Anal Bioanal Chem DOI 10.1007/s00216-016-9820-5

PAPER IN FOREFRONT

Contaminant screening of wastewater with HPLC-IM-qTOF-MS and LC+LC-IM-qTOF-MS using a CCS database Susanne Stephan 1 & Joerg Hippler 1 & Timo Köhler 1 & Ahmad A. Deeb 2 & Torsten C. Schmidt 2,3 & Oliver J. Schmitz 1

Received: 6 June 2016 / Revised: 11 July 2016 / Accepted: 20 July 2016 # Springer-Verlag Berlin Heidelberg 2016

Abstract Non-target analysis has become an important tool in the field of water analysis since a broad variety of pollutants from different sources are released to the water cycle. For identification of compounds in such complex samples, liquid chromatography coupled to high resolution mass spectrometry are often used. The introduction of ion mobility spectrometry provides an additional separation dimension and allows determining collision cross sections (CCS) of the analytes as a further physicochemical constant supporting the identification. A CCS database with more than 500 standard substances including drug-like compounds and pesticides was used for CCS data base search in this work. A non-target analysis of a wastewater sample was initially performed with high performance liquid chromatography (HPLC) coupled to an ion mobility-quadrupole-time of flight mass spectrometer (IM-qTOF-MS). A database search including exact mass (±5 ppm) and CCS (±1 %) delivered 22 different compounds. Furthermore, the same sample was analyzed with a twodimensional LC method, called LC+LC, developed in our group for the coupling to IM-qTOF-MS. This four dimensional separation platform revealed 53 different compounds, identified over exact mass and CCS, in the examined wastewater Electronic supplementary material The online version of this article (doi:10.1007/s00216-016-9820-5) contains supplementary material, which is available to authorized users. * Oliver J. Schmitz [email protected] 1

Applied Analytical Chemistry, University of Duisburg-Essen, Universitaetsstr. 5, 45141 Essen, Germany

2

Instrumental Analytical Chemistry, University of Duisburg-Essen, Universitaetsstr. 5, 45141 Essen, Germany

3

Centre for Water and Environmental Research (ZWU), University of Duisburg-Essen, Universitaetsstr. 2, 45141 Essen, Germany

sample. It is demonstrated that the CCS database can also help to distinguish between isobaric structures exemplified for cyclophosphamide and ifosfamide. Keywords LC+LC . Waste water . Non-target . Ion mobility . IM-qTOF-MS . 2D-LC

Introduction Anthropogenic compounds such as pharmaceuticals, pesticides, or personal-care products are released into the environment and these substances are found in the water cycle. The complexity of different aqueous matrices and the occurrence of thousands of different contaminants require analytical methods for a non-target screening and identification of many compounds in a single run [1]. Liquid chromatography coupled to mass spectrometry (LC-MS) allows the detection of drug residues in ultratrace levels without prior derivatization compared with GC-MS [2]. Hence, LC-MS-based methods are widespread for the analysis of pharmaceuticals and other pollutants as well as their metabolites and transformation products in wastewater [3–5] and environmental water such as rivers [6]. High resolution mass spectrometer (HRMS) like time-of-flight (TOF) and Orbitrap MS combine sensitive full-spectrum data with high mass resolution and high mass accuracy, resulting in high-quality data. Different analytical approaches using LC-HRMS methods can be (1) target analysis, which aims to detect and quantify specific substances, (2) suspect screening for the identification of known compounds from a suspect target list, or (3) complete non-target screening [7]. In non-target screening approaches, no information about the compounds are available before analysis. A huge

S. Stephan et al.

number of unspecified signals are recorded and an ion list is generated. Via the exact mass, possible molecular formulas can be calculated and databases have to be searched for possible structures. For a reliable identification, further MS/MS experiments or other hyphenated methods are required. Another possibility to solve this problem is ion mobility spectrometry (IMS) coupled to mass spectrometry (MS). An increasing number of research papers have been published since the 1990s [8, 9]. In 2006, Waters introduced the Synapt HDMS system, a quadrupole/ traveling-wave IMS/TOF (TWIMS) system, which was the first commercially available IM-MS [10]. A system using drift time IMS (DTIMS) was released in 2014 by Agilent (Agilent 6560 IM-qTOF) [11]. In IMS, ions are separated according to their size/shape-to-charge ratio and their collision cross sections (CCS or Ω), which are specific values for certain substances, can be determined [12]. The calculation of CCS (Ω) using TWIMS, where a non-uniform electric field is used, can only be realized after an analyte-dependent calibration [13], but in DTIMS proportionality of CCS (Ω) and drift times of the ions is given by the Mason-Schamp equation (Eq. 1) [14, 15]: rffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 3ze 2π 1 1 t d E 760 torr T Ω¼ ⋅ ⋅ þ ⋅ ⋅ ⋅ 16 N 0 k b T mB mA L p 273; 2 K ð1Þ Herein, z is the ion charge state, e is the elementary charge, N0 is the buffer gas density at standard temperature, kb is the Boltzmann constant, T is the temperature in the drift tube, mB and mA are the masses of the drift gas molecule and the analyte ion, respectively, L is the length of the drift tube, td is the drift time, E is the electric field strength, and p is the pressure in the drift tube. The CCS value of an ion can be calculated directly from its drift time observed in DTIMS according to Eq. 1. CCS values for different substance classes like peptides [15–18], N-Glycans [13], a few drug-like compounds [12], metabolites [19], lipids [20], or different biomolecules [21] are given in the literature. The application of HPLC coupled to IM-MS for the nontarget analysis of water samples can be a promising tool since ion mobility offers an additional separation dimension nested between LC and MS without increasing the analysis time, as it was already shown for different real samples [22, 23]. The combination of a two-dimensional LC system combined with IM-MS leads to a separation in four dimensions, resulting in a very high separation power [24]. Beside an increased peak capacity in HPLC-IM-MS or even more in 2D-LC-IM-MS, also co-eluting isobaric compounds with the same m/z may be separated according to their shape (CCS). Additionally, in qualitative analysis of unknown samples, the direct measurement of CCS values offers the possibility to identify substances

