ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Pharmacy 252

Determination of theEnvironmental Fate of DrugSubstances and the Matrix Effectsof Complex Samples in SFC/ESI-MS

ALFRED HAGLIND

ISSN 1651-6192ISBN 978-91-513-0281-2urn:nbn:se:uu:diva-346164

Dissertation presented at Uppsala University to be publicly examined in B42, BMC,Husargatan 3, Uppsala, Wednesday, 9 May 2018 at 09:15 for the degree of Doctor ofPhilosophy (Faculty of Pharmacy). The examination will be conducted in English. Facultyexaminer: Associate Professor Deirdre Cabooter (KU Leuven).

AbstractHaglind, A. 2018. Determination of the Environmental Fate of Drug Substances and theMatrix Effects of Complex Samples in SFC/ESI-MS. Digital Comprehensive Summaries ofUppsala Dissertations from the Faculty of Pharmacy 252. 44 pp. Uppsala: Acta UniversitatisUpsaliensis. ISBN 978-91-513-0281-2.

Awareness of the potential problems caused by drug compounds in the environment hasincreased over the last decade, both among researchers and with the public. This thesisdescribes the development of analytical methods and their application to wetlands constructedfor purification of wastewater from e.g. drug compounds. Different wetlands were investigatedusing microcosm-models, to determine their biodegradation. An enantioselective and sensitiveSFC/ESI-QqQ method was developed and validated for the enantiomeric separation of atenolol,metoprolol, propranolol and metoprolol acid. It was applied measuring the enantiomeric fractionof the compounds in three different microcosm-models. The same microcosms were also usedto investigate the transformation products formed in these wetlands. In this work, LC/ESI-QToF was used to identify the transformation products using standard references, the originalcompounds or analogs, comparing their accurate mass and product ions. One not previouslyobserved major transformation product identified from propranolol were 1-naphthol. Severalminor transformation products were also identified, showing how diverse the formation mightbe in wetlands.

A second part compares the matrix effect of ESI/MS using SFC and reversed phase LC,utilizing general screening methods for drug compounds in plasma, horse urine and influent/effluent wastewater. These matrices are known to suffer from matrix effects when using theESI-source, and if SFC would suffer less than LC it could be a great benefit. The matrix profilesshowed that this is likely not the case: although SFC was affected by different interferences thenLC. One example is the formation of clusters causing major ion suppression. This unique SFC-phenomenon was investigated further, showing that metal ions were separated and eluted atdifferent retention times, forming clusters in the ion source between metal ions and the organicmodifier and/or make-up solvent.

In conclusion, the first part of this thesis describes analytical methods for determinationof drug compounds in the environment, using LC and SFC, connected to both high andlow resolving MS. The second part focuses on fundamental analytical chemistry, comparingthe matrix effects of SFC/ESI-MS with LC/ESI-MS, and investigates the cluster phenomenaobserved for samples containing alkali ions and an organic modifier in the mobile phase in SFC/ESI-MS.

Keywords: SFC-MS, matrix effect, ion cluster, Supercritical fluid chromatography; ESI, chiralseparation, transformation products, LC-QToF, wetland microcosms

Alfred Haglind, Department of Medicinal Chemistry, Analytical Science, Box 574, UppsalaUniversity, SE-751 23 Uppsala, Sweden.

© Alfred Haglind 2018

ISSN 1651-6192ISBN 978-91-513-0281-2urn:nbn:se:uu:diva-346164 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-346164)

Gör gott och var glad

Carl von Linné

Do good and be happy

Carl von Linné

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I SvanⅠ, A., Hedeland, M., Arvidsson, T., Jasper, J.T., Sedlak, D.L. and Pettersson, C.E.* (2015) Rapid chiral separation ofatenolol, metoprolol, propranolol and the zwitterionic metoprolol acid using supercritical fluid chromatography–tandem massspectrometry – Application to wetland microcosms. Journal ofChromatography A, 1409, 251-258.

II SvanⅠ, A., Hedeland, M., Arvidsson, T., Jasper, J.T., Sedlak, D.L. and Pettersson, C.E.* (2016) Identification of transfor-mation products from β-blocking agents formed in wetland mi-crocosms using LC-Q-ToF. Journal of Mass Spectrometry, 51,207-218

III SvanⅠ, A.,* Hedeland, M., Arvidsson, T. and Pettersson, C.E. (2018) The differences in matrix effect between supercritical fluid chromatography and reversed phase liquid chromatography coupled to ESI/MS. Analytica Chimica Acta, 1000, 163-171.

IV Haglind, A.,* Hedeland, M., Arvidsson, T. and Pettersson, C.E. (2018) Major signal suppression from metal ion clusters in SFC/ESI-MS – Cause and Effects. Journal of Chromatography B, 1084, 96-105.

*Corresponding authorⅠBirth name

Reprints were made with permission from the respective publishers.

Contents

Introduction ..................................................................................................... 9 Drug substances in the environment .......................................................... 9 

Occurrence ............................................................................................. 9 Concern ................................................................................................ 10 Determination ...................................................................................... 12 

Supercritical fluid chromatography and the ESI source ........................... 13 Supercritical fluid chromatography – A new revival? ......................... 13 ESI source and the sample matrix ....................................................... 15 

Aims .............................................................................................................. 17 

Part 1. Wetlands ............................................................................................ 18 The wetlands ............................................................................................ 18 Chiral separation ...................................................................................... 20 Identification using HRMS ...................................................................... 22 

Part 2. Sample Matrix ................................................................................... 26 Comparison of LC and SFC ..................................................................... 26 Metal ion clusters in SFC/ESI-MS ........................................................... 30 

Conclusions ................................................................................................... 32 

Summary in Swedish/ Populärvetenskaplig sammanfattning ....................... 34 

Acknowledgements ....................................................................................... 37 

References ..................................................................................................... 39 

Abbreviations

API Active pharmaceutical ingredient CSP ESI EF ELSD GC HRMS LC ME MS PCI PCIMP PEA RPLC SFC SPE SRM UHPLC

Chiral stationary phase Electrospray ionization Enantiomeric fraction Evaporative light scattering detector Gas chromatography High resolution mass spectrometry Liquid chromatography Matrix Effect Mass spectrometry Post-column infusion Post-column infusion matrix profile Post-extraction addition Reversed phase liquid chromatographySupercritical fluid chromatography Solid phase extraction Selected reaction monitoring Ultra high performance liquid chroma-tography

9

Introduction

Drug substances in the environment Occurrence The commonly used NSAID diclofenac has been detected in environmental samples in over 50 countries all over the world (2015), and is one of the top five most detected active pharmaceutical ingredients (APIs) in all five UN regions.[1] Unfortunately, it is not only widely spread in the environment, it is also known to be toxic for many species. For example, it is believed to have caused the >95% population decline of the white-rumped vulture on the Indian subcontinent.[2] Thanks to the increasing volume of reports in recent decades, the occurrence of drug compounds in our aquatic environment has now be-come well established, which has been made possible by the development of efficient and sensitive analytical methods.[3–5] The major emission source of these compounds to the environment is generally urban wastewater (Fig. 1), through both the intended use and the incorrect disposal of pharmaceuti-cals.[1,4] This also makes it possible to estimate the drug consumption of e.g. illicit drugs such as cocaine and doping compounds in communities through measurements in wastewater and rivers, called environmental forensics.[6–8]

In comparison to the human pharmaceutical consumption, veterinary med-icine products are used in smaller quantities and the number of APIs are less diverse.[4] Although found in lower amounts, the classes of APIs used in vet-erinary medicine, e.g. antibiotics, can in combination with a direct emission into the environment be a source of concern. Emission from the manufacturing of APIs is contributing to local but often severe water contaminations in the areas of production, with examples from both developing- and industrialized countries.[9] An aggravating factor is the relocation of manufacturing sites of APIs mainly to Asia, often to countries maintaining less strict regulations and non-satisfactory wastewater processing, as shown in several studies from e.g. India[10,11] and China[12]. In 2016, a review concluded that measurements of pharmaceuticals in the environment are reported from 71 different countries.[1]

10

Figure 1. Schematic picture of the suggested flow of drug substances into the aquatic environment.

Concern The concern regarding the effects from APIs and related compounds on our environment is logical due to the nature of drug compounds. They are gener-ally designed to affect a biological system at low concentrations, often through relatively well-studied proteins in humans, e.g. a receptor. What impact these compounds have in a non-target biological system such as e.g. fish, which often contains the same or similar target proteins, is largely unknown how-ever. Some cases connecting APIs and adverse effects on animals, such as the death of vultures on the Indian subcontinent caused by diclofenac[2], the ef-fects of low concentrations of synthetic estrogen on minnow[13] and benzodi-azepine effect on European perch[14] can be found in research reports. Per-forming interventional studies of the effect on concerned species for all the approximately three thousand authorized APIs available in EU[4] is, however, not realistic. One solution is the use of prediction models to identify com-pounds with a higher risk that should be prioritized and further tested. Since 1 December 2006, an environmental risk assessment must accompany market approval applications for pharmaceuticals within the European Union. These assessments use the principle of prediction followed by generation of experi-mental data, if proven necessary by the prediction.[15] The assessments only concern new drug substances though, and have been criticized, e.g. for not managing the chronic exposure and mixtures of compounds.[16,17] Stereoselec-tive quantification of chiral compounds is also often neglected, despite the generally different effects obtained for the enantiomers.[18] Enantiomeric sep-aration has however been applied in a few environmental studies using the enantiomeric fraction (EF) as a tool for determination of sewage origin in sur-face water[19] and for studying the degradation mechanism.[20,21]

A study by Scheurell et al.[22] detecting transformation products of diclo-fenac in Pakistan, provides an example of contamination from households lacking wastewater treatment and in situ transformation, creating a product

Manufacturing

HospitalsHouseholds

Wastewater treatment plant

Animal use

Aquatic environment

Soil

11

with higher toxicity than the parent compound (8-chlorocarbazole-1-yl-etha-noic acid). It hereby concerns two important topics: the necessity of a suffi-cient wastewater treatment and the occurrence of metabolites and transfor-mation products. The fate of APIs in wastewater treatment plants, where most drug compounds end up in communities with functional sewer systems, has been an important research area. By comparing influent and effluent levels of APIs, it is now known that the degree of degradation largely differs between drug compounds and between different wastewater treatment plants.[23,24] Sev-eral methods to increase the removal rate have been investigated, mainly through implementation of an additional removal step, e.g. through membrane bioreactors, activated carbon and ozonation.[23,25] Although several techniques successfully remove a majority of the APIs during the water treatment, they would demand some modifications to today’s wastewater treatment plants, with increasing costs and energy consumption as a result. As a cheaper alter-native due to its lower operating cost, constructed wetlands are being investi-gated as an alternative to the other removal techniques.[26] The elimination of organic contaminants in wetlands is believed to occur through three major mechanisms: sorption, biotransformation and photolysis. As the degradation of most APIs varies in different wetland types, the combination of wetlands and mechanisms is being investigated and optimized. The biotransformation performed by microorganisms is for example dependent on the offered micro-environment, which decides what combination of microorganisms will flour-ish. Differently vegetated wetlands affect the biofilm, generally being more important than planktonic microorganisms for the degradation, and may create zones of both aerobic and anaerobic conditions.[27,28]

The second subject exemplified through the study by Scheurell et al.[22] concerned toxic transformation products formed in situ in the environment, often co-existing with the parent compound and metabolites. The human me-tabolites from APIs are nowadays included in the environmental risk assess-ment demanded for new market approvals by EMA, but do not include the metabolites from older existing APIs such as diclofenac.[15] The combination of APIs with similar effects, older APIs, metabolites and transformation prod-ucts, is still a largely unexplored area.[29] Several reports highlight this phe-nomenon, often referred to as the “cocktail effect”, i.e. the effects of similar compounds in combination.[29–31] Estimating this effect is complicated though, since many compounds, both known and unknown, may contribute to the ef-fect on a target. To determine this, knowledge is needed concerning what com-pounds are present from emission and degradation and what the impact is of these compounds in combination.

