Chromatography A (ISSN 0021-9673). For 2000 Vols. 862-901 are scheduled for
publication. Subscription prices for Journal of Chromatography A+B (combined), or for section A or B are available upon
request from the Publisher or from the Regional Sales Office nearest you or from this journal's website (http: //www.elsevier.nl/
locate/chrom). Further information is available on this journal and other Elsevier Science products through Elsevier's website-
(http://www.elsevier.nl). Subscriptions are accepted on a prepaid basis only and are entered oh a calendar year basis. Issues
are sent by standard mail (surface within Europe, air delivery outside Europe). Priority rates are available upon request Claims
for missing issues should be made within six months of the date of dispatch.
Orders, claims, and product enquiries: please contact the Customer Support Department at the Regional Sales Office nearest
NeJCT5)r.k: E'sevier Science. P-O- Box 945, New York, NY 10159-0945, USA; phone: (+1) (212) 633 3730 [toll free number for
North American customers: 1-888-4ES-INFO (437-4636)]; fax: (+1) (212) 633 3680; e-mail: usinfo-f@elsevier.com
Amsterdam: Elsevier Science, PO Box 211,1000 AE Amsterdam, The Netherlands; phone: (+31) 20 4853757- fax- (+31) 20
4853432; e-mail: nlinfo-f@elsevier.nl ' '
Tokyo: Elsevier Science, 9-15 Higashi-Azabu 1-chome, Minato-ku, Tokyo 106-0044, Japan; phone: (+81) (3) 5561 5033- fax-
(+81) (3) 5561 5047; e-mail: info@elsevier.co.jp
Singapore: Elsevier Science, No. 1 Temasek Avenue, #17-01 Millenia Tower, Singapore 039192; phone: (+65) 434 3727- fax-
(+65) 337 2230; e-mail: asiainfo@elsevier.com.sg . '
Ri?,d^^?J?,eJl?: Elseviei; Science, Rua Sete de Setembro 111/16 Andar, 20050-002 Centra, Rio de Janeiro - RJ, Brazil; phone:
(+55) (21) 509 5340; fax: (+55) (21) 5071991; e-mail: elsevier@campus.com.br [Note (Latin America): for orders, claims and
help desk information, please contact the Regional Sales Office in New York as listed above].
Enquiries concerning manuscripts and proofs: .questions arising after acceptance of the manuscript, especially those relatinq
to proofs, should be directed to: : .
Elsevier Science B.V., Author Support Department; Fax (+31 20)485-3752; e-mail authorsupport@elsevier.nl
For information on editorial matters (including submission, reviews and revision of manuscripts) please contact: Editorial Office
Journal of Chromatography, P.O. Box 681,1000 AR Amsterdam, The Netherlands; Tel. (+31-20)485-2794- Fax (+31-20)485-
2304; e-mail chrom-eo@elsevier.nl '
Journal of Chromatography A
PUBLICATION SCHEDULE
MONTH
Journal of
ChromatographyA
Bibliography Section
Sept
855/1
855/2
856/1+2
861/1
Oct.
857/1+2
858/1
858/2
859/1
859/2
Nov
862/1
862/2
863/1
863/2
Dec.
851/1+2"
864/1
864/2
865/1+2
861/2
Jan. 2000
866/1
866/2
867/1+2
868/1
The publication schedule
for further issues will be
published later
Vol. 851 contains the Cumulative Author, Subject and Compound Indexes and is. published out of sequence.
-------
144
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
Table 1
oi-Oxocarboxylates examined in this study and retention times for the methyl esters of their oximes (two geometric isomers)
Analyte anion
[CAS RN]
Oxoethanoate
[563-96-2]"
2-Oxopropanoate
[113-24-6]°
Oxopropanedioate
[7346-13-6]d
2-Oxobutanoate
[600-18-0]°
2-Oxopentanoate
[13022-83-8]f
Synonyms
Glyoxylate
Fonnylformate
Pyruvate
2-Methylglyoxylate
Ketomalonate
Oxomalonate
Mesoxalate
2-Ketobutyrate
ot-Ketobutyrate
2-Ketovalerate
ot-Ketovalerate
Formula of acid
HC(O)CO,H
CH3C(0)C02H
HO2CC(O)COaH
CH3CH,C(O)CO,H
CH3CH,CH2C(O)CO2H
Retention time (min)"
Isomer 1
12.28
12.55
19.90s
14.43
16.52
Isomer 2
13.14
14.29
15.62
17.48
" Some workers have applied the (B)/(Z) nomenclature system to these geometric isomers. That notwithstanding, these compounds are not
alkenes and the priority for an electron pair (on the oxime nitrogen) is not defined, although one could use the protonated form for the
purpose of naming. Regardless, applying (E) and (Z) descriptors is not useful because: (1) the relative retention of the two isomers is
unknown and (2) the sum of the peak areas is used for quantitation. Accordingly, the two will simply be listed in this report as isomer 1 and
isomer 2, in order of elution from the column.
