Characterization of carbonyl by-products during Uniblu-A ozonation by liquid chromatography/hybrid quadrupole time-of-flight/mass spectrometry

Article abstract of DOI:10.1002/rcm.5045

The structural elucidation of carbonyl-containing by-products arising from Uniblu-OH ozonation has been investigated by liquid chromatography/electrospray ionization tandem mass spectrometry (LC/ESI-MS/MS) employing a quadrupole time-of-flight mass spectrometer. The by-products were derivatized with 2,4-dinitrophenylhydrazine, allowing the formation of [M-H]^- ions of the derivatives in the electrospray source. Exact mass measurements of both the [M-H]^- ions and their product ions allowed the elemental formulae and related structures of ten by-products to be determined confidently. The main degradation pathway were decarboxylation followed by further oxidation. It is noteworthy that the experimental procedure employed allowed the identification of both nitrogen- and sulphur-containing carbonyl by-products during Uniblu-OH ozonation. This result is of environmental relevance for monitoring the balance of organic nitrogen and sulphur during the ozonation of organic pollutants. These atoms, in fact, do not undergo complete mineralization. Copyright

Full text of DOI:10.1002/rcm.5045

Research Article
Received: 8 February 2011
Revised: 5 April 2011
Accepted: 7 April 2011
Published online in Wiley Online Library
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811
(wileyonlinelibrary.com) DOI: 10.1002/rcm.5045
Characterization of carbonyl by‐products during Uniblu‐A
ozonation by liquid chromatography/hybrid quadrupole
time‐of‐flight/mass spectrometry
*
A. Amorisco, V. Locaputo and G. Mascolo
Istituto di Ricerca Sulle Acque, Consiglio Nazionale delle Ricerche, Viale F. De Blasio 5, 70132 Bari, Italy
The structural elucidation of carbonyl‐containing by‐products arising from Uniblu‐OH ozonation has been
investigated by liquid chromatography/electrospray ionization tandem mass spectrometry (LC/ESI‐MS/MS) employing
a quadrupole time‐of‐flight mass spectrometer. The by‐products were derivatized with 2,4‐dinitrophenylhydrazine,
allowing the formation of [M–H]− ions of the derivatives in the electrospray source. Exact mass measurements of both the
[M–H]− ions and their product ions allowed the elemental formulae and related structures of ten by‐products to be
determined confidently. The main degradation pathway were decarboxylation followed by further oxidation. It is
noteworthy that the experimental procedure employed allowed the identification of both nitrogen‐ and sulphur‐
containing carbonyl by‐products during Uniblu‐OH ozonation. This result is of environmental relevance for monitoring
the balance of organic nitrogen and sulphur during the ozonation of organic pollutants. These atoms, in fact, do not
undergo complete mineralization. Copyright # 2011 John Wiley & Sons, Ltd.
Textile dyes and other dyestuffs represent a major source of
industrial water pollution in many countries due to their
persistence, acute toxicity or mutagenic effects to exposed
aquatic organisms. A major issue of discussion and regulation
in the developed countries is the removal of color from textile
wastewater, also including the requirement to recycle the
rinsing water from dye baths.
Conventional textile wastewater treatment plants that rely
on biological oxidation (biodegradation)[1] and/or simple
physicochemical (adsorption)[2] strategies are ineffective in
achieving complete decolorization and the removal of dye
residues because most dyestuffs are both resistant to biological
degradation and highly water soluble. It follows that the
employment of more powerful chemical methods, namely
those known as advanced oxidation processes (AOPs), is
necessary to reach the target legal limit for effluent discharge.
AOPs are chemical methods based on the generation of
hydroxyl radicals (•OH) able to oxidize contaminants in a
non‐selective manner.[3–8]
It has been reported that ozonation is one of the most
effective technologies for the remediation of textile wastewater
containing bio‐recalcitrant dyes and other persistent organic
compounds.[9,10] Several studies have been carried out that
demonstrate the efficiency of ozone treatment for removal of
dyes and some of them were also aimed at identifying the
resulting by‐products.[11–15] Due to its high oxidation poten-
tial, ozone can effectively break down the conjugated double
bonds of the chromophores of dyestuffs as well as other
functional groups such as their complex aromatic rings.[16–18]
This ultimately leads to the formation of smaller molecules
with fewer conjugated double bonds, thereby decreasing the
color of effluent. However, although the removal of dyes by
ozonation is often successfully achieved, their complete
oxidation (i.e., the total mineralization of organic carbon to
carbon dioxide) is seldom obtained. Therefore, ozonation by‐
products should usually be expected in ozonation effluents.
Reactive dyestuffs based on the anthraquinone structure
have long been used in cellulose fiber dyeing. They are more
resistant to biodegradation due to their fused aromatic
structures than azo‐based dyes.[19] The anthraquinone‐based
dyes represent an environmental problem as it is not possible
to recover them from spent baths because the reaction carried
out to link them to fibers leads to the formation, through a side
reaction, of the hydrolyzed dyes. As the fixation rate between
the reactive dye and the fiber is usually below 90% it follows
that the resulting spent waters cannot be further used and
should be disposed of properly. However, little information
about the ozonation products of anthraquinone reactive dyes
is available.[20–25] In general, as degradation by‐products are
more polar than the parent compounds their identification is
often carried out by liquid chromatography/electrospray
ionization mass spectrometry (LC/ESI‐MS), often employing
accurate mass measurement.[26,27] It is known that such by‐
products are transient compounds as they are further
degraded, leading to the formation of low molecular weight
carbonyl compounds which, in turn, are finally degraded to
low molecular weight organic acids. However, the identifica-
tion of low molecular weight aldehydes and ketones only on
the basis of comparison with authentic standards fails to
account for the total organic carbon present in the ozonated
aqueous solution.[28] It follows that unknown low molecular
* Correspondence to: G. Mascolo, Istituto di Ricerca Sulle
Acque, Consiglio Nazionale delle Ricerche, Viale F. De
Blasio 5, 70132 Bari, Italy.
E-mail: giuseppe.mascolo@ba.irsa.cnr.it
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811
Copyright # 2011 John Wiley & Sons, Ltd.

