I
Separation and determination of selected organic pharmaceuticals
in waters by means of natural flat membranes, GC, HPLC and
mass spectrometry
Der Fakultät für Naturwissenschaften
Department Chemie
der Universität Paderborn
zur Erlangung des Grades eines
Doktors der Naturwissenschaften
Dr. rer. nat.
vorgelegte Dissertation
von
Mahmoud Bataineh, M.Sc. in Chemistry
aus Irbid / Jordanien
Paderborn 2005
II
The current work has been carried out during September 2001 and December 2004 at the
Institute for Analytical Sciences (ISAS-Dortmund) under supervision from Prof. Dr. M. Grote
University of Paderborn (Facility of Science, Chemistry Department) and Dr. J. Nolte (ISAS-
Dortmund).
1. Referent Prof. Dr. M. Grote
2. Referent Prof. Dr. G. Fels
Eingereicht am 19.01.2005
Tag der mündlichen Prüfung: 15.02.2005
III
Acknowledgements
The present study was carried out at the Institute for Analytical Sciences, Dortmund,
Germany.
I wish to express my sincere gratitude to my supervisor, Prof. Dr. Manfred Grote, for giving
me the opportunity to be a graduate student in his research group and for his scientific
guidance, the critical review of the thesis, continuous support and encouragement.
My thanks and appreciations to my adviser at the Institute for Analytical Sciences, Dr. Jürgen
Nolte, for scientific guidance, a careful reviewing of the thesis, continuous moral support
during the last years and his friendly understanding at all time.
My thanks go to Helma Geltenpoth and Rita Fobbe for their continuous technical assistance
and also Dr. Michael Schilling, Ulrich Marggraf, Walter Nigge, Rolf Bandur and Rüdiger
Kuckuk for the helpful and inspiring discussions
I also wish to extend my thank to all members of the Institute for Analytical Sciences for
welcoming me at the institute, making me feel at home and teaching me some German
culture. It is truly an honour to know each and everyone of them.
I also wish to thank Dr. Birgit Kuhlmann from the Institute for Water Research in Schwerte-
Geisecke for realizing the evaluation of the target drugs in model systems and for providing
real water samples.
Special thanks to Bedia Haciosmanoglu from the University of Paderborn for sharing her
knowledge with me.
Finally I extend my warmest thanks to my wonderful family for being a source of inspiration
and encouragement.
This work was financially supported by the Ministerium für Umweltschutz und Naturschutz,
Landwirtschaft und Verbraucherschutz des Landes NRW, the Ministerium für Schule,
Wissenschaft und Forschung des Landes NRW, and the Bundesministerium für Bildung und
Forschung.
I
Table of contents
1. Introduction and aim of the study ............................................................................ 1
1.1. Pharmaceuticals and their effects on the aquatic environment .................................... 1
1.2. Aim of the study ........................................................................................................... 2
1.3. Introduction of the drugs under study .................................................................... ...... 3
1.4. State of the art in drugs analysis of aquatic samples .................................................... 3
1.5. State of the art with regards to the scope in water treatment ........................................ 4
2. Fate of the selected drugs after medical application.............................................. 10
2.1. Fate of the target drugs in human bodies .................................................................... 10
2.2. Fate of the drugs in aquatic environment.................................................................... 13
3. Results and discussion............................................................................................... 15
3.1. Methodical approach ................................................................................................ 15
3.2. Membrane Extraction............................................................................................... 17
3.2.1. Stability of the drugs stock solutions and monitoring of drug transport..................... 17
3.2.2. Natural flat membranes............................................................................................... 18
3.2.3. Natural flat membrane characteristics......................................................................... 21
3.2.3.1. Membrane stability..................................................................................................... 21
3.2.3.2. Reproducibility of permeation properties .................................................................. 21
3.2.3.3. Membrane surface structure and pore size................................................................. 22
3.2.4. Influences on permeation characteristics ................................................................... 23
3.2.4.1. Influence of general parameters on the permeation process ...................................... 23
3.2.4.2. Influence of additives on the permeation process...................................................... 25
3.2.4.3. Influence of aqueous matrix on the permeation process............................................ 27