by the combination of their exact mass and their specific size, as long as enough CCS database values are available. Such a database was used in this work with standards from different compound classes, including many pharmaceuticals and other environmentally relevant substances. The applicability of this database for non-target analysis of a real wastewater sample is shown using a standard HPLC-IM-qTOF method as well as a 2D-LC method, developed for the coupling to IM-MS, allowing the evaluation of complex data in four dimensions [24]. Since the method works as a continuous multi-heart cutting and does not meet all requirements of a comprehensive two-dimensional liquid chromatography (LCxLC) [25], we call it LC+LC [24].

Material and methods Chemicals All solvents and mobile phases were of LC-MS grade. Methanol and acetonitrile were purchased from VWR (Leuven, Belgium) and formic acid was from Merck (Darmstadt, Germany). Ultrapure water was generated with a water purification system from Sartorius Stedim (Goettingen, Germany). Standards for the establishment of a CCS database were collected from different working groups. For mass calibration of the MS as well as for direct CCS calibration, a low concentration tuning mix (G1969-85000; Agilent Technologies, Santa Clara, CA, USA) was used. Sampling and sample preparation The examined wastewater sample was collected as a 24-h composite in February 2015 at the municipal wastewater treatment plant (WWTP) located in Duisburg-Vierlinden (Germany). 200 mg/6 mL Oasis MAX and Oasis MCX cartridges from Waters (Eschborn, Germany) were used for solid phase extraction. Detailed information about development and validation of the SPE method have recently been published [26].The two cartridges were conditioned and equilibrated with 2 × 3 mL methanol and 2 × 3 mL water and connected together in a tandem mode in which Oasis MAX was the cartridge connected directly to the sample reservoir, while Oasis MCX was the subsequent one. 1 L of the wastewater sample was filtered with a bottle-top vacuum filtration unit through a glass microfiber filter (GF/F, 0.7 μm average pore size, 47 mm diameter) and left without pH adjustment. The extraction was carried out on a vacuum manifold via large volume adapters by applying a vacuum suction (maximum of 65 kPa) at a flow rate of ~15 mL/min. After the extraction, the cartridges were dried under vacuum for 30 min. Washing and elution steps were carried out for each cartridge

Contaminant screening of wastewater

Fig. 1 Interface for the LC+LC system with an additional pump for dilution and split of the first dimension column effluent. DAD: diode array detector; IM-qTOF: ion mobility quadrupole time-of-flight mass spectrometer

individually. Oasis MAX was washed with 2 mL waterammonia solution (95:5, v/v) mixture and eluted with 6 mL methanol-ethyl acetate-formic acid (69:29:2, v/v/v) mixture. The Oasis MCX washing and elution solvents were 2 mL water-formic acid (98:2, v/v) and 6 mL methanol-ethyl acetate-ammonia solution (67.5:27.5:5, v/v/v) mixtures, respectively. The gathered eluates from both cartridges were mixed together and the reduced solvent was evaporated under vacuum. The final solution was filled up to 1 mL with water. Instrumentation An Agilent 1290 Infinity two-dimensional liquid chromatography system (Agilent Technologies, Waldbronn, Germany) was used, consisting of a 1290 Infinity binary pump (G4220A) using a Jet Weaver V35 mixer for each the first and the second dimension, a 1290 Infinity Flexible cube solvent management module (G4227A), a 1290 Infinity HiP sampler (G4226A), and a 1290 Infinity Thermostatted Table 1

IM-qTOF-MS instrument parameters

Parameter

CCS of standards

LC- and LC+ LC-IM-qTOF

ESI Gas temperature [°C] Gas flow [L/min] Nebulizer pressure [psig]

Positive 200 °C 5 L/min 20 psig

Positive 200 °C 5 L/min 20 psig

Sheath gas temperature [°C] Sheath gas flow [L/min] Trap fill time [μs] Trap release time [μs] Drift voltage [V]

275 °C 8 L/min 40,000 μs 150 μs 1000–1700 (in 100 V steps) Nitrogen 60 ms 40–1700

325 °C 12 L/min 30,000 μs 150 μs 1700

Drift gas Max. drift time [ms] Mass range (m/z)