12

Determination The determination of APIs in the environment is made possible due to devel-opments and advances within analytical chemistry. Identification and quanti-fication of APIs and transformation products occurring at low concentrations in complex matrices is however still a demanding task, exemplified e.g. in a recent internal calibration between five Swedish laboratories.[32] Sampling and sample handling, sample preparation and separation-detection techniques are generally dependent on concentrations, sample matrix, compounds and the aim of the analysis. To obtain large concentration factors, solid phase extrac-tion (SPE) has been the main technique applied for sample preparations in environmental studies, often using mixed-mode copolymers such as Oasis® HLB for multi-residue applications.[33] Matrix effects (defined on p.15) are both commonly occurring and often challenging when determining drug sub-stances in environmental matrices. When for example calculating a removal rate for a wastewater treatment plant, the more contaminated influent water may suffer from higher ion suppression than the treated effluent water. A neg-ative calculated removal rate might therefore be presented, and while this is often explained by deconjugation of metabolites or inappropriate sampling, this might actually be due to the matrix effect.[34]

Historically, the separation technique for low concentrations and unknowns was mainly gas chromatography (GC) due to its chromatographic resolution and compatibility with mass spectrometry (MS). Although GC is still used for some hydrophobic compound classes, the main technique for separations is today reversed phase liquid chromatography (RPLC) using electrospray ioni-zation (ESI) for MS detection.[35] RPLC has benefits in environmental analy-sis: it provides compatibility with the water-based samples and medium polar metabolites often encountered in environmental applications analyzing drug substances. Modern LC provides high separation efficiency through the ultra high performance liquid chromatography- (UHPLC) instruments utilizing sub-2 µm particle columns, combined with high detection sensitivity when using ESI-MS/MS. The most commonly used mass analyzer for environmen-tal applications is the triple quadrupole (QqQ) instrument detecting in the se-lected reaction monitoring (SRM)-mode, generally using two product ions per precursor and thereby providing both sensitive and selective detection. SRM detection is only applicable to targeted analysis, however, and is therefore used for quantification of specific groups of compounds and some multi-resi-due methods.[36,37] For screening purposes and untargeted search for com-pounds, high resolution mass spectrometry (HRMS) has successfully been ap-plied within the environmental field, often using hybrid instruments with the combination of a quadrupole and a time-of-flight instrument (Q-ToF)[38,39] or Orbitrap[40].

13

In comparison to the biomedical field, chiral separations within the envi-ronmental field is still a relatively recent interest.[41] The macrocyclic antibi-otic-, protein- and polysaccharide-based chiral stationary phases (CSPs) have been the most commonly used chiral selectors in the environmental field, us-ing LC and reversed phase or ion-polar mode to support the MS ionization.[41] The increasing interest in supercritical fluid chromatography (SFC) has so far not been matched in the environmental field.

Supercritical fluid chromatography and the ESI source Supercritical fluid chromatography – A new revival? Supercritical fluid chromatography (SFC) has been around for a long time, initially starting as an alternative to gas chromatography over fifty years ago said to have “liquid-like solvation power and gas-like viscosity”.[42] It has his-torically been more “GC-like” than today, i.e. using mainly open-tubular ca-pillary columns and detectors associated with GC. The composition of the critical phase was initially not fixed; except from the currently dominating carbon dioxide, several other solvents have been evaluated, such as ammonia, water, nitrous oxide etc. The low critical point (Tc=304.13 K, Pc=73.773 bar), low toxicity and flammability, availability, low cost and overall performance has however made CO2 the natural choice for SFC.[42,43] CO2 is also preferred for its ability to solve more polar solvents: pure CO2 has a polarity correspond-ing to hexane, but through the miscibility of e.g. methanol, the solubility of polar compounds can be dramatically increased. The use of polar organic modifiers also allows the inclusion of small amounts of water and polar addi-tives (e.g. ammonia, acetic acid, ammonium formate), often with a high im-pact on elution and peak shape for acidic and basic compounds, increasing the technique’s application range (Fig. 2). The conditions commonly used today, with gradients of polar organic solvent and polar additives, often forms a sub-critical phase instead of a supercritical phase. The concern and potential prob-lem of this conversion has been the formation of two separate phases and a modification of analyte solubility. However, through the use of a backpres-sure-regulator, a high pressure is maintained and the drop in temperature and formation of a subcritical phase does not seem to make the system lose its “supercritical abilities.”[44]

14

Figure 2. A schematic picture of the polar coverage of SFC in comparison to differ-ent forms of LC (based on reference [43]).

Although SFC has been known since 1962, the theory regarding retention mechanisms and factors controlling this is still an ongoing debate. This has, in combination with the increased complexity in comparison to LC and GC, contributed to decreasing the spread and development of the technique. It has however always stayed popular in some niches where it possesses certain ad-vantages, such as preparative scale chromatography and chiral separations. In recent years, SFC has started to face a new dawn, with several new instru-ments being commercially available, providing higher performance and com-patibility with sub-2 µm i.d. particle columns.[45] The new instruments have been evaluated and are naturally compared with the state of the art-technique for pharmaceuticals: UHPLC in combination with MS.[46–48] The results from these studies show that SFC in combination with MS is comparable with LC-MS, even when analyzing polar compounds. SFC also shows orthogonality to RPLC and they therefore often fail for different compounds, hence showing potential usage as complimentary techniques.

The hyphenation between SFC and MS was first described 1983, using a chemical ionization source.[49] When most open-tubular capillary columns were replaced by packed columns, electron impact and chemical ionization sources were replace by atmospheric pressure ionization sources, mainly elec-trospray ionization (ESI) and atmospheric pressure chemical ionization (APCI).[50] However, some difficulties exist when connecting SFC to MS: when the pressure drops, CO2 will decompress and enter gas form, with a sub-sequent decrease in solubility and temperature. The interface therefore needs to be appropriately designed to avoid blockage from ice formation or precipi-tated salts in the tubing or probe capillary. A make-up solvent, e.g. methanol or ethanol, is generally added to counteract these problems and will also im-prove the ionization efficiency and increase the precision of repeated experi-ments.[50,51]

Increasing Polarity

SFC: Pure CO2

SFC: CO2 + modifier

Normal-phase LC

HILIC

Reversed-phase LC

Ion chromatography

SFC: CO2 + modifier + additive

SFC: CO2 + modifier + additives including H2O

15

ESI source and the sample matrix The discovery of the electrospray ionization technique[52] has, together with the MALDI ionization, revolutionized the field of mass spectrometry. Alt-hough the exact ionization mechanism for ESI is still not fully understood, its straight hyphenation to LC, soft ionization and wide mass range have led to the popularity it possesses today. The high signal sensitivity for polar com-pounds also applies to interferences and thereby increasing matrix effects (ME), i.e. a concentration-dependent alteration of the signal intensity caused by components in the sample matrix.[53] Although the mechanism behind ME is not fully understood, especially not for ion enhancement, the proposed causes are summarized in Figure 3.[54–56] The ESI source is more affected than the closely related atmospheric pressure ionization source (APCI), especially when analyzing samples containing a high amount of interferences and a low degree of sample cleanup.[57] Biological matrices such as blood plasma is an example of a common sample matrix often suffering from ME. Hence, the investigation of ME has e.g. since 2001 been included in the FDA guideline for validation of bioanalytical methods.[58] ME is also a common problem within environmental analysis, with for example a potential underestimation of concentrations in complex matrices such as influent wastewater.[34] It is also important to be aware of the relative matrix effect, i.e. the variations in ME of a sample matrix from different sources.[59] Wastewater for example, is a highly variable matrix, depending on factors such as rainfall, season and the dis-charges from served community and industry. This might cause problems when developing and validating a method and later applying it to wastewater of a different origin. Especially if combined with the use of surrogate stand-ards, i.e. extrapolating the internal standard correction through the use of dif-ferent compounds with often different retention times as standards, the relative ME might be problematic. This might explain some negative mass-balances seen in the published literature, when a previously developed method is de-veloped[60] and then applied[61] without any (presented) method validations us-ing the new sample matrix.

To visualize and measure ME, two different methods have generally been used: post-extraction addition (PEA) and post-column infusion (PCI). PEA is quantitative and provides the amount of ME in % for a selected signal at the retention time of the compound, by comparison of the signal in sample matrix and neat standard.[59] PCI is performed by post-column infusions of the analyte while injecting a blank sample matrix, thereby providing information about signal intensity for the infused compound throughout the chromatogram and providing qualitative ME information.[62] A third, more recent and so far not as commonly used method has also been described,[63] which merges the ben-efits from both the PEA and PCI methods, receiving both quantitative and qualitative information of the ME. By performing post-column infusions with several injections of neat solvent and sample matrix (as for a PCI) but with

16

data treatment afterwards, the post-column infusion matrix profile (PCIMP) is created.

Figure 3. The proposed mechanism behind matrix effects in electrospray ioniza-tion.[54–56] The experimental data behind the described mechanisms are generated us-ing LC-ESI/MS. However, using make-up solvent and organic modifier, the same mechanisms are most likely also valid for SFC/ESI-MS.