3.2.4.4. Influence of additional extraction steps on the permeation process........................... 28
3.2.5. Comparison between natural flat membranes and technical membranes ................... 30
3.2.6. Conclusion.................................................................................................................... 31
3.3. Methods development based on LC/MS and GC/MS........................................... 32
3.3.1. Real water sample preparation................................................................................... 32
3.3.2. GC/MS ........................................................................................................................ 40
3.3.2.1. Determination of underivatized analytes.................................................................... 40
3.3.2.2. Methylation by diazomethane.................................................................................... 40
3.3.2.3. Fragmentation pattern of the methylated analytes ..................................................... 41
3.3.2.4. Identification by SIM ................................................................................................. 42
3.3.2.5. LOD and LOQ............................................................................................................ 42
3.3.2.6. Calibration................................................................................................................... 43
3.3.3. LC/MS........................................................................................................................ 45
3.3.3.1. Separation characteristics........................................................................................... 45
3.3.3.2. Mass spectrometer parameters ................................................................................... 46
3.3.3.3. Confirmation by MS/MS............................................................................................ 47
3.3.3.4. Analyte fragmentation pattern.................................................................................... 48
II
3.3.3.5. LOD and LOQ............................................................................................................. 53
3.3.3.6. Calibration.................................................................................................................. 54
3.3.4. Validation of the developed methods.......................................................................... 56
3.3.5. Comparison between the methods developed............................................................. 58
3.3.6. Applications of the developed methods for a real surface water sample.................... 58
3.3.7. Conclusion................................................................................................................... 60
3.4. Investigation of the biodegradation of drugs in different model systems............ 61
3.4.1. Purity investigation of the standard drug materials..................................................... 61
3.4.2. Batch experiments....................................................................................................... 66
3.4.2.1. The behaviour of the target drugs ............................................................................... 66
3.4.2.2. The behaviour of the synthesised metabolites............................................................. 67
3.4.3. Column experiments ................................................................................................... 69
3.4.3.1. The behaviour of the drugs.......................................................................................... 70
3.4.4. Attempts to identify of degradation products.............................................................. 72
3.4.4.1. Degradation products of the target drugs under batch conditions............................... 72
3.4.4.2. Degradation products of the target drugs under column conditions ........................... 72
3.4.4.3. Degradation products of the synthesised metabolites under batch condition ............. 72
3.4.5. Conclusion.................................................................................................................... 80
4. Summary and conclusions........................................................................................ 81
5. Experimental ............................................................................................................. 83
5.1. Chemicals and materials.............................................................................................. 83
5.2. Sample preparation...................................................................................................... 83
5.3. Instrumentation parameters......................................................................................... 85
5.4. The conditions of the batches and biofilm reactors…................................................. 87
6. References.................................................................................................................. 90
7. Publications and presentations of the present work.............................................. 98
III
Abbreviations
AA Ammonium acetate