Nitrogen 50 ms 50–1700

Column compartment (G1316C) with a 2D-LC-Quick Change Valve (2 Pos/4 Port duo valve, G4236A). Measurements of the wastewater by LC-IM-qTOF-MS were carried out on a Kinetex (Phenomenex, Aschaffenburg, Germany) C18 column (100 × 3.0 mm, 2.6 μm particle size) with water containing 0.1 % formic acid (solvent A) and methanol containing 0.1 % formic acid (solvent B) as mobile phase, using a flow of 500 μL/min. A linear gradient was applied starting with 5 % B held for 3 min, increasing to 90 % B in 20 min and holding this for 5 min, and 20 μL of the prepared wastewater sample was injected. For 2D-LC measurements (LC+LC), the interface between first and second dimension was set up with a dilution and flow splitting system using an additional 1290 Infinity binary pump. For the two-dimensional setup (Fig. 1), pump head B of the additional binary pump is used for diluting the effluent (100 μL/min) of the first dimension column with 0.1 % formic acid in water (300 μL/min) to avoid high amounts of organic solvent being injected to the second dimension, resulting in broad peaks and bad resolution [27]. The flow was directed into a Jet Weaver V35 mixer to get a proper mixing of the eluate and the water. After this, a T-piece and Pump Head A of the additional binary pump are used for an active flow splitting, resulting in a total flow of 20 μL/min transferred to the loops. With a sampling time of 4 min and a loop volume of 100 μL, the loops are filled up to 80 % with the diluted fractions coming from the first dimension column before injection of the complete loop volume onto the second dimension. For the second dimension, a third binary pump was used. For more details about LC+LC, see the recently published work from Stephan et al. [24]. In the first dimension, a 150 × 2.0 mm Luna CN column packed with 3.0 μm particles (Phenomenex, Aschaffenburg, Germany) was used with water containing 0.1 % formic acid and acetonitrile as mobile phase (gradient from 5 to 80 % ACN). As second dimension column a 50 × 3.0 mm Kinetex C18 with 2.6 μm core-shell particles (Phenomenex) was used. The mobile phase consisted of water and methanol, both

S. Stephan et al.

Fig. 2 HPLC-IM-qTOF-MS measurement of a wastewater sample. Heatmap showing IM drift time [ms] versus LC retention time [min], intensities by colors (101–106 counts) (a); drift spectrum (red), mass spectrum (blue), and resulting 2D plot extracted from 10.72–10.75 min (b)

containing 0.1 % formic acid. In the second dimension with a flow of 500 μL, a shifted gradient [28] was used, starting with a gradient from 10 to 60 % methanol in the first sampling cycle and reaching a gradient from 50 to 90 % methanol in the last sampling cycle. The injection volume of the prepared sample to the first dimension was 20 μL. IM-qTOF measurements were performed with an Agilent 6560 IM-qTOF System, equipped with a dual Agilent Jet Stream electrospray ionization (AJS ESI) source. Nitrogen was used as buffer gas in the drift tube at a pressure of about 3.95 Torr. For mass calibration of the MS as well as for direct CCS calibration, a low concentration tuning mix (G196985000; Agilent Technologies, Santa Clara, USA) was used. The instrument parameters for the CCS measurements of standards and for the LC- or LC+LC-IM-qTOF-MS measurements are listed in Table 1.

measured and CCS values were calculated with the following method. Standard solutions of 10 mg/L in methanol/water (50/50) were injected eight times to the IM-qTOF instrument using the HiP sampler and the binary pump of the HPLC for an in-flow injection. For each of the eight injections, different drift tube entrance voltages between 1000 and 1700 V (in 100 V steps) were applied. With a constant drift tube exit voltage of 250 V and a drift tube length of 78 cm, the resulting electric fields were between 9.615 and 18.590 V/cm. Corrected drift times td were calculated by subtracting the flight time through the interfacing IM-MS ion optics from the total observed drift times [21]. The calculation of the CCS according to MasonSchamp [14] was done by the Agilent IM-MS Browser B.07.01 software. If available, different adduct ions in positive and negative mode were considered. CCS measurement in real samples and Database search

CCS measurement of standards To collect collision cross section values of different compounds for establishing a database, more than 500 standards were

Data evaluation of LC- and LC+LC-IM-qTOF measurements of the wastewater sample was carried out with the Agilent IMMS Browser B.07.01 software. In this software version,

Contaminant screening of wastewater Table 2 Compounds found by 4D feature extraction after HPLC-IM-qTOF measurement that could be identified according to their exact mass (Δ m/z ± 5 ppm) and their CCS (Δ CCS ± 1 %) with an in-house database. Shown exact mass and CCS are the database values Adduct

CCS [Å2]

Δ CCS [%]

Substance

Group

Chemspider ID or CAS

0.0

H+

263.1885 231.1008 263.1885 290.1379 265.1579 263.1885 267.1834 278.0837

0.8 –3.4 0.4 2.1 3.0 –0.4 –2.2 1.8

H+ H+ H+ H+ H+ H+ H+ H+

133.6 161.1 149.9 161.1 174.0 159.7 161.1 172.2 163.5

0.0 0.4 –0.2 0.5 –0.3 –0.1 –0.2 –0.1 0.2

Fenuron Tramadol 4–Formylaminoantipyrine Tramadol Trimethoprim Mirtazapine Tramadol Metoprolol Sulfadimidine

Herbicide Analgesic Metabolite Analgesic Antibiotic Antidepressant Analgesic Beta-blocker Antibiotic

7279 5322 65525 5322 5376 4060 5322 4027 5136

12.70

748.5085

–2.7

H+

12.73 13.10 13.86 15.57 15.66 16.14 16.31 16.54 17.87 18.12 18.56

325.2253 277.2042 424.1799 292.0979 236.0950 198.1732 241.1361 191.1310 836.5246 440.1597 270.1620 270.1620 295.0167 290.0822

3.1 –0.4 2.1 0.7 –0.4 1.0 1.2 –4.7 –1.0 –2.3 0.4 0.4 2.4 0.7

H+ H+ H+ H+ Na+ H+ H+ H+ Na+ H+ H+ H+ H+ H+

263.4 187.8 171.4 202.7 173.2 161.9 148.9 159.4 143.2 278.3 196.3 165.5 166.3 159.1 175.2

