LC/Q-TOF Workflows for

Comprehensive Micropollutant
Analysis
Targeted Quantification, Suspect Screening, and
Unknown Compound Identification

Application Note
Environmental

Authors Abstract
Christoph Moschet and This application note presents three complementary LC/Q-TOF workflows
Thomas M. Young designed to provide comprehensive analysis of micropollutants in surface waters:
Department of Civil and • Targeted quantification
Environmental Engineering,
• Suspect screening with and without MS/MS spectra
University of California,
• Unknown compound identification
Davis, CA, USA
The first two workflows rely on the Agilent All Ions MS/MS accurate mass
Tarun Anumol
capabilities of the Agilent LC/Q-TOF system, Agilent MassHunter Qualitative
Agilent Technologies Inc.
Analysis software, and Agilent Personal Compound Database and Libraries
2850 Centerville Rd,
(PCDLs) to detect and confirm compound identities, with or without reference
Wilmington, DE, USA
standards. Agilent Molecular Structure Correlator (MSC) software is shown to aid
in the identification of unknown compounds, in this case transformation products
(TPs).

The targeted quantification workflow was validated using 32 reference standards.

The workflow detected and quantified 25 compounds in at least one of the
51 surface water samples tested. The suspect screening workflow generated an
expanded list of 85 possible pollutants, of which 73 were subsequently positively
identified with an authentic standard (67 compounds) or by matching MS/MS
spectra (six compounds). The unknown compound identification workflow
identified five TPs that had not been identified using the targeted quantification or
suspect screening workflows.
Introduction For compounds not identified through targeted analysis, the
suspect screening workflow, which uses Agilent All Ions
Comprehensive analysis of micropollutants in waste, surface, MS/MS and Personal Compound Database and Libraries
and drinking waters is necessary to confidently assess (PCDLs), allows accurate mass determination and MS/MS
exposure and risk. Traditional targeted screening workflows, fragment confirmation of compounds. The screening workflow
such as triple quadrupole MS methods, monitor and quantify allows the analyst to presumptively identify compounds
a predefined list of compounds using analytical reference without analytical reference standards with a high degree
standards. However, a targeted approach will miss pollutants of confidence. The unknown compound identification
not on the target list, and may underestimate exposure when workflow attempts to identify compounds for which there
unexpected pollutants are present. are not MS/MS fragment entries in the selected PCDLs, nor
are analytical standards readily available. In this workflow,
While triple quadrupole LC/MS systems are well suited for Agilent Molecular Structure Correlator (MSC) software
targeted analysis, their methods require standards that are compares MS/MS fragments for compounds of interest
not always easily obtained. In addition, they are not useful to plausible candidate structures in additional databases
when trying to identify new or unknown compounds in the such as Chemspider, Pubchem, or custom databases. MSC
sample. High-resolution, accurate-mass (HRAM) quadrupole calculates scores based on the quality of the matches of the
time-of-flight (Q-TOF) LC/MS analysis allows complementary experimental fragments with those predicted.
suspect screening and unknown compound identification
workflows, which, when used together with targeted This application note describes the use of targeted
quantification, provide a more complete picture of the quantification, suspect screening, and unknown compound
chemical profile of the sample analyzed (Figure 1). identification with LC/Q-TOF data acquisition and analysis
workflows to achieve broader, more confident characterization
of micropollutants, including transformation products (TPs),
in surface water samples. The study LC- and GC-QTOF-MS
Extract water samples with validated as Complementary Tools for a Comprehensive Micropollutant
extraction protocol Analysis in Aquatic Systems provides a detailed description of
the analytical results and their implications for environmental
monitoring [1]. Because not all of the compounds studied are
Perform All Ions LC/Q-TOF/MS Data amenable to LC/MS analysis, the research also describes
Acquisition, + and –ESI modes a complementary GC/Q-TOF workflow that provides a
comprehensive chemical profile of the sample.
Quantify target compounds using
Agilent MassHunter Quantitative 25 Targeted Experimental
Analysis, [M+H]+ or [M−H]− quantifier pesticides detected
and two MS/MS qualifiers
Target compounds and standards
For the targeted quantification workflow, 32 LC/MS‑amenable
Screen data for suspects using 85 Compounds detected, pesticides were chosen for analysis of the surface water
Agilent MassHunter Find by Formula 73 identified with
standards or MS/MS samples (Table 1). The target compounds were chosen
and Pesticide and Water Contaminant
PCDLs spectra matches to include compounds used in the area of the waters
sampled, and to represent pesticides of different classes and
physicochemical properties. Seventeen of the targets had
Discover unknowns using predicted better sensitivity in positive ESI mode, while 15 had better
Transformation Product Custom
PCD/PDCL, and Agilent MassHunter
5 TPs identified sensitivity in negative ESI mode. Eleven isotopically-labeled
Find by Formula and MSC internal standards were used for accurate quantification.