1. Competition for access to the droplet surface for gas-phase emission be-tween matrix components (M) and analyte ions (A) that are co-eluting in the sprayed solution.

2. Competition between analyte and matrix component for available charge. 3. Modification of droplet surface tension and thereby alternation of the abil-

ity of analytes to reach gas phase. May cause both enhancement and ion suppression.

4. Nonvolatile matrix component precipitates with analyte. 5. Neutralization of analyte by acid/base reactions or by ion pairing with ma-

trix components or mobile phase additive. Can likely work the opposite way, and thereby cause enhancement.

6. Matrix component accumulation on e.g. sample cone. The ion source de-sign will also have an impact on ME.

Inlet flow →

e-

1.2.

AM

A A

AM

M

M MM

MA

+

+

++ +

+

++

MA

M

+

+

+

+

AM+

+ +

+

3.

A+ + M → A + M+

orA+ + M- → A+M-

4. 5.

AM A

A

A

MM

M M

M

MA

+

+

++

+

+

++

+

M+

M+

M+

M+

M+

M+

6.

A

Orifice

17

Aims

The overall aim of the thesis project was: To utilize analytical chemistry in a degradation study of drug compounds in wetlands. And to study fundamental analytical aspects analyzing complex samples with SFC/ESI-MS and thereby contribute to the development and use of these tech-niques.

The specific aims of the included papers were:

I To develop analytical tools able to detect a change in enantio-meric ratio during the degradation of chiral drug substances in wetlands and thereby potentially confirm a biological degrada-tion mechanism. The method should be MS-compatible and sensitive enough to monitor the degradation as long as possible before reaching LOD.

II To explore the transformation occurring in wetland microcosms adapted for biodegradation: How diverse is the transformation? Are there differences between similar parent compounds? And how can previously undescribed transformation products be identified using high resolution MS when lacking references?

III Is SFC suffering from less matrix effects than LC? Comparison

of the matrix effect generated for SFC and LC, using the same ESI/MS interface, four complex sample matrices and com-monly used screening conditions for both techniques.

IV To investigate the alkali metal clusters causing ion suppression in SFC/ESI-MS: how they arise, what they consist of, their im-pact on MS detection and how to possibly avoid them.

18

Part 1. Wetlands

In summary, three different microcosm treatments adapted for biodegradation were used. Three β-blocking agents were added, and five samples were taken regularly during a seven-day period. Two things were measured; the change of enantiomeric ratio during degradation and the formation of transformation products, preceded by analytical method development. From the enantiomeric ratio results, no clear conclusions could be drawn, as no clear trends could be observed. Several different transformation products could be described, show-ing large differences between the β-blockers, and diversity in transformations for the wetlands.

The wetlands

“I’m a strong proponent of the restoration of the wetlands, for a lot of reasons. There’s a practical reason, though, when it comes to hurricanes: The stronger the wetlands, the more likely the damage of the hurricane.”

George W. Bush (43rd President of the United States)

The wetland experiments were performed by Justin Jasper within Prof. David Sedlak’s research group at the University of California at Berkeley, USA, and are therefore only briefly described here. The microcosms used in the two studies (Paper I-II) were made from constructed pilot-scale wetlands located in Discovery Bay, California. They were later (2015) drained and the scale was increased to larger wetlands and experiments have been performed to ver-ify the conclusions drawn through the pilot-scale wetlands.[64] The pilot-scale wetlands received effluent wastewater from a treatment plant and were con-structed for the evaluation of different wetland types (e.g. open-water and bul-rush grown wetlands) facilitating different removal mechanisms. To optimize the removal of both nutrients and organic contaminants, different wetland types were evaluated solely and in combinations (Fig. 4). They have, for ex-ample, been evaluated for their ability to remove nitrate[65] and organic con-taminants through phototransformation[66] and biotransformation[67].

19

Figure 4. An engineering diagram of the pilot-scale wetlands in Discovery Bay, and an example of how the cells could be connected for combinational studies. POP= open-water wetland, WWTP= wastewater treatment plant.

To identify biotransformation as a degradation mechanism, the enantiomeric fraction (EF) was measured for three of the compounds used to evaluate the wetlands: atenolol, metoprolol and propranolol (Fig. 5). Since metoprolol acid was previously known as the major transformation product for both metopro-lol and atenolol in wetlands, it was also included in the EF measurement. This provides the opportunity to measure EF for both precursor and the formed transformation product (Paper I). To be able to exclude phototransformation as degradation mechanism and control the conditions through a controlled en-vironment, the experiments were performed in microcosms of the wetlands. Materials from two different varieties of wetlands were used to create the mi-crocosms: cattail and open-water wetlands. The open-water wetland was used in two different conditions mimicking day and night-conditions, providing the three different biotransformation models used in both Paper I and Paper II. The cattail microcosms were kept in the dark together with one set of open-water wetland microcosms, mimicking night-time conditions. The second set with open-water microcosms were kept under red light, using a wavelength (635 nm) allowing algal photosynthesis and thereby mimicking daytime con-ditions but without photodegradation. The microcosm experiments were per-formed over the course of seven days and in two batches, one with a low con-centration (1 µg L-1) of atenolol, metoprolol and propranolol used for the en-antiomeric separation (Paper I) and one using a higher concentration (100 µg L-1) to facilitate transformation product identification (Paper II). The condi-tions were otherwise the same for both batches, and a microcosm set for each

20

treatment without added compounds was used as a control. Samples contain-ing 5 ml were collected regularly, five times during the seven-day incubation period, stored frozen and sent from Berkeley to Uppsala stored with cool packs. The cool packs were still partly frozen at arrival, and the water was kept frozen (-26 °C) before and between analyses. For more details regarding the wetland microcosms, see Papers I-II or the biotransformation study per-formed by Jasper et al.[67]

Figure 5. The compounds studied in the wetland microcosm experiments, Paper I and II. Asterisks mark their chiral center.

Chiral separation

“And should not I spare Nineveh, that great city, wherein are more than sixscore thousand persons that cannot discern between their right hand and their left hand; and also much cattle.”

Book of Jonah, Chapter 4, Verse 11 (KJB)

Following EF during the degradation of the β-blocking agents with as many samples as possible requires not only a method capable of enantiomeric sepa-ration, but also a sensitive one. The literature contains examples of MS-com-patible methods for chiral separation of the β-blockers, but none that includes the zwitterionic metoprolol acid using MS detection. After some initial exper-iments using LC-MS and several different CSPs (data not published) the chro-

21

matographic technique was changed to SFC. By applying solid phase extrac-tion (SPE) using the mixed-mode copolymer column Oasis HLB and a simple protocol using mild washing, the sample solvent was exchanged from water to methanol, an organic solvent acceptable for SFC and able to solve all four compounds. Applying an extraction also allowed a pre-concentration of the samples, beneficial in Paper I, since sensitivity was essential to receive as much data as possible. As known compounds was analyzed during the EF-experiments, matching deuterium-labeled compounds could be used as inter-nal standards (IS).

As discussed in Paper I, the focus during the method development was the zwitterionic metoprolol acid, since its enantiomers did not resolve in methods capable of fully enantioseparating its precursors atenolol and metoprolol. The CSP used in the final method, Chiralpak® IB-3 contains cellulose tris (3,5-dimethylphenylcarbamate) immobilized on silica. The same phase was ini-tially evaluated using HPLC and its coated reversed-phase form (Chiralcel® OD-R) using an acidic mobile phase, without success. Studies has shown that the enantioresolution might vary due to the influence of CO2,[68] which might explain the different result using SFC.

The final method described in Paper I consisted of dual columns: the achi-ral 2-ethylpyridine based column added to avoid isobaric interferences in the low resolving triple quadrupole, and the CSP Chiralpak® IB-3 (Fig. 6). The mobile phase consisted of carbon dioxide with 18% methanol containing 0.5% (v/v) of the additives trifluoroacetic acid and ammonia in a 2:1 molar ratio and a post-column make-up flow of methanol containing 0.1% (v/v) formic acid. The method was able to successfully resolve (RS>1.5) all the enantiomers within 8 min, with an limit of detection (LOD) range of 9.2-81 ng L-1 using MS in SRM-detection mode and solid phase extraction (SPE). The method was validated according to the aim and application of Paper I, including pre-cision, accuracy, selectivity, detection limit, recovery and linearity.

After development and validation of the method described in Paper I, it was applied to the low-concentration microcosm samples. The rapid degrada-tion of atenolol, especially in the open-water microcosms kept in dark, led to an inclusion of the high concentration samples spiked with atenolol in the en-antioselective analysis. Through this, more data points could be collected, showing a change from racemic EF over time.

22

Figure 6. Enantiomeric separation of the three β-blocking agents and the transfor-mation product metoprolol acid, formed from both atenolol and metoprolol. This chromatogram shows the separation using SFC and the chiral column (Chiralpak IB-3) solely, providing a fast separation with good chromatographic resolution but un-fortunately a co-elution of S-metoprolol and E1-metroprolol acid. Using low con-centration samples and ToF-MS, as in this chromatogram, no isobaric interferences occurred. For high-concentration samples and triple quadrupole MS, interference be-tween these two compounds were seen, requiring a separation prior to the MS.

Identification using HRMS

“We might know what we are, but not what we may be” William Shakespeare, Hamlet Act 4, Scene 5

When investigating transformation products from the wetland samples (Paper II), the use of sample extraction was excluded, in order to avoid a discrimi-nating step that could remove transformation products yet to be discovered. The samples were instead concentrated through evaporation overnight using a gentle N2-gas flow and a temperature of 40 ᵒC. The choice of pre-concentra-tion method was the result of a pilot study, controlling the recovery for the three precursors using different evaporation methods (data not published). Some samples were initially analyzed directly without the second filtering and concentration step. This could theoretically have resulted in satisfying limit of detection without any risk of discriminating any transformation products. The MS signal is well known to be non-linear under certain circumstances and

0

50

100

0 0.5 1 1.5 2 2.5 3 3.5 4

Inte

nsity

%

min

R(+)-metoprolol S(-)-metoprolol

0

50

100

0 0.5 1 1.5 2 2.5 3 3.5 4

Inte

nsity

%

min

E1-metoprolol acid

268.1543 0.02 Da

E1-metoprolol acid

0

50

100

0 0.5 1 1.5 2 2.5 3 3.5 4

Inte

nsity

%

min

R(+)-propranolol

260.1645 0.02 Da

S(-)-propranolol

0

50

100

0 0.5 1 1.5 2 2.5 3 3.5 4

Inte

nsity

%

min

R(+)-atenolol

267.1703 0.02 Da

S(-)-atenolol

268.1907 0.02 Da

23

sensitive to matrix components, resulting in i.e. the “dilute-and-shoot” ap-proach, providing an increase in signal despite a dilution of the sample. The tenfold higher concentrations of the samples did however provide an increased signal in these experiments, clearly shown by the control samples: the micro-cosms contained water from the pilot-scale wetlands, which received effluent water from a wastewater treatment plant, containing small amounts of APIs. The control from day 0 therefore provided a peak for propranolol after the sample treatment and concentration, not visible using direct injection. This highlights a common problem within the field of environmental analyses: the blank matrix is often not blank. In this study, this problem only affected matrix matching performed for the recovery and matrix effect experiments (Paper I) and the quantification of 1-naphthol (Paper II). It was handled through the use of wetland water known to contain a negligible amount of the actual com-pounds. The water for matrix effect and recovery experiment (Paper I) was, for example, taken from a blank microcosm at the last time-point, known to not contain any precursors (albeit some metoprolol acid as noted in the matrix effect experiment).