A. Amorisco, V. Locaputo and G. Mascolo
weight carbonyl compounds are present in these ozonation
products. The unequivocal identification of these compounds
should be possible by LC/ESI‐MS combined with accurate
mass measurement, both in single and tandem MS, using
quadrupole time‐of‐flight mass spectrometry (QqTOFMS). In
order to achieve such a goal it is mandatory to employ a
preliminary derivatization step, for example by using 2,4‐
dinitrophenylhydrazine (DNPH) to form hydrazone deriva-
tives, as reported in many applications such as the study
of automobile and cigarette exhaust samples,[29] beer,[30]
ozonated drinking waters and outdoor swimming pools,[31]
and biological matrices including urine[32–34] and exhaled
breath.[35]
The aim of the present investigation was the unequivocal
characterization of aldehydes and ketones, formed during the
ozonation of hydrolized Uniblu‐A (Uniblu‐OH) at longer
contact times, by accurate mass MS and MS/MS determina-
tion using LC/ESI‐QqTOFMS after the derivatization of the
compounds with DNPH.
purging with air. They were then subjected to derivatiza-
tion with DNPH according to a published procedure[36,37]
in order to determine the aldehydes and ketones by LC/
MS. Briefly, 50 mL of an aqueous sample was buffered to
pH 3 with citrate buffer and 3 mL of a DNPH aqueous
solution (3 g/L) was added. The compounds were deriva-
tized at 40°C for 1 h. The aqueous phase was then
analyzed by LC/MS.
LC/ESI‐QqTOFMS and LC/ESI‐QqTOFMS/MS analysis
A QSTAR QqTOFMS/MS system (AB Sciex, Foster City,
CA, USA) equipped with a TurboIonSpray source oper-
ated in negative ion mode was used throughout this
work. High‐purity nitrogen gas was used as the curtain
and collision gas, while high‐purity air was used as the
nebulizer and auxiliary heated gas. The TurboIonSpray
interface conditions were as follows: nebulizer voltage,
−4200 V; declustering potential, −20 V; focusing potential,
−100 V; nebulizer gas flow rate, 1 L/min; curtain gas flow
rate, 1 L/min; auxiliary gas (air) flow delivered by a turbo
heated probe, at 8 L/min at 350°C. Samples injected by a
series 200 autosampler (Perkin‐Elmer, Waltham, MA,
USA), equipped with a 9010 Rheodyne valve and a 200
μL loop, were eluted at 0.6 mL/min through an Alltima
5 mm C18 reversed‐phase column (250 × 3.2 mm; Alltech,
Deerfield, IL, USA) and C18 precolumn (4 × 2.0 mm;
Phenomenex, Macclesfield, UK) with the following gra-
dient, delivered by three series 200 micro‐LC pumps
(Perkin‐Elmer): from 85:5:10 (water/CH3CN/CH3COOH
1% in CH3CN) to 10:80:10 in 20 min, which was then held
for 5 min. The flow from the LC pumps was split 1:1, by
means of a zero dead volume T‐piece, to allow one‐half to
enter the TurboIonSpray interface. A few analyses were
also performed using two C18 Chromolith columns
(Merck, Darmstadt, Germany) connected in series at
1 mL/min using the same above‐reported gradient. In
this case the flow from the LC pumps was split 1:2.
Accurate mass measurements (four decimal places) were
carried out at a mass resolution higher than 5000 (full
width at half maximum) by obtaining averaged spectra
from chromatographic peaks and by then recalibrating
them using the [M–H]− ions of polyacrylic acid injected
EXPERIMENTAL
Chemicals
Uniblu‐A (95% purity; Aldrich, Milwaukee, WI, USA) was
used without further purification. Uniblu‐OH was obtained
by hydrolyzing an aqueous solution of Uniblu‐A (500 mg/L)
at pH 12 and 50°C for 1 h (hydrolysis yield >90%) as reported
below.
O
NH2
O
NH2
SO3Na
SO3Na
OH-
O
NH
O
NH
SO2CH=CH2
SO2CH2CH2OH
Uniblu-A
Uniblu-OH
Solvents used for high‐pressure liquid chromatography
(HPLC) were HPLC‐MS grade and purchased from Sigma‐
Aldrich (St. Louis, MO, USA). Water used for liquid
chromatography as well as for preparing all standard aqueous
solutions (18.2 MΩcm, organic carbon content ≤ 4 µg/L) was
obtained from a Milli‐Q Gradient A‐10 system (Millipore,
Billerica, MA, USA). 2,4‐Dinitrophenylhydrazine (DNPH;
purity ≥ 99%) was purchased from Aldrich. All other reagents
were analytical grade and purchased from VWR (West
Chester, PA, USA) or Carlo Erba (Milan, Italy).
−
−
post‐column (C9H13O4−, C12H17O6 and C15H21O8 at m/z
185.08193, 257.10306 and 329.12419, respectively). Accu-
rate mass product ion measurements (four decimal places)
were carried out by fragmenting the target (precursor)
[M–H]− ions of the selected DNPH derivative at an
optimized collision energy (CE) and at a collision gas
Ozonation experiments
pressure of
4 mTorr. Each averaged spectrum was
A volume of 500 mL of a Uniblu‐OH aqueous solution (the
pH was adjusted to 7 with HCl) was placed into a
Normag reactor (Normag, Limenau, Germany). Ozone
was produced by a Fisher 502 ozonator (Meckenheim,
Germany) fed with oxygen. The oxygen/ozone mixture
(1 L/min, 16 mg/L) was split to allow just 0.1 L/min to be
bubbled into the Normag reactor. The ozone output was
monitored before each experiment by determining (by
titration with sodium thiosulfate) the amount of free
iodine liberated from a potassium iodide solution. Sam-
ples (10 mL) were withdrawn from the reaction mixtures
at scheduled times and the residual ozone was stripped by
recalibrated using, as a lock mass, the precursor ion mass
obtained in single MS mode or the product ion at m/z
182.0207 ([M–H]− ion of 2,4‐dinitroaniline) which is the
typical product ion deriving from the derivatization agent,
or both these masses. The CE was optimized for each
deprotonated DNPH derivative by‐product in order to
obtain, where possible, spectra showing acceptable signal‐
to‐noise (S/N) ratios in the MS/MS spectra, with both the
[M–H]− ion and the highest number of product ions being
of high enough abundance for accurate mass determina-
tion. All MS data handling was performed using Analyst
QS software (AB Sciex).