Figure 1. Complementary targeted quantification, suspect screening, For method validation and quality control analyses, reference
and unknown compound identification LC/Q-TOF workflows for standards for the target compounds were prespiked before
comprehensive micropollutant analysis. extraction and post-spiked before injection. Procedural blanks
were also used to look for blank contamination.

2
Table 1. Target Pesticides and Targeted Quantification Results

ESI Matrix Method detection Absolute

Target analyte mode factor1 limit (ng/L)2 recovery (%)3 Accuracy (%)4 Precision (%)5
2,4-D – 2.6 2.6 95 % 160 % 18 %
2-Phenylphenol – 1.2 1.2 75 % 170 % 6%
Azoxystrobin + 1.7 0.2 95 % 128 % 1%
Boscalid – 1.6 0.3 97 % 111 % 3%
Chlorantraniprole + 2.7 2.7 95 % 87 % 3%
Clomazone + 2.7 1.3 76 % 190 % 1%
Cyprodinil + 3.1 0.3 91 % 118 % 3%
DEET + 2.2 0.2 76 % 78 % 4%
Difenoconazole + 1.7 0.9 95 % 104 % 1%
Dimethoate + 3.8 0.9 92 % 62 % 5%
Diuron – 1.7 0.2 92 % 102 % 1%
Fipronil – 1.2 0.1 97 % 96 % 6%
Fipronil-desulfinyl – 1.2 0.1 101 % 78 % 4%
Fipronil-sulfide – 1.1 0.1 97 % 74 % 18 %
Fipronil-sulfone – 1.2 0.1 96 % 113 % 1%
Hexazinon + 3.0 0.3 91 % 117 % 3%
Imidacloprid – 4.2 2.1 93 % 152 % 5%
MCPA – 2.9 1.1 96 % 112 % 1%
Methomyl + 2.7 13 93 % 104 % 2%
Methoxyfenozide – 1.3 0.1 99 % 72 % 4%
Metolachlor + 1.7 0.2 80 % 108 % 2%
Novaluron – 1.2 0.6 74 % 91 % 2%
Pendimethalin + 1.8 1.8 71 % 74 % 3%
Propanil – 1.1 1.2 98 % 138 % 4%
Propoxur + 2.6 1.3 76 % 83 % 1%
Pyriproxyfen + 2.3 0.2 89 % 98 % 7%
Simazine + 7.2 1.8 89 % 77 % 3%
Thiacloprid + 4.1 1.0 93 % 97 % 6%
Thiamethoxame + 2.0 1.0 92 % 108 % 3%
Thiobencarb + 1.6 1.6 77 % 99 % 2%
Triclocarban – 1.4 0.1 92 % 97 % 1%
Triclosan – 1.2 1.3 89 % 89 % 2%

Area STD 100 ng/mL

Equation 1. Matrix factor =
Area extract post spiked 100 ng/mL – Area extract unspiked

Instrument detection limit (ng/mL) × Matrix factor (–)

Equation 2. Method detection limit (ng/L) =
Concentration factor (mL/L)

Area prespiked extract – Area unspiked extract

Equation 3. Absolute recovery (%) =
Area post spiked extract – Area unspiked extract

Concentration prespiked extract – Concentration unspiked extract

Equation 4. Accuracy (%) =
100 ng/mL

Standard deviation concentration prespiked extract (triplicate)

Equation 5. Precision (%) =
Average concentration prespiked extract (triplicate)