–0.5 –0.5 –0.1 –0.3 0.3 –0.2 0.1 –0.3 –0.5 –0.7 –0.3 –0.5 0.0 –0.2 0.3

Azithromycin Bisoprolol Venlafaxine Clindamycin Climbazole Carbamazepine Cycluron Terbutryn DEET (Diethyltoluamide) Roxithromycin Candesartan Estron Trenbolon Diclofenac Chloroxuron

Antibiotic Beta-blocker Antidepressant Antibiotic Antifungal Analgesic Herbicide Herbicide Insect repellent Antibiotic Cardiovascular Steroid Anabolic agent Anti-inflammatory Herbicide

10482163 2312 5454 393915 34752 2457 15694 12874 4133 5291557 2445 5660 23383 2925 15299

RT [min]

Exact Mass [Da]

2.81

164.0950

6.82 8.24 8.37 8.67 9.86 10.25 10.62 11.21

20.26 20.49

Δ m/z [ppm]

features can be found in a chromatographic run coupled to IM-qTOF-MS with information for each feature about retention time, drift time, m/z, and abundance (4D feature extraction). When a direct CCS calibration with a known tuning mix (G1969-85000; Agilent Technologies, Santa Clara, CA, USA) is performed on the same day, CCS values can be calculated directly from the measured drift time for each feature. This calibration was derived directly from Mason-Schamp [29]. 4D feature extraction was performed for the measurements of a wastewater sample and the complete list of features was exported as a *.csv file. CCS values of the standards were entered into a database using an in-house programmed software tool (CCS-Database by AAC vers. 1.2.0.22). Besides the CCS of different adduct ions—measured with the method described above—the database contains information for each substance about its name, IUPAC name, CAS no., molecular formula, and exact mass. With our software tool, the database can be searched for substances by their exact mass and their CCS for different adducts like [M + H]+ or [M + Na]+ in a certain range (Δ m/z [ppm] and Δ CCS [%]). For database search of the wastewater

sample, the *.csv file generated before was imported to the CCS-Database software. The complete database was then searched for substances by the m/z and CCS values of each feature.

Results and discussion CCS Database More than 500 standards of different compound classes like amino acids, lipids, and sugars as well as drug-like compounds, pesticides, or other environmentally relevant substances, including about 350 exogenous compounds, were measured with ion mobility qTOF-MS for the establishment of a database containing CCS. Solutions of single standards or in some cases mixtures of several standards were injected eight times to the IM-qTOF system and each time, a different drift tube entrance voltage was applied as described above. Previous studies in our group (results not shown) revealed that more reliable results are obtained when a wide range of drift

S. Stephan et al.

Fig. 3 LC+LC-IM-qTOF-MS measurement of a wastewater sample. Heatmap showing IM drift time [ms] versus retention time [min], intensities by colors (101–106 counts) (a); heatmap of one modulation –

fraction collected in the loop between 16 and 20 min – injected to the second dimension and separated there between 20 and 24 min (b)

voltages is covered. The CCS was then directly calculated by the Mason-Schamp equation from the observed drift time for each injection. The average of these eight injections was added as CCS value to our in-house database. Relative standard deviations were always below 0.5 %. For quality control of the system, colchicine was used as a known reference standard. Ten measurements of a colchicine standard on different days resulted in an average CCS for the [M + H]+ adduct of Ω = 196.8 Å2 ± 0.27 Å2 (RSD 0.14 %), which shows a very good interday repeatability. The deviation from a value published in the literature [12] represents 0.41 %, which is also quite small (Ω(colchicine) = 196.2 Å2). Every time CCS values are determined in our lab, colchicine is measured again and the calculated CCS is checked to be in a range of ±0.5 % of the value ascertained above.

acquisition time) reveals a lot of spots in this complex sample and points out that compounds with the same retention time can be separated in the ion mobility dimension by their collision cross section. At each retention time a 2D plot of drift time versus m/z can be extracted (Fig. 2b). Via 4D feature extraction a list of features, which are characterized by their retention time, drift time, m/z, and abundance, was generated. CCS values of all features were calculated directly from the according drift times. The list of extracted features from the wastewater sample, containing information about CCS and m/z of detected ions, was subsequently screened for entries in the CCS database. Table 2 summarizes hits for 22 different compounds according to the database values with their CCS ± 1 % and their exact mass ±5 ppm. In this case, 20 compounds were found as [M + H]+-adducts and two could be identified by their [M + Na]+-adduct. All identified substances can be grouped to pharmaceutical or pesticide structures that are relevant as possible contaminants in wastewater. The identification via the combination of exact mass and specific CCS of a substance does not only

LC-IM-qTOF analysis At first a non-target analysis of a wastewater sample, prepared with SPE, was performed using the LC-IM-QTOF-MS system. The resulting heat map in Fig. 2a (drift time versus

Contaminant screening of wastewater Table 3 Compounds found by 4D feature extraction after LC+LC-IM-qTOF measurement that could be identified according to their exact mass (Δ m/z ± 5 ppm) and their CCS (Δ CCS ± 1%) with an in-house database. Shown exact mass and CCS are the database values RT [min] Exact Mass [Da] Δm/z [ppm] Adduct CCS [Å2] ΔCCS [%] Substance