° Oxoethanoic acid monohydrate, Aldrich, Milwaukee, WI, USA.
° Sodium 2-oxopropanoate, Aldrich. The name 2-methylgIyoxyIate requires a locant of either 2 or a and should be written without a space
to avoid confusion with the ester formed from glyoxylic acid and methanol.
d Disodmm Oxopropanedioate, Sigma, St. Louis, MO, USA. Although the disodium salt can be made anhydrous, both anionic forms and
the acid exist as gem-diols at carbon 2. Thus, oxopropanedioic acid actually exists mostly in the form of dihydroxypropanedioic acid:
C(OH)2(CO2H),. The same would be true for the deprotonated anions in aqueous solution.
° 2-Oxobutanoic acid, Aldrich.
r Sodium 2-oxopentanoate, Aldrich. ' •
8 Carbon 2 of Oxopropanedioate is not chiral; therefore, only one geometric isomer is formed upon derivatization.
lytes into tert.-butyl methyl ether. (2) It introduces a
functional group (C6FS-) that increases sensitivity
by making electron-capture detection (BCD) pos-
sible. Accordingly, variations of this procedure have
been used for quantitatively determining short-chain
a-oxocarboxylates found as byproducts from the
ozonation of potable water supplies [2—4]; however,
reaction conditions are unspecified or varied between
laboratories. Under the dilute concentrations (<100
(juM) found in post-ozonation drinking water sys-
tems, they exist >99.9% as the ionized anions rather
than the parent carboxylic acids; therefore, they can
also be determined by ion chromatography (1C) [5].
These species are listed in Table 1.
The GC-ECD method continues to be relied upon
for the measurement of ozonation byproducts of
natural waters; therefore, it .was deemed prudent to
assess its ruggedness (resistance to matrix effects),
reliability (day-to-day variability), and reproducibil-
ity (precision). Because this method relies on car-
bonyl oximation, high concentrations of aldehydes
and/or ketones interfere by competing for the de-
rivatizing agent, PFBOA, which could become a
limiting reagent in waters with sufficiently high
organic matter concentrations. In addition, any ma-
terial that inhibits partitioning and extraction (e.g., a
surfactant) can also be expected to interfere.
2. Experimental1
2.1. Analyte standards and test solutions
An aqueous standard was prepared at 1000
'Mention of specific brand names or manufacturers should not be
construed as an endorsement of products or companies by the
United States government.
-------
ELSEVIER
Journal of Chromatography A, 867 (2000) 143-149
JOURNAL OF
CHROMATOGRAPHY A
www.elsevier.com/locate/chroma
Comparative methodology in the determination of
a-oxocarboxylates in aqueous solution
Ion chromatography versus gas chromatography after oximation,
extraction and esterification^
Edward T. Urbansky3'*, W. Jerry Basheb
"United States Environmental Protection Agency (EPA), Office of Research and Development, National Risk Management Research
Laboratory, Water Supply and Water Resources Division, Treatment Technology Evaluation Branch, Cincinnati, OH 45268, USA
bZW & Associates, Inc., Andrew W. Breidenbach Environmental Research Center, Cincinnati, OH 45268, USA
Received 19 July 1999; received in revised form 27 October 1999; accepted 2 November 1999
Abstract
The a-oxocarboxylates (a-ketocarboxylates) and the corresponding a-oxoacids (ot-ketoacids) have been reported as
byproducts of ozonation of potable water supplies. Some of these species also occur in biophysiological systems. Five
analytes were investigated in this study: oxoethanoate (glyoxylate), 2-oxopropanoate (pyruvate), 2-oxobutanoate (2-
ketobutyrate), 2-oxopentanoate (2-ketovalerate) and oxopropanedioate (ketomalonate, mesoxalate). Ion chromatography (1C)
and gas chromatography (GC) were evaluated for the quantitation of these analytes at concentrations £200 ng ml"'. For the
1C method, the samples are run directly with minimal to no pre-treatment. For the GC method, the analytes must be
derivatized with O-(2,3,4,5,6-pentafluorobenzyl)oxylamine to form oximes. The oximes are extracted into tert.-butyl methyl
ether and the carboxylic acid is esterified (methylated) with diazomethane. It was concluded that the ion chromatographic
determination is significantly superior to the gas chromatographic method for these analytes. Published by Elsevier Science
B.V.
