Journal of Chromatography A, 1605 (2019) 360350

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Towards the revision of the drinking water directive 98/83/EC.

Development of a direct injection ion chromatographic-tandem mass
spectrometric method for the monitoring of fifteen common and
emerging disinfection by-products along the drinking water
supply chain
Maria Concetta Bruzzoniti a,∗ , Luca Rivoira a , Lorenza Meucci b , Martino Fungi b ,
Maria Bocina b , Rita Binetti b , Michele Castiglioni a
a
Department of Chemistry, University of Turin, via P. Giuria 7, Torino, Italy
b
SMAT S.p.A., Research Centre, C.so Unità d’Italia 235/3, Torino, Italy

a r t i c l e i n f o a b s t r a c t

Article history: According to the recent proposal released by the European Commission for the revision of the 98/83/EC
Received 21 April 2019 Directive, water suppliers will be requested to monitor the nine bromine- and chlorine congeners of
Received in revised form 28 June 2019 haloacetic acids, HAAs, as well as the oxyhalides chlorite and chlorate, as disinfection by-products (DBPs)
Accepted 2 July 2019
originated during the potabilization process.
Available online 29 July 2019
In this work, we propose a direct-injection method based on ion chromatography and mass spectro-
metric detection for the determination of the mentioned DBPs as well as bromate (already included in the
Keywords:
98/83/EC), implemented also for the following emerging HAAs monoiodo-, chloroiodo- and diiodo-acetic
Drinking water directive
Haloacetic acids
acids. The method was optimized to include the fifteen compounds in the same analytical run, tuning the
Ion-chromatography chromatographic (column and gradient) and detection conditions (suppression current, transitions, RF
Mass spectrometry lens settings and collision energies). To avoid matrix effect and to manage the instrumental conditions,
Plant monitoring optimization was performed directly in drinking water matrix. The method quantitation limits satisfy
the new limits imposed by the future directive and range from 0.08 ␮g/L (monobromoacetic acid) to
0.34 ␮g/L (trichloroacetic acid). The performance of the method was checked along different strategic
sampling points of three potabilization plants serving the city of Turin (Italy), including intermediate
treatments and finished waters. Recovery was checked according to the ±30% limit of acceptability set
by EPA regulations. The effect of disproportionate concentrations of chlorite and chlorate in respect to
HAAs on HAA signals was studied; this aspect is underestimated in literature. The method is routinely
applied by the potabilization plant of the city of Turin to confirm the effectiveness of all control measures
in abstraction, treatment, distribution and storage. This study represents the first example in Italy of
development and use of a cutting-edge technique for HAAs analysis along the potabilization processes.
© 2019 Elsevier B.V. All rights reserved.

1. Introduction European Citizens’ Initiative (ECI) Right2Water, the European Com-

mission adopted on 1 February 2018 a proposal for the revision
Within the European Community, the quality and safety of water of the Drinking Water Directive [2]. In the upcoming proposal for
intended for human consumption is currently disciplined by the the revision of the 98/83/EC Directive, attention is devoted to the
so-called 98/83/EC Drinking Water Directive [1]. disinfection by-products (DBPs) originated during potabilization
As a result of the Regulatory Fitness and Performance pro- process. More in detail, the directive requires the monitoring of the
gramme (REFIT) evaluation and of the follow up actions to the nine bromine- and chlorine congeners of haloacetic acids, HAAs,
(monochloro-, dichloro-, and trichloro-acetic acid, mono- and
dibromo-acetic acid, bromochloroacetic acid, bromodichloroacetic
acid, dibromochloroacetic acid and tribromoacetic acid), which
∗ Corresponding author at: Department of Chemistry, University of Turin, Via P. must not be present at concentrations higher than 80 ␮g/L as a
Giuria 5, 10125 Turin, Italy. sum. After trihalomethanes, HAAs are the second most prevalent
E-mail address: mariaconcetta.bruzzoniti@unito.it (M.C. Bruzzoniti).

https://doi.org/10.1016/j.chroma.2019.07.004
0021-9673/© 2019 Elsevier B.V. All rights reserved.
2 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350