3
Sample preparation The Q-TOF mass spectrometer was operated in both
Fifty-one 1-L surface water samples collected from the positive and negative ESI modes to maximize compound
Sacramento-San Joaquin River Delta in Northern California detection. Agilent MassHunter Workstation software was
were extracted using a mixed-mode solid phase extraction used to acquire data (version B.07.00). The All Ions MS/MS
(SPE) cartridge. The cartridges were eluted sequentially with acquisition mode with collision energies (CEs) of 0, 10, 20,
6 mL of 50/50 methanol/ethyl acetate with 0.5 % ammonia, and 40 V was used to obtain both precursor and fragment
3 mL of methanol/ethyl acetate with 1.7 % formic acid, ion data for all species. The All Ions MS/MS acquisition
and 2 mL of methanol, per Moschet; et al. [2]. The eluent mode simultaneously collects high- and low-CE scans. The
was evaporated to 0.2 mL and reconstituted to 1 mL with low CE scans allow the user to obtain precursor information,
nanopure water to obtain a 20 %/80 % methanol/water ratio while the higher CE scans provide fragment information that
for injection into the LC/MS system. enhances compound identification and confirmation when
comparing experimental spectra to those in the PCDL. Table 3
lists the Q-TOF mass spectrometer parameters.
LC/Q-TOF analysis
LC/MS analysis of the reference standards and sample To facilitate identification of TPs not identified in the targeted
extracts was performed using an Agilent 1260 Infinity LC quantification or suspect screen workflows, the samples
coupled to an Agilent 6530 Accurate-Mass Quadrupole with the highest abundances of plausible TPs were rerun in
Time‑of-Flight (Q-TOF) LC/MS system equipped with an targeted MS/MS mode with the collision energy set to 20 V.
Agilent Jet Stream dual electrospray ionization (ESI) source.
The HPLC system included a binary pump, Agilent 1260
Autosampler, and an Agilent ZORBAX Eclipse Plus C-18, Table 3. Q-TOF Mass Spectrometer Parameters
2.1 × 100 mm, 1.8 µm column (959758-902). Table 2 lists the Parameter Value
LC parameters. Mass spectrometer Agilent 6530 Accurate Mass Q-TOF-LC/MS with
Agilent Jet Stream Technology
Table 2. LC Parameters. (+): Positive ESI Mode; (-): Negative ESI Mode Ionization mode(s) Positive (+) and negative (–) ESI
Instrument mode 2 GHz extended dynamic range
Parameter Value
Mass range 50–1,050 m/z
Liquid chromatograph Agilent 1260 Infinity Binary LC
Gas temperature 300 °C
Analytical column Agilent ZORBAX Eclipse Plus C-18,
2.1 × 100 mm, 1.8 µm Drying gas flow 12 L/min

Injection volume 20 µL Nebulizer 25 psig

Column temperature 30 °C Sheath gas temperature 350 °C

Mobile phase A) Water + 0.1 % formic acid (+)/ Sheath gas flow 11 L/min
Water + 1 mM ammonium fluoride (–) Capillary voltage 3,500 V (+), 3,000 V (–)
B) Acetonitrile + 0.1 % formic acid (+)/ Fragmentor voltage 110 V
Acetonitrile (–) Scan speed 4.0 spectra/sec
Flow rate 0.35 mL/min Collision energies Agilent All Ions MS/MS: 0, 10, 20, 40 V
Gradient Time (min) %A Targeted MS/MS: 20 V
 0.0 98 Reference ions* (+) 121.0509 and 922.0098
 1.5 98 (–) 112.9855 and 1033.9881
16.5 0
21.5 0 * The positive mass reference ions were not used in some instances
because of interferences in some of the samples. The user can check for
Equilibration time 3.0 minutes
interferences and use alternate reference ions if required.

4
Data processing and analysis The suspect screening workflow used the Agilent Pesticide
Target compounds were quantified using Agilent MassHunter and Water Screening PCDLs, in combination with the
Quantitative Analysis software (version B.07.00). The [M+H]+ MassHunter Qualitative Analysis software (version B.07.00)
or [M−H]− ion within the exact mass window of ±10 ppm Find by Formula algorithm. Find by Formula automatically
was used as the quantifier ion. The two most abundant extracts precursor ions from the All Ions MS/MS data using
unique MS/MS fragments for each compound were selected the accurate-mass database in the PCDLs. When available
from the MS/MS library spectra in the Agilent PCDLs in the accurate mass library, the corresponding MS/MS
to use as qualifiers. The Agilent Pesticide PCDL for TOF fragments are also extracted from the data. Precursor and
or Q-TOF LC/MS systems (1,684 compounds, 914 with corresponding fragment ion peaks are plotted to score the
MS/MS spectra) and the Agilent Water Screening PCDL quality of their correlation for each compound. The suspect
(1,451 compounds, 1,157 with MS/MS spectra) were used. screening workflow data analysis parameters used are
provided in Figure 2.