9-AC 9 Hydroxymethyl-10-carbamoyl acridan
Ac-SFM N-4-Acetyl-sulfamethoxazole
API Atmospheric pressure ionisation
BSA N,O-Bis(trimethylsilyl)acetamide
C-18 Octadecasilane
CA-HA Carboxyhydratropic acid
CBZ Carbamazepine
CBZ-2-OH 2-Hydroxy-carbamazepine
CBZ-3-OH 3-Hydroxy-carbamazepine
CBZ-10-OH 10, 11-Dihydro-10-hydroxycarbamazepine
8-CCA 8-Chlorocarbazole-1-aceticacid
β-CD β-Cyclodextrine
CH2N2 Diazomethane
CID Collision induced dissociation
CX-IBU Carboxyibuprofen
DCF Diclofenac
Diol-CBZ 10,11-Dihydro-10,11-dihydroxycarbamazepine
DW Distilled water
EDTA Ethylenediaminetetraacetic acid
EI Electron impact ionization
EP-CBZ 10,11-Dihydro-10,11-epoxycarbamazepine
ESI Electrospray ionization
EtOAC Ethyl acetate
GC Gas chromatography
Glucu-SFM N-1-Glucuronide sulfamethoxazole
GW Ground water
HO13 Hohlohsee13
HPLC High performance liquid chromatography
HUS Humic substances
IBU Ibuprofen
IDL Instrument detection limit
IMINO Iminostilbene
i.d. Internal diameter
ISAS Institute for analytical sciences
Ka Dissociation constant
Kow Distribution coefficient octanol/water
LC Liquid chromatography
Lichrolut EN Ethylvinylbenzene-di-vinylbenzene-copolymer
LOQ Limit of quantitation
Me Methyl group
MDL Method detection limit
MeOH Methanol
Min Minute
MS Mass spectrometry
MS/MS Tandem mass spectrometry
MWCO Molecular weight cutoff
m/z Mass to charge ratio
n Number of samples
IV
n.d. Not detected
NIST National Institute of Standards & Technology (USA)
Oasis HLB Poly (divinylbenzene-co-N-vinylpyrrolidone)
OECD Organization for Economic Co-operation and Development
OH-DCF 4`-Hydroxy-diclofenac
OH-IBU 2-Hydroxyibuprofen
PCB 169 3,3',4,4',5,5'-hexachlorobiphenyl
PEG Polyethylene glycol
PFBB Pentafluorobenzyl bromide
PH The negative logarithm of the hydrogen ion (H+) concentration
PhACs Pharmaceutically active compounds
PTFE Poly (tetra fluoro ethylene)
r2 Regression coefficient
R.I. Refractive index
RSD Relative standard deviation
RPM Rotation per minute
Rt Retention time
SDS Sodium dodecylsulfate
SFM Sulfamethoxazole
SIM Selected ion monitoring
SPE Solid-phase extraction
SRM Selected reaction monitoring
STD Standard
STP Sewage treatment plants
SW Surface water
SWM Surface water affected by the run off of a wastewater sewage plant
TIC Total ion current
TMS Trimethylsilyl group
UV Ultraviolet
VM Venner Moor
V/V Volume/volume
W/W Weight/weight
WW Wastewater
1
1. Introduction and aim of the study
It was reported in the last decade in many investigations about the emerging environmental
residuals in aquatic compartments, especially those of toxic/carcinogenic pesticides and
industrial intermediates displaying persistence in the environment. Other diverse groups of
bioactive chemicals receiving comparatively little attention as potential environmental
pollutants include the pharmaceuticals and active ingredients in personal care products [1-6].
In this work pharmaceutical compounds and their metabolites are collectively termed
pharmaceutically active compounds (PhACs), though not all metabolites must be active.
The occurrence of the PhACs in the aquatic environment has been investigated in several
studies in many countries. More than 80 PhACs from various prescription classes have been
detected up to the µg/L level in sewage, surface water and ground water [1].
1.1. Pharmaceuticals and their effects on the aquatic environment
Sources and pathways
Depending on their use, pharmaceuticals enter the environment on different pathways, as
outlined in (Fig. 1.1) [7]. Drugs applied by human are excreted via urine and faeces as a
mixture of metabolites and/or as the unchanged substance to enter the sewer system. The
waste drugs unused are assumed to be disposed of in the sewer system as surplus medical
substance.
Subsequently, they are released via the effluents of sewage treatment plants (STPs) into the
aquatic environment. The proportion of a drug that is retained in sewage treatment either due
to transformation or by adsorption to sludge strongly depends on its chemical structure and
physico-chemical properties, but also on the specific conditions within the respective plant.
Water temperature, residence times corresponding to flow rates, dilutions with rainwater and
sludge age (and thus adaptation of microbial communities) were found to have influence on
elimination efficiencies [8-9]. If the drugs are not eliminated in STPs, they may enter the
aquatic environment and eventually reach drinking water Fig. 1.1. Observed elimination rates
ranged from more than 80 % for ibuprofen to less than 10 % for carbamazepine [2,10].
Comparison of the elimination rates of different municipal STPs gives no reliable results due
to varying influent concentrations and operating parameters. [11]
In many cases, veterinary pharmaceuticals are directly released into the environment by its
use in agriculture, the dispersion of manure from treated livestock on fields or the therapeutic
treatment of livestock on meadows [1,7,106].
Effects
Recent European studies have revealed the presence of a wide array of non-hormonal, non-
antibiotic pharmaceuticals and antiseptics in surface waters, STP effluents and even in
drinking water, often at a level comparable to traditionally recognized “priority pollutants”
[1].