9.01

272.1195

–1.1

H+

11.53 11.53 13.85

279.1471 279.1471 263.1885 263.1885

1.4 4.3 –1.0 0.0

H+ Na+ Na+ H+

17.33 17.47 17.49 18.44 18.58

249.0572 290.1379 231.1008 236.1161 424.1799 424.1799

–4.4 1.7 –1.7 –2.3 –0.9 2.6

Na+ H+ H+ Na+ Na+ H+

21.47 21.56 21.76 21.98 22.32

245.1164 238.1106 331.1332 278.0837 260.0248 260.0248 260.0248

–0.8 –3.3 –1.2 –0.4 2.5 –1.5 –1.5

H+ H+ H+ H+ Na+ H+ H+

22.45 22.50 25.72 26.19 26.21

260.0248 164.0950 263.1885 748.5085 325.2253 325.2253

3.2 –1.2 0.8 –0.8 –1.4 1.8

Na+ H+ H+ H+ Na+ H+

29.71

265.1579 253.0521 188.0950 188.0950

1.9 –3.9 –1.9 –2.6

H+ H+ Na+ H+

311.1092 133.0640 133.0640 252.0899

2.2 2.2 2.2 –0.8

H+ H+ H+ H+

37.63 38.08 46.74 46.81 54.28 58.58 58.65 61.25 61.84 66.23 66.57 66.59 70.53 74.54

119.0483 277.2042 250.1569 241.1361 295.0627 198.1732 191.1310 267.1834 269.1416 292.0979 206.1419 733.4612 747.4769 836.5246 836.5246

2.5 1.8 –2.2 2.9 –2.7 –1.5 1.6 2.2 –1.1 0.7 0.0 0.7 0.3 –2.6 –1.4

H+ H+ Na+ H+ H+ H+ H+ H+ H+ H+ Na+ H+ H+ H+ Na+

74.75

230.0943

1.6

Na+

29.72 30.21 30.25 37.60

167.5 159.2 173.6 176.8 167.0 164.4 174.0 149.9 165.9 207.3 202.7 155.2 151.4 177.1 163.5 158.7 147.3 147.7 155.2 133.6 161.1 263.4 178.8 187.8 159.7 155.1 154.8 138.8 172.1 125.1 125.6 161.3

0.4 –0.3 0.2 0.2 0.1 0.4 0.1 –0.1 0.8 –0.5 0.0 0.0 –0.1 0.5 0.4 0.1 –0.5 –0.3 0.0 0.1 0.1 0.1 0.1 –0.2 0.0 0.2 –0.1 0.1 0.9 –0.5 –0.1 –0.1

122.1 171.4 166.9 159.4 170.8 148.9 143.2 172.2 168.7 173.2 159.4 257.8 266.1 276.2 278.3 161.3

0.2 0.3 –0.7 –0.1 –1.0 0.2 –0.3 0.5 –0.4 0.8 0.1 0.6 0.2 0.2 –0.1 0.6

Sotalol Metalaxyl Metolachlor carbonic acid Desvenlafaxine Desvenlafaxine Sulfapyridine Trimethoprim 4-Formylaminoantipyrine Carbetamide Clindamycin Clindamycin 4-Acetamidoantipyrin 10,11–Dihydrocarbamazepine Ciprofloxacin Sulfadimidine Ifosfamide Cyclophosphamid Ifosfamide Cyclophosphamid Fenuron Tramadol Azithromycin Bisoprolol Bisoprolol Mirtazapine Sulfamethoxazol Phenazone Phenazone Fenamidone 4-MBTZ 5-MBTZ Carbamazepine-10,11Epoxide BTA Venlafaxine Gemfibrozil Terbutryn N4-Acetyl-Sulfamethoxazole Cycluron DEET Metoprolol Mepronil Climbazole Isoproturon Erythromycin Clarithromycin Roxithromycin Roxithromycin Naproxen

Group

Chemspider ID or CAS

antihypertensive agent fungicide metabolite antidepressant

5063 38839 21170688 111300

antibiotic antibiotic metabolite herbicide antibiotic

5145 5376 65525 25761 393915

metabolite metabolite antibiotic antibiotic antineoplastic agent antineoplastic agent

59166 18028 2662 5136 3562 2804

herbicide analgesic antibiotic beta-blocker

7279 5322 10482163 2312

antidepressant antibiotic analgesic

4060 5138 2121

fungicide anticorrosive anticorrosive metabolite

8578637 29878-31-7 136-85-6 2458

anticorrosive antidepressant antilipemic drug herbicide metabolite herbicide insect repellent beta-blocker fungicide antifungal agent herbicide antibiotic antibiotic antibiotic

6950 5454 3345 12874 58771 15694 4133 4027 37994 34752 33695 12041 10342604 5291557

anti-inflammatory

137720

S. Stephan et al. Table 3 (continued) RT [min] Exact Mass [Da] Δm/z [ppm] Adduct CCS [Å2] ΔCCS [%] Substance

78.15

277.1830

–4.7

H+

78.44 78.60 79.24 81.56 82.54 82.58

422.1622 278.0861 270.1620 270.1620 324.1638 440.1597 403.1168

2.6 –4.7 –2.6 –2.6 –0.6 1.1 1.2

H+ H+ H+ H+ H+ H+ Na+

83.01

435.9387

0.7

Na+

83.12

295.0167 295.0167 376.2000 236.0950

2.0 –2.5 0.3 0.0

H+ Na+ H+ Na+

83.53 93.83

Group

Chemspider ID or CAS

167.6 200.8 163.3 165.5 166.3 181.8 196.3 196.9

0.9 0.8 0.8 –0.5 –0.1 0.7 0.5 0.2

Amitriptyline Losartan Triphenylphosphan-oxid Estron Trenbolon Citalopram Candesartan Azoxystrobin

antidepressant 2075 antihypertensive agent 3824 12549 steroid estrogene 5660 anabolic agent 23383 anti-depressant 2669 cardiovascular drug 2445 fungicide 2298772