DBP class generated in disinfected waters, and their toxicological mochloroacetic acid and tribromoacetic acid, chlorite, chlorate, as
effects are well ascertained [3]. well as bromate (already included in the 98/83/EC), and monoiodo-,
The upcoming revision of the 98/83/EC Directive is also going chloroiodo- and diiodo-acetic acids as emerging HAAs.
to regulate the presence of chlorate and chlorite, which are pre- Before applying the developed method to real samples of dif-
dominantly formed when the disinfectants used are hypochlorite ferent provenience, the robustness of the method was checked
and/or chlorine dioxide solutions. According to WHO recommen- evaluating the recovery of analytes in samples withdrawn from
dations, the guideline value allowed for chlorite and chlorate in different points of three potabilization plants, characterized by
drinking water is 0.7 mg/L. According to indications provided by matrix composition at different complexity. Quantitation limits and
the European Food Safety Authority (EFSA) on toxicological refer- acceptance criteria of the method fully comply future regulatory
ence value for chronic risk assessment provided for chlorate, the requirements. The method is currently routinely applied for the
EU Commission is going to regulate the presence of both chlorate analysis of fifteen DBPs by the laboratory in charge of supplying
and chlorite at the stricter level of 0.25 mg/L, overcoming the fact and monitoring drinking water in the Italian city of Torino.
that current EU drinking water directive does not set any specific This study represents the first example of simultaneous analysis
limits in drinking water. of the DBPs included in the forthcoming Drinking Water Directive
For those regions whose drinking water sources are impacted by revision and a rare example in Italy of development and application
sea water intrusion and thus contain relatively high concentrations of direct injection IC/MS-MS technique for the analysis of organic
of Br− and I− ions, besides brominated compounds, the presence and inorganic disinfection by-products along the drinking water
of iodinated (emerging) DBPs in finished drinking waters could supply chain (raw, treated and distributed waters).
also be observed [4]. Monoiodoacetic acid inhibits glyceraldehyde-
3phosphate dehydrogenase (GAPDH) activity in a greater extent 2. Materials and methods
than bromo- and chloro-analogous [5].
Bromate occurrence in drinking water is ascribed to the oxida- 2.1. Chemical standards and reagents
tion of Br− naturally occurring in water during ozonation process,
even if bromate could be present, as a contaminant, in commercial Acetonitrile, ammonium chloride, monoiodoacetic acid (MIAA),
solutions of sodium hypochlorite used for disinfection of drinking as well as the following isotopically enriched internal stan-
water [6]. Bromate is considered a probable human carcinogen, it dards monobromoacetic acid-1-13 C (MBAA-13 C), dichloroacetic
was listed in B2 Group by IARC and its presence is regulated in acid-2-13 C (DCAA-13 C), trichloroacetic acid-2-13 C (TCAA-13 C), were
drinking waters by US EPA and 98/83/EC Directive which both set from Sigma Aldrich (St. Louis, MO, USA). Iodoacetic acid-D3
a limit of 10 ␮g/L. (MIAA-D3), diiodoacetic acid (DIAA) and chloroiodoacetic acid
Regarding the analytical determination of DBPs, the methods (CIAA), were from Chemical Research (Rome, Italy). Inorganic
most used for this purpose are based on gas (GC) and liquid chro- anions were purchased in a standard mixture of 1000 mg/L from
matographic (LC) techniques. GC is employed for HAA analysis after Ultra Scientific (Bologna, Italy). The nine bromo- chloro- HAA
a preliminary derivatization step [7], as recommended by EPA 552.2 congeners (monochloro- MCAA, dichloro-DCAA, and trichloro-
method [8]. Detection can be accomplished with ECD or MS [9] acetic acid TCAA, mono- MBAA and dibromo-acetic acid DBAA,
detectors at ␮g/L levels. bromochloroacetic acid BCAA, bromodichloroacetic acid BDCAA,
LC methods are mainly based on the anion-exchange methods, dibromochloroacetic acid DBCAA and tribromoacetic acid TBAA)
exploiting, if possible, the ionic nature of the DBPs. Ion chromatog- were purchased from Restek (Bellefonte, PA, USA) in a mixture
raphy coupled to MS-MS detection allows to achieve detection containing 1000 mg/L of each HAA in MTBE. Deionized water
limits for selected HAAs at fractions of ␮g/L without sample pre- (18.2 M cm resistivity) for eluent preparation and for dilution of
treatment [10,11]. Hundreds ng/L detection limits levels can be stock standard solutions was obtained by an EMD Millipore Milli-Q
achieved for HAAs enriching acidified sample onto functionalized Direct Water Purification System (Millipore, Bedford, MA, USA).
graphene/alumina nanocomposites [12]. So far, only few emerging
iodinated HAAs have been monitored in waters, using GC–MS [13] 2.2. Instrumental equipment and operating conditions
and LC–MS [4] methods after sample pretreatment or direct large
volume injection [14]. Oxyhalide DBPs (chlorite, chlorate, perchlo- A Thermo Fisher Scientific (Waltham, MA USA) ICS-5000 IC sys-
rate and bromate) are easily determined in drinking water using ion tem was used throughout this work. The system includes a DP dual
chromatography with suppressed conductivity as recommended pump module for analytical and capillary applications, a CD con-
by EPA methods 300.1 [15] and 314.0 [16], colorimetry [17] and in ductivity detector, an AS autosampler, and a Reagent-Free (RFIC)
few cases by mass or mass tandem spectrometry [18]. eluent generator EG-5000 with ECG III cartridges KOH to provide
In view of the upcoming revision of the Drinking Water Direc- the gradient of KOH (mobile phase) using deionized water from
tive 98/83/EC, water suppliers that treat and supply drinking water an AXP-MS pump (Thermo Fisher Scientific). For sample injections
as well as institutions in charge to control safety of the distributed (120 ␮L), two autosamplers without (AS-DV) and with sample tray
water must be ready to measure all the above-mentioned com- temperature control (AS-AP) set at 9 ± 1 ◦ C were used; both were
pounds in a routine basis, to meet future legislative requirements. from Thermo Fisher Scientific. Separations were performed on an
The aim of this work is to develop a sensitive, accurate method IonPac AS24 (250 x 2 mm i.d.) coupled with a guard column IonPac
without sample pretreatment for the determination of DBPs, AG24 (50 x 2 mm i.d.) both from Thermo Fisher Scientific, ther-
including emerging iodinated HAAs, in one chromatographic run, to mostatted at 15 ◦ C in order to minimize the degradation at high pH
be used by the water supplier laboratories for the routine controls values for MBAA, CDBAA and TBAA. Eluent gradient (0.3 mL/min)
required for the upcoming Drinking Water Directive. was set as follows, 7 mM KOH: t = 0–15.1 min; 7–15.5 mM KOH:
With the aim of satisfying currently accepted EPA standards t = 15.1–25.8 min; 60 mM KOH: t = 25.9 min, keep until 46 min;
[10] in an analytical method of wider applicability, an ion chro- 7 mM KOH; t = 47–58 min.
matographic method with tandem mass spectrometry was here To remove trace anion contaminants from hydroxide eluent and
optimized for the simultaneous determination of the DBPs sub- to minimize base line shifts during gradient operation, an elec-
jected to the attention of the future legislation, i.e. monochloro-, trolytically continuously regenerated trap column (CR-ATC, 8% DVB
dichloro-, and trichloro-acetic acid, mono- and dibromo-acetic crosslinking, 55 ␮m particle size) was installed in the eluent line
acid, bromochloroacetic acid, bromodichloroacetic acid, dibro- after the pump prior to the sample injection. After eluent gener-
 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350 3

ation and before the separation column, Electrolytic suppression species [10,20]. In this work, the [M−H]− was preferred over the
was accomplished using an ASRS 500 (2-mm) from Thermo Fisher [M−COOH]− species due to the difference in signal response which
Scientific. was as high as 104 ([M−H]− /[M−COOH]− ).
A TSQ Endura triple-stage quadrupole mass spectrometer with For each precursor ion, the three most abundant product ions
ESI interface (HESI-II) was employed for detection. A diverter valve were monitored. Transitions to halide substituent were found to
was used to waste the anion interfering species from matrix, thus be the most abundant for HAAs containing one and three halo-
preventing inorganic anions to enter the MS equipment. After the IC gen atoms, i.e. MCAA, MBAA, MIAA, BDCAA, CDBAA, TBAA, except
suppressor and before the ESI inlet, acetonitrile (CH3 CN) was added for TCA, for which transition to the [M−COOH]− ion is preferred.
to the eluate at 0.3 mL/min through an additional AXP-MS pump. For HAAs containing two halogen atoms, except for CIAA, the
The addition of CH3 CN leads to higher efficiency in gas phase ion [M−COOH]− ion is also preferred. These findings are coherent with
generation during the ESI process [19], enhancing analyte sensitiv- literature reports [23]. CIAA exhibits the most abundant transi-
ity [20]. The MS spectrometer was tuned and calibrated through tion to the I− ion in agreement with detection studies conducted
the software TSQ Endurance Tune Application 2.1 (Thermo Fisher in reversed phase liquid chromatography and tandem mass spec-
Scientific) by direct infusion of polytyrosine-1,3,6 (Thermo Fisher trometry [14].
Scientific). Performance was checked every two weeks using the Precursor ion was used as quantifier ion, whereas product ion
same polytyrosine-1,3,6 solution. was used as qualifier ion.