Library spectrum
110 372.09787
100 PCDL spectra at CE: 20 eV
• Database/Library Search 90
• Agilent Pesticide PCDL 80
70
• Agilent Water Screening PCDL
Counts

Find By 60 344.10297
Formula 50 301.08591
40
30 329.07950
20 216.06552
10 134.06004 172.03931
0
120 160 200 240 280 320 360 400
Mass-to-charge (m/z)

60
12.289
• Mass tolerance: 10 ppm 50
Counts (%)

• Allowed adducts: [M+H]+, [M+Na]+; [M-H]–, [M+F]– 40 Azoxystrobin

Precursor • Allowed charge state: 1 Mass: 403.1159
30 Score: 97.0
identification
20
10
0
12.0 12.2 12.4 12.6 12.8 13.0 13.2 13.4
Acquisition time (min)
×10 5

1.2
• Mass accuracy (weight: 100) 1.1
1.0 404.12303
• Isotope spacing (weight: 60) 0.9 ([C22H17N3O5]+H)+
0.8
Counts

• Relative isotope abundance (weight: 60) 0.7

Precursor ion 0.6
• Minimum overall score: 70 405.12695
assessment 0.5
0.4 ([C22H17N3O5]+H)+
0.3
0.2
0.1
0
404.0 404.4 404.8 405.2 405.6 406.0
Mass-to-charge (m/z)
1.4
1.2
• Minimum number of fragments present: 1
1.0
• Minimum coelution score: 85
Counts (%)

Fragment ion 0.8

confirmation 0.6
0.4
0.2
0
12.2 12.3 12.4 12.5 12.6 12.7
Acquisition time (min)

Figure 2. Suspect screening workflow: data analysis parameter settings and example. The precursor mass of the fungicide azoxystrobin
(403.1168 m/z) is found at retention time 12.3, and its fragment ions are confirmed.

5
The unknown compound identification workflow can be Method detection limits for all target analytes ranged from
used to identify compounds without a priori knowledge 0.1 and 13 ng/L in water. All targets had absolute recoveries
using accurate mass and fragment data along with in silico between 70–110 %. The precision, calculated as the relative
fragment prediction solfware. In this case, it relied on a standard deviation (%RSD) of triplicate injections on the
custom PCD of potential TPs, which contained accurate LC/Q-TOF system, was <10 % for 30 of the target analytes,
masses and molecular formulas. The Eawag Pathway while the accuracy for 26 analytes was between 70 and
Prediction System (EAWAG‑PPS) [3] was used to generate 130 %.
1,409 possible TP structures for the pesticides detected in the
study. After eliminating 71 structures due to the implausibility Targeted screening detected 25 of the target compounds
of being ionized by ESI, the molecular formulas of the among the 51 water samples tested (Figure 1).
plausible structures were added to build the custom PCD. The
MassHunter Qualitative Analysis software Find by Formula Suspect screening
feature was used to screen the data for these plausible The All Ions MS/MS sample data collected in both positive
molecular formulas. Because there were no MS/MS spectra and negative ESI modes were screened for compounds
for these compounds in the custom PCD, the exact mass, present in the Agilent Pesticide and Water Screening PCDLs
isotope score, number of detections across the 51 samples, using the Find by Formula algorithm. Based on specified
and retention time (RT) were used to produce a short list of adducts (Figure 2), the software automatically searched
plausible matches. the acquired data for the presence of the precursor ions of
compounds stored in the PCDL, and assigned a cumulative
Water samples with the highest abundances of the short list score to matches based on mass accuracy, isotopic spacing,
of plausible candidates were rerun in targeted MS/MS mode and relative isotope abundance.
to obtain MS/MS spectra for processing by MSC software
(version, B.07.00). MSC software was used to search the A score threshold of >70 was chosen (Figure 2) for
custom PCD for compounds with the same exact mass as compounds to be considered for further evaluation.
the isolated mass. MSC software automatically compares Compound matches with higher scores were subsequently
predicted in silico fragments of the structures in the custom evaluated for the presence of MS/MS fragment ions
PCD (or in a web-based database such as ChemSpider or matching the compounds’ MS/MS spectra in the PCDL (when
PubChem) with the measured MS/MS spectra. All measured MS/MS spectra were present in the PCDL). Compounds
MS/MS fragments that can be explained by each structure without MS/MS spectra in the PCDL were tentatively
were listed and scored based on a weighted match. It identified.
is important to note that this workflow requires that the
matching structures are present in the custom PCD. The The validity of fragment ion matches was evaluated by
structures can be manually uploaded into the PCD using scoring their coelution with their corresponding precursor
mol‑files. ions (coelution score). Using the spiked analytes, it was
determined that the presence of one fragment with a
To aid in the identification of unknown TPs, CFM-ID coelution score >85 was sufficient to identify compounds
(http://cfmid.wishartlab.com/predict) was used to predict the with a low false positive rate, while producing a manageable
MS/MS spectra of the plausible TPs [4]. amount of data to process and review (Figure 2). Figure 2
also provides an example of the identification of the fungicide
Results and Discussion azoxystrobin using the suspect screening workflow. Figure 3
provides a more detailed comparison of the theoretical (in the
Targeted quantification PCDL) and measured isotope pattern, as well as the coelution
of the main fragments (from the PCDL) of the herbicide
Table 1 shows the LC/Q-TOF positive and negative ESI fluridone.
quantification results for the 32 target pesticide standards.
The pesticide standards were run on the LC/Q-TOF system
to obtain method validation parameters and to determine
method suitability for targeted quantification. The method
validation parameters, including method detection limits,
absolute recovery, accuracy, precision, and matrix factors, are
presented. The parameter details and their calculation have
been described by C. Moschet; et al. [1].