Keywords: Oxocarboxylates; Ketocarboxylates; Ketoacids
*This paper was prepared by a United States government em-
ployee in the course of his official duties. All work contained
herein was performed by a US government employee or under a
contract held by the US government; consequently, this paper is
not subject to copyright restrictions. Ion chromatography work
was performed under EPA Office of Research and Development
contract 68-C6-0079 and subjected to EPA review. All data and
results are the property of the US government and are therefore
exempt from copyright.
*Corresponding author. Fax: +1-513-5697-658.
E-mail address: urbansky.edward@epa.gov (E.T. Urbansky)
1. Introduction
• ' •.' ' • ''•• i
As a class, a-oxocarboxylates can be readied for
gas chromatography (GC) analysis by a two-step
process [1]. First, they are derivatized with. O-
(2,3,4,5,6-pentafluorobenzyl)oxylamine (PFBOA).
Second, the carboxylic acid moieties, are esterified
(or alternately silylated). Formation of the oxime has
two advantages: (1) it permits extraction (and thus
pre-concentration) of the otherwise hydrophilic ana-
0021-9673/00/$ - see front matter Published by Elsevier Science BY.
PII: 80021-9673(99)01158-9
-------
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
145
ml in each analyte by dissolving the commercially
available reagents into doubly deionized water (see
Table 1); this solution was used for both the GC and
1C tests. In addition, replicate standards were made
from the solids on several instances and used to
verify that the original standard had not deteriorated.
Volumes of this stock standard were diluted 1/50 to
produce a working standard 20.0 [xg ml~l in each
analyte; this solution was confirmed to be usable for
30 days without loss of instrument response by 1C.
Both stock and working standards were kept in
polypropylene bottles hi a laboratory refrigerator at
4±1°C. The working standard was injected via
microliter syringes or Eppendorf pipettor (Brink-
mann, Westbury, NY, USA) into 20.0-ml portions of
doubly deionized water to produce test solutions
containing up to 200 ng ml"1 of each analyte.
Blanks (no analytes added) were also prepared. Test
solutions were prepared directly in pre-cleaned 40-ml
glass US Environmental Protection Agency (EPA)
vials with screw-caps and PTFE-lined septa obtained
from Supelco, Bellefonte, PA, USA or Nalge Nunc
(I-Chem), International, Rochester, NY, USA.
2.2. Assessment strategy
2.2.1. Precision
To assess the reproducibility on a single day for
both the GC and 1C methods, replicate standard test
solutions were prepared from the stock solutions and
subjected to the GC or 1C method.
2.2.2. Sample holding time
Standard test solutions were prepared from the
stock solutions and stored in our sample storage
facility (cold room) in the dark at 7±2°C during the
holding time. Subsequently, the test solutions were
brought to ambient temperature and subjected to the
GC or 1C method.
2.2.3. Extract integrity
Although our laboratory is usually able to produce
derivatized extracts of these samples within a short
period, large numbers of samples which undergo GC
analysis often result in a delay between the sample
treatment and the instrumental analysis. Consequent-
ly, it was important to determine whether storing the
extracts resulted in reduced performance. A set of
extracts was run and then stored in a freezer at
— 15°C for 13 days to determine how much degra-
dation occurred; this is the temperature of a typical
laboratory freezer. Extracts of triplicate standards at
0 (blank), 10, 20, 50, 100 and 200 ng ml"1 (all
analytes together) were used for this test. Additional
test solutions were placed in a freezer at -80°C for
up to 7 days; this temperature is the standard for
sensitive biochemicals.