2.3. Preparation of standard solutions and water samples 3.2. Optimization of ion chromatographic conditions

Standard solutions were prepared in 5-mL vials directly in the Separation column. The fifteen DBPs and the main common
autosampler. Ten levels of standard solutions were used for the con- anions in drinking water are characterized by different chemical
struction of the calibration curve which was comprised between properties, hence their simultaneous separation in matrix is a chal-
0.25 and 20 ␮g/L starting from a 1 mg/L standard mixture of DBPs lenging task. Gradient conditions are often required to provide
in water. To each standard solution, 500 ␮L of 1000 mg/L NH4 Cl elution in reasonable analysis time and baseline resolution for ana-
were added to reach a final concentration of 100 mg/L NH4 Cl as lytes belonging to different classes. The elution of chlorite, chlorate
well as and 50 ␮L of internal standard solution (0.4 mg/L) to reach and bromate is usually accomplished with isocratic runs on high
a final concentration of 4 ␮g/L. capacity carbonate selective columns, such as IonPac AS9−HC, and
Water samples were withdrawn from the treatment train of more recently IonPac AS23 [25] which ensure baseline resolution
the water plant and filtered in Millex Gv filters (0.22-␮m, Milli- of oxyhalides even at high matrix ion content. However, carbon-
pore). Water was sampled into 100 mL glass flasks containing 10 mg ate selective columns are not recommended for gradient elution,
NH4 Cl and immediately analysed. since baseline drift is too severe, hence hydroxide selective columns
are the election choice. Hydroxide selective column such as IonPac
3. Results and discussion AS19 have shown improved sensitivity and allows the detection of
chlorite, chlorate and bromate at lower concentrations in respect
3.1. Optimization of MS/MS conditions to the carbonate selective IonPac AS23 column [26].
On the other hand, hydroxide selective columns of even high
Starting key MS/MS conditions were set as follows. Ion source capacity are best suited for HAAs monitoring in drinking waters
polarity was in the negative ion mode, spray voltage: 3200 V, vapor- where common ions can be present in concentrations as high as
izer gas pressure (N2 ): 45 units, auxiliary gas pressure (N2 ): 10 250 mg/L Cl− and SO4 2- [20].
units, capillary temperature: 200 ◦ C, vaporizer temperature: 200 ◦ C, At the light of the above considerations, for the simultaneous
collision gas (Ar) pressure: 1.5 mTorr, ion cycle time: 0.5 s. To max- elution of the fifteen DBPs, the column chosen was the IonPac AS24,
imise the peak response for the analytes, capillary and vaporizer which is as yet the best hydroxide selective high-capacity column
temperatures were further optimized in the range 200–230 ◦ C (cap- available in the market for the elution of nine Cl-, Br- HAA con-
illary T) and 200–260 ◦ C (vaporizer T) by the injection of analyte geners. The separation for all the fifteen DBPs in the presence of Cl− ,
mixtures at 5 ␮g/L. Best conditions were achieved with capillary SO4 2- , NO3 − and CO3 2- ions must be preliminarily checked with
temperature of 220 ◦ C and vaporizer temperature of 250 ◦ C. Further conductivity detection (see below). A good separation of matrix
increase of these values lead to decreased peak signals especially ions from analytes of interest is important to reduce matrix effects
for HAAs due to analyte degradation [21]. and to preserve the ESI source, through eluate diversion to the
RF lens settings and collision energies (CE) for each transi- waste. In fact, it has been shown that in the absence of matrix diver-
tion were specifically optimized for each analyte, by infusion of sion, recoveries for species eluting close to Cl− ion can be reduced
500 ␮g/Lof each HAA and isotopically enriched internal standard to 77 ± 10% in finished drinking waters [23].
(Table 1). According to literature data [22,23], [M−H]− , resulting Elution conditions. Gradient profile proposed by column manu-
from deprotonation of molecular ion, is the predominant precur- facturer (Eluent #1, Table 2) was initially tested in drinking water
sor ion for haloacetic acids containing one or two halogen atoms, distributed in Turin, Italy (15 mg/L Cl− , 20 mg/L NO3 − , 35 mg/L
whereas [M−COOH]− precursor is preferred for haloacetic acids SO4 2- , 250 mg/L HCO3 − ), spiking 5 ␮g/L of each analyte. Although
containing three halogen atoms. Dimer ions can even be formed the fifteen analytes could be separated from matrix interferent,
increasing infusion concentration (>1 ␮g/L) [24]. In this work, each diversion to waste could not avoid the enhancement of chlorate
precursor ion was selected based on literature information on the signal by carbonate ion and the suppression of DIAA signal by
most abundant species formed in ESI detection [10,24]. In detail, for sulfate ion. This suppression can be avoided changing the selectiv-
HAAs, the selected precursor ion is the one deriving from deproto- ity coefficient DIAA/sulfate ion. Taking advantages of the fact that
nation ([M−H]− ) of molecular ion for MCAA, MIAA, DCAA, MBAA, changes in counter-ion eluent concentration (OH− ) have greater
BCAA, DBAA, CIAA, DIAA and TCAA of the acid, whereas for BDCAA, effects on divalent ions rather than on monovalent ions, as pre-
CDBAA and TBAA, the precursor ion selected is the one resulting dicted by the ion-exchange mechanisms [27], the instantaneous
from decarboxylation ([M−COOH]− ) of the acid. For TCAA, even if eluent change to 60 mM KOH was anticipated just after the elution
many authors suggest the selection of [M−COOH]− as the precur- of DCAA (Eluent #2, Table 2), keeping constant the slope of gra-
sor ion [22,23], it is not infrequent the selection of the [M−H]− dient after the first 15 min of elution. As expected, the increase of
4 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350

Table 1
Optimised MS transitions for each compound of this study.

Analyte Assigned internal standard Precursor ion (m/z) Product ion (m/z) RF lens (V) CE (V)

MCAA MBAA-13 C 93.113 35.444 55.18 10.253

Chlorite MBAA-13 C 67.262 51.286 65.49 13.64
MBAA-13 C – 137.848 79.058 50.629 10.253
MBAA MBAA-13 C 136.991 79.04 53.663 10.253
MIAA-D3 186.862 126.946 53.36 13.89
MIAA MIAA-D3 184.878 126.889 51.236 10.253
Bromate MBAA-13 C 126.9 110.929 131.933 22.792
DCAA-13 C – 128 84.04 66.101 10.253
DCAA DCAA-13 C 127.052 83.04 73.382 10.253
BCAA DCAA-13 C 172.87 128.889 61.551 10.253
CIAA MIAA-D3 218.862 126.911 64.28 21.78
DBAA DCAA-13 C 216.83 172.778 64.888 10.253
Chlorate DCAA-13 C 83.162 67.125 95.83 20.01
DIAA MIAA-D3 310.725 266.679 70.65 10.25
TCAA TCAA-13 C 160.839 116.946 43.652 10.253
TCAA-13 C – 161.909 117.946 40.92 10.25
DCBAA TCAA-13 C 162.839 81.071 57 10.253
DBCAA TCAA-13 C 207.052 79.04 70.652 11.77
TBAA TCAA-13 C 252.726 81.071 83.393 19.809