6
 330.11003
×104 ([C19H14F3NO]+H)+
330.1100
310.1038 11.69 9
1.0
309.0960 8
0.9 290.0976 7 Isotope match score: 98
294.0725 6

Counts
0.8 5
4
0.7 3 331.11384
([C19H14F3NO]+H)+
2
0.6
1
Counts (%)

0
0.5 330.0 330.5 331.0 331.5 332.0
Mass-to-charge (m/z)
0.4

0.3

0.2

0.1

11.55 11.6 11.65 11.7 11.75 11.8 11.85 11.9 11.95 12 12.05 12.1 12.15 12.2 12.25 12.3
Acquisition time (min)

Figure 3. Details of the theoretical and measured isotope pattern, along with molecular and fragment ions, identified for
fluridone using All Ions MS/MS and the PCDL.

After manual inspection, the suspect screening workflow Comparison of the abundance pattern of the 33 plausible
detected 85 compounds (53 in positive ESI, 26 in negative TPs with the concentration pattern of their potential parent
ESI, and six in both positive and negative ESI). Of these, 67 compounds in the water samples reduced the list of those
could be confirmed unambiguously by a reference standard, tentatively identified to 14.
six could be confirmed tentatively with high confidence by
matching MS/MS spectra, and 12 compounds were rejected After rerunning the samples in targeted MS/MS mode,
because they were not confirmed by a reference standard evaluating the MS/MS spectra using the MSC software,
or due to implausible MS/MS fragments. The latter was the comparing measured fragments to those predicted by CFM‑ID
case when there were no MS/MS spectra in the PCDL for the and manually inspecting the results, seven compounds
suspected compound. were eliminated for having implausible MS/MS spectra
(that is, having fragments that could not be explained by
Unknown compound identification the molecular structure). Thus, seven plausible TPs were
identified using the workflow. Two of them were already
The MassHunter Qualitative Analysis software Find by detected and confirmed using the targeted quantification
Formula feature was used to identify TPs in the water sample or suspect screening workflows. Therefore, five new TPs
data using plausible TP structures that had been added to could uniquely be identified by the unknown compound
the custom PCD. This step yielded 110 matches. Inspection identification workflow. Three of them were able to be
of peak shape, signal-to-noise ratio (S/N), RT plausibility, and confirmed unambiguously by a reference standard, and two
whether the detected compound is theoretically ionizable in of them remained tentatively identified because no reference
the selected ionization mode left 33 plausible compounds. standard was commercially available.