2.3. Gas chromatographic method
2.3.1. Oximation
Solutions of the derivatizing agent PFBOA, were
prepared fresh each day at 10 mg ml"1 of the
hydrochloride salt, PFBOA-HC1, Sigma, St. Louis,
MO, USA. A 1.0-ml aliquot of PFBOA solution, and
a 1.0-ml aliquot of 1.0 M total phosphate buffer
(0.50 M NaH2PO4+0.50 M Na2HPO4) were added
to each 20.0 ml standard hi the 40-ml vials; salts
were obtained from Fluka, Buchs, Switzerland. The
vials were placed into a forced air oven thermostated
at 45±2°C for 90 min. Subsequently, the vials were
placed into an ice bath. After cooling, a few drops of
0.25% (w/w) FD&C Blue No. 1 aqueous solution
(improves visibility of phase separation) and 1.0 ml
of 9.0 M H2SO4(aq.) were added to the derivatized
test solutions. This' dye contains no nucleophilic
moieties and thus should not interfere in the oxima-
tion. In addition, it is ionic and partitions almost
exclusively into the water phase; it cannot be ob-
served by GC-MS of the extracts. Dye was obtained
from Warner Jenkinson, St. Louis, MO, USA; 18 M
(98%, w/w) sulfuric acid was from J.T. Baker,
Phillipsburg, NJ, USA. The test solutions were then
extracted with a 4.0-ml aliquot of pesticide residue
analysis (PRA)-grade tert>butyl methyl ether
(MTBE), Aldrich, Milwaukee, WI, USA. The ex-
tracts were dried with Tracepur Na2SO4 from EM
Science, Gibbstown, NJ, USA.
2.3.2. Esterification
Methylation of the carboxylic acid functionalities
was done using a flowing stream of diazomethane in
argon. This minimizes the risks associated with
concentrated CH2N2 solutions and the uncertainty
associated with adding a volume of diazomethane
solution to an unknown volume of recovered extract.
-------
146
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
Solutions of AT-memyl-Ar-nitroso-p-toluenesul-
fonamide [80-111-5] (Diazald, Aldrich) was pre-
pared fresh prior to methylation by combining 3.0 g
N-methyl-AT-nitroso-p-toluenesulfonamide and 30 ml
of solvent (8 ml USP EtOH+22 ml PRA-grade
MTBE); total volume was scaled up or down as
needed. A solution of 33% (w/w) NaOH(aq.) was
prepared by diluting 50% (w/w) solution obtained
from Fisher Scientific, Pittsburgh, PA, USA.
The apparatus used was similar to that of EPA
Method 552 [6]. Immediately preceding the
methylating apparatus, the argon stream was satu-
rated with MTBE vapor by passing it through a
sintered glass dispersion tube immersed in MTBE.
Extracts were transferred to 100X16 mm disposable
borosilicate glass test tubes for the methylation.
After 30-45 s of exposure to the gas stream, the
yellow color indicated the presence of CH2N2; the
color was used as the criterion for providing excess
diazomethane. After 30 min, reaction was considered
to be complete, and the methylated extracts were
transferred to autosampler vials and stored at — 80°C
prior to GC-ECD analysis.
2.3.3. GC-ECD analysis
Derivatized and methylated extracts were analyzed
on a Hewlett-Packard (Palo Alto, CA, USA) 6890
GC-ECD system equipped with an HP 7673 auto-
injector. Using splitless injections, volumes of 3.0 jo-l
were loaded onto a J&W Scientific (Folsom, CA,
USA) DB-5 MS column (30 mX250 |xm I.D., 0.25
|xm film) at constant (high purity) helium flow of 1.0
ml min"1; inlet and detector temperatures: 270°C.
Temperature program: hold 60°C for 2.0 min; ramp
20.0°C min"1 to 120°C, hold for 1.0 min; ramp
4.0°C min'1 to 130°C, hold for 2.0 min; ramp 4.0°C
min"1 to 150°C; ramp 5°C min'1 to 200, hold for
1.0 min; ramp 20°C min"1 to 260°C. Retention times
are given in Table 1.
2,4. Ion chromatographic analysis
Samples were placed into 5.5-ml autosampler vials
and analyzed on a Dionex (Sunnyvale, CA, USA)
DX-300 ion chromatograph with conductivity de-
tection (all parts were obtained from Dionex). We
used the method of Kuo [5] without modification
save instrument model; for experimental details see
Table 1 of Ref. [5]. Care was taken to avoid
carbonate contamination; the eluent was prepared
fresh on the day of analysis from 50% (w/w)
NaOH(aq.), and air exposure was minimi7p.fl. We
obtained similar retention times to Kuo; refer to the
chromatogram in Kuo's Fig. 1 [5].