From these studies, we can recognize the magnitude of the problems caused by an increase of
the concentrations of some non-degradable PhACs in aquatic environment during the last
years. Little information is available on the long-term effect of the active substance on
organisms in the aquatic and terrestrial environment [1]. High concentrations of some
compounds, i.e. in the mg/L range have been found to produce effects in environmental
2
organisms. However, an effect on Daphnia, algae and bacteria has also been demonstrated
using low concentration in chronic tests [1].
Since 1997 the interest in the occurrence and behaviour of pharmaceuticals in the aquatic
environment has significantly increased [2-3, 6-7, 12-14]. One motivation for this attention is
the fact that these chemicals are designed to trigger specific biological effects. Thus, they can
be expected to interfere with the respective receptors, enzymes or hormonal systems of
unintentionally exposed organisms.
In the case of antimicrobial agents, the possible maintenance and spread of bacterial
resistance is of major concern. In industrialized countries most human use antimicrobials and
other pharmaceuticals, which reach the aquatic environment unchanged or transformed,
mainly via discharge of effluent from municipal STPs. The residual concentrations of these
bioactive compounds in the treated effluents depend on their removal during wastewater
treatment. They can potentially pose a hazard for aquatic organisms if the removal is
incomplete. In addition, exposure via sewage sludge disposal on land could cause a hazard for
soil organisms.
General drug classification
Human medicine based on its medical function is classified in analgesics and anti-
inflammatories, antibiotic/bacteriostatics (antibacterial drugs), antiepileptics, beta-blockers,
blood lipid regulators, contrast media and cytostatic drugs.
1.2. Aim of the study
To solve environmental problems sensitive analytical methods based on an enrichment step,
chromatographic separation and mass spectrometric determination with and without
derivatization should be developed. In this study carbamazepine (CBZ), diclofenac (DCF),
ibuprofen (IBU) and sulfamethoxazole (SFM) were chosen based on their high application
quantity in medicine and high concentrations found in aquatic environment in previous
studies and further criteria are their great varieties of physical and chemical characteristics,
such as pKa value and polarity. That means, for analytical determination and quantification
mass spectrometric methods must be coupled to both a GC and a LC separation. So the
developed methods should allow application for many residues of drugs considering their
features.
Specifically, the aims of the study were:
1. Within the enrichment procedures the applicability of particular natural flat membranes
from animal intestines should be tested to discharge drug residues in water treatment and
in sample preparation for the analytical determination of selected target drugs and their
metabolites (Table 1.1).
2. Develop a sensitive analytical methods based on mass spectrometry for determination of
the target drugs and some of their main metabolites.
3. The fate of the selected drugs and some of their main metabolite should be investigated in
surface and ground water under certain pilot model systems such as batch and biofilm
reactors applying the developed methods.
3
1.3. Introduction of the drugs under study
In order to get an impression, how many tons of the selected drugs (Table 1.1) are used each
year as human medicine, T. Ternes estimated the sold quantities of the selected
pharmaceuticals in 1997 in Germany in tons [15]: CBZ 80, DCF 75, IBU 180 and SFM 60.
Table 1.2 gives an overview of the concentration of the targeted drugs as reported in the
literature for different water sources in the last years.
Skimming the dates in Table 1.2, notable differences can be observed among diverse studies
and countries. These can be attributed to many reasons, such as the consumption pattern and
rates, elimination rates in different STPs, operating parameters and seasonal differences.
1.4. State of the art in drugs analysis of aquatic samples
The measurement of trace concentration of pharmaceuticals in the aquatic environment has
paid much attention in the last years, but the problem was only a small number of species can
be analysed because there is a lack in reliable analytical techniques. The conventional
analytical methods often used have a lack in sensitivity and specificity needed to quantify the
subpart per billion concentrations of pharmaceuticals typically present in environmental
samples. Furthermore, many pharmaceuticals containing functional groups which render the
compounds difficult to extract from water or even to measure. In addition, many methods are
subject to significant matrix interference when used to analyse wastewater effluent.