7
Figure 4 shows an example of a transformation product
of the herbicide dithiopyr (CAS 128294-56-4), which was
tentatively identified with high confidence in several samples.
This compound, along with several MS/MS fragment ion
structures, was identified using MSC (score of 92.4). MSC
was able to identify and elucidate structures for 96.7 % of the
ions in the MS/MS spectra for this TP.

1. Predicted molecular structure of

compound (MSC score: 92.4)

2. All the ions identified in the MS/MS

spectra with predicted structure scores
and potential candidates

3. Possible structures of selected

fragment ion with score and
potential fragment formula.

B F
F O
–
4. C+ H 6. F + H3C
1. H3C S F C
×10 4 3 C
+
CH3 CH3
N
8 F F
N
O 4 2. C+
F OH 236.0134 – F
+C O
7 F C F S
263.0368 F F F
CH3
N
6 CH3 F
F
F O CH 5.
F O F + H3C
5 3
F F S 5 3. F C+ C CH3
Counts

1 CH3 +
N
4
6 N CH2
F
C+ C O
278.0611
3 46.9968 F F F +

306.0573 F F
2 2
1 326.0637
68.9964 210.0436
96.9584 168.0059 370.1863
0
40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440
Mass-to-charge (m/z)

Figure 4A. Agilent MSC software window showing MS/MS fragments of a dithiopyr TP with CAS 128294-56-4 (MSC score 92.6). 1) predicted by MSC software;
2) list of all ions identified with predicted structure scores and candidates; 3) possible structures and formulas of selected fragment ions. B) Mass
spectrum of a dithiopyr TP (§) predicted by MSC software; (*) predicted by CFM-ID.

8
Conclusions References
LC/Q-TOF targeted quantification, suspect screening, and 1. C. Moschet, et al. LC- and GC-QTOF-MS as Complementary
unknown compound identification workflows were applied Tools for a Comprehensive Micropollutant Analysis in
to the analysis of surface water samples. The targeted Aquatic Systems. Environ. Sci. Technol. 51(3), 1553–1561
quantification workflow was validated using 32 pesticide (2017).
standards. Twenty-five of them were detected among the
2. C. Moschet, et al. Alleviating the reference standard
samples tested using the targeted quantification workflow.
dilemma using a systematic exact mass suspect screening
The suspect screening workflow generated an expanded list
approach with liquid chromatography-high resolution mass
of 85 possible pollutants, 73 of which were subsequently
spectrometry. Anal. Chem. 85(21), 10312−20 (2013).
positively identified. When applied to herbicide and pesticide
TPs, the unknown compound identification workflow 3. EAWAG-BBD Pathway Prediction System website; http://
identified five plausible TPs not identified using the targeted eawagbbd.ethz.ch/predict/ (accessed 06/10/2016).
quantification or suspect screening workflows. Compared
4. F. Allen, et al. Competitive fragmentation modeling of
to targeted analysis alone, the complementary workflows
ESI-MS/MS spectra for putative metabolite identification.
enabled by the Agilent 6530 Accurate-Mass Q-TOF LC/MS
Metabolomics 11(1), 98−110 (2015).
system provided more comprehensive, higher confidence
characterization of the micropollutants present, including
potential pesticide and herbicide TPs. Several more
compounds were identified that would have been missed if
only a targeted approach were used.

The Agilent Pesticide and Water-Screening PCDLs, combined

with the Agilent All Ions MS/MS accurate mass capabilities
of the Q-TOF LC/MS system and Agilent MassHunter
Qualitative Analysis software, enabled presumptive
matching of acquired spectra with library spectra to confirm
compound identities, without the need to source standards.
Agilent Molecular Structure Correlator software aided
identification of TPs by correlating the unknown MS/MS
spectrum against multiple candidate structures in compound
databases.

9
For More Information
These data represent typical results. For more information on
our products and services, visit our Web site at
www.agilent.com/chem.

www.agilent.com/chem
Agilent shall not be liable for errors contained herein or for incidental or consequential
damages in connection with the furnishing, performance, or use of this material.

Information, descriptions, and specifications in this publication are subject to change

without notice.

© Agilent Technologies, Inc., 2017

Printed in the USA
October 2, 2017
5991-8459EN