195.1 159.1 167.5 187.6 161.9

0.3 0.6 0.6 –0.3 0.4

Fipronil Diclofenac Diclofenac Enalapril Carbamazepine

insecticide anti-inflammatory

give a higher certainty but also enables to differentiate between isobaric structures. For example, tramadol, which has the same molecular formula as desvenlafaxin, is found in the sample, and high resolution mass spectrometry alone would not help to distinguish the two possibilities because the exact mass (263.1885 Da) is the same for both structures. But since they can be separated in ion mobility and, furthermore, since their different CCS values are included in the database, the compound detected in the sample can clearly be assigned to tramadol. Interestingly, tramadol was found as a matching result for three different features at different retention times

3235 2925

antihypertensive agent 4534998 analgesic 2457

(6.82, 8.37, and 10.25 min). Deviations of the database values differ in Δ m/z and ΔCCS between –0.4 and +0.8 ppm, and –0.2 and +0.5 %, respectively. Perhaps the sample contains some substances with structures very similar to tramadol, which are not yet contained in our database. A limitation of this system is revealed in one feature, which was found at a retention time of 18.56 min. Here, two isobaric substances with very similar database CCS values, estron (Ω = 165.5 Å 2 ) and trenbolon (Ω = 166.3 Å 2 ), are suggested. Although the detected CCS is closer to trenbolon (Δ 0.0 %) than to estron (Δ 0.5 %), a certain identification is not possible without

Fig. 4 Extracted ion chromatogram for the [M + Na]+-adduct (m/z = 283.0140) of ifosfamid and cyclophosphamide. For both chromatographic peaks, the extracted 2D plot (drift time versus m/z) and the resulting CCS value are shown

Contaminant screening of wastewater

further experiments (e.g., MS/MS) because both results are within a CCS-range of below 1 % and there is no difference in retention time of both standards. LC+LC-IM-qTOF analysis Recently, we developed a two-dimensional HPLC method for the coupling to IM-qTOF, called LC+LC [24]. To simplify the data evaluation of this complex four dimension instrumental set-up, a long sampling time (4 min) is used, so that peaks eluting from the first dimension column are not modulated several times as it is common in LC×LC. This allows to view the data of a 2D-LC separation on a single time axis similar to a simple 1D chromatogram without the need to generate a contour plot. Hence, a heat map with drift time versus acquisition time above for LC-IM-qTOF can be generated, in which each compound, except a few ones that elute from the first dimension between two modulations, should appear as one spot (Fig. 3a). Here, every four min a fraction from the first dimension is injected for separation on the second dimension column. Figure 3b shows the drift time versus acquisition time on the second dimension during one modulation cycle. The described method reaches a peak capacity of 7828 (for calculation see Electronic Supplementary Material, ESM). After 4D feature extraction and database search, 53 different compounds were found within a range of m/z ±5 ppm and CCS ±1%. Table 3 only lists the hits found in two independent analyses with the described method of the same sample that did not occur in a blank. With the exception of chloroxuron, all substances found in the HPLC-IM-qTOF measurement are also obtained after the LC+LC-IM-qTOF analysis. Because of the separation power of the LC+LC-system and, therefore, lower ion suppression and pure MS spectra, additional 31 compounds could be identified. Unfortunately, a differentiation between the isobars estron and trenbolon is still not possible with LC+LCIM-qTOF-MS (feature with RT = 79.24 min), although the CCS value is closer to the database entry of trenbolon (Δ CCS = 0.1 %). 4-Methylbenzotriazole (4-MBTZ) and 5MBTZ (RT = 30.25 min) are also too similar in structure to be separated in IM under the given conditions as it is known from standard measurements. Other examples like desvenlafaxin and tramadol show that the CCS value helps to distinguish between known isobaric pairs. They are separated after 2D-LC (RT 13.85 and 25.72 min), but in a non-target analysis without standards it would not be possible to decide which peak belongs to which substance because they have the same exact mass. A database search with CCS shows clearly that the first peak belongs to desvenlafaxin (which was not detected in the simple HPLCIM-qTOF run) with ΔCCS = 0.1 % for the [M + H]+ and ΔCCS = 0.2 % for the [M + Na]+-adduct. For the second peak, tramadol is found with ΔCCS = 0.1 % ([M + H]+). Perhaps the dilution with water of the eluate after the first dimension is the reason for only one detected feature for a signal with the m/z and

CCS of tramadol in comparison with the results by HPLC-IMqTOF analysis, which would also explain why chloroxuron has not been detected here. The identification of tramadol in the sample was finally determined by comparison with the retention time of a standard (data not shown). The importance of considering, if possible, different adducts is shown in the following. The [M + H]+-adducts of ifosfamide and cyclophosphamide have the same exact mass and also identical CCS values. But for the sodium-adduct, two features are found which differ in CCS, allowing an identification of these two peaks with slightly different retention times (22.32 and 22.45 min in Table 3). This is also visualized in Fig. 4, where an extracted ion chromatogram of the [M + Na]+-adduct of ifosfamide and cyclophosphamide (m/z = 283.0140) shows two peaks eluting close together. The first of these peaks has a slightly higher drift time, resulting in a calculated CCS value of Ω = 158.6 Å2, than the second peak with Ω = 155.2 Å2. This difference enables to identify the first peak as ifosfamide (database value is Ω([M + Na]+) = 158.7 Å2), whereas the second peak must be cyclophosphamide (database value is Ω([M + Na]+) = 155.2 Å2). A direct comparison of both applied methods is not easy because of the difference in total run time (120 min in LC+LC versus 30 min in LC), but in this work, we chose two different chromatographic methods that could be run as standardized methods to focus on the applicability of the CCS database. An approach of Leonhardt et al. [30] was to compare 1D- and 2D separations by the total number of detected peaks. Here, the higher number of hits after the LC+LC run compared with the LC-IM-qTOF can be explained with a better separation attributable to the additional HPLC dimension and, therefore, less ion suppression and pure MS spectra, so that more features can be found and identified. Thus, the developed LC+LC-IM-qTOF method requires a longer analysis time, but also leads to a higher content of information than a shorter LC-IM-qTOF method.