3. Results and discussion
3.1. Precision—reproducibility on a single day
As shown by Table 2, the GC method is capable
of satisfactory precision under optimal conditions.
Oxoethanoate and oxopropanedioate show the lowest
precision, with relative standard errors of 9.3% and
7.1%, respectively, in the slopes of their calibration
lines. Nonetheless, the 1C method has better preci-
sion for all of the analytes except 2-oxopentanoate,
which experiences chloride interference due to over-
lap of the two peaks. For the ot-oxocarboxylates
normally encountered as ozonation byproducts, anal-
ysis by 1C should be more precise. The effect of the
chloride peak overlap with that of 2-oxopentanoate is
most pronounced at analyte concentrations below 30
"ng ml"1; however, our laboratory has not found this
compound in ozonation byproduct formation studies.
For the other analytes, agreement remains excellent
(R2>0.99) for concentrations ranging from 5 to 200
ng ml"1.
3.2. Reliability-reproducibility from day to day
The ion chromatographic method far outperforms
the gas chromatographic method in this area. The
derivatization and methylation steps are probably
responsible for the variability in the GC method.
However, we do not observe such variability in the
determination of aldehydes using EPA Method 556
[7]. During 9 months of testing, recoveries of quality
control samples (oxoethanoate, 2-oxopropanoate, 2-
oxobutanoate and oxopropanedioate) made from
fresh standards have varied less than 5% for the 1C
method.
The most reliable results were obtained for 2-
oxopropanoate, 2-oxobutanoate and 2-oxopentanoate.
For these anions, the normalized (relative to injection
-------
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
147
Table 2
Reproducibility on a single day for calibration plots (peak area vs. concentration)
Analyte
Oxoethanoate
2-Oxopropanoate
2-Oxobutanoate
2-Oxopentanoated
Oxopropanedioate
Method
GC°
IC°
GC
1C
GC
1C
GC
1C
GC
1C
Slope (ml ng~')
43+4
60900±1400
321 ±7
114500+500
265±6
88600+1200
202+4
86 000+5000
28±2
117 000+1600
y-Intercept (unittess)
Ob
(-4±3)-105
Ob
(-6±10)-104
0"
(-26+25)- 10"
Ob
(2±1>10S
0"
(-4±3)-105
R-
0.907
0.995
0.991
0.9998
0.992
0.998
0.992
0.974
0.930
0.998
'Four replicates were prepared at each of the following concentrations: 0 (blank), 50, 100, 150 or 200 ng ml~' (aE analytes in each
standard sample).
b Intercepts are statistically indistinct from zero.
° Calibration curves are based on duplicate runs of a mixed standard (containing all analytes) at concentrations of 0 (blank), 5,10, 20, 30
40, 50, 80, 100, 150 or 200 ng mT1.
Interference from traces of chloride reduces precision and accuracy in determining 2-oxopentanoate; the chloride peak overlaps the
analyte peak.
volume) least squares slopes varied less than ± 10%
over a 6-month period, with at least 10 sets of
standards. In fact, 2-oxopentanoate is generally more
reliably determined by the GC method since chloride
cannot be adequately excluded from the samples and
co-elutes in the 1C method. Oxoethanoate quantisa-
tion was less reproducible than the others by GC.
Least-squares slopes of the calibration curve could
vary by as much as ±30% from day to day, without
any apparent trend or reason. Nonetheless, for that
day, the result was quite precise as previously
indicated in Table 2 and by the error bars for the
slope values in Fig. 1. .
Assaying oxopropanedioate was highly unreliable
by this GC method. Least-squares slopes varied by
factors of 2 to 5, with some test solutions giving
appropriate results and others showing no signal or
less than 10% of the expected peak area. On some
days, the method failed utterly for this analyte,
producing data with so much scatter that no line
could reasonably be drawn through a plot of peak
area against concentration (R2<0.3). As described in
the previous section, it was possible to obtain highly
precise results on some occasions; however, the
method performed unsatisfactorily most of the time
for oxopropanedioate.
Generally, Oxoethanoate, 2-oxopropanoate and
oxopropanedioate have been found in this laboratory
and elsewhere as ozonation byproducts. Occasion-
ally, we have also observed 2-oxobutanoate, but not
2-oxopentanoate2. It cannot be argued that degra-
dation of the analytes is an issue because this is not
observed with the 1C method. In addition, there is
not a steady loss of instrumental response, but rather
random fluctuation.