wileyonlinelibrary.com/journal/rcm Copyright # 2011 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811

Identification of carbonyl by‐products of Uniblu‐OH ozonation by QqTOF
RESULTS AND DISCUSSION
by‐product 2 is probably due to its low intensity. Close
examination of the product ion spectrum of the deprotonated
The reaction of Uniblu‐OH with ozone led to the quick
disappearance of the parent compound. At the same time, a
number of by‐products were formed as a result of oxidation
reactions of the amino group, of the aromatic rings of the
anthraquinone nucleus and/or of the benzenesulfoxide
groups. Competitive reactions such as the hydrolysis of the
2‐hydroxyethylsulfonyl and benzenesulfoxide groups can be
invoked to explain the formation of the already detected by‐
products.[28] It is worth noting that such by‐products were
mainly identified at short ozonation times and disappeared
completely at reaction times longer than 15 min. At such
reaction times no by‐product was detectable by HPLC‐UV‐MS
but several low molecular weight DNPH‐aldehydes were
detected on the basis of authentic standards (formaldehyde,
acetaldehyde, glyoxal, propanal, 2‐propenal, propanone,
butanal, butanone, 2‐butenal). However, these compounds
failed to account for the whole organic matter left in the
reaction mixture suggesting the presence of unidentified
aldehydes and ketones. Indeed, some of them are also likely
to be sulfur‐containing aldehydes, on the basis of the finding
that the amount of mineralized sulfur (i.e. inorganic sulfate)
fails to account for the total organic sulphur of Uniblu‐OH
present at the beginning of the reaction. Employing the
derivatization procedure with DNPH ten unknown by‐
products were identified on the basis of accurate mass
measurements of both the [M–H]− ions and their main product
ions, as listed in Table 1. As expected, all the derivatized by‐
products gave a characteristic product ion at m/z 182,
corresponding to the [M–H]− ion of 2,4‐dinitroaniline, as a
consequence of the derivatization procedure with DNPH.
Therefore, this ion was usefully employed as a lock mass, as
mentioned in the Experimental section. It is worth noting that
the mass measured [M–H]− ions reported in Table 1 often have
associated errors of a few ppm due to the employed QqTOFMS
instrumentation. However, this did not prevent the elemental
composition being unequivocally assigned. In fact, if no atom
constraints were set several possible elemental compositions
are possible for each by‐product within 6 ppm error for mass
accuracy. Instead, by considering the additional information
that each by‐product had previously been derivatized by
DNPH it was possible to apply proper atom and DBE
constraints (C ≥ 6, H ≥ 3, O ≥ 4, N ≥ 4, S ≥ 0, DBE ≥ 7.5)
leading to obtaining a single or a couple of possible elemental
compositions for each compound (Supplementary Fig. S1,
Supporting Information). Further, when two elemental com-
positions were obtained, we found that one was not
chemically consistent due to the low number of hydrogen
atoms. Therefore, for all the derivatized by‐products a single
elemental composition was confidently obtained although the
mass accuracy error was about 6 ppm.
by‐product 1 (Fig. 1(a)) reveals other minor product ions at m/z
223.0512 and 257.0130, corresponding to C8H7N4O4 and
–
C8H7N3O5S–, respectively. The ion at m/z 223 is formed through
the loss of SO3 from the precursor ion. This ion then gives rise
to the [M–H]– ion of 2,4‐dinitroaniline (m/z 182) through loss
of acetonitrile. The radical anion at m/z 257 represents
the competitive loss of nitrous acid from the deprotonated
molecule followed by a proton shift. All the information
obtained from the product ion spectrum was rationalized in the
fragmentation pathway depicted in Fig. 2, leading to the
conclusion that by‐product 1 is a 2‐oxoethanesulfonic acid,
DPNH derivatized in position 2. It is formed by loss of 2‐
hydroxyethanesulfonic acid from Uniblu‐OH and then oxida-
tion of the ethoxy group.
The product ion mass spectrum of deprotonated by‐product
2 (Supplementary Fig. S1, Supporting Information) revealed
only an additional product ion at m/z 96.9600, assigned to
–
HSO4 (error −1.1 ppm). On the basis of such few observed
product ions, the structure of by‐product 2 is proposed as 3‐
oxopropane‐1‐sulfonic acid, DPNH derivatized in position 3
(Supplementary Fig. S2, Supporting Information). This com-
pound is formed by cleavage of the 2‐hydroxyethylsulfonyl
group of Uniblu‐OH and oxidation of the hydroxyl moiety.
From Table 1 it can be seen that many of the by‐products (3–8)
formed product ions through neutral loss of 44 Da, generally
indicating the presence of a carboxylic acid group. The [M–H]–
ions of by‐products 3 and 4 were at m/z 253.0208 and 267.0364
–
with elemental compositions of C8H5N4O6 (error −2.6 ppm)
−
and C9H7N4O6 (error −2.6 ppm), respectively. Their struc-
tures were assigned as the DNPH derivatives of 2‐oxoacetic
acid and 2‐oxopropanoic acid, i.e. an aldehyde and a saturated
ketone, respectively. The product ion spectra of these by‐
products were mainly dominated by the m/z 182 ion
(Supplementary Figs. S3 and S5, Supporting Information)
and had similar fragmentation pathways. Specifically, the
deprotonated molecules yield product ions at m/z 209.0363