The default MS mode used was MSE, with acquisition using collision en-ergy every second scan, facilitating e.g. extraction of characteristic product ions of the parent compounds from the collision trace.[69] However, as no pre-cursor selection is performed, the data acquired using MSE is often very noisy (Fig. 7). A confirmation analysis using precursor selection in the quadrupole was therefore generally performed in Paper II, comparing sample and refer-ences. The real challenge though is to identify transformation products lack-ing references: what proofs are needed for a positive identification? In Paper II, standards were used for confirmation when possible, comparing retention time, accurate mass, product ion formation and when necessary and possible; isotopic pattern and product ion ratio. When comparing to a standard, the iden-tification criteria formulated in the EU Commission decision concerning resi-dues in animal production[70] are often applied within environmental analysis. These are not adapted to modern HRMS though, since they e.g. do not provide any criteria for mass accuracy, and additions to the criteria have been proposed to solve this.[71] What can safely be stated without a reference standard is, however, not as easy to answer. To confirm a molecule structure with high certainly, an additional NMR analysis is generally required, not easily per-formed in environmental analysis studying trace amounts. In Paper II, several transformation products are confirmed using standard references. In cases where a standard was missing, the data of the transformation products were instead compared with the product ion formation of their precursors or ana-logs, providing a probable structure. Even if this approach can generate struc-tures with high reliability using HRMS, it is not as certain as a comparison with a standard. In Paper III, comparison with databases and literature were performed using accurate mass and product ions to identify the interfering components in mainly plasma and urine sample matrices. Since these sample

24

matrices are well studied, and the interfering components causing ion suppres-sion exist at high concentrations, the identification of them is fairly confident, although still tentative without a comparison to reference compounds.

Figure 7. A comparison of the product ions of N-desisopropyl metoprolol from a wetland-sample using the MSE-collision trace (top) and precursor scan analyzing reference (below). The MSE-mode, with acquisition using collision energy every second scan, creates a large amount of data without the need of extra acquisitions. However, as illustrated here, the spectra becomes more difficult to interpret, since no precursor is selected in the quadrupole.

The different degrees of confidence for an HRMS identification should there-fore be communicated. This is however seldom done today, and the confi-dence scale looks different depending on the field of research. A proposal of criteria applicable within the environmental field has been published by Schy-manski et al., based on the existing criteria published by the Metabolomic Standard Initiative.[72] A confidence scale covering both the criteria from the Metabolomic Standard Initiative and the EU Commission (2002/657/EC) has also been published,[73] meant to be a more general scale for all fields using HRMS identification of unknowns. Although it is clear that a way of communicating identification confidence is needed, which system this will be and wheter it will be applied by all HRMS fields, only the future will tell.

In summary, Paper II describes how several transformation products were identified, several not previously described as transformation products formed in wetland models or other environmental samples (Fig. 8). Differences in di-versity and type of transformations were seen between the different micro-cosm treatments, but mainly between the different compounds. Atenolol was

0.E+00

1.E+04

2.E+04

3.E+04

50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240

Inte

nsit

y

m/z

Sample - MSE

77.0392

79.0540

91.0542

93.0698

105.0700115.0540

121.0652

133.0646141.0682

148.0759

159.0806 191.1086

103.0543

0.E+00

4.E+05

7.E+05

1.E+06

1.E+06

50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240

Inte

nsit

y

m/z

Reference - Product ion scan

79.0549

91.0551

93.0705

103.0548

105.0705

115.0548

121.0652

133.0653

148.0765159.0814

191.1071141.0703

77.0392

25

confirmed to transform exclusively to metoprolol acid, also being a transfor-mation product from metoprolol, although not as dominant as for atenolol. Hydroxylations occurred on both metoprolol and metoprolol acid, though the positions of these were difficult to precisely localize. Using MS/MS product ion data from precursors and deuterium standards, a position on the aromatic ring was suggested. Metoprolol also formed several other transformation products, showing a more diverse pattern in comparison to atenolol and pro-pranolol.

Figure 8. The identified transformation products from atenolol, metoprolol and pro-pranolol formed in the wetland microcosms (Paper II). The superscripted number in-dicates the level of confidence, according to Schymanski et al. [72], where 1 indicates a confirmed structure and 2b a probable structure, from diagnostic data.

Metoprolol1

OH-O-desmethyl metoprolol2b

O-desmethyl metoprolol1

Metoprolol acid1

Deaminated metoprolol1

N-dealkylated metoprolol1

α-OH-metoprolol1 X-OH-metoprolol2b

Atenolol1 Metoprolol acid1OH-metoprolol acid2b

Propranolol11-naphthol1 Propranolol sulfate2b

26

Part 2. Sample Matrix

Comparison of LC and SFC

“Unfortunately, no one can be told what the Matrix is. You have to see it for yourself.”

Morpheus, from the movie Matrix, 1999 When starting to work with environmental analysis of drug substances using ESI-MS detection, you quickly get to know the matrix effect. Analyzing e.g. wastewater, the signal for drug compounds is often heavily affected by ME, even after a generic SPE-extraction using the within the field commonly used HLB-columns (Paper III). An example of the impact of ME can be found in a master’s thesis in environmental forensics, which developed an analytical method using wastewater for determination of pharmaceutical substances without marker approval in Sweden.[74] Since the targeted substances were likely present at very low concentrations, 200 mL of influent wastewater was sample prepared using polymer SPE columns. The high concentration factor was however counteracted by the ME; for several compounds, an ion suppres-sion over 95% was concluded, and only highly abundant and previously known compounds could be identified. The matrix effect is clearly a factor to take into consideration when analyzing complex sample matrices, demanding different approaches for e.g. sample preparation.

Even if good chromatographic selectivity and longer run times might de-crease the impact of ME, there is no guarantee analyzing complex matrices. ME is therefore generally compensated for by using isotopically labeled inter-nal standards (IL-IS), since removing all ME using sample preparation meth-ods and e.g. matrix-matching might often be problematic.[34] IL-IS is however often expensive and not always available for all analytes. Even when perfectly matched, the increase of SRM-channels will decrease dwell time, and can to-gether with induced self-suppression cause less sensitivity.[75] And in the end, even a perfect IL-IS will not reduce ME, only compensate for it.

Considering the factors believed to cause ME (Fig. 3), it is reasonable to as-sume that the difference between LC and SFC will create different ME behav-ior, even using the same ion source and mass spectrometer. The few studies including ME-evaluation published prior to the experiments resulting in Paper

27

III usually indicated less effect from ME using SFC. If this generally would be the case, in combination with an overall comparable performance between modern LC and SFC, it would be a great benefit for SFC when analyzing com-plex samples. The results presented in Paper III indicate that SFC is not less affected by ME, but it does act differently (Fig. 9). Why some previous re-search has seen less ME for SFC than for LC[76,77] might have a logical expla-nation: in e.g. Paper I, in order to change the sample solvent and make it more beneficial for SFC (and to concentrate the sample), a sample extraction was performed. This, and the change in solubility achieved when going from a water-based sample to an organic one, will of course affect the composition of the sample, and thereby also the interferences entering the ion source. This is also the case for several of the previous studies showing low impact from ME.[77–79] Since SFC is sensitive to other interferences than LC (Paper III), this might explain why a low degree of ME was seen for SFC in several cases, where e.g. polar contaminants were removed. That would mean that SFC is not less affected by ME. It is, however, less affected than LC by the interfer-ences remaining after sample treatment in the two studies comparing the tech-niques published prior Paper III[76,77].

It is also worth noticing that different methods were used for measuring ME in Paper III compared to previous studies. The two recent studies com-paring LC and SFC both use post extraction addition to measure ME, as one parameter when comparing methods. This evaluation method provides a quan-tification of ME for a specific compound at its retention time. For a targeted, stable and validated method, this is generally accurate. However, an evalua-tion made with post-column infusion provides more qualitative information, as the chosen compounds are measured in the whole chromatogram and not only during peak elution. If combined with injections of standards and some data treatment (described in Papers III-IV), it is also possible to receive both qualitative and quantitative information from the post-column infusion exper-iments (Fig. 9). This creates a more complete picture of ME, and is easier to compare with other methods and techniques, since it is not dependent on the retention time of specific compounds.

One additional study comparing ME for SFC and LC has been published after Paper III,[80] by the same group as the previous two comparisons dis-cussed above. In this study, three different SFC columns are used and com-pared with an RPLC method for blood plasma and urine-samples, ME evalu-ated using post-extraction addition. The authors concludes that SFC is gener-ally less vulnerable to ME in comparison to reversed phase LC, with some exceptions. Unfortunately, they only evaluates ME using PEA, and hereby only having measurements for their chosen analytes at their retention times. For SFC, the compounds are generally more spread over the chromatogram for the three methods used, in comparison with the RPLC method where nearly all analytes elutes during one minute. The whole comparison will there-fore depend on what interferences that elutes during this minute, giving e.g.

28

the slope of the LC gradient and choice of organic modifier a very high impact. It would have been interesting to compare their results with Paper III, if their evaluation method had been PCIME or post column infusions. Since ME is concentration dependent[76], it would also have been beneficial to use the same concentrations in SFC and LC, and to include more than basic compounds in the comparison. It could however be noted that ion suppression was mainly seen for the 2-PIC column using SFC, which is in line with Paper III. It is therefore interesting that enhancement and suppression were observed for the BEH bare silica and HSS C18 columns, since ion suppression was generally observed in Paper III. Clearly, there is more to discover within this topic.