3.3. Lower limits of detection
Data and results from the 1C method permit a
fairly uncomplicated estimation of the lower limits of
detection (LLODs). Using the signal of the blank
plus three-tunes the noise of the blank gives the
following lower limits of detection (ng ml"1): oxo-
ethanoate 8, 2-oxopropanoate 3, 2-oxobutanoate 6,
2-oxopentanoate 60, and oxopropanedioate 7.
Calculation of LLODs for the GC method is not
straightforward. Because there are so many potential
chemical problems with this method, identifying
the mam sources of error is difficult. For 2-oxo-
2Although the results are not reported in this paper, we have used
both the GC and 1C methods for samples generated from
ozonation experiments in our laboratories and from public water
supplies that use ozone as a disinfectant.
-------
148
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
4.0
3.0
2.0
u
1.0
0.0
I i I
I i I i I
2 4 .6 8 10 12 14
holding time/days
Fig. 1. Least-squares slopes of calibration lines obtained for the
a-oxocarboxylates by gas chromatography. Key: oxoethanoate
(•), 2-oxopropanoate (A), 2-oxobutanoate (•), 2-oxopentanoate
(T), oxopropanedioate (4). Much of the variability observed here
is believed to result from deficiencies in method reliability rather
than analyte decomposition as explained in the text.
propanoate, 2-oxobutanoate and 2-oxopentanoate,
which are reliably and reproducibly measured, cali-
bration curves are well-behaved to the 5-10 ng ml"1
region. Oxoethanoate achieves an LLOD of 10-15
ng ml"1; however, oxopropanedioate is quite unreli-
able and can range from 10 ng ml"1 on a good day
to as much as 100-200 ng ml"1 when method
performance is poor. This is unacceptable as the
concentrations measured in ozonated drinking water
samples usually fall below 15 ng ml"1.
3.4. Holding time
Table 3 gives the 1C results for samples held at
7±2°C for a period of days subsequent to calibration
with standard solutions prepared from the same stock
standard; all of the solutions were made on the same
day as the calibration standards. Some variation is
observed, but the greatest is 23%.
The 1C samples showed some loss after only 24 h;
therefore, some adsorption to the container wall is
suspected. Nevertheless, switching to polypropylene
tubes did not alter this initial loss. Further loss may
be microbially mediated rather than a simple chemi-
cal decomposition. It is worth pointing out that Kuo
[5] demonstrated the stability of the analytes when
preserved with benzalkonium chloride. We have
continued to use mercury (II) chloride, but we do not
observe a problem with quantitatmg 2-oxopropanoate
as Kuo reported.
The slopes obtained for test solutions subjected to
the GC method are highly variable from day to day,
as demonstrated in Fig. 1. Based on the ion chroma-
tography results in Table 2, we feel that much of
Table 3
Recoveries relative to day zero for standard samples held N days"
and then subjected to the ion chtomatographic method1"
Analyte
Oxoethanoate
2-Oxopropanoate
2-Oxobutanoate
Oxopropanedioate
N (days)
1
95±3d
103 ±2
84±5
87±7
5
77.5+7
104±1
77±10
93±3
12
93 ±3
102.5±3
78±7
96±3
"N=0 was the starting date. N=l was 24 h later, etc.
b Values on day 0 established a calibration curve that was used
to compute the concentrations on days 1, 5, and 12 from peak
areas; day 0 values were set to 100.0% recovery. Duplicate
standards at 5, 10, 20, 30, 40 and 50 ng ml"' were used for
calibration; recoveries are based on duplicate standards at the
same concentrations.
c 2-Oxopentanoate was omitted due to the chloride interference.
d Estimated standard deviation of the mean (standard error) is
the uncertainty in the average: esdm=(estimated standard
deviation)//!1'2, n=6; these values were computed before round-
ing of the estimated standard deviations.
-------
E.T. Urbansky, W.J. Bashe I J. Chromatogr. A 867 (2000) 143-149
149
behavior shown in Fig. 1 is actually due to de-
ficiencies in the GC method reliability and not
analyte degradation.