and 223.0504, neutral loss of CO2 and then the base peak (the
[M–H]− ion of 2,4‐dinitroaniline at m/z 182) through CH3CN
loss and hydrogen shift (Fig. 3 and Supplementary Fig. S4,
Supporting Information). This product ion, in turn, gives rise
to an ion at m/z 152, an odd‐electron ion consistent with an
elemental composition of C6H4N2O3−⋅, formed by loss of a NO
radical (30 Da). It has been reported that product ions
originating from the DNPH moiety might be affected by the
structure of the respective carbonyl compound, making it
possible to distinguish between aliphatic aldehydes (product
ion at m/z 163 and less abundant ion at m/z 179), aldehydes and
ketones (intense m/z 152 product ion) and dicarbonyl
compounds (intense m/z 182 product ion).[38,39] The product
ion spectra of the DNPH derivatives of by‐products 3 and 4
(Supplementary Figs. S3 and S5, Supporting Information)
revealed a base ion peak at m/z 182 and product ions at m/z 152
and 163 of very low intensity. It follows that the above
reported features cannot be applied to compounds having
multiple functional groups. Specifically, when the carboxylic
group is present together with an aldehyde or a ketone, a very
intense m/z 182 product ion (base peak) is observed. This
means that in such a situation the dicarbonyl character of the
compound prevails in the product ion spectrum over the other
characteristics. The product ion at m/z 163.0279 of by‐product
By‐products 1 and 2 showed [M–H]− ions at m/z 303.0038 and
317.0217, consistent with elemental compositions C8H7N4O7S−
and C9H9N4O7S– (error −1.0 and 6.2 ppm, respectively)
indicating a methylene group difference between the two com-
pounds. Their MS/MS spectra (Fig. 1(a) and Supplementary
Fig. S1, Supporting Information) show product ion at m/z
–
80.9648 and 80.9681 attributed to HSO3 (error −4.8 and
35.8 ppm for by‐product 1 and 2, respectively), suggesting that
these compounds are sulfur‐containing by‐products. The high
−
error in the measurement of the HSO3 product ion from
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811 Copyright # 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

A. Amorisco, V. Locaputo and G. Mascolo
Table 1. Accurate mass measurements and elemental compositions of degradation products of Uniblu‐OH derivatized with
DNPH and their product ions using LC/QqTOF‐MS and LC/QqTOF‐MS‐MS analysis
Structures of by‐products
derivatized by DNPH
Measured mass
CE
(eV)
Major
ions
Elemental
composition
Calculated
mass
Error
(ppm)
[M–H]−
303.0038
20
303.0038
257.0130
223.0512
182.0207
80.9648
C8H7N4O7S−
C8H7N3O5−S−·
C8H7N4O4−
C6H4N3O4
303.0041
257.0112
223.0473
182.0207
80.9652
−1.0
7.0
17.5
LM
−4.8
(1)
NO2
N
N
SO3H
−
HSO3
O2N
H
317.0217
15
317.0217
182.0207
96.9600
80.9681
C9H9N4O7−S−
C6H4N3O4
317.0197
182.0207
96.9601
80.9652
6.2
LM
(2)
NO2
−
HSO4−
−1.1
N
N
HSO3
35.8
N
N
SO3H
O2N
H
H
−
253.0208
267.0364
15
20
253.0208
209.0363
182.0145
163.0279
152.0240
C8H5N4O6−
253.0215
209.0316
182.0207
163.0261
152.0227
−2.6
22.3
−34.2
10.7
8.3
(3)
NO2
C7H5N4O4−
C6H4N3O4−
C6H3N4O2−·
C6H4N2O3
O
O
OH
O2N
−
267.0364
223.0504
197.0328
182.0209
179.0232
167.0335
163.0288
152.0248
C9H7N4O6−
267.0371
223.0473
197.0316
182.0207
179.0211
167.0336
163.0261
152.0227
−2.6
13.9
5.9
(4)
NO2
C8H7N4O4−
C6H5N4O4−
C6H4N3O4−
C6H3N4O3−·
C6H5N3O3−
C6H3N4O2−·
C6H4N2O3
N
N
0.9
OH
11.9
−0.8
16.3
13.5
O2N
CH3
−
323.0288
15
323.0288
279.0360
182.0207
96.0064
C11H7N4O8−
323.0269
279.0371
182.0207
96.0091
5.8
−3.9
LM
(5)
NO2
C10H7N4O6
O
−
C6H4N3O4
N
N
−
N
O
A
C4H2NO2
−28.1
O2N
H
O
OH
NO2
O
B
N
O
CH3
O2N
O
OH
−
297.0101
311.0286
20
15
297.0101
253.0212
209.0306
182.0207
C9H5N4O8−
297.0113
253.0215
209.0316
182.0207
−4.0
−1.2
−4.9
LM
(6)
C8H5N4O6−
C7H5N4O4−
C6H4N3O4
NO2
O
N
N
OH
O2N
HO
O
−
311.0286
267.0387
223.0476
182.0207
C10H7N4O8
311.0269
267.0371
223.0473
182.0207
5.3
5.9
1.4
LM
OH
OH
(7)
A
NO2
O
H
−
C9H7N4O6−
N
C8H7N4O4−
C6H4N3O4
N
O2N
O
OH
NO2
O
N
B
N
O2N
HO
O
(Continues)
wileyonlinelibrary.com/journal/rcm Copyright # 2011 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811

Identification of carbonyl by‐products of Uniblu‐OH ozonation by QqTOF
Table 1. (Continued)
Structures of by‐products
derivatized by DNPH
Measured mass
CE
(eV)
Major
ions
Elemental
composition
Calculated
mass
Error
(ppm)
[M–H]−
−
325.0442
15
325.0442
281.0536
253.0184
182.0207
179.0219
163.0219
153.0301
C11H9N4O8−
325.0426
281.0528
253.0215
182.0207
179.0211
163.0261
153.0306
5.0
3.0
NO2
(8)
C10H9N4O6
O
OH
−
N
N
C8H5N4O6−
−12.2
LM
O
C6H4N3O4−
C6H3N4O3−
C6H3N4O2−
C6H5N2O3
A
O2N
H
OH
4.6
−26.1
− 3.0
NO2
O
N
N
O
OH
B
O2N
OH
NO
−
252.0366
260.0425
15
10
252.0366
209.0318
182.0169
166.0283
163.0247
C8H6N5O5−
252.0374
209.0316
182.0207
166.0258
163.0261
−3.3
0.8
−21.0
14.9
−8.9
NO2
C7H5N4O4−
C6H4N3O4−
C6H4N3O3−
C6H3N4O2
(9)
N
N
O2N
H
−
260.0425
223.0471
182.0198
C10H6N5O4
260.0425
223.0473
182.0207
−0.1
−0.9
–5.1
−
(10)
C8H7N4O4−
NO2
C6H4N3O4
N
N
N
O2N
LM: lock mass used to recalibrate the averaged spectrum.
3, typical for aliphatic DNPH‐aldehydes, is formed by oxalic
acid elimination and a cyclization reaction which stabilizes the
formed product ion (Supplementary Fig. S4, Supporting
Information). In addition, the formation of an ion at m/z
between the two proposed DNPH structures, namely 3,4,5‐
trioxopentanoic acid and 2,3,4‐trioxopentanoic acid.
By‐products 6 and 7 showed similar fragmentation patterns.