Table 2 Table 3
Eluent gradient optimization for the separation and detection of DBPs in drinking Limits of detection (LOD) and quantitation (LOQ) for the fifteen DBPs.
water matrix.
Analyte Regression equation R2 LOD (␮g/L) LOQ (␮g/L)
Analyte Eluent
Chlorite 0.0646x+0.0073 0.9999 0.036 0.110
#1a #2b MCAA 0.0461x+0.011 0.9999 0.134 0.405
tr (min) MBAA 0.3863x+0.0038 0.9999 0.026 0.078
MIAA 0.5391x+0.0059 0.9999 0.045 0.136
MCAA 12.9 12.7 Bromate 0.3860x+0.0628 0.9998 0.042 0.127
ClO2 − 13.1 13.0 DCAA 0.3451x+0.0109 0.9999 0.059 0.177
MBAA 14.7 14.3 BCAA 0.2764x+0.0008 0.9999 0.037 0.111
MIAA 15.4 15.0 DBAA 0.5612x+0.0005 0.9999 0.055 0.166
BrO3 − 15.7 15.1 CIAA 0.0745x+0.0047 0.9999 0.085 0.256
Cl− 20.3 20.2 Chlorate 0.0334x+0.0095 0.9999 0.188 0.569
DCAA 24.9 24 DIAA 1.1879x+0.0861 0.9999 0.036 0.109
BCAA 26.6 25.5 TCAA 0.3032x+0.1699 0.9999 0.113 0.342
CIAA 28.6 27.3 DCBAA 0.0136x+0.0052 0.9999 0.099 0.301
DBAA 28.9 27.5 DBCAA 0.0108x+0.0048 0.9998 0.108 0.326
CO3 2− 32.7 28.5 TBAA 0.0167x+0.0051 0.9995 0.037 0.111
ClO3 − 33.0 29.1
SO4 2− 35.3 28.5
DIAA 35.9 29.8
NO3 − 37.1 31.9
(i) low ionic strength, (ii) the absence of inorganic non-volatile
TCAA 39.5 32.8 salts and (iii) the presence of the analyte as an ion in solution [28].
DCBAA 41.6 37.8 Chemical suppression is a necessary step to meet these conditions;
DBCAA 44.6 40.7 the efficiency of eluent suppression affects the sensitivity of the
TBAA 48.8 43.5
MS detection, since excessive background conductivity causes MS
a
Eluent #1: 7 mM KOH: t = 0–15 min; 7–18 mM KOH: t = 15.1–30.8 min; 60 mM signal suppression. The suppressor current value was optimized
KOH: t = 31 min, keep until 46 min; 70 mM KOH; t = 47–58 min. Diversion valve to through the injection of HAA mixture and the evaluation of limits of
the waste:19–24 min,35.1–35.6 min,37.4–38.2 min.
b
Eluent #2: 7 mM KOH: t = 0–15 min; 7–15 mM KOH: t = 15.1–23.8 min; 60 mM
detection (LODs) and quantitation (LOQs) according to Shrivastava
KOH: t = 23.9 min, keep until 46 min; 7 mM KOH; t = 47–58 min. Diversion valve to and Gupta [29]. The current range explored was varied between
the waste: 18–23 min, 28.3–28.8 min.30.5–32.3 min. 45 mA and 70 mA, which corresponds to the recommended range
for current setting at the higher KOH concentration reached in the
gradient. Data obtained show that the lowest quantitation limits
eluent strength shifted the divalent SO4 2- ion more than the mono- can be achieved setting the suppressor current at 50 mA; higher
valent DIAA, moving SO4 2- ion close to carbonate ion which could current values enhance the background noise. The best improve-
be both diverted to waste (Table 2). Therefore, the following time ments of quantitation limits were observed for DBAA, DCBAA e
intervals for eluate diversion to waste were set: 18–23 min (Cl− ), DBCAA and in a less extent for TBAA. At this current value, total
28.3–28.8 min (CO3 2- , SO4 2- ), 30.5–32.3 min (NO3 − ) which allow us conductivity within the imposed gradient conditions varies from
to detect all the fifteen DBPs. The optimized diverter times elimi- 0.8 to 3.0 ␮S.
nate the suppression effect on chlorate due to carbonate ion, which
in drinking water samples was about 35%. 3.4. Figures of merit of the method
The optimized separation of the fifteen DBPs is shown in Fig. 1.
Total analysis time is 60 min and includes the re-equilibration of Linearity, limits of detection and quantitation. Linearity was eval-
the column to the starting gradient conditions. uated over two orders of magnitude, correcting peak response of
each analyte with the relative response factor of the internal stan-
3.3. Optimization of suppressor current dard, as assigned in Table 1. Table 3 collects the results obtained,
as well as the LOD and LOQ values [29].
Factors known to favour ionization process at atmospheric pres- A comparison of LOD values with EPA 557 method is not pos-
sure, besides organic solvents such as methanol or acetonitrile, are: sible for all the analytes, since this study also includes oxyhalide
 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350 5

Fig. 1. IC-MS/MS separation of fifteen DBPs and isotopically enriched internal standards (2 ␮g/L each).

DBPs (chlorite, chlorate) and emerging iodoacetic (monoiodo-, ies presented by Lifongo et al. [31]. The limits of detections
chloroiodo- and diiodo-acetic) acids not included in the above- obtained within this work at controlled autosampler and elu-
mentioned standard. However, the optimization carried out tion temperature conditions (Table 3) were compared with those
allowed to get improved (from 2 to 3 times) detection limits for obtained by Wu et al. [23], who eluted HAAs at alkaline condi-
MCAA, MBAA, BCAA and TBAA, but higher (from 2 to 3.5 times) tions, thermostatting the column at 45 ◦ C, without any control
for DBAA, BDCAA, DBCAA and bromate. Comparable LODs were of injection temperature. In this regard, the limits presented
obtained for DCAA and TCAA. [23] for some analytes seem surprisingly low (MCAA: 0.041 ␮g/L,
As regards iodoacetic acids, when comparisons are possible, bromate: 0.0051 ␮g/L, TCAA: 0.03 ␮g/L) in consideration of the
detection limits are improved in respect to the IC-ICP/MS approach above-mentioned discussion on compound stabilities and of the
[30], and comparable with those recently obtained by IC-tandem limits obtained in this work and current literature [10].
mass spectrometry methods [11]. The better detection limits Accuracy and precision. Recovery (R) for all analytes were deter-
obtained for MIAA (2.3 ng/L) and DIAA (0.9 ng/L) by Li et al. [14] mined at five concentration levels spiking known concentrations
should be ascribed to the ten-fold higher amount of sample injected from 0.25 to 20 ␮g/L for each analyte in ultrapure water in the
in respect to this work. presence of 100 mg/L NH4 Cl. Each concentration level was analysed
As regards oxyhalides (chlorite, chlorate and bromate) our LODs with 24 repetitions for each DBP and 57 repetitions for internal
are more than 20 times better than conductivity detection in standards. The following equation was used [10]:
hydroxide selective columns [26] and comparable for chlorite and
bromate to those shown by the few studies based on IC-MS for oxy- (A − B)
R = 100 ·
halides [18]. The slightly better LOD obtained for chlorate in respect C
to this work (0.045 vs 0.188 ␮g/L) is explained with the pretreat-
where A = measured concentration in the fortified sam-
ment of drinking water samples with OnGuard cartridges for matrix
ple; B = measured concentration in the unfortified sample;
removal, which is effective also for carbonate ions.
C = fortification concentration.
Effect of refrigeration. Current literature dealing with HAAs
According to the data obtained (Table 4), recovery is within
determination underline the possibility of degradation of MBAA,
±50% of the true value for 0.25 ␮g/L (which corresponds to the low-
DBCAA and TBAA with temperature at high pH value, thus rec-
est calibration level of the calibration curve) and within ±30% of the
ommending the injection of samples at refrigerated conditions
true value for the other levels, thus fulfilling the requirement set
and elution at sub-ambient temperature. Differently from what
by EPA [10].
expected, refrigeration was found also beneficial for the enhance-
Precision ranged from 1.3% (MIAA) to 12% (MCAA) for the lowest
ment of signal intensity for TCAA (+73%) > DCAA (+62%) > BCAA
calibration level and from 1% (MIAA) to 5.4% (chlorate) for the high-
(+58%) > DBAA (+41%) > bromate (40%). The easier degradation
est calibration level. These data fully satisfy precision requirements
of tri-substituted haloacids agrees with the degradation stud-
set by EPA according to which seven replicates in the midrange of
6 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350