As a result of the widely varying properties of pharmaceuticals no single method or
instrument is capable of measuring all of the compounds of interest. In fact, accurate and
precise measurement of pharmaceuticals and many other organic compounds often require
compromise methods focused on a small number of compounds with similar properties.
Recent improvements in analytical techniques have expanded the compound range that is
amenable to specific identification. Using these improved tools, much has been learned about
the occurrence, fate and transport of organic pharmaceuticals and many other pollutants in the
aquatic environment.
Because water samples, in particular wastewater samples, usually contain a high loading of
organic material and suspended particles, filtration is the first step of sample preparation. The
filtration step is particularly necessary when a subsequent extraction based on solid-phase
extraction (SPE) should be performed, because suspended solids can easily clog the adsorbent
bed.
In fact, there are many techniques used for sample pre-concentration applied in environmental
analysis as liquid-liquid extraction (LLE), solid phase extraction (SPE) and membrane
extraction (ME).
SPE is now the most common sampling technique in environmental, pharmaceutical, clinical,
food and industrial analysis. Over time various sampling formats and sorbents have been
developed to facilitate a convenient processing of different sample types and to extend the
scope of the application.
Silica-based chemically bonded sorbents are the most widely used sorbents for SPE but are
unsuitable for some applications. Silica-based sorbents contain a low concentration of ionised
4
silanol groups capable of retaining basic solutes by an ion-exchange mechanism [16]. Silica-
based sorbents are unstable at extreme pH values pH<2 or >8. Porous polymer sorbents offer
a high potential in solving problems. They are, in general, stable throughout the full pH range
and do not possess ionised silanol groups. Modern porous polymer sorbents are copolymer of
styrene and divinylbenzene processed to enhance their properties for SPE [16]. While with
C18-silicas retention is achieved by van der Waals forces and eventually by hydrogen
bonding between residual silanol groups of the silica sorbent base and functional groups of
the analyte, polymer sorbents additionally offer π-π interactions.
The polymeric sorbents have been successfully used for the extraction of the whole range of
organic contaminants [93-94]. They proved to be especially suitable for medium to high polar
substances, where they showed substantially higher recovery rates than alkyl-silica sorbents
[95-96].
From a chemical point of view, pharmaceuticals comprise a complex variety of structure
increments, often combining moieties of different polarities in one molecule. A common
feature of most substituents of pharmaceuticals is their hydrophilic character. Hydroxy-,
carboxy- and amino groups are frequent constituents of pharmacological active substances,
necessary either for the intended effect or for the transport to the place of action.
The most common methods of quantifying pharmaceuticals in surface and waste waters
involve the use of hyphenated techniques such as GC/MS or HPLC/MS, however, even those
sophisticated methods need efficient sample pretreatment to minimize effects of the sample
matrix and to enrich the analytes [16-17].
For GC methods, it is often necessary to derivatise the compounds prior to analysis. For
HPLC/MS methods, sample clean up or selective extraction may be used to control
interferences from organic matter [92]. For both techniques, MS/MS is often used to control
interferences from the organic matter present in the water samples.
GC/MS is the method of choice for semi-acidic, phenolic and non-polar target analytes with
or without derivatization yielding better separations and lower detection limits. A further
advantage for the EI ionization technique used mostly in GC/MS is the presence of the NIST
database, while other ionization methods such as chemical ionization (CI) or atmospheric
pressure ionization usually results in a quasimolecular ion only. The nature of these ions
depends on the reactant gas in GC-MS or on the eluent/buffer composition and ionization
conditions in LC-MS. In a similar way LC/MS can successfully be used in environmental
analysis of drugs and their metabolites in water resources. This technique provides the
opportunity especially for polar, unstable and high molecular mass compounds. Further
advantages in comparison to GC/MS are the reduced run time and the fact that derivatization
becomes redundant. However, lower resolution and especially the suppression of signals in
the electrospray interface by matrix impurities are responsible for limitation of the
application.
1.5. State of the art with regards to the scope in water treatment
Residues of pharmaceuticals, their active compounds and their metabolites can be found in all
aquatic environments. As a consequence, there is an urgent need to improve the purification
of water and wastewater and to monitor the drug input of the different waterways, particularly
of surface and groundwater by means of sensitive chromatographic methods.