They exhibited [M–H]− ions at m/z 297.0101 and 311.0286,
−
–
179.0232 of by‐product 4 was assigned to C6H3N4O3 (error
associated with elemental compositions of C9H5N4O8 and
–
11.9 ppm). Its formation can be rationalized by consecutive
C10H7N4O8 (error −4.0 and 5.3 ppm), respectively. Both
−
loss of ethene (leading to the ion at m/z 197.0328, C6H5N4O4
,
product ion spectra showed two consecutive carbon dioxide
eliminations, compatible with a double carboxylation on the
DNPH derivative, leading to the formaldehyde derivative at
m/z 209.0306 (error −4.9 ppm) and the acetaldehyde derivative
at m/z 223.0476 (error 1.4 ppm) for by‐products 6 and 7,
respectively (Figs. 1(b) and 1(c)). These ions then undergo
further fragmentation leading to the ion at m/z 182 (Figs. 4 and
5). For by‐product 6 the accurate mass measurement result
supported the assignment of a malonic acid derivative with a
carbonyl group at the 2‐position. For by‐product 7, on the
other hand, on the basis of the product ion spectrum alone it
was not possible to distinguish between two possible
dicarboxylic acid structures derivatized at the oxo position,
namely 2‐formylmalonic acid and 2‐oxosuccinic acid. The
second possible structure can be also correlated with by‐
product 4 through an oxidative decarboxylation.
error 5.9 ppm) and water from the product ion at m/z 223.0504
(Fig. 3), assigned to acetaldehyde‐DNPH, this hypothesis
being supported by published data.[38] In addition, the
proposed structure of by‐product 4 was also consistent with
its chromatographic behaviour, showing a double peak, as
typically found for asymmetric DNPH‐ketones.
By‐product 5 showed a [M–H]− ion at m/z 323.0288
−
consistent with elemental composition C11H7N4O8 (error
5.8 ppm) for which two possible chemical structures can be
proposed (Table 1). The product ion spectrum (Supplementary
Fig. S6, Supporting Information) showed, in addition to the
intense ion at m/z 182, the formation of a product ion at m/z
–
279.0360 whose elemental composition is C10H7N4O6
(error −3.9 ppm). This ion can be assigned to deprotonated
2,3‐dioxobutanal‐DNPH, formed by CO2 neutral loss, sup-
porting the presence of a mono‐carboxylic acid group on this
by‐product. This ion fragmented further, leading to the ion at
m/z 182 and competitively to the ion at m/z 96.0064 assigned to
By‐product 8 showed a [M–H]− ion at m/z 325.0442 consistent
−
with the elemental composition C11H9N4O8 (error 5.0ppm).
The product ion spectrum easily allowed us to assign this by‐
product as a carboxylated derivative but, once again, it was not
possible to unequivocally assign a single structure (Supple-
mentary Fig. S8, Supporting Information). In fact, two product
–
C4H2NO2 (error −28.1 ppm), as depicted in Supplementary
Fig. S7 (see Supporting Information). However, in this case
the fragmentation pattern did not allow us to distinguish
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811 Copyright # 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

A. Amorisco, V. Locaputo and G. Mascolo
NO2
80.9648
NO2
182.0207
O
434
400
N
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
N
a
b
N
N
OH
SO3H
O2N
H
O2N
350
300
250
200
150
100
50
HO
O
By-product 1
By-product 6
303.0048
297.0101
253.0212
209.0306
182.0207
257.0130
223.0512
0
80 100 120 140 160 180 200 220 240 260 280 300 320
m/z
80 100 120 140 160 180 200 220 240 260 280 300
m/z
OH
NO2
O
NO2
182.0169
182.0207
N
N
NO
2.2e5
2.0e5
1.8e5
1.6e5
1.4e5
1.2e5
1.0e5
8.0e4
6.0e4
4.0e4
2.0e4
0.0
N
OH
N
1.4e4
1.3e4
N
d
c
O2N
O2N
H
O
O
1.2e4
1.1e4
1.0e4
O2N
H
OH
NO2
By-product 9
N
9000.0
8000.0
7000.0
6000.0
5000.0
4000.0
3000.0
2000.0
1000.0
0.0
HO
O
By-product 7
267.0387
223.0476
209.0318
311.0281
166.0283
163.0247
252.0374
240 260
80 100 120 140 160 180 200 220 240 260 280 300 320
m/z
80
100
120
140
160
180
200
220
m/z
Figure 1. Calibrated product ion spectra of (a) by‐product 1 (20 eV), (b) by‐product 6 (20 eV), (c) by‐product 7 (15 eV), and (d)
by‐product 9 (15 eV). Ions at m/z 182.0207 (by‐products 1, 6 and 7) and 252.0374 (by‐product 9) were used as lock mass.
NO2
- C8H6N4O4
-
HSO3
N
N
80.9652
m/z
SO3H
O2N
H
-
by-product 1
, C8H7N4O7S
303.0041
- SO3
m/z
NO2
NO2
O
S O
O
N
N
N
CH3
N
O2N
H
H
-
C8H7N3O5S -
257.0112
C8H7N4O4
- CH3CN
NH
223.0473
m/z
m/z
NO2
O2N
-
C6H4N3O4
182.0207
m/z
Figure 2. Proposed fragmentation pathway of by‐product 1 from LC/QqTOF‐
MS/MS data at 20 eV. Accurate mass measurements of product ions are
reported in Table 1.
ions are present, at m/z 281.0536 and 253.0184, consistent with a
mass measurement of the latter ion made it possible to remove
ambiguities between CO and C2H4 elimination from [M–H]−,
both of which have the same nominal mass of 28Da. A
CO2 loss (error 3.0 ppm) and C2H4 elimination (error
−12.2ppm) from [M–H]−, respectively. Interestingly, accurate
wileyonlinelibrary.com/journal/rcm Copyright # 2011 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811

Identification of carbonyl by‐products of Uniblu‐OH ozonation by QqTOF
NO2
O
N
N
OH
O2N
, C9H7N4O6
CH3
- CO2
-
by-product 4
267.0371
m/z
NO2
NO2
NH
- CH3CN
H shift
N
N
H
O2N
C6H4N3O4
O2N
CH3
-
-
C8H7N4O4
182.0207
223.0473
m/z
m/z
- NO
- C2H2
NO2
NO2
NO2
N
- H2O
NH2
N
HN
N N
O
O2N
O
-
-
-
C6H4N2O3
152.0227
C6H5N4O4
197.0316
C6H3N4O3
m/z
- NO
179.0211
m/z
m/z
- H2O2
NO2
NO2
N
O
NH2
N
N N
-
C6H5N3O3
m/z
167.0336
-
C6H3N4O2
163.0261
m/z
Figure 3. Proposed fragmentation pathway of by‐product 4 from LC/QqTOF‐
MS/MS data at 20 eV. Accurate mass measurements of product ions are
reported in Table 1.
NO2
O
NO2
O
N
further fragmentation pathway that is common to that of the
2‐oxoacetic acid‐DNPH derivative (by‐product 3) is depicted in
Supplementary Fig. S9 (see Supporting Information).