3.5. Extract integrity
After 13 days at -15°C, the degradation of the
methyl esters of the pentafluorobenzyloximes of the
analytes was assessed by the change in the slopes of
the calibration curves. Oxoethanoate (—18 ±10%)
and oxopropanedioate (-34±5%) showed the great-
est effect (loss), while 2-oxopropanoate showed a
gain of questionable significance (4±3%). 2-Oxo-
butanoate and 2-oxopentanoate did not experience a
statistically significant effect during this time. At
—80°C, much of the residual water actually precipi-
tates (particulate ice is visible), although the MTBE
does not freeze. None of the analytes show a
significant difference in quantitation when held at
this temperature for up to 7 days. These particular
times were chosen based on limitations in resources
(instrumentation and personnel) for our laboratory
relative to sample load. Although we did not sys-
tematically examine the stability of the esterified
oximes at room temperature, we have anecdotally
observed losses of up to 20% in peak area overnight.
As expected, losses are greater when the laboratory
is warmer. For this reason, precautionary measures
such as chilling the autosampler rack and running
extracts immediately are advisable.
4. Conclusion
Although the 1C method does not work well for
2-oxopentanoate, this species is usually not observed
as an ozonation byproduct, and we do not routinely
monitor for it. For the other a-oxocarboxylates, the
1C method is demonstrably better in terms of re-
liability and precision. In terms of practical compara-
tive methodology, we have found the ion chromato-
graphic analysis to be substantially superior to the
multi-step GC method. The GC method requires
time-consuming and potentially error-contributing
derivatization, extraction, washing, methylation, and
hour-long GC analysis, whereas the 1C method
allows multiple injections of a sample to be run
directly with less volume and essentially no preparat-
ory steps. Moreover, the 1C method takes less time
per analysis. Although the 1C method does suffer
from migrating retention times as carbonate infil-
trates the eluent stock solutions, we suspect that this
problem can be eliminated if the hydroxide is
generated electrolytically (as with the Dionex EG-
40).
Because we can collect only a limited number of
samples during field studies and often cannot repeat
the sampling, it is imperative that the analytical
method work reliably for the three species of greatest
interest: oxoethanoate, 2-oxopropanoate and oxo-
propanedioate (dihydroxopropanedioate). Primarily
for this reason, we have abandoned the GC analysis
of a-oxocarboxylates in favor of the 1C analysis in
studies of ozonation byproducts.
Acknowledgements
We thank Warner JerJdnson Co., Inc., for pro-
viding a complimentary sample of FD&C Blue No. 1
for use in this study. We acknowledge EPA techni-
cian Kenneth Kropp, who assisted in the analysis of
the gas chromatography data and chemist H. Paul
Ringhand, who performed some of the preliminary
experimentation under the EPA's Senior Environ-
mental Employment Program.
References
[1] K. Kobayashi, E. Fukui, M. Tanaka, S. Kawai, J. Chroma-
togr. 202 (1980) 93-98.
[2] Y. Xie, D.A. Reckhow, Ozone Sci. Eng. 14 (1992) 269-275.
[3] Y. Xie, D.A. Reckhow, in: AWWA Proceedings, Annual
Conference, Denver, CO, USA, 1992.
[4] J.F. Garcia-Araya, J.P. Croue, FJ. Beltran, B. Legube, Ozone
Sci. Eng. 17 (1995) 647-657.
[5] C.-Y. Kuo, J. Chromatogr. A 804 (1998) 265-272.
[6] J.W. Hodgeson, J. Collins, R.E. Earth, Method 552, De-
termination of Haloacetic Acids in Drinking Water by
Liquid-Liquid Extraction, Derivatization, and Gas Chroma-
tography with Electron Capture Detection, United States
Environmental Protection Agency, July 1990, Note that this
is not the most current version, which uses acidic methanol
instead of diazomethane.
[7] J.W. Munch, D.J. Munch, S.D. Winslow, S.C. Wendelken,
B.V. Pepich, Method 556, Determination of Carbonyl Com-
pounds in Drinking Water by Pentafluorobenzylhydroxy-
lamine Derivatization and Capillary Gas Chromatography
with Electron Capture Detection, Rev. 1.0, United States
Environmental Protection Agency, June 1998.