5
Organic contaminants present in municipal wastewater, such as pharmaceuticals, may be
removed or transformed by a variety of mechanisms. In conventional STPs, pharmaceuticals
can be removed by sorption to particles or by biotransformation [101]. Sorption of
pharmaceuticals to particles present in wastewater treatment plants can occur either via
hydrophobic or electrostatic interactions, e.g. ion exchange, surface complexation [101]. For
hydrophobic interaction, the octanol/water partition coefficient is a good predictor of the
affinity of the compound for the solid phase. Under the condition encountered in conventional
wastewater treatment plants, only those compounds with an octanol/water partition coefficient
greater than approximately 100 will be removed to an appreciable degree [18]. A few
pharmaceuticals meet this criterion. Therefore, we do not expect a substantial removal of
pharmaceuticals by this mechanism. Sorption of pharmaceuticals via other interactions
usually requires the presence of acidic, phenolic, or amino functional groups. Although many
of the pharmaceuticals contain such functional groups, removal via these mechanisms is not
expected to be significant under the conditions encountered in municipal wastewater [101].
Removal of pharmaceuticals also can occur via biotransformation. Available data from full-
scale STPs suggest that certain compounds, such as IBU, are readily degraded while other
compounds, such as CBZ are removed to a much smaller extent [2].
Conventional treatment systems require a large amount of space, a long treatment time and
are often unable to produce effluent that meets quality levels needed for discharge because a
lot of organic compounds are not eliminated during the cleaning process. Therefore, a need
for other technologies is obviously which are more efficient than the standard primary and
secondary treatment processes currently utilised.
Further improvement of water quality after conventional secondary treatment is achieved
through tertiary treatments such as sand or activated carbon filtration, nitrification /
denitrification, coagulation flocculation, membrane filtration, electrodialysis and reverse
osmosis [99,100].
Though many separation procedures both for technical water treatment and for analytical
purpose as well have been applied, an overall satisfying method is still missing. Meanwhile,
technical solid membranes have led to a standard method for water cleaning processes. But
synthetic flat membranes with a favourable pore size often have problems with clogging
caused by particles or dissolved macromolecular substances [99]. Moreover, the operation
cost in a technical way is still substantial.
Therefore, this work focused on the investigation of other kind of membranes for the
depletion of target drugs in water treatment and for sample preparation prior to laboratory
analysis.
6
Sewage
Treated sewage water
Soil
Ground water
Drinking water
Surface water
STP-sludge
Excretion
Waste disposal
Veterinary
Fish farms Excretion
Pharmaceuticals
Human
STP
Sewage
Treated sewage water
Soil
Ground water
Drinking water
Surface water
STP-sludge
Excretion
Waste disposal
Veterinary
Fish farms Excretion
Pharmaceuticals
Human
STP
Fig. 1.1: Sources, distribution and sinks of pharmaceuticals in the environment [7]
7
Table 1.1: pKa, logKow, structure and application of selected pharmaceuticals and some of their main metabolites
Analyte Abbreviation pKa1/ pKa2 Log Kow[98] Chemical structure Application
1. Carbamazepine
CBZ
13.94[98]
2.67 N
ONH2
Antiepileptic
2. 10,11-Dihydro-10,11-dihyd-
roxycarbamazepine
Diol-CBZ*
-
-
N
ONH2
OH OH
-
3. Diclofenac
DCF
4.15[97]
5.44 N
HCl
HOOC
Cl
Anti-inflammatory
4. Ibuprofen
IBU
4.91[97]
3.72
CH3
COOH
CH3CH3
Anti-inflammatory
5. 2-Hydroxyibuprofen
OH-IBU*
-
-
CH3
COOH
CH3CH3
OH
-