Table 4
Mean percentage recovery and relative standard deviation (n = 24) at different concentration levels for the fifteen DBPs. Inter-day (4 days, 57 replicates) and intra-day
precision (15 replicates in one day) for 4 ␮g/L internal standards is also shown.

Analyte Recovery % (RSD%, n = 24) Inter-day precision Intra-day precision

0.25 0.5 1 10 20
␮g/L

Chlorite 71.3 ± 8.0 (11) 74.5 ± 8.4 (11) 105 ± 4.3 (4.1) 103 ± 2.4 (2.4) 94.5 ± 1.4 (1.5)
MCAA 88.1 ± 10.6 (12) 99.7 ± 4.6 (4.6) 107 ± 3.2 (3.0) 106 ± 1.7 (1.6) 95.0 ± 2.0 (2.0)
MBAA-13 C 4.2 2.0-3.7
MBAA 88.7 ± 5.6 (6.3) 96.8 ± 3.6 (3.7) 104 ± 3 (2.9) 101 ± 1.2 (1.2) 95.5 ± 1.4 (1.5)
MIAA-D3 4.1 1.6-6.1
MIAA 105 ± 1.4 (1.3) 99.7 ± 1.7 (1.7) 101 ± 1.2 (1.2) 104 ± 0.7 (0.7) 104 ± 1.0 (1.0)
Bromate 85.6 ± 5.2 (6.1) 95.6 ± 3.6 (3.7) 102 ± 2.1 (2.1) 99.1 ± 1.4 (1.4) 92.3 ± 1.3 (1.4)
DCAA-13 C 1.9 0.9-2.1
DCAA 105 ± 2.0 (1.9) 97.7 ± 4.7 (4.8) 96.1 ± 1.2 (1.3) 99.2 ± 1.1 (1.1) 98.8 ± 1.1 (1.1)
BCAA 102 ± 4.7 (4.6) 96.3 ± 2.2 (2.3) 96.1 ± 1.4 (1.4) 99.3 ± 1.1 (1.1) 99.8 ± 1.2 (1.2)
DBAA 109 ± 2.9 (2.7) 99.4 ± 1.5 (1.5) 95.9 ± 1.2 (1.2) 99.6 ± 0.9 (1.1) 99.5 ± 1.2 (1.2)
CIAA 134 ± 4.5 (3.4) 112 ± 34 (30) 103 ± 1.5 (1.5) 102 ± 1.4 (1.4) 103 ± 1.3 (1.2)
Chlorate 88.41 ± 5 (5.7) 85.4 ± 3.0 (3.5) 104 ± 3.0 (2.9) 99.6 ± 2.6 (2.6) 94.6 ± 2.3 (2.4)
DIAA 105 ± 1.7 (1.6) 111 ± 5.0 (4.6) 97.6 ± 1.0 (1.0) 100 ± 2.2 (2.2) 99.0 ± 1.6 (1.6)
DCAA-13 C 2.2 1.4-29
TCAA 120 ± 6.7 (5.6) 106 ± 6.2 (5.9) 108 ± 14 (13) 103 ± 1.9 (1.9) 100 ± 2.1 (2.1)
DCBAA 105 ± 37 (36) 102 ± 8.2 (8.0) 99.0 ± 8.1 (8.1) 108 ± 2.7 (2.5) 102 ± 3.4 (3.3)
DBCAA 106 ± 25 (24) 90.7 ± 17 (19) 111 ± 13 (12) 106 ± 9.1 (8.6) 99.9 ± 4.9 (4.9)
TBAA 106 ± 10 (9.4) 103 ± 8.8 (8.5) 105 ± 8.2 (7.8) 99.6 ± 19 (19) 96.8 ± 3.2 (3.3)

calibration curve should be ≤20% [10]. Since a unique method for

the determination of the disinfection by-products considered in
this work is not available in literature, comparisons are possible
only for classes of analytes determined with different analytical
approach. For the nine Cl- and Br- congeners, mean recoveries
for the same fortification levels are comparable or even improved
(DCAA, DCBAA) in respect to other IC-MS/MS methods [10]. For
iodinated DBPs, better mean recoveries were obtained within this
work for MCAA in respect to the ones obtained by reversed-phase
LC–MS/MS with large volume injection [14]. It should be remarked
that the above-mentioned methods were tested for limited num-
bers of replicates (n = 4–15) in respect to our study.
Inter-day, evaluated in 4 different days by 57 replicates, and
intra-day precision, evaluated within the same day by 15 replicates,
was studied using internal standards at 4 ␮g/L concentration. The
satisfactory data obtained (Table 4) indicate the robustness of the
method developed.

3.5. Application to drinking water supply chain

Before applying the developed method to the analysis of drink- Fig. 2. Effect of chlorite concentration on the suppression of MCAA signal. Continu-
ing water samples, the robustness of the method was checked ous line: spiked MCAA concentration (5 ␮g/L); dotted lines: ±30% requisite (3.5 and
evaluating the recovery of analytes in samples withdrawn from 6.5 ␮g/L).