Another factor likely to be of importance for how ME behaves in SFC is the hyphenation to MS. For the instrument used in this thesis (Waters UPC2), a split of the post-column flow is generally introduced into the MS after the addition of a make-up flow. For other instruments, e.g. the Nexera UC (Shi-madzu) and 1260 Infinity (Agilent), a splitless configuration is also used. This will probably have an impact on ME, although how is still to be investigated.

29

Figure 9. Comparison of the matrix effect from blood plasma using ESI/ToF with LC and SFC (Paper III). The colored lines mark the ME% for three drug com-pounds, infused post-column prior to the MS during injections of sample matrices and blank solvent. The black circles mark areas with high ion suppression, and the red capital letters show any (often tentatively) identified interference with a match-ing elution to the ion suppression. The chromatographic peaks marked by numbers are added from injections of standard mixtures, arbitrarily set to 100% intensity.

-1000

100

05

10

%

min

A, B

, CD

DD

DD

DD

DE,

F, G

-1000

100

05

10

%

min

Ate

nolo

lA

milo

ride

Enal

april

DG

FE

BC

HI

LC SFC

12

67

83

45

910

/11

1. A

ceta

zola

mid

e 2.

Ate

nolo

l3.

Am

ilorid

e4.

Met

opro

lol

5. P

ropr

anol

ol6.

Ena

lapr

il7.

Car

bam

azep

ine

8. F

luox

etin

e

9. M

icon

azol

e10

. Mef

enam

ic ac

id11

. Gem

fibro

zil

119

105

67

84

23

1

Infu

sions

:

A.T 0

–Eg

. ED

TA, N

a-ED

TA, c

reat

ine

D. P

EG ((

C 2H4) nH

2O)

G.G

lyce

roph

osph

ocho

lines

(PC)

B.

Ace

tylc

arni

tine

E. L

yso-

glyc

erop

hosp

hoch

olin

esH

.K+

-clu

ster

C.Ca

rniti

ne

F. S

phin

gom

yelin

s I.

Na+ -c

luste

r

30

Metal ion clusters in SFC/ESI-MS

“In the natural sciences, and particularly in chemistry, generalities must come after the detailed knowledge of each fact and not before it”

Louis Joseph Gay-Lussac (1778-1850), French chemist and physicist A major cause of the ion suppression identified in Paper III was cluster for-mations, later found to partly consist of alkali-ions. Through the accurate m/z and calculations of the elemental composition it was concluded that these clus-ters had not been previously described in the literature. The elution and sepa-ration of inorganic ions in a SFC screening method and the unknown compo-sition of the clusters observed in SFC/ESI-MS was interesting enough to ini-tiate the work presented in Paper IV. In addition to standards of different salts, urine and blood plasma were the sample matrices in focus since they naturally contain alkali metals. Wastewater samples, included in Paper III, are gener-ally pre-concentrated using reversed phase SPE, and thus the content of inor-ganic ions is very low. It was however possible to identify Cl3Fe in effluent wastewater, using accurate m/z and isotopic pattern. Iron chloride is used in the treatment process of the sewage water in the wastewater treatment plant, hence the detection of this metal cluster in effluent but not in influent wastewater. It did however not affect the already high ion suppression seen (Paper III).

Except from Na+ and K+, Mg2+ was also identified as a cluster forming metal ion causing ion suppression (Paper III). Although easy to identify using the isotopic pattern and clearly separated from the other clusters (Fig. 10), this metal ion was not included in Paper IV. The reason behind this is partly the increased complexity of the spectra received from the doubly charged magne-sium ion. An inclusion of one more metal ion in Paper IV would also only add more examples of the same principle. If the aim had been to separate and quantify the metal ion content in samples using SFC, it would have been in-teresting to include as many metal ions as possible. This has recently been done using SFC-ELSD by Foulon et al. using several different inorganic anion and cations.[81] The ions are evaluated chromatographically, using four differ-ent polar columns and different mobile phase compositions. Although a de-tailed comparisons between Paper IV and the work by Foulon et al. is difficult to make, as they exclude Na+ from their results, some similarities exist; for example, the elution order of K+ and Mg2+ and chromatographic behavior of cations using a silica-phase column without additives.[81]

Where the clusters are formed is not always obviously; even if the detector allows the detection of the cluster, they could have been formed previously within the chromatographic system. Most evidence suggests that the clusters

31

presented in Paper IV are formed within the ionization source during the elec-trochemical process: the overlapping retention time of the different clusters, the formation of iron-clusters for Cl- when no cation elutes, and finally the formation of clusters using the make-up solvent, added post-column.

Even if these clusters likely have appeared and caused ion suppression in other recently published studies using methanol modifiers and samples containing alkali ions, they have not been previously reported. This might have an expla-nation in the detection: most recent work where this phenomena could have arisen uses triple quadrupole MS and SRM detection[48,76,80,82,83]. Using this targeted detection mode, there is a risk that any compound co-eluting with the metal clusters are so suppressed that they simply disappear and might there-fore be excluded from the method.

Figure 10. Analysis of a urine sample using SFC/ESI-MS ToF and the 2-PIC col-umn, with three m/z extracted and arbitrarily set to 50% intensity, marking the elu-tion of three metal ions. To the left, a piece of a spectra covering the m/z and iso-topic pattern for [Mg(CH3O)3+Mg]+ from the urine sample is added, together with the exact mass and theoretical isotopic ratio.

0

20

40

60

80

100

0 2 4 6 8

%

(min)

Na+ Mg2+

140 145 150

m/z

Theoretic Isotopic ratio and m/z [Mg(CH3O)3+Mg]+

141.0247100%

142.025929%

143.022831%

144.02405%

K+140 145 150

m/z

Urine sample141.0240100%

143.023029%

142.024923% 144.0265

6%

32

Conclusions

The studies included in this thesis are divided into two different parts. One part with a clear connection to environmental application: the use of analytical chemistry to study the removal of drug substances through wetlands. The sec-ond part is of a more fundamental nature, investigating the consequences of the hyphenation of modern SFC and ESI/MS.

The aims of both Paper I and II included in this doctoral thesis seek to develop methods to increase the knowledge of biodegradation in wetlands, through the use of microcosms and three β-blocking agents. In this perspective, the aim has been successfully met. In Paper I, a chiral method was developed and ap-plied to measure EF during the degradation in microcosms, and in Paper II several transformation products were successfully identified using HRMS. In Paper I, the application showed no clear change in EF, with one exception for atenolol, though this was not large nor followed by a change of EF for its transformation product metoprolol acid. The interpretation of this result is that biological mechanisms may be involved, but probably no stereoselective ones, and they are therefore not discernible from other mechanisms. The experi-ments presented in Paper II demonstrated one new major transformation prod-uct, 1-naphthol formed from propranolol, although this product was not as dominant as e.g. metoprolol acid from atenolol. Through the use of QToF, a diverse formation of low-concentration transformation products were identi-fied, showing the potential of this technique.

Paper III compares the matrix effect using ESI/MS between SFC and LC, the chromatographic techniques used in Papers I and II, respectively. The few studies available discussing this topic prior to our study indicated that SFC could be less affected by ME than LC. While our study could not confirm this to be generally true, it did find that ME behaved differently for the two tech-niques. Another earlier observation, that SFC suffered more from ion suppres-sion and LC from ion enhancement, could however be confirmed. One sur-prising finding in Paper III was the source of the clusters causing areas of major ions suppression: alkali ions. This phenomenon was investigated in Pa-per IV, finding that the clusters originated from the inorganic ions in combi-nation with counter ions formed from the methanol modifier. Since methanol is a widely used and useful modifier for SFC, especially for more polar ana-lytes and sample matrices, it would be difficult to abandon. Instead, our work

33

will hopefully make SFC-MS users aware of this phenomenon when analyz-ing samples containing inorganic metal ions, so that proper precautions can be taken. Whether SFC can be used for detection and quantification of inorganic ions without the use of chelators or other complex binding additives, the future will tell.

Although different aims between the two parts included in this thesis, the sec-ond part is clearly a child of the first, and will hopefully be beneficial for future applications. This might still be a privilege of working in academia: being able to follow up unexpected findings, spend time on fundamental research and let curiosity guide you.

34

Summary in Swedish Populärvetenskaplig sammanfattning

Denna avhandling är uppdelad i två delar: först en sammanfattande och dis-kuterande del, som i sin tur är baserat på de fyra artiklarna som sedan följer. Artiklarna kan i sin tur delas upp i två teman: artikel I och II handlar om ana-lytiska metoder för att spåra läkemedels nedbrytning i våtmarker. Artikel III och IV handlar om hur komplexa prover kan påverka signalen i en masspektrometer, när denna kombineras med separationstekniken superkritisk vätskekromatografi.

Att läkemedel som används också släpps ut i miljön är ett faktum som många är medvetna om idag. I dagstidningar förekommer reportage som handlar om tvåkönade fiskar, farliga värktabletter (framförallt diklofenak) och utsläpp av antibiotika som bidrar till resistensspridning. Många läkemedelssubstanser påverkar djur, eftersom de mål som substanserna är tänkta att verka på hos människa ofta har en motsvarighet i olika djur, generellt med okända effekter. Även om vissa studier har gjorts för specifika läkemedel och djurslag, så är det svårt och mycket kostsamt att studera alla läkemedelssubstansers effekt på natur och miljö. I praktiken behöver man titta på vilka substanser som passerar reningsverken, både de mänskliga och kommunala, och inte bara bryts ned direkt. I ett nästa steg behöver man också undersöka effekten av nedbrytnings-produkterna som bildas, eftersom många av dessa också har effekt, ibland mer än själva ursprungssubstansen. Man behöver också ta hänsyn till vilka kon-centrationer det rör sig om: ett läkemedel som äts av enbart några få individer kommer antagligen aldrig uppnå en farlig koncentration i miljön.

För att kunna studera förekomsten och effekten av läkemedelsutsläpp i mil-jön krävs bra mätmetoder som sen kan ligga till grund för vilka åtgärder som bör utföras. Eftersom det generellt är svårt att mäta så låga koncentrationer i så komplexa prover som till exempel avloppsvatten, ställs det höga krav på analysmetodiken. Den ökade kunskapsnivån som idag finns kring läkemedel i miljön är därför beroende av utvecklingen som skett av mätinstrument och mätmetoder. Dessa mätningar bidrar sedan till mer kunskap, som vidare kan leda till beslut om det är värt att investera i ytterligare rening av avloppsvatten för att bli av med läkemedelsresterna och vilken metod man i så fall bör an-vända.