-------
-------
JOURNAL OF CHROMATOGRAPHYA
INFORMATION FOR AUTHORS
(Detailed Instructions to Authors were published in J. Chromatogr. A Vol. 866, pp. 307-315. A free reprint can be obtained by
application to the publisher, Elsevier Science B.V., P.O. Box 330, 1000 AH Amsterdam, The Netherlands.) The instructions can
also be found on the Analytical Separation Home Page on the World Wide Web: access under http://www.elsevier.nl/locate/
chrom or http://www.elsevier.com/locate/chrom
| Journal of Chromatography has no page charges I
Types of Contributions. The following types of papers are published: Regular research papers (full-length papers), Review
articles, Short Communications, Discussions and Letters to the Editor. Short Communications are usually descriptions of short
investigations, or they can report minor technical improvements of previously published procedures; they reflect the same
quality of research as full-length papers, but should preferably not exceed five printed pages. Discussions (one or two pages)
should explain, amplify, correct or otherwise comment substantively upon an article recently published in the journal. Letters to
the Editor (max. two printed pages) bring up ideas, comments, opinions, experiences, advice, disagreements, insights, etc. For
Review articles, see inside front cover under Submission of Papers.
Submission. Every paper must be accompanied by a letter from the senior author, stating that he/she is submitting the paper
for publication in the Journal of Chromatography A.
Authors in Japan please note: Upon request, Elsevier Science Japan will provide authors with a list of people who can check
and improve the English of their paper (before submission). Please contact our Tokyo office: Elsevier Science Japan, 9-15
Higashi-Azabu 1-chorne, Minato-ku, Tokyo 106-0044; Tel. (03)-5561-5032, Fax (03)-5561-5045.
Manuscripts. Manuscripts should be typed in .double spacing on consecutively numbered pages of uniform size. The
manuscript should be preceded by a sheet of manuscript paper carrying the title of the paper and the name and full postal
address of the person to whom the proofs are to be sent. As a rule, papers should be divided into sections, headed by a
caption (e.g., Abstract, Introduction, Experimental, Results, Discussion, etc.). All illustrations, photographs, tables, etc., should
be on separate sheets. Electronic manuscripts: Electronic manuscripts have the advantage that there is no need for the
rekeyirig of text, thereby avoiding the possibility of introducing errors and resulting in reliable and fast delivery of proofs. For
the initial submission of manuscripts for consideration, hardcopies are sufficient. Upon revision, your disk and (exactly
matching) printed version (printout, hardcopy) should be submitted together to the-accepting editor or Editorial Office
according to their request. Double density (DD) or high density (HD) diskettes (31/2 or 51/4 inch) are acceptable. It is
important that the file saved is in the native format of the wordprocessor program used. Label the disk with the name of the
computer and wordprocessing package used, your name, and the name of the file on the disk. Further information may be
obtained from the Publisher.
Abstract. All articles should have an abstract of 50-100 words which clearly and briefly indicates what is new, different and
significant. No references should be given.
Introduction. Every paper must have a concise introduction mentioning what has been done before on the topic described, and
stating clearly what is new in the paper now submitted.
Experimental conditions should preferably be given on a separate sheet, headed "Conditions". These conditions will, if
appropriate, be printed in a block, directly following-the heading "Experimental".
Illustrations. The figures should be submitted in a form suitable for reproduction, drawn in Indian ink on drawing or tracing
paper. Each illustration should have a caption, all \he-captions being typed (with double spacing) together on a separate
sheet. If structures are given in the text, the original drawings should be provided. Coloured illustrations are reproduced at the
author's expense. The written permission of the author'and publisher must be obtained for the use of any figure already
published. Its source must be indicated in the legend.
References. References should be numbered in the order in which they are cited in the text, and listed in numerical sequence
on a separate sheet at the end of the article. Please check a recent issue for the layout of the reference list. Abbreviations for
the titles of journals should follow the system used by Chemical Abstracts. Articles not yet published should be given as "in
press" (journal should be specified), "submitted for publication" (journal should be specified), "in preparation" or "personal
communication".
Vols. 1-651 of the Journal of Chromatography, Journal of Chromatography, Biomedical Applications and Journal of
Chromatography, Symposium Volumes should be cited as J. Chromatogr. From Vol. 652 on, Journal of Chromatography A
should be cited as J. Chromatogr. A and Journal of Chromatography B as J. Chromatogr. B.
Dispatch. Before sending the manuscript to the Editor please check that the envelope contains four copies of the paper
complete with references, captions and figures. One of the sets of figures must be the originals suitable for direct
reproduction. Please also ensure that permission to publish has been obtained from your institute.
Proofs. One set of proofs will be sent to the author to be carefully checked for printer's errors. Corrections must be restricted to
instances in which the proof is at variance with the manuscript.
Reprints. Fifty reprints will be supplied free of charge. Additional reprints can be ordered by the authors. An order form
containing price quotations will be sent to the authors.
For publication schedule see inside front cover.
-------
------- |