different points of the treatment train, characterized by matrix

composition at different complexity. NH4 Cl, were fortified with 5 ␮g/L HAAs and bromate and analysed.
Three drinking water plants (DW1, DW2, DW3) were consid- The data obtained (Table 5) clearly show that all HAAs (except
ered and analysed. The first two, DW1 and DW2, are conventional MCAA) satisfy the ±30% requisite of the EPA regulation. MCAA is at
treatment plants, including dynamic separation basins (DSB) for the lower limit of acceptability of the above-mentioned requisite
the removal of slurry from clarified waters, in which coagulant, in DSB1-2 and in E1.
hypochlorite and chlorine dioxide solutions are dosed. The third, This behaviour is explained by the suppression effect of chlorite
DW3, is an advanced treatment plant, dosing ozone as oxidant and ion, which in DSB1-2 and in E1 samples is present in dispropor-
performing biological treatment and extended activated carbon fil- tionate concentrations (about 350 ␮g/L, respectively) in respect to
tration; samples were taken at the outlet of a clarification basin MCAA (5 ␮g/L).
(CB3) in which coagulant and hypochlorite solutions are added. To this purpose, the effect of chlorite on MCAA signal suppres-
Effluents from DW plants (E1, E2, E3), which represent distributed sion in drinking water samples is reported in Fig. 2, where the
waters, were also analysed. continuous line represents the spiked MCAA concentration (5 ␮g/L)
Due to the unbalanced amounts of HAAs and bromate in respect and the two dotted lines represent the ±30% requisite (3.5 and
to chlorite and chlorate ions (which derive from reagent conver- 6.5 ␮g/L).
sion), recoveries of analytes were determined in DSB1, DSB2, E1, E2 Data show that the limit set for chlorite (250 ␮g/L) by the
(for DW1 and DW2 plants), and in CB3 and E3 (for DW3 plant) for revision of the Drinking Water Directive 98/83/EC allows the deter-
HAAs and bromate. The five samples withdrawn from each treat- mination of MCAA with the required accuracy. Concentrations of
ment stage (DSB1, DSB2, E1, E2, CB3 and E3), added with 100 mg/L chlorite as high as 1 mg/L still allow the quantitation of MCAA with
 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350 7

Table 5
Recovery of 5 ␮g/L HAAs (including emergent compounds) and 5 ␮g/L bromate spiked on five water samples withdrawn from intermediate and final purification stages of
three potabilization plants.

Intermediate treatments Finished waters

DSB1 DSB2 CB3 E1 E2 E3

MCAA 69 ± 5.9 (8.5) 69.1 ± 5.9 (8.5) 113 ± 3.6 (2.3) 62.2 ± 5.7 (9.2) 79.6 ± 4.7 (6.0) 94.8 ± 6.1 (6.5)
MBAA 101 ± 2.1 (2.1) 100 ± 3.7 (3.7) 104 ± 2.9 (2.8) 103 ± 2.6 (2.5) 103 ± 2.6 (2.5) 100 ± 1.3 (1.3)
MIAA 100 ± 1.0 (1.0) 99.7 ± 1.1 (1.2) 98.9 ± 0.9 (1.0) 100 ± 0.9 (0.9) 99.3 ± 1.2 (1.2) 99.2 ± 1.3 (1.3)
Bromate 111 ± 3.7 (3.3) 113 ± 3.4 (3.0) 115 ± 2.9 (2.5) 116 ± 2.4 (2.1) 116 ± 5.1 (4.4) 115 ± 3.8 (3.3)
DCAA 96.2 ± 8.1 (8.4) 99.5 ± 5.0 (5.0) 101 ± 2.4 (2.4) 102 ± 1.0 (1.0) 101 ± 1.9 (1.9) 102 ± 1.0 (1.0)
BCAA 96.0 ± 6.0 (6.2) 98.6 ± 6.4 (6.5) 100 ± 2.5 (2.5) 100 ± 3.8 (3.8) 101 ± 2.6 (2.6) 101 ± 1.2 (1.1)
DBAA 97.7 ± 3.9 (3.9) 97.7 ± 2.6 (2.7) 99.3 ± 2.7 (2.7) 97.8 ± 1.6 (1.6) 98.9 ± 2.0 (2.1) 99.7 ± 1.0 (1.0)
CIAA 98.7 ± 1.9 (2.0) 99.5 ± 2.9 (3.0) 99.6 ± 0.9 (0.9) 98.9 ± 2.2 (2.2) 98.0 ± 2.2 (2.3) 98.4 ± 2.7 (2.7)
DIAA 97.9 ± 1.6 (1.6) 97.4 ± 3.7 (3.8) 97.1 ± 1.0 (1.1) 97.9 ± 2.4 (2.5) 96.4 ± 2.6 (2.7) 95.5 ± 2.4 (2.5)
TCAA 91.4 ± 9.1 (9.9) 93.2 ± 9.5 (10.2) 94.2 ± 1.5 (1.6) 91.8 ± 2.2 (2.4) 93.5 ± 3.5 (3.7) 95.9 ± 1.3 (1.3)
DCBAA 107 ± 12 (11) 92.4 ± 12 (13) 103 ± 6.9 (6.7) 105 ± 1.3 (1.3) 104 ± 4.1 (3.9) 105 ± 4.7 (4.4)
DBCAA 93.5 ± 8.9 (9.6) 91.5 ± 17 (18) 89.2 ± 25 (29) 108 ± 23 (21) 111 ± 21(19) 110 ± 25 (23)
TBAA 81.6 ± 18 (22) 85.2 ± 15 (17) 95.9 ± 9.5 (9.9) 87.2 ± 9.8 (11) 93.7 ± 14 (15) 104 ± 4.3 (4.1)

Mean chlorite concentration, mg/L (n = 5): DSB1: 395; DSB2: 340; E1: 350; E2: 230, E3: 85.
Mean chlorate concentration, mg/L (n = 5): DSB1: 420; DSB2: 440; CB3: 80; E1: 470; E2: 450, E3: 165.

Table 6
Concentrations (expressed in ␮g/L) of the fifteen DBPs along the treatment train of three potabilization plants evaluated by the method developed.

Day Treatment stage Analyte (␮g/l)