Conclusions A two-dimensional LC+LC-IM-qTOF analysis leads to a better separation and, therefore, a higher number of identified compounds compared with a shorter standard LC-IM-qTOF method. The determination of the CCS value and the exact mass of the analytes allows also the identification of even isobaric compounds, if the isobars differ in their shape. The size of the database still limits the number of identified features. Thus, more and more standards have to be measured. Acknowledgments The authors thank Agilent Technologies for the third Infinity pump system and Phenomenex for the HPLC-columns. They thank IUTA (Duisburg), LANUV (Duesseldorf), and the working group of Professor G. Scriba (Jena) for providing standards. The research project no. 18861 of the research association Institut für Energie- und Umwelttechnik e. V. (IUTA) has been funded via AiF within the agenda for the promotion of industrial cooperative research and development

S. Stephan et al. (IGF) by the German Federal Ministry of Economics and Technology based on a decision of the German Bundestag.

15.

Compliance with Ethical Standards

16.

Conflict of interest There is no conflict of interest. 17.

References 1.

2.

3.

4.

5.

6.

7.

8. 9.

10.

11. 12.

13.

14.

Hernández F, Ibáñez M, Bade R, Bijlsma L, Sancho JV. Trends in Analytical chemistry investigation of pharmaceuticals and illicit drugs in waters by liquid chromatography-high-resolution mass spectrometry. Trends Anal Chem. 2014;63:140–57. Perez S, Barceló D. Application of advanced MS techniques to analysis and identification of human and microbial metabolites of pharmaceuticals in the aquatic environment. Trends Anal Chem. 2007;26:494–514. Gracia-lor E, Sancho JV, Serrano R, Hernández F. Chemosphere occurrence and removal of pharmaceuticals in wastewater treatment plants at the Spanish Mediterranean area of Valencia. Chemosphere. 2012;87:453–62. Mendoza A, Aceña J, Pérez S, De Alda ML, Barceló D, Gil A. Pharmaceuticals and iodinated contrast media in a hospital wastewater : a case study to analyse their presence and characterise their environmental risk and hazard. Environ Res. 2015;140:225–41. Nurmi J, Pellinen J. Multiresidue method for the analysis of emerging contaminants in wastewater by ultra performance liquid chromatography-time-of-flight mass spectrometry. J Chromatogr A. 2011;1218:6712–9. Loos R, Gawlik BM, Locoro G, Rimaviciute E, Contini S, Bidoglio G. EU-wide survey of polar organic persistent pollutants in European river waters. Environ Pollut. 2015;157:561–8. Krauss M, Singer H, Hollender J. LC-high resolution MS in environmental analysis: from target screening to the identification of unknowns. Anal Bioanal Chem. 2010;397:943–51. Mukhopadhyay R. IMS/MS: its time has come. Anal Chem. 2008;80:7918–20. Lapthorn C, Pullen F, Chowdhry BZ. Ion mobility spectrometrymass spectrometry (IMS-MS) of small molecules: separating and assigning structures to ions. Mass Spectrom Rev. 2013;32:43–71. Pringle SD, Giles K, Wildgoose JL, Williams JP, Slade SE, Thalassinos K, et al. An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/ travelling wave IMS/oa-ToF instrument. Int J Mass Spectrom. 2007;261:1–12. May JC, McLean JA. Ion mobility-mass spectrometry: timedispersive instrumentation. Anal Chem. 2015;87:1422–36. Campuzano I, Bush MF, Robinson CV, Beaumont C, Richardson K, Kim H, et al. Structural characterization of drug-like compounds by ion mobility mass spectrometry: comparison of theoretical and experimentally derived nitrogen collision cross sections. Anal Chem. 2012;84:1026–33. Hofmann J, Struwe WB, Scar CA, Scrivens JH, Harvey DJ, Pagel K. Estimating collision cross sections of negatively charged Nglycans using traveling wave ion mobility-mass spectrometry. Anal Chem. 2014;86:10789–95. Kanu AB, Dwivedi P, Tam M, Matz L, Hill HHJ. Ion mobility-mass spectrometry. J Mass Spectrom. 2008;43:1–22.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Tao L, McLean JR, McLean JA, Russell DH. A collision crosssection database of singly-charged peptide ions. J Am Soc Mass Spectrom. 2007;18:1232–8. Counterman AE, Valentine SJ, Srebalus CA, Henderson SC, Hoaglund CS, Clemmer DE. High-order structure and dissociation of gaseous peptide aggregates that are hidden in mass spectra. J Am Soc Mass Spectrom. 1998;9:743–59. Valentine SJ, Counterman AE, Clemmer DE. A database of 660 peptide ion cross sections: use of intrinsic size parameters for bona fide predictions of cross sections. J Am Soc Mass Spectrom. 1999;10:1188–211. Bush MF, Hall Z, Giles K, Hoyes J, Robinson CV, Ruotolo BT. Collision cross sections of proteins and their complexes: a calibration framework and database for gas-phase structural biology. Anal Chem. 2010;82:9557–65. Paglia G, Williams JP, Menikarachchi LC, Thompson JW, Tyldesley-Worster R, Halldórsson S, et al. Ion mobility-derived collision cross-sections to support metabolomics applications. Anal Chem. 2014;86:3985–93. Paglia G, Angel P, Williams JP, Richardson K, Olivos HJ, Thompson JW, et al. Ion mobility-derived collision cross section as an additional measure for lipid fingerprinting and identification. Anal Chem. 2015;87:1137–44. May JC, Goodwin CR, Lareau NM, Leaptrot KL, Morris CB, Kurulugama RT, et al. Conformational ordering of biomolecules in the gas phase: nitrogen collision cross sections measured on a prototype high resolution drift tube ion mobility-mass spectrometer. Anal Chem. 2014;86:2107–16. Liu X, Valentine SJ, Plasencia MD, Trimpin S, Naylor S, Clemmer DE. Mapping the human plasma proteome by SCX-LC-IMS-MS. J Am Soc Mass Spectrom. 2007;18:1249–64. Malkar A, Devenport NA, Martin HJ, Patel P, Turner MA, Watson P, et al. Metabolic profiling of human saliva before and after induced physiological stress by ultra-high performance liquid chromatography-ion mobility-mass spectrometry. Metabolomics. 2013;9:1192–201. Stephan S, Jakob C, Hippler J, Schmitz OJ. A novel fourdimensional analytical approach for analysis of complex samples. Anal Bioanal Chem. 2016;408:3751–9. Marriott PJ, Wu Z, Schoenmakers P. Nomenclature and conventions in comprehensive multidimensional chromatography—an update. LCGC Eur. 2012;25:266–75. Deeb AA, Schmidt TC. Tandem anion and cation exchange solid phase extraction for the enrichment of micropollutants and their transformation products from ozonation in a wastewater treatment plant. Anal Bioanal Chem. 2016;408:4219–32. Stoll DR, Talus ES, Harmes DC, Zhang K. Evaluation of detection sensitivity in comprehensive two-dimensional liquid chromatography separations of an active pharmaceutical ingredient and its degradants. Anal Bioanal Chem. 2014;407:265–77. Li D, Schmitz OJ. Use of shift gradient in the second dimension to improve the separation space in comprehensive two-dimensional liquid chromatography. Anal Bioanal Chem. 2013;405:6511–7. Kurulugama RT, Darland E, Kuhlmann F, Stafford G, Fjeldsted J. Evaluation of drift gas selection in complex sample analyses using a high performance drift tube ion mobility-QTOF mass spectrometer. Analyst. 2015;140:6834–44. Leonhardt J, Teutenberg T, Tuerk J, Schluesener MP, Ternes TA, Schmidt TC. Analytical methods two-dimensional liquid chromatographic approaches coupled to high resolution mass spectrometry for the analysis of complex samples. Anal Method. 2015;7:7697– 706.