35

Ett sätt att rena vatten från läkemedel som undersöks är med hjälp av våt-marker. Tanken är att man genom att leda det utgående renade vattnet från reningsverk genom naturens egna reningsverk, våtmarker, skulle kunna få ett ytterligare kostnadseffektivt reningssteg. Miniatyrvåtmarker har därför byggts upp och undersökts, bland annat av vår samarbetspartner i delarbete I och II, som studerar denna metod i Kalifornien, USA. För att kunna optimera våtmar-kernas nedbrytningsförmåga testas olika varianter av våtmarker, till exempel sådana som är bevuxna av kaveldun, säv eller ingen vass. I avhandlingens delarbete I och II analyseras prover från två olika varianter av våtmarker, un-der natt- och dagbetingelser. Målet i delarbete I var att se om nedbrytningen av läkemedel var biologiskt eller ej. Ett sätt att undersöka detta är att se om de så kallade spegelisomererna av ett läkemedel bryts ned olika snabbt: spegel-isomerer av en molekyl är extremt lika, och skiljer sig enbart i sin 3D-konfi-guration, då de är varandras spegelbilder. Om nedbrytningen sker av t.ex. sol-ljus eller genom reaktion med exempelvis syre, kommer ingen skillnad mellan spegelisomererna kunna ses, eftersom dessa enbart skiljer sig som höger och vänster hand. En biologisk nedbrytning skulle dock kunna skilja spegelisome-rerna åt, eftersom denna kan vara mycket specifik. På samma sätt måste den mätmetod man använder för att kunna skilja de båda spegelisomererna åt vara väldigt specifik, speciellt då metoden i denna studie skulle ha förmågan att mäta spegelbildisomererna för tre olika substanser samt en nedbrytningspro-dukt samtidigt. En sådan metod utvecklades och presenteras i delarbete I, där superkritisk vätskekromatografi med koldioxid används för att separera iso-mererna åt. Eftersom syftet var att kunna följa nedbrytningen till så låga kon-centrationer som möjligt, används en mycket känslig masspektrometer som detektor. Förutom att följa nedbrytningen av spegelisomererna undersöktes också vilka nedbrytningsprodukter som bildades av de tre olika läkemedels-substanserna, vilket presenteras i delarbete II. Kunskap om nedbrytningspro-dukter kan vara viktigt när det gäller läkemedel i miljön, eftersom nedbryt-ningsprodukter kan ha liknande farmakologisk effekt som modersubstansen. I delarbete II kartläggs nedbrytningsprodukterna, och arbetet är ett exempel på hur man kan få information om strukturerna hos nedbrytningsprodukterna med hjälp av högupplösande masspektrometri och deuteriummärkta standar-der.

Den andra delen av avhandlingen, som innehåller delarbete III och IV, är av en annan karaktär och är mer inriktade på analysmetodik än tillämpning. Här undersöks superkritisk vätskekromatografi, en teknik som använder koldioxid under högt tryck och temperatur. Det gör att koldioxiden föreligger i ett su-perkritiskt tillstånd, något som kan liknas vid ett vätsketillstånd med gasegen-skaper. Detta är en separationsteknik som fått en ökande popularitet de senaste åren, och ett bredare applikationsområde. I delarbete III jämförs denna teknik med den idag dominerande tekniken för läkemedel i miljön, vätskekromato-grafi, i kombination med detektionsmetoden masspektrometri. Syftet var att

36

jämföra den påverkan på signalen man kan se i masspektrometri när man ana-lyserar olika prover med hög andel störande ämnen, så kallade matriseffekter. Om man analyserar exempelvis blodplasma utan att först ha renat provet kan en förstärkning eller en minskning av signal erhållas i jämförelse med om man analyserar rent lösningsmedel. Om den separationsteknik man använder till masspektrometern skapar mindre matriseffekter, skulle detta kunna vara en stor fördel för denna teknik. Tyvärr visade inte delarbete III att detta var fallet. Snarare var effekten annorlunda, men inte mindre förekommande. Något som var oväntat och ej tidigare observerat för superkritisk vätskekromatografi var en stor påverkan som kunde kopplas till innehållet av metalljoner i proverna, något som referensmetoden vätskekromatografi generellt inte påverkas av. Detta resulterade i delarbete IV, som undersöker ursprunget för denna påver-kan och under vilka förhållanden den uppkommer. Tyvärr visade sig upp-komsten bero på det lösningsmedel som tillsätts koldioxiden, vanligast meta-nol, och som i allmänhet är nödvändigt för att kunna skapa bra separations-metoder. Lösningsmedlet bildar tillsammans med metalljonerna ett kluster i masspektrometern, vilket gör att signalen av andra ämnen som analyseras just då försvinner. Det finns sannolikt ingen enkel och enskild åtgärd som skulle lösa problemet, istället kommer lösningen behöva anpassas till den aktuella metoden och frågeställningen.

Sammanfattningsvis innehåller avhandlingen fyra delarbeten, varav två som beskriver analytiska metoder för detektion av läkemedel i miljön (I-II). Dessa metoder användes för att erhålla information om nedbrytning i våtmarker, med mål att kunna optimera våtmarkernas design och därmed förmåga att bryta ner läkemedel. Tyvärr gav undersökningarna inte något tydligt svar angående ned-brytningsmekanismen, men tillämpningen visar att den analytiska metodiken fungerar.

Delarbete III och IV behandlar frågeställningar som uppkom under den första delen av avhandlingsarbetet, nämligen hur signalen i masspektrometri påverkas av komplexa prover vid analys med superkritisk vätskekromatografi. Genom dessa arbeten har kunskapen om denna kombination av tekniker ökat, vilket kommer vara av värde när användandet av superkritisk vätskekromato-grafi i kombination av masspektrometri ökar.

37

Acknowledgements

The work included in this thesis was carried out at the Division of Analytical Pharmaceutical Chemistry, Department of Medicinal Chemistry at Uppsala University. I would like to thank some people who have contributed to my development as a PhD-student and making my time at the department as pleas-ant as possible.

My main supervisor Prof. Curt Pettersson, for accepting me as a PhD-student and providing me this opportunity. Even when your workload has been high, you have helped me with arisen situations both within the project and outside (i.e. teaching). You have believed in my ability, even when I have not.

My assistant supervisors Prof. Mikael Hedeland and Prof. Torbjörn Arvidsson, for your great knowledge and enthusiasm. You have contributed with feedback whenever I have asked for it, from experimental plans to re-viewer comments.

You have all three contributed to my knowledge of science and the scien-tific community, from faculty business to conference dinners.

Dr. Olle Gyllenhaal and Dr. Ylva Hedeland for introducing me to SFC and sharing your knowledge within the analytical field, during both seminars and the fika. A special thanks to Olle for introducing the SFC community for me and for your help with proofreading.

My fellow PhD-students, both previous and present: Alexander Hellqvist and Axel Rydevik, for your introduction into PhD-student life. Åke Stenholm for arranging educational visits and for nice company. Ida Erngren for being both relaxed, independent and a professional colleague. Kristian Pirttilä for your enthusiasm for science and analytical chemistry, and time spent on our shared hobbies (should have started earlier). You have all been an appreciated and important company, both socially and scientifically. A special thanks to Annelie Hansson and Albert Elmsjö: whose company I have enjoyed both dur-ing endless lab courses and during conferences that ended too quickly. My time as a PhD-student would not have been the same without you.

Dr. Jakob Haglöf and Dr. Mikael Engskog for your help in teaching and good company during these years. I hope you will find time to play board games with future colleagues.

Dr. Victoria Barclay, who both trained me as my supervisor during my master thesis, and advised me during my PhD-studies.

38

Colleagues at the department: The people at the Pharmacognosy for pleasant chats during coffee and lunch breaks. A special thanks to Elisabet Vikeved, a fateful fika-companion in every weather. Biomarker Discovery group (Daniel, Caroline, Mario, Neeraj and Louis) for bringing internationalization to our lab. The people at Organic Chemistry, especially for the time in the new lunch-room (we should have been jointed earlier). Birgitta, Gunilla, Karin S, Linda, Maj, Ulrika R, Sandra and Sorin: thanks for your support in all different things that surrounds research.

To my former MSc-students Maria Pettersson and Joakim Tam for you ambitious performance in the project I supervised you in. Thank also to the “roomies” I had (Adam, Albin, Ann B, Ires and Malin) who respected my need of silence and “medium level” of social conversation.

Prof. David Sedlak and Dr. Justin Jasper, for our cooperation and welcome when I visited Berkeley.

Anna Maria Lundins Travel Grant (Smålands Nation) and The Swedish Pharmaceutical Society (Apotekarsociteten) are both acknowledged for giving me the opportunity to attend scientific conferences. Jag vill även tacka vänner och familj som på olika sätt har stöttat mig, både under min doktorandtid och under tidigare studier:

Mina vänner, som även om de flesta inte bidragit direkt till mitt arbete med avhandlingen, hjälper mig på många andra sätt: genom gott umgänge, humor, middagar och perspektiv. Till min ingifta familj, som utöver sitt inkluderande i sin gemenskap också visat både intresse och nyfikenhet för min forskning. Mina föräldrar (Britt-Marie och Bengt) och syskon (Sofie och Anton), som ställer upp när jag behöver er, på alla de sätt ni kan. Till min dotter Iris, som genom sin tillkomst effektiviserade den sista tiden av arbetet med min avhandling. Min älskade fru Ulrika, som alltid funnits där för konsultationer, alltid gett mig stöd och råd, och haft förståelse, oavsett vad det handlat om. Mitt liv hade varit fattigt utan dig.

Alfred Uppsala, 15 mars 2018

39

References

[1] T. aus der Beek, F.-A. Weber, A. Bergmann, S. Hickmann, I. Ebert, A. Hein, A. Küster. Pharmaceuticals in the environment-Global occurrences and perspectives. Environ. Toxicol. Chem., 2016, 35, 823–835.

[2] J. L. Oaks, M. Gilbert, M. Z. Virani, R. T. Watson, C. U. Meteyer, B. a Rideout, H. L. Shivaprasad, S. Ahmed, M. J. Iqbal Chaudhry, M. Arshad, S. Mahmood, A. Ali, A. Ahmed Khan. Diclofenac residues as the cause of vulture population decline in Pakistan. Nature, 2004, 427, 630–633.

[3] T. A. Ternes. Occurrence of drugs in German sewage treatment plants and rivers. Water Res., 1998, 32, 3245–3260.