ClO2 − MCAA MBAA MIAA BrO3 − DCAA BCAA CIAA ClO3 − DBAA DIAA TCAA DCBAA DBCAA TBAA Suma

Raw river water 0.49 nd nd <LOQ nd nd <LOQ nd <LOQ nd <LOQ <LOQ nd nd nd –

1 DSB1 550 nd 0.24 <LOQ 0.91 2.56 1.79 nd 350 0.87 nd 1.53 0.86 nd 2.25 10.1
1 E1 297 nd nd <LOQ 0.52 0.26 0.33 nd 580 <LOQ nd 0.72 0.41 nd 0.88 1.3
1 CB3 – nd <LOQ <LOQ 1.04 0.97 0.78 nd 89 0.37 nd 0.38 nd 2.69 0.89 6.1
1 E3 100 nd nd <LOQ 0.45 <LOQ 0.11 nd 211 <LOQ nd Nd nd nd nd 0.1
2 DSB1 408 nd 0.16 <LOQ 0.55 1.64 1.20 nd 380 0.55 nd 0.85 0.44 nd 1.42 6.3
2 E1 313 nd nd <LOQ 0.49 0.34 0.33 nd 446 <LOQ nd 0.56 nd nd 1.28 2.5
3 DSB1 330 nd <LOQ <LOQ 0.41 1.47 1.16 nd 451 0.62 nd 0.41 0.19 nd 1.69 5.5
3 E1 418 nd nd <LOQ 0.49 0.23 0.29 nd 431 <LOQ nd 0.89 nd nd 1.12 2.5
3 DSB2 340 nd <LOQ <LOQ 0.65 1.84 1.34 nd 395 0.47 nd 1.29 0.83 nd 2.08 7.8
3 E2 267 nd nd <LOQ 0.59 0.42 0.35 nd 367 <LOQ nd 0.66 nd nd 1.48 2.9
3 CB3 – <LOQ <LOQ <LOQ 0.84 0.94 0.67 nd 57 0.40 nd 0.5 nd nd nd 2.5
3 E3 113 nd nd <LOQ 0.43 <LOQ 0.12 nd 136 Nd nd Nd nd 1.07 nd 1.2
4 DSB1 248 nd <LOQ <LOQ 0.19 1.73 1.25 nd 459 0.53 nd 0.75 0.38 nd 1.88 6.5
4 E1 369 nd nd <LOQ 0.52 0.30 0.30 nd 762 <LOQ nd 0.62 0.64 nd 1.53 6.1
4 DSB2 279 nd <LOQ <LOQ 0.59 1.78 1.30 nd 508 0.56 nd 0.84 0.75 nd 2.29 7.5
4 E2 231 nd <LOQ <LOQ 0.63 0.50 0.11 nd 645 <LOQ nd 0.42 <LOQ nd 1.15 2.2
4 CB3 – <LOQ <LOQ <LOQ 1.27 1.02 0.81 nd 133 0.44 nd 0.35 nd nd 0.53 3.1
4 E3 90 nd nd <LOQ 0.51 <LOQ nd nd 147 Nd nd Nd nd nd 0.26 0.3
4 DSB1 344 nd <LOQ <LOQ 0.16 1.59 1.11 nd 488 0.52 nd 0.58 0.39 nd 0.96 5.2
4 E1 336 nd nd <LOQ 0.14 0.36 <LOQ nd 560 <LOQ nd 0.52 0.29 nd <LOQ 1.2
4 DSB2 328 nd <LOQ <LOQ 0.14 1.62 1.17 nd 408 0.52 nd 0.68 <LOQ nd 1.36 5.3
4 E2 215 nd <LOQ <LOQ 0.17 0.53 0.40 nd 408 <LOQ nd 0.33 <LOQ nd 0.94 2.2
4 CB3 – <LOQ <LOQ <LOQ 1.03 0.81 0.78 nd 88 0.40 nd <LOQ nd nd nd 2.0
4 E3 30 nd nd <LOQ 0.46 <LOQ nd nd 132 <LOQ nd Nd nd nd nd –

nd: not detected.

a
Sum of the nine HAAs as foreseen by the proposal for the revision of the Drinking Water Directive.

standard addition method (20% recovery for MCAA). It is worth Table 7

Analysis of drinking waters of different origins by the method developed. A,B: houses
mentioning that the effect of chlorite on MCAA detection is not
(Turin, Italy) ; C: house (Monte Carlo, Principality of Monaco); D: drinking fountain
investigated in current literature [14,23], since only Cl− , NO3 − , (Imperia, Italy). Concentrations are expressed in ␮g/L.
SO4 2- , HCO3 − are considered in the matrix. Moreover, current EPA
Analyte A B C D
method [10] does not allow the determination of MCAA in waters
containing chlorite. Chlorite nd nd nd 173
The method developed was hence used to check the drinking MCAA nd nd nd nd
MBAA <LOQ <LOQ <LOQ nd
water supply chain in the main stages of treatment for each DW
MIAA nd nd <LOQ <LOQ
plant on a daily basis (Table 6), as well as in domestic tap water Bromate nd nd <LOQ nd
samples of different provenience (Table 7). DCAA nd <LOQ nd nd
As far as the plant is concerned, the presence of DBPs in BCAA <LOQ 0.29 0.28 nd
CIAA nd nd nd nd
DSB1-2 and CB3 is coherent with the addition of the hypochlo-
DBAA <LOQ <LOQ 0.72 nd
rite solution. This intermediate disinfection stage is of low impact Chlorate 13.0 15.3 nd 7.00
in HAAs formation, since the sum of the compounds subjected DIAA nd nd nd nd
to regulation is well below the limit established for finished TCAA <LOQ <LOQ nd <LOQ
waters (80 ␮g/L). The subsequent filtration stages are efficient DCBAA nd nd <LOQ nd
DBCAA nd nd nd nd
in the reduction of HAA9 since these compounds are present in
TBAA nd nd nd nd
the distributed waters at concentrations below 6 ␮g/L. The fre-
8 M.C. Bruzzoniti et al. / J. Chromatogr. A 1605 (2019) 360350