- CO2
N
N
N
OH
OH
O2N
O2N
H
-
HO
O
By‐products 9 and 10 were shown to be nitrogen‐containing
compounds. They exhibited [M–H]− ions at m/z 252.0366
and 260.0425, associated with elemental compositions of
C8H5N4O6
-
by-product 6, C9H5N4O8
253.0215
m/z
297.0113
m/z
−
−
C8H6N5O5 and C10H6N5O4 (error −3.3 and −0.1 ppm),
respectively. Their structures were rationalized as the DNPH
derivatives of 2‐nitrosoacetaldehyde and 4‐oxobut‐2‐enenitrile,
respectively. Their identification confirmed the assumption
that nitrogen‐containing by‐products must be present during
the ozonation of Uniblu‐OH. In fact, previous experiments
employing total nitrogen measurements showed that almost no
nitrogen mineralization occurs during the entire reaction
time.[28] The product ion spectrum of deprotonated by‐product
9 (Fig. 1(d)) revealed the presence, in addition to the ion at m/z
182 (base peak), of ions at m/z 209 and 163 typical of aldehyde‐
DPNH derivatives as discussed above (Fig. 6). The product ion
spectrum of deprotonated by‐product 10 (Supplementary
- C3HNO4
- CO2
NO2
NO2
- HCN
NH
N
N
CH2
O2N
O2N
-
-
C6H4N3O4
182.0207
C7H5N4O4
209.0316
m/z
m/z
Figure 4. Proposed fragmentation pathway of by‐product 6
from LC/QqTOF‐MS/MS data at 20 eV. Accurate mass
measurements of product ions are reported in Table 1.
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811 Copyright # 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

A. Amorisco, V. Locaputo and G. Mascolo
OH
NO2
O
OH
O
A
B
NO2
O
H
N
N
N
or
N
OH
O2N
O2N
HO
O
-
by-product 7
, C10H7N4O8
311.0269
m/z
- CO2
O
- C4H3NO4
OH
NO2
NO2
NH
N
N
O2N
O2N
C9H7N4O6
267.0371
H
-
-
C6H4N3O4
182.0207
m/z
- CO2
- CH3CN
CH3
m/z
NO2
N
N
O2N
H
-
C8H7N4O4
223.0473
m/z
Figure 5. Proposed fragmentation pathway of by‐product 7 from LC/QqTOF‐
MS/MS data at 15 eV. Accurate mass measurements of product ions are
reported in Table 1.
NO2
NO2
NO
N
N
N
O2N
H
O
NH
-
by-product 9
, C8H6N5O5
252.0374
-
C6H4N3O3
m/z
m/ z
NO
166.0258
-
CH2CN
NO2
NO2
NO2
NH
N
N
N
CH2
N N
O2N
O2N
-
-
-
C6H4N3O4
182.0207
C7H5N4O4
m/z
C6H3N4O2
209.0316
163.0261
m/z
m/z
Figure 6. Proposed fragmentation pathway of by‐product 9 from LC/QqTOF‐
MS/MS data at 15 eV. Accurate mass measurements of product ions are
reported in Table 1.
Fig. S10, Supporting Information ) showed a diagnostic ion at
insights about their formation mechanism, and the measured
profiles are depicted in Fig. 7. The identified by‐products are
consistent with further oxidation of the anthraquinone
moiety and/or of the benzenesulfoxide group of by‐products
previously detected in the early stage of Uniblu‐OH
ozonation.[28] It follows that through hydroxylation and
subsequent decarboxylation reactions, very polar low molec-
ular weight carbonyl compounds, such as keto‐acids and
aliphatic aldehydes, are formed. These include newly identi-
fied by‐products (Table 1) as well as a number of aldehydes of
known chemical structure. Specifically, by‐products 6, 7 and 8
m/z 223.0471 (base peak) with elemental composition of
C8H7N4O4− (error −0.9 ppm) assigned to acetaldehyde‐DNPH
that further decomposes to the ion at m/z 182 (Supplementary
Fig. S11, Supporting Information).
Formation mechanism of identified low molecular weight
carbonyl by‐products
The identified low molecular weight carbonyl by‐products
were monitored over a long ozonation time in order to gain
wileyonlinelibrary.com/journal/rcm Copyright # 2011 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811

Identification of carbonyl by‐products of Uniblu‐OH ozonation by QqTOF
8.0e+5
by-product 1
by-product 3
by-product 4
by-product 6
by-product 9
by-product 2
by-product 5
by-product 7
by-product 8
by-product 10
50000
40000
30000
20000
10000
0
6.0e+5
4.0e+5
2.0e+5
0.0
0
20
40
60
80
100
120
0
20
40
60
80
100
120
ozonation time (min)
ozonation time (min)
Figure 7. Formation/degradation profiles of identified carbonyl by‐products (extracted ion peak area from
HPLC‐ESI‐QqTOFMS) during ozonation of Uniblu‐OH.
are likely to be correlated since their structures differ by one
methylene group. The most probable mechanism begins with
the decarboxylation of by‐products 5 and 8 followed by
oxidation of the terminal carbon of the formed intermediate,
leading ultimately to by‐product 7. Such an oxidation is likely
to proceed through hydrogen homolytic abstraction by a
hydroxyl radical forming a peroxy radical, and then a
hydroperoxide. This mechanism can further proceed giving
rise to by‐product 6 as shown in the degradation pathway
(Fig. 8). In addition, the formation of this compound also
occurs through degradation of by‐product 5 by decarboxyla-
tion and oxidation of both the aldehyde and the methyl group
(structure A) or homolytic cleavage of the terminal acetyl
group and further oxidation (structure B). In addition,
cleavage between the keto groups in positions 2 and 3
followed by hydrolysis produces by‐product 4 (Fig. 8). It
follows that three of the by‐products all lead to a single
compound, namely by‐product 6. This is also consistent with
the formation/decomposition profiles of the identified by‐
products (Fig. 7), showing that by‐product 6 is formed at much
higher abundance than by‐products 5, 7 and 8. In addition, a
further decarboxylation of by‐product 6 may lead to by‐
product 3. Again, this is consistent with the formation/
decomposition profiles, showing the latter compound to be
persistent at longer reaction time with high abundance.