[4] BIO Intelligence Service (2013). Study on the environmental risks of medicinal products, Final Report prepared for Executive Agency for Health and Consumers. 2013.

[5] A. Kuster, N. Adler. Pharmaceuticals in the environment: scientific evidence of risks and its regulation. Philos. Trans. R. Soc. B Biol. Sci., 2014, 369, 20130587–20130587.

[6] E. Zuccato, C. Chiabrando, S. Castiglioni, D. Calamari, R. Bagnati, S. Schiarea, R. Fanelli. Cocaine in surface waters: a new evidence-based tool to monitor community drug abuse. Environ. Heal. A Glob. Access Sci. Source, 2005, 4, 14.

[7] T. L. Jones-Lepp, D. a. Alvarez, J. D. Petty, J. N. Huckins. Polar Organic Chemical Integrative Sampling and Liquid Chromatography-Electrospray/Ion-Trap Mass Spectrometry for Assessing Selected Prescription and Illicit Drugs in Treated Sewage Effluents. Arch. Environ. Contam. Toxicol., 2004, 47, 427–439.

[8] A. Causanilles, V. Nordmann, D. Vughs, E. Emke, O. de Hon, F. Hernández, P. de Voogt. Wastewater-based tracing of doping use by the general population and amateur athletes. Anal. Bioanal. Chem., 2018, DOI 10.1007/s00216-017-0835-3.

[9] D. G. J. Larsson. Pollution from drug manufacturing: review and perspectives. Philos. Trans. R. Soc. B Biol. Sci., 2014, 369, DOI 10.1098/rstb.2013.0571.

[10] J. Fick, H. Söderström, R. H. Lindberg, C. Phan, M. Tysklind, D. G. J. Larsson. Contamination of surface, ground, and drinking water from pharmaceutical production. Environ. Toxicol. Chem., 2009, 28, 2522.

[11] D. G. J. Larsson, C. de Pedro, N. Paxeus. Effluent from drug manufactures contains extremely high levels of pharmaceuticals. J. Hazard. Mater., 2007, 148, 751–755.

[12] D. Li, M. Yang, J. Hu, L. Ren, Y. Zhang, K. Li. Determination and fate of oxytetracycline and related compounds in oxytetracycline production wastewater and the receiving river. Environ. Toxicol. Chem., 2008, 27, 80.

40

[13] J. L. Parrott, B. R. Blunt. Life-cycle exposure of fathead minnows (Pimephales promelas) to an ethinylestradiol concentration below 1 ng/L reduces egg fertilization success and demasculinizes males. Environ. Toxicol., 2005, 20, 131–141.

[14] T. Brodin, J. Fick, M. Jonsson, J. Klaminder. Dilute concentrations of a psychiatric drug alter behavior of fish from natural populations. Science, 2013, 339, 814–5.

[15] EMEA/CHMP/SWP/4447/00 corr 21*. Guideline on the environmental risk assessment of medicinal products for human use. 2006, 1–12.

[16] H. Schmitt, T. Boucard, J. Garric, J. Jensen, J. Parrott, A. Péry, J. Römbke, J. O. Straub, T. H. Hutchinson, P. Sánchez-Argüello, A. Wennmalm, K. Duis. Recommendations on the environmental risk assessment of pharmaceuticals: Effect characterization. Integr. Environ. Assess. Manag., 2010, 6 Suppl, 588–602.

[17] M. Ågerstrand, C. Berg, B. Björlenius, M. Breitholtz, B. Brunström, J. Fick, L. Gunnarsson, D. G. J. Larsson, J. P. Sumpter, M. Tysklind, C. Rudén. Improving Environmental Risk Assessment of Human Pharmaceuticals. Environ. Sci. Technol., 2015, 49, 5336–5345.

[18] B. Kasprzyk-Hordern. Pharmacologically active compounds in the environment and their chirality. Chem. Soc. Rev., 2010, 39, 4466–503.

[19] L. J. Fono, E. P. Kolodziej, D. L. Sedlak. Attenuation of wastewater-derived contaminants in an effluent-dominated river. Environ. Sci. Technol., 2006, 40, 7257–7262.

[20] V. K. H. Barclay, N. L. Tyrefors, I. M. Johansson, C. E. Pettersson. Trace analysis of fluoxetine and its metabolite norfluoxetine. Part II: Enantioselective quantification and studies of matrix effects in raw and treated wastewater by solid phase extraction and liquid chromatography-tandem mass spectrometry. J. Chromatogr. A, 2012, 1227, 105–14.

[21] H. Buser, T. Poiger, M. D. Müller. Occurrence and Environmental Behavior of the Chiral Pharmaceutical Drug Ibuprofen in Surface Waters and in Wastewater. Environ. Sci. Technol., 1999, 33, 2529–2535.

[22] M. Scheurell, S. Franke, R. M. Shah, H. Hühnerfuss. Occurrence of diclofenac and its metabolites in surface water and effluent samples from Karachi, Pakistan. Chemosphere, 2009, 77, 870–876.

[23] R. L. Oulton, T. Kohn, D. M. Cwiertny. Pharmaceuticals and personal care products in effluent matrices: A survey of transformation and removal during wastewater treatment and implications for wastewater management. J. Environ. Monit., 2010, 12, 1956–78.

[24] P. Falås, H. R. Andersen, A. Ledin, J. L. C. Jansen. Occurrence and reduction of pharmaceuticals in the water phase at Swedish wastewater treatment plants. Water Sci. Technol., 2012, 66, 783.

[25] C. Prasse, D. Stalter, U. Schulte-Oehlmann, J. Oehlmann, T. a. Ternes. Spoilt for choice: A critical review on the chemical and biological assessment of current wastewater treatment technologies. Water Res., 2015, 87, 237–270.

[26] J. T. Jasper, M. T. Nguyen, Z. L. Jones, N. S. Ismail, D. L. Sedlak, J. O. Sharp, R. G. Luthy, A. J. Horne, K. L. Nelson. Unit Process Wetlands for Removal of Trace Organic Contaminants and Pathogens from Municipal Wastewater Effluents. Environ. Eng. Sci., 2013, 30, 421–436.

[27] J. R. White, M. a Belmont, C. D. Metcalfe. Pharmaceutical compounds in wastewater: wetland treatment as a potential solution. ScientificWorldJournal., 2006, 6, 1731–6.

41

[28] P. N. Carvalho, J. L. Araújo, A. P. Mucha, M. C. P. Basto, C. M. R. Almeida. Potential of constructed wetlands microcosms for the removal of veterinary pharmaceuticals from livestock wastewater. Bioresour. Technol., 2013, 134, 412–6.

[29] M. I. Vasquez, a. Lambrianides, M. Schneider, K. Kümmerer, D. Fatta-Kassinos. Environmental side effects of pharmaceutical cocktails: What we know and what we should know. J. Hazard. Mater., 2014, 279, 169–189.

[30] H. L. Osachoff, M. Mohammadali, R. C. Skirrow, E. R. Hall, L. L. Y. Brown, G. C. van Aggelen, C. J. Kennedy, C. C. Helbing. Evaluating the treatment of a synthetic wastewater containing a pharmaceutical and personal care product chemical cocktail: Compound removal efficiency and effects on juvenile rainbow trout. Water Res., 2014, 62, 271–280.

[31] B. Styrishave, B. Halling-Sørensen, F. Ingerslev. Environmental risk assessment of three selective serotonin reuptake inhibitors in the aquatic environment: A case study including a cocktail scenario. Environ. Toxicol. Chem., 2011, 30, 254–261.

[32] O. Svahn, E. Björklund. Internkalibrerad Läkemedelsanalys 2017 - Ett Samarbetsprojekt För Ökad Analyskvalité, 2017.

[33] W. W. Buchberger. Current approaches to trace analysis of pharmaceuticals and personal care products in the environment. J. Chromatogr. A, 2011, 1218, 603–18.

[34] M. Petrovic. Methodological challenges of multi-residue analysis of pharmaceuticals in environmental samples. Trends Environ. Anal. Chem., 2014, 1, e25–e33.

[35] D. Fatta-Kassinos, S. Meric, A. Nikolaou. Pharmaceutical residues in environmental waters and wastewater: current state of knowledge and future research. Anal. Bioanal. Chem., 2011, 399, 251–75.

[36] M. Seifrtová, L. Nováková, C. Lino, A. Pena, P. Solich. An overview of analytical methodologies for the determination of antibiotics in environmental waters. Anal. Chim. Acta, 2009, 649, 158–79.

[37] R. Grabic, J. Fick, R. H. Lindberg, G. Fedorova, M. Tysklind. Multi-residue method for trace level determination of pharmaceuticals in environmental samples using liquid chromatography coupled to triple quadrupole mass spectrometry. Talanta, 2012, 100, 183–95.

[38] S. Herrera-Lopez, M. D. Hernando, E. García-Calvo, A. R. Fernández-Alba, M. M. Ulaszewska. Simultaneous screening of targeted and non-targeted contaminants using an LC-QTOF-MS system and automated MS/MS library searching. J. Mass Spectrom., 2014, 49, 878–93.

[39] J. Wang, P. R. Gardinali. Identification of phase II pharmaceutical metabolites in reclaimed water using high resolution benchtop Orbitrap mass spectrometry. Chemosphere, 2014, 107, 65–73.

[40] M. Favier, R. Dewil, K. Van Eyck, A. Van Schepdael, D. Cabooter. High-resolution MS and MSninvestigation of ozone oxidation products from phenazone-type pharmaceuticals and metabolites. Chemosphere, 2015, 136, 32–41.

[41] A. R. Ribeiro, A. S. Maia, Q. B. Cass, M. E. Tiritan. Enantioseparation of chiral pharmaceuticals in biomedical and environmental analyses by liquid chromatography: An overview. J. Chromatogr. B, 2014, 968, 8–21.

[42] A. Tarafder. Metamorphosis of supercritical fluid chromatography to SFC: An Overview. TrAC - Trends Anal. Chem., 2016, 81, 3–10.

[43] T. A. Berger. Supercritical Fluid Chromatography. 2015, 1–170.

42

[44] E. Lesellier, C. West. The many faces of packed column supercritical fluid chromatography – A critical review. J. Chromatogr. A, 2015, 1382, 2–46.

[45] L. Nováková, A. Grand-Guillaume Perrenoud, I. Francois, C. West, E. Lesellier, D. Guillarme. Modern analytical supercritical fluid chromatography using columns packed with sub-2μm particles: A tutorial. Anal. Chim. Acta, 2014, 824, 18–35.