quency of occurrence of haloacetic DBPs roughly followed the order [7] D.W. Hawker, J.L. Cumming, A. Watkinson, M.E. Bartkow, The occurrence of
DCAA > TBAA > BCAA > TCAA»DBAA»DCBAA»MBAA. Emerging iodi- the herbicide dalapon (2,2-dichloropropionate) in potable water as a
disinfection by-product, J. Environ. Monitor. 13 (2011) 252–256.
nated compounds were not detected. [8] US Environmental Protection Agency, Method 552.2: Determination of
Regarding the domestic tap water samples, two of them were Haloacetic Acids and Dalapon in Drinking Water by Liquid–liquid Extraction,
withdrawn from houses served by the plant here studied (samples Derivatization and Gas Chromatography With Electron Capture Detection,
Environmental Monitoring and System Laboratory, Cincinnati, OH, 1995.
A,B, Turin, Italy), one from a house located in Monte Carlo (sample [9] Y. Xie, Analyzing haloacetic acids using gas Chromatography/Mass
C, Principality of Monaco) and one from a drinking fountain of the spectrometry, Water Res. 35 (2001) 1599–1602.
province of Imperia (sample D, Italy). Samples C and D were chosen [10] US Environmental Protection Agency, Method 557: Determination of
Haloacetic Acids, Bromate, and Dalapon in Drinking Water by Ion
since their sampling areas correspond to municipalities located in
Chromatography Electrospray Ionization Tandem Mass Spectrometry
coastal zones and hence vulnerable to the presence of brominated (IC-ESI-MS/MS), EPA Document no. 815-B-09-012, 2009.
and iodinated compounds. [11] R. Xue, A. Donovan, H. Shi, J. Yang, B. Hua, E. Inniss, T. Eichholz, Rapid
simultaneous analysis of 17 haloacetic acids and related halogenated water
The results on tap waters sampled in houses located in the plant
contaminants by high-performance ion chromatography-tandem mass
area considered confirms the absence of any criticality. Waters spectrometry, Anal. Bioanal. Chem. 408 (2016) 6613–6622.
sampled from the coastal area are not affected by the presence [12] Z. Zhong, G. Li, Y. Shao, B. Zhu, Z. Liu, J. Deng, J. Mo, Amino-functionalized
of iodinated HAAs, even if a signal below the quantitation limit graphene oxide/neutral alumina nanocomposite based solid-phase extraction
coupled with ion chromatography-mass spectrometry for the determination
could be ascribed to MIAA. In one case, the presence of brominated of trace haloacetic acids in drinking water, Anal. Methods-UK 9 (2017)
species (BCAA, DBAA) at very low concentration levels (sum 1 ␮g/L) 2425–2432.
was revealed. [13] S.D. Richardson, F. Fasano, J.J. Ellington, F.G. Crumley, K.M. Buettner, J.J. Evans,
B.C. Blount, L.K. Silva, T.J. Waite, G.W. Luther, A.B. McKague, R.J. Miltner, E.D.
Wagner, M.J. Plewa, Occurrence and mammalian cell toxicity of iodinated
disinfection byproducts in drinking water, Environ. Sci. Technol. 42 (2008)
4. Conclusions
8330–8338.
[14] Y. Li, J.S. Whitaker, C.L. McCarty, Analysis of iodinated haloacetic acids in
This paper reports the first chromatographic method to fulfil drinking water by reversed-phase liquid chromatography/electrospray
ionization/tandem mass spectrometry with large volume direct aqueous
the upcoming revision of the Drinking Water 98/83/EC Directive,
injection, J. Chromatogr. A 1245 (2012) 75–82.
allowing the simultaneous determination of the nine HAAs and the [15] U.S. Environmental Protection Agency, Method 300.1: Determination of
three oxyhalides ions listed in the regulation. The method already Inorganic Anions in Drinking Water by Ion Chromatography, Revision 1.0,
includes three additional emerging iodinated acids (not yet con- Cincinnati, OH, 1997.
[16] U.S. Environmental Protection Agency, Method 314.0 Determination of
sidered by the revision). The method, validated directly in waters Perchlorate in Drinking Water Using Ion Chromatography. Revision 1.0,
withdrawn from strategic points of the potabilization plant, is a Cincinnati, OH, 1999.
powerful tool for water suppliers which are asked to put in place [17] H.P. Wagner, B.V. Pepich, D.P. Hautman, D.J. Munch, US Environmental
Protection Agency Method 326.0, a new method for monitoring inorganic
operational, supply-specific monitoring programmes intended to oxyhalides and optimization of the postcolumn derivatization for the selective
confirm the effectiveness of all control measures in abstraction, determination of trace levels of bromate, J. Chromatogr. A 956 (2002) 93–101.
treatment, distribution and storage. [18] S.A. Snyder, B.J. Vanderford, D.J. Rexing, Trace analysis of bromate, chlorate,
iodate, and perchlorate in natural and bottled waters, Environ. Sci. Technol.
39 (2005) 4586–4593.
Declaration of Competing Interest [19] Y. Hsieh, Potential of HILIC-MS in quantitative bioanalysis of drugs and drug
metabolites, J. Sep. Sci. 31 (2008) 1481–1491.
[20] R. Slingsby, C. Saini, C. Pohl, The determination of haloacetic acids in real
The authors declare that they have no known competing finan- world samples using IC-ESI-MS-MS, J. Chromatogr. Sci. 47 (2009) 523–528.
cial interests or personal relationships that could have appeared to [21] C. Yang, S. Henday, Application Note 454. Analysis of Haloacetic Acids in
Drinking Water by IC-MS/MS, AN62963 E 01/09S, Thermo Fisher Scientific,
influence the work reported in this paper. Sunnyvale, CA, USA, 2009.
[22] R. Loos, D. Barceló, Determination of haloacetic acids in aqueous
environments by solid-phase extraction followed by ion-pair liquid
Acknowledgements chromatography–electrospray ionization mass spectrometric detection, J.
Chromatogr. A 938 (2001) 45–55.
[23] S. Wu, T. Anumol, J. Gandhi, S.A. Snyder, Analysis of haloacetic acids, bromate,
Financial support from Ministero dell’Istruzione, dell’Università
and dalapon in natural waters by ion chromatography–tandem mass
e della Ricerca (MIUR ex 60%) is gratefully acknowledged. M. Cas- spectrometry, J. Chromatogr. A 1487 (2017) 100–107.
tiglioni is grateful to SMAT for the scholarship granted. The authors [24] Q. Luo, D. Wang, Z. Wei, Z. Wang, Optimized chromatographic conditions for
are grateful to Mr. Maurizio Politi (SMAT, Turin) for his assistance separation of halogenated acetic acids by ultra-performance liquid
chromatography–electrospray ionization-mass spectrometry, J. Chromatogr.
in laboratory activities. A 1277 (2013) 26–34.
[25] Thermo Scientific, Product Manual for Dionex IonPac AS23-4␮m, Document
n. 065711-01, June 2016, 2016.
References [26] B. De Borba, J. Rohrer, Application Note 184, Determination of Trace
Concentrations of Chlorite, Bromate, and Chlorate in Bottled Natural Mineral
[1] European Commission, Council Directive 98/83/EC of 3 November 1998 on Waters, AN70407-EN 08/16S, Thermo Fisher Scientific, 2008, Sunnyvale, CA,
the Quality of Water Intended for Human Consumption, available at, 1998 USA.
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:31998L0083. [27] M.C. Bruzzoniti, E. Mentasti, C.A. Pohl, J.M. Riviello, C. Sarzanini, Effect of
[2] European Commission, Proposal for a Directive of the European Parliament ion-exchange site and eluent modifiers on the anion-exchange of carboxylic
and of the Council on the Quality of Water Intended for Human Consumption, acids, J. Chromatogr. A 925 (2001) 99–108.
2018 https://eur-lex.europa.eu/legal-content/EN/TXT/ [28] J.J. Conboy, J. Henion, M.W. Martin, J.A. Zweigenbaum, Ion
?qid=1519210589057&uri=CELEX:52017PC0753. Chromatography/Mass spectrometry for the determination of organic
[3] S.D. Richardson, M.J. Plewa, E.D. Wagner, R. Schoeny, D.M. DeMarini, ammonium and sulfate compounds, Anal. Chem. 62 (1990) 800–807.
Occurrence, genotoxicity, and carcinogenicity of regulated and emerging [29] A. Shrivastava, V. Gupta, Methods for the determination of limit of detection
disinfection by-products in drinking water: a review and roadmap for and limit of quantitation of the analytical methods, Chron. Young Sci. 2 (2011)
research, Mutat. Res.-Rev. Mutat. 636 (2007) 178–242. 21–25.
[4] Y. Pan, W. Li, H. An, H. Cui, Y. Wang, Formation and occurrence of new polar [30] H. Shi, C. Adams, Rapid IC–ICP/MS method for simultaneous analysis of
iodinated disinfection byproducts in drinking water, Chemosphere 144 iodoacetic acids, bromoacetic acids, bromate, and other related halogenated
(2016) 2312–2320. compounds in water, Talanta 79 (2009) 523–527.
[5] J.A. Pals, J.K. Ang, E.D. Wagner, M.J. Plewa, Biological mechanism for the [31] L.L. Lifongo, D.J. Bowden, P. Brimblecombe, Photodegradation of haloacetic
toxicity of haloacetic acid drinking water disinfection byproducts, Environ. acids in water, Chemosphere 55 (2004) 467–476.
Sci. Technol. 45 (2011) 5791–5797.
[6] H.S. Weinberg, C.A. Delcomyn, V. Unnam, Bromate in chlorinated drinking
waters: occurrence and implications for future regulation, Environ. Sci.
Technol. 37 (2003) 3104–3110.