Contaminant screening of wastewater Susanne Stephan is currently a Ph.D. student in the working group of Professor Oliver J. Schmitz for Applied Analytical Chemistry at the University of Duisburg-Essen, Germany. She majored in Food Chemistry at the University of Wuppertal in 2012 with emphasis on analytical chemistry. During her Ph.D. study, she worked on CCS measurements with IM-MS and analysis of complex samples with multidimensional separation methods.

Joerg Hippler is a senior researcher in Applied Analytical Chemistry at the University of Duisburg-Essen. For many years he has worked intensively in the field of analytical chemistry, using modern LC- and GC-MS techniques for the analysis of complex samples. Over the years he has developed a CCS database software and different applications of instrumental analytical methods; most recently with a 2D-LC coupled to an IM- QTOF-MS system.

Timo Köhler is a student research assistant in the working Group of Professor Oliver J. Schmitz for Applied Analytical Chemistry (AAC) at the University of Duisburg-Essen and is currently studying chemistry (M.Sc.) at the University of Duisburg-Essen, Germany. He received his B.Sc. degree in 2015 from the University of DuisburgEssen, with a Bachelor Thesis focused on analytical chemistry. His responsibility in the working group is the measurement of CCS values with IM-MS for a CCS database.

Ahmad A. Deeb is a Ph.D. student in the Department o f Instrumental Analytical Chemistry at the University of Duisburg-Essen, Germany. His research focusses on sample preparation, target and suspect screening of organic micropollutants, and their transformation products in aqueous samples.

Torsten C. Schmidt is Head of the Department of Instrumental Analytical Chemistry and the C e n t e r f o r Wa t e r a n d Environmental Research (ZWU) at the University of DuisburgEssen, and Scientific Director at t h e I W W Wa t e r C e n t r e i n Muelheim an der Ruhr. He is currently President of the German Water Chemistry Society. In 2013, he received the Fresenius Award of the German Chemical Society. His main research interests include the development and application of analytical methods with focus on separation techniques (GC, LC), sample preparation and compound-specific stable isotope analysis, process-oriented environmental chemistry, and oxidation processes in water technology.

Oliver J. Schmitz is a Full Professor at the University of Duisburg-Essen and Chair of the Applied Analytical Chemistry. In 2009 he cofounded the company iGenTraX UG, which develops new ion sources and units to couple separation techniques with mass spectrometers. The research fields are the development of ion sources, use and optimization of comprehensive LC and GC, and coupling analytical techniques with mass spectrometers.