[46] A. Grand-Guillaume Perrenoud, J.-L. Veuthey, D. Guillarme. Comparison of ultra-high performance supercritical fluid chromatography and ultra-high performance liquid chromatography for the analysis of pharmaceutical compounds. J. Chromatogr. A, 2012, 1266, 158–67.

[47] E. Lemasson, S. Bertin, P. Hennig, E. Lesellier, C. West. Comparison of ultra-high performance methods in liquid and supercritical fluid chromatography coupled to electrospray ionization – mass spectrometry for impurity profiling of drug candidates. J. Chromatogr. A, 2016, 1472, 117–128.

[48] M. K. Parr, B. Wuest, E. Naegele, J. F. Joseph, M. Wenzel, A. H. Schmidt, M. Stanic, X. de la Torre, F. Botrè. SFC-MS/MS as an orthogonal technique for improved screening of polar analytes in anti-doping control. Anal. Bioanal. Chem., 2016, 408, 6789–6797.

[49] R. D. Smith, J. Fjeldsted, M. L. Lee. Supercritical fluid chromatography - mass spectrometry. Int. J. Mass Spectrom. Ion Phys., 1983, 46, 217–220.

[50] A. Grand-Guillaume Perrenoud, J.-L. Veuthey, D. Guillarme. Coupling state-of-the-art supercritical fluid chromatography and mass spectrometry: from hyphenation interface optimization to high-sensitivity analysis of pharmaceutical compounds. J. Chromatogr. A, 2014, 1339, 174–84.

[51] J. Duval, C. Colas, V. Pecher, M. Poujol, J. F. Tranchant, E. Lesellier. Hyphenation of ultra high performance supercritical fluid chromatography with atmospheric pressure chemical ionisation high resolution mass spectrometry: Part 1. Study of the coupling parameters for the analysis of natural non-polar compounds. J. Chromatogr. A, 2017, 1509, 132–140.

[52] M. Yamashita, J. B. Fenn. Electrospray ion source. Another variation on the free-jet theme. J. Phys. Chem., 1984, 88, 4451–4459.

[53] A. Furey, M. Moriarty, V. Bane, B. Kinsella, M. Lehane. Ion suppression; a critical review on causes, evaluation, prevention and applications. Talanta, 2013, 115, 104–22.

[54] P. J. Taylor. Matrix effects: the Achilles heel of quantitative high-performance liquid chromatography–electrospray–tandem mass spectrometry. Clin. Biochem., 2005, 38, 328–334.

[55] P. Panuwet, R. E. Hunter, P. E. D’Souza, X. Chen, S. a. Radford, J. R. Cohen, M. E. Marder, K. Kartavenka, P. B. Ryan, D. B. Barr. Biological Matrix Effects in Quantitative Tandem Mass Spectrometry-Based Analytical Methods: Advancing Biomonitoring. Crit. Rev. Anal. Chem., 2016, 46, 93–105.

[56] H. Trufelli, P. Palma, G. Famiglini, A. Cappiello. An overview of matrix effects in liquid chromatography-mass spectrometry. Mass Spectrom. Rev., 2011, 30, 491–509.

[57] R. King, R. Bonfiglio, C. Fernandez-Metzler, C. Miller-Stein, T. Olah. Mechanistic investigation of ionization suppression in electrospray ionization. J. Am. Soc. Mass Spectrom., 2000, 11, 942–950.

[58] FDA. “Guidance for Industry Bioanalytical Method Validation,” Available at: http://www.fda.gov/downloads/Drug/Guidances/ucm070107.pdf, 2001.

[59] B. K. Matuszewski, M. L. Constanzer, C. M. Chavez-Eng. Strategies for the Assessment of Matrix Effect in Quantitative Bioanalytical Methods Based on HPLC−MS/MS. Anal. Chem., 2003, 75, 3019–3030.

43

[60] B. Petrie, J. Youdan, R. Barden, B. Kasprzyk-Hordern. Multi-residue analysis of 90 emerging contaminants in liquid and solid environmental matrices by ultra-high-performance liquid chromatography tandem mass spectrometry. J. Chromatogr. A, 2016, 1431, 64–78.

[61] E. Archer, B. Petrie, B. Kasprzyk-Hordern, G. M. Wolfaardt. The fate of pharmaceuticals and personal care products (PPCPs), endocrine disrupting contaminants (EDCs), metabolites and illicit drugs in a WWTW and environmental waters. Chemosphere, 2017, 174, 437–446.

[62] R. Bonfiglio, R. King, T. Olah, K. Merkle. The effects of sample preparation methods on the variability of the electrospray ionization response for model drug compounds. Rapid Commun. Mass Spectrom., 1999, 13, 1175–1185.

[63] H. Stahnke, T. Reemtsma, L. Alder. Compensation of matrix effects by postcolumn infusion of a monitor substance in multiresidue analysis with LC−MS/MS. Anal. Chem., 2009, 81, 2185–2192.

[64] S. E. Bear, M. T. Nguyen, J. T. Jasper, S. Nygren, K. L. Nelson, D. L. Sedlak. Removal of nutrients, trace organic contaminants, and bacterial indicator organisms in a demonstration-scale unit process open-water treatment wetland. Ecol. Eng., 2017, 109, 76–83.

[65] J. T. Jasper, Z. L. Jones, J. O. Sharp, D. L. Sedlak. Nitrate Removal in Shallow, Open-Water Treatment Wetlands. Environ. Sci. Technol., 2014, 48, 11512–11520.

[66] J. T. Jasper, D. L. Sedlak. Phototransformation of wastewater-derived trace organic contaminants in open-water unit process treatment wetlands. Environ. Sci. Technol., 2013, 47, 10781–90.

[67] J. T. Jasper, Z. L. Jones, J. O. Sharp, D. L. Sedlak. Biotransformation of trace organic contaminants in open-water unit process treatment wetlands. Environ. Sci. Technol., 2014, 48, 5136–44.

[68] N. Bargmann-Leyder, C. Sella, D. Bauer, A. Tambute, M. Caude. Supercritical Fluid Chromatographic Separation of .beta.-Blockers on Chyrosine-A: Investigation of the Chiral Recognition Mechanism Using Molecular Modeling. Anal. Chem., 1995, 67, 952–958.

[69] Waters Corporation. “An Overview of the Principles of MSE, The Engine that Drives MS Performance,” 2011.

[70] 96/23/Ec Commission Decision. 96/23/EC COMMISSION DECISION of 12 August 2002 implementing Council Directive 96/23/EC concerning the performance of analytical methods and the interpretation of results (notified under document number C(2002) 3044)(Text withEEA relevance) (2002/657/EC). 96/23/Ec Comm. Decis., 2002, 29.

[71] M. W. F. Nielen, M. C. van Engelen, R. Zuiderent, R. Ramaker. Screening and confirmation criteria for hormone residue analysis using liquid chromatography accurate mass time-of-flight, Fourier transform ion cyclotron resonance and orbitrap mass spectrometry techniques. Anal. Chim. Acta, 2007, 586, 122–9.

[72] E. L. Schymanski, J. Jeon, R. Gulde, K. Fenner, M. Ruff, H. P. Singer, J. Hollender. Identifying Small Molecules via High Resolution Mass Spectrometry: Communicating Confidence. Environ. Sci. Technol., 2014, 48, 2097–2098.

[73] B. Rochat. Proposed Confidence Scale and ID Score in the Identification of Known-Unknown Compounds Using High Resolution MS Data. J. Am. Soc. Mass Spectrom., 2017, 28, 709–723.

44

[74] J. Tan. Development of an Analytical Method for Identification of Drug Substances and Their Major Metabolites in Raw Wastewater Samples by UPLC-Q-TOF, Uppsala University, 2016.

[75] H. R. Liang, R. L. Foltz, M. Meng, P. Bennett. Ionization enhancement in atmospheric pressure chemical ionization and suppression in electrospray ionization between target drugs and stable-isotope-labeled internal standards in quantitative liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 2003, 17, 2815–2821.

[76] L. Nováková, M. Rentsch, A. Grand-Guillaume Perrenoud, R. Nicoli, M. Saugy, J. Veuthey, D. Guillarme. Ultra high performance supercritical fluid chromatography coupled with tandem mass spectrometry for screening of doping agents. II: Analysis of biological samples. Anal. Chim. Acta, 2015, 853, 647–659.

[77] V. Desfontaine, L. Nováková, F. Ponzetto, R. Nicoli, M. Saugy, J. L. Veuthey, D. Guillarme. Liquid chromatography and supercritical fluid chromatography as alternative techniques to gas chromatography for the rapid screening of anabolic agents in urine. J. Chromatogr. A, 2016, 1451, 145–155.

[78] R. a Coe, J. O. Rathe, J. W. Lee. Supercritical fluid chromatography-tandem mass spectrometry for fast bioanalysis of R/S-warfarin in human plasma. J. Pharm. Biomed. Anal., 2006, 42, 573–80.

[79] T. Berg, L. Kaur, A. Risnes, S. M. Havig, R. Karinen. Determination of a selection of synthetic cannabinoids and metabolites in urine by UHPSFC-MS/MS and by UHPLC-MS/MS. Drug Test. Anal., 2015, n/a-n/a.

[80] V. Desfontaine, F. Capetti, R. Nicoli, T. Kuuranne, J. Veuthey, D. Guillarme. Systematic evaluation of matrix effects in supercritical fluid chromatography versus liquid chromatography coupled to mass spectrometry for biological samples. J. Chromatogr. B, 2018, 1079, 51–61.

[81] C. Foulon, P. Di Giulio, M. Lecoeur. Simultaneous determination of inorganic anions and cations by supercritical fluid chromatography using evaporative light scattering detection. J. Chromatogr. A, 2018, 1534, 139–149.

[82] D. Wolrab, P. Frühauf, C. Gerner. Direct coupling of supercritical fluid chromatography with tandem mass spectrometry for the analysis of amino acids and related compounds: Comparing electrospray ionization and atmospheric pressure chemical ionization. Anal. Chim. Acta, 2017, 981, 106–115.

[83] V. Pauk, T. Pluháček, V. Havlíček, K. Lemr. Ultra-high performance supercritical fluid chromatography-mass spectrometry procedure for analysis of monosaccharides from plant gum binders. Anal. Chim. Acta, 2017, 989, 112–120.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Pharmacy 252

Editor: The Dean of the Faculty of Pharmacy

A doctoral dissertation from the Faculty of Pharmacy, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofPharmacy. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Pharmacy”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-346164

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018