The same mechanism of decarboxylation and further
oxidation discussed above can be applied to by‐product 2 in
order to explain the formation of by‐product 1. This is, once
again, consistent with the formation/decomposition profiles
showing by‐product 1 to be persistent and to be formed at
much higher abundance than the other compounds. The
finding that sulfur‐containing by‐products are still present at
O
O
O
NH2
SO3H
NO
SO3Na
H
H
ozonation
intermediate
organic
acids
by-product 1
by-product 9
by-products[28]
O
NH
O
O
SO3H
SO2CH2CH2OH
N
Uniblu-OH
H
H
by-product 2
by-product 10
O
O
O
O
O
HO
H
HO
OH
OH
OH
H3C
CH3
O
O
O
O
O
O
O
O
by-product 4
by-product 8 (B)
by-product 5 (A)
by-product 5 (B)
- CO2,
- CO2,
oxidation
oxidation
oxidation
- CO2,
oxidation
- CO2,
oxidation
O
O
O
O
organic
acids
HO
OH
HO
OH
H
OH
O
O
O
O
by-product 7 (B)
by-product 6
by-product 3
Figure 8. Proposed pathway for Uniblu‐OH degradation by ozone leading to carbonyl by‐products.
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811 Copyright # 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm

A. Amorisco, V. Locaputo and G. Mascolo
[3] P. R. Gogate, A. B. Pandit. review of imperative
technologies for wastewater treatment II: hybrid methods.
Adv. Environ. Res. 2004, 8, 553.
[4] M. Pera‐Titus, V. García‐Molina, M. A. Baños, J. Giménez, S.
Espulgas. Degradation of chlorophenols by means of
advanced oxidation processes: a general review. Appl. Catal.
B 2004, 47, 219.
A
longer ozonation time confirms that the sulphur present in
Uniblu‐OH does not undergo complete mineralization. The
concentration of inorganic sulfate alone, in fact, fails to
account for the entire organic sulfur balance initially present
in the Uniblu‐OH.
[5] S. Devipriyas, S. Yesodharan. Photocatalytic degradation of
pesticide contaminants in water. Sol. Energy Mater. Sol. Cells
2005, 86, 309.
[6] J. J. Pignatello, E. Oliveros, A. MacKay. Advanced oxidation
processes for organic contaminant destruction based on the
Fenton reaction and related chemistry. Crit. Rev. Environ.
Sci. Technol. 2006, 36, 1.
[7] C. Comninellis, A. Kapalka, S. Malato, S. A. Parsons,
I. Poulios, D. Mantzavinos. Perspective. Advanced oxidation
processes for water treatment: advances and trends for R&D.
J. Chem. Technol. Biotechnol. 2008, 83, 769.
[8] M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis,
B. J. Mariñas, A. M. Mayes. Science and technology for
water purification in the coming decades. Nature 2008,
452, 301.
[9] O. J. Hao, H. Kim, P. C. Chiang. Decolorization of waste-
water. Crit. Rev. Environ. Sci. Technol. 2000, 30, 449.
[10] S. Mondal. Methods of dye removal from dye house effluent –
an overview. Environ. Eng. Sci. 2008, 25, 383.
[11] J. N. Wu, T. W. Wang. Ozonation of aqueous azo dye in a
semi‐batch reactor. Water Res. 2001, 35, 1093.
[12] M. Koch, A. Yediler, D. Lienert, G. Insel, A. Kettrup.
Ozonation of hydrolyzed azo dye Reactive Yellow 84 (CI).
Chemosphere 2002, 46, 109.
[13] A. Lopez, H. Benbelkacem, J. S. Pic, H. Debellefontaine.
Oxidation pathways for ozonation of azo dyes in a semi‐
batch reactors: a kinetic parameters approach. Environ.
Technol. 2004, 25, 311.
CONCLUSIONS
QqTOF‐ESI‐MS both in single and tandem MS mode,
coupled to HPLC, was employed for the characterization of
the several polar by‐products generated at longer reaction
time during the degradation of Uniblu‐OH by ozonation.
Specifically, the obtained results demonstrated the effective-
ness of accurate mass measurements for the identification
of by‐product structures as well as for obtaining detailed
information about their fragmentation patterns. Most of
the by‐products were characterized by single or double CO2
loss, consistent with assignment to aldheyde or keto‐
carboxylic acid structures. The employed experimental
approach allowed the identification of both nitrogen‐ and
sulfur‐containing carbonyl by‐products during Uniblu‐OH
ozonation. This result is of environmental relevance for
the balance of both organic nitrogen and sulfur during
the ozonation of Uniblu‐OH. These atoms, in fact, do not
undergo complete mineralization during AOPs. Owing to the
complexity of the reactions occurring during ozonation, it
is difficult to draw exhaustive reaction schemes explaining
the formation of all the formed by‐products. However, it
was possible to assess the presence of correlations between
different identified by‐products, such as aldehyde acids
and keto acids. Most of these compounds in turn decom-
pose, through decarboxylation and further oxidation, to
smaller homologues. Finally, formation/decomposition pro-
files confirm that most by‐products were further degraded
probably leading to the formation of low molecular weight
organic acids whose identification is still in progress in
our laboratory.
[14] A. Lopez, G. Ricco, G. Mascolo, G. Tiravanti, A. C. Di
Pinto, R. Passino. Biodegradability enhancement of refrac-
tory pollutants by ozonation: a laboratory investigation
on an azo‐dyes intermediate. Water Sci. Technol. 1998, 38,
239.
[15] F. F. Zhang, A. Yediler, X. M. Liang. Decomposition
pathways and reaction intermediate formation of the
purified, hydrolyzed azo reactive dye C.I. Reactive Red 120
during ozonation. Chemosphere 2007, 67, 712.
[16] A. Reife, H. S. Freeman. Environmental Chemistry of Dyes and
Pigments, John Wiley, New York, 1996.
SUPPORTING INFORMATION
[17] F. J. Beltran, Ozone Reaction Kinetics for Water and Wastewater
Systems, Levis Publishers, Florida, 2004 (imprint of CRC
Press LLC).
Additional supporting information may be found in the
online version of this article.
[18] O. S. G. P. Soares, J. J. M. Órfão, D. Portela, A. Vieira,
M. F. R. Pereira. Ozonation of textile effluents and dye
solutions under continuous operation: influence of oper-
ating parameters, J. Hazard. Mater. 2006, 137, 1664.
[19] D. Deng, J. Guo, G. Zeng, G. Sun. Decolorization of
anthraquinone, triphenylmethane and azo dyes by a new
isolated Bacillus cereus strain DC11. Int. Biodeter. Biodegr.
2008, 62, 263.
Acknowledgement
This work was partially supported by Apulia Region by
funding Rela‐Valbior project within the Scientific Research
Framework Program 2007–2013.
[20] J. L. Liu, H. J. Luo, C. H. Wei. Degradation of anthraquinone
dyes by ozone. Trans. Nonferr. Metal Soc. 2007, 17, 880.
[21] Z. Q. He, L. L. Lin, S. Song, M. Xia, L. J. Xu, H. P. Ying,
J. M. Chen. Mineralization of C.I. Reactive Blue 19 by
ozonation combined with sonolysis: performance optimiza-
tion and degradation mechanism. Sep. Purif. Technol. 2008,
62, 376.
[22] R. Lall, R. Mutharasan, Y. T. Shah, P. Dhurjati. Decolorization
of the dye, Reactive Blue 19, using ozonation, ultrasound,
and ultrasound‐enhanced ozonation. Water Environ. Res.
2003, 75, 171.
REFERENCES
[1] L. S. Andrade, L. A. M. Ruotolo, R. C. Rocha‐Filho, N. Bocchi,
S. R. Biaggio, J. Iniesta, V. Garcìa‐Garcia, V. Montiel. On the
performance of Fe and Fe, F doped Ti–Pt/PbO2 electrodes in
the electrooxidation of the Blue Reactive 19 dye in simulated
textile wastewater. Chemosphere 2007, 66, 2035.
[2] W. J. Epolito, Y. H. Lee, L. A. Bottomley, S. G. Pavlostathis.
Characterization of the textile anthraquinone dye Reactive
Blue 4. Dyes Pigm. 2005, 67, 35.
wileyonlinelibrary.com/journal/rcm Copyright # 2011 John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811

Identification of carbonyl by‐products of Uniblu‐OH ozonation by QqTOF
[23] Y. C. Hsu, Y. F. Chen, J. H. Chen. Decolorization of dye RB‐19
solution in a continuous ozone process. J. Environ. Sci. Health.
A 2004, 39, 127.
[24] J. M. Fanchiang, D. H. Tseng. Decolorization and transfor-
mation of anthraquinone dye Reactive Blue 19 by ozonation.
Environ. Technol. 2009, 30, 161.
[25] J. M. Fanchiang, D. H. Tseng. Degradation of anthraquinone
dye C.I. Reactive Blue 19 in aqueous solution by ozonation.
Chemosphere 2009, 77, 214.
[26] M. Constapel, M. Schellentrager, J. M. Marzinkowski, S. Gäb.
Degradation of reactive dyes in wastewater from the textile
industry by ozone: Analysis of the products by accurate
masses. Water Res. 2009, 43, 733.
mass aldehydes in human urine. J. Chromatogr. B 2010, 878
(7–8), 653.
[33] R. Andreoli, P. Manini, M. Corradi, A. Mutti, W. M. A. Niessen.
Determination of n‐hexane metabolites by liquid
chromatography/mass spectrometry 1. 2,5‐Hexanedione
and other phase I metabolites in untreated and hydrolyzed
urine samples by atmospheric pressure chemical ioniza-
tion. Rapid Commun. Mass Spectrom. 2003, 17, 637.
[34] M. Eggink, M. Wijtmans, R. Ekkebus, H. Lingeman,
I. J. P. de Esch, J. Kool, W. M. A. Niessen, H. Irth. Devel-
opment of a selective ESI‐MS derivatization reagent: synthe-
sis and optimization for the analysis of aldehydes in
biological mixtures. Anal. Chem. 2008, 80, 9042.
[27] A. Detomaso, G. Mascolo, A. Lopez. Characterization of
carbofuran photodegradation by‐products by liquid chro-
matography/hybrid quadrupole time‐of‐flight mass spec-
trometry. Rapid Commun. Mass Spectrom. 2005, 19, 2193.
[28] G. Mascolo, A. Lopez, A. Bozzi, G. Tiravanti. By‐products
formation during the ozonation of the Reactive Dye Uniblu‐A.
Ozone Sci. Eng. 2002, 24, 439.
[29] M. S. van Leeuwen, L. Hendriksen, U. Karst. Determination
of aldehydes and ketones using derivatization with 2,4‐
dinitrophenylhydrazine and liquid chromatography–
atmospheric pressure photoionization‐mass spectrometry. J.
Chromatogr. A 2004, 1058, 107.
[30] L. M. Gonçalves, P. J. Magalhães, I. M. Valente, J. G. Pacheco,
P. Dostálek, D. Sykora, J. A. Rodrigues, A. A. Barros. Analysis
of aldehydes in beer by gas‐diffusion microextraction: Char-
acterization by high‐performance liquid chromatography–
diode‐array detection–atmospheric pressure chemical
ionization–mass spectrometry. J. Chromatogr. A 2010, 1217,
3717.
[31] S. D. Richardson, T. V. Caughran, T. Poiger, Y. B. Guo, F.
G. Crumley. Application of DNPH derivatization with LC/
MS to the Identification of polar carbonyl disinfection by‐
products in drinking water. Ozone Sci. Eng. 2000, 22, 653.
[32] C. E. Baños, M. Silva. Liquid chromatography–tandem
mass spectrometry for the determination of low‐molecular
[35] P. Manini, R. Andreoli, S. Sforza, C. Dall’Asta, G. Galaverna,
A. Mutti, W. M. A. Niessen. Evaluation of Alternate Isotope‐
Coded Derivatization Assay (AIDA) in the LC–MS/MS
analysis of aldehydes in exhaled breath condensate. J.
Chromatogr. B 2010, 878, 2616.
[36] Methods for the Determination of Organic Compounds
in Drinking Water, Method 554. Determination of
Carbonyl Compounds in Drinking Water by DNPH
Derivatization (HPLC), U.S. Environmental Protection
Agency, 1992.
[37] E. Grosjean, P. G. Green, D. Grosjean. Liquid
Chromatography analysis of carbonyl (2,4‐dinitrophenyl)
hydrazones with detection by diode‐array ultraviolet
spectroscopy and by atmospheric pressure negative
chemical ionization mass spectrometry. Anal. Chem. 1999,
71, 1851.
[38] S. Kolliker, M. Oheme. Structure elucidation of 2,4‐
dinitrophenylhydrazone derivatives of carbonyl compounds
in ambient air by HPLC/MS and multiple MS/MS using
atmospheric chemical ionization in the negative ion mode.
Anal. Chem. 1998, 70, 1979.
[39] C. Zwiener, T. Glauner, F. H. Frimmel. Method optimization
for the determination of carbonyl compounds in disinfected
water by DNPH derivatization and LC‐ESI‐MS‐MS. Anal.
Bioanal. Chem. 2002, 372, 615.
Rapid Commun. Mass Spectrom. 2011, 25, 1801–1811 Copyright # 2011 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/rcm