6. Sulfamethoxazole
SFM
5.7[20]
1.8[21]
-0.38
S
O
O
N
O
NCH3
H
NH2
Antibiotic
7
8
*: Synthesised at the University of Paderborn [M. Grote, M. Borges, K. Borchert, unpublished]
Analyte Abbreviation pKa1/ pKa2 Log Kow Chemical structure Application
7. N-4-Acetylsulfamethoxazole
Ac-SFM*
5.0 [20]
- S
O
O
N
O
NCH3
N
CH3
O
H
H
-
8. N-1-Glucuronide-
sulfamethoxazole
Glucu-SFM*
-
-
O
S
O
O
N
O
N
CH3
NH2
OH
OH
OH
COOH
-
8
9
Table 1.2: Survey of the concentrations of target pharmaceuticals and some of their main
human metabolites detected in different water sources as reported in literature
WW : wastewater
SWM : surface water adjacent to discharge of effluents from STP
GW : ground water
TW : tap water
EP-CBZ : 10,11-dihydro-10,11-epoxycarbamazepine
CBZ-2OH : 2-hydroxycarbamazepine
CBZ-3OH : 3-hydroxycarbamazepine
CBZ-10OH : 10,11-dihydro-10- hydroxycarbamazepine
n.d. : not detected
* : [22, 24-29, 81]
Analyte
WW-Influent
[µg/L]
Effluent-STP
[µg/L]
SWM
[µg/L]
Rivers
[ng/L]
GW
[ng/L]
TW
[ng/L]
DCF
3.02[6]
0.359-28.4[*]
0.81[2]
0.026-0.194[*]
489[31]
2298[34]
20-150[33]
150[2]
1-6[31]
IBU 3.4[22]
1.5[6]
1.885-24.6[*]
0.37[2]
0.064-0.790[*]
139[31]
n.d.-80[33]
70[2]
OH-IBU 0.92[6]
CBZ 6.3[22]
0.368[24]
0.635[6]
0.426[24]
2.1[2]
0.020-0.650[*]
250[30]
30-250[33]
250[2]
0.5-7.8[7]
Diol-CBZ 1.571[24] 1.325[24] 0.002-0.002[*]
EP-CBZ 0.047[24] 0.052[24] n.d.[*]
CBZ-2OH 0.121[24] 0.132[24] n.d.[*]
CBZ-3OH 0.094[24] 0.101[24] n.d.[*]
CBZ-10OH 0.0085[24] 0.0093[24] n.d.[*]
SFM 0.243-0.871[22]
343[23]
0.008[*]
0.352[23]
0.099[*] 1000[32]
Ac-SFM 0.518[23] 0.082[23]
10
2. Fate of the selected drugs after medical application
The fate of the drugs from medical applications should be evaluated because metabolism can
lead to produce new and possibly more toxic species [36]. Drug metabolites have special
importance as environmental pollutants, because they are known to be the main excretion
products of most active pharmaceuticals. A few data are available in literature concerning the
fate and effects of the drugs after the medication.
To answer the question for the fate of the drugs, we have to consider different pathways. First
in the human, the major route in human metabolism leads to a series of compounds in varying
concentrations [37]. Other drugs have one or two major metabolic pathways that dominate
their metabolism, but several minor pathways can produce at least a metabolite too. After
ingestion most drugs undergo substance-specific metabolisation distinguished between phase
I and phase II metabolites. Phase I reactions usually include oxidation, reduction or
hydrolysis, and the products are often more reactive and sometimes more toxic than the
respective parent compounds [7]. Phase II reactions involve conjugation mainly with
glucuronic or sulfuric acid, but also with acetic acid, glutathion and taurine. Both, phase I and
phase II metabolisation renders the parent compound more water soluble [38]. While phase I
metabolites may also possess a pharmacological activity that sometimes is even higher than
that of the parent drug [39], phase II metabolites are usually inactive. During sewage
treatment and in manure cleavage of the conjugates was observed [7]. Secondly in water
environments, the degradation might be caused by enzymatic activities, hydrolysis or
photodegradation. An other possibility for the metabolism could happen during the biological
treatment in the STP induced by biodegradation as described in pilot systems for IBU by C.
Zwiener et al. [35].
2.1. Fate of the target drugs in human bodies
CBZ
Thirty three metabolites of CBZ have been identified from human and rat urine [40]. The
main metabolic pathway of CBZ is oxidation to EP-CBZ, then hydration to Diol-CBZ and
conjugation of Diol-CBZ with glucuronide (Fig. 2.1). The hydrolysis of EP-CBZ to Diol-CBZ
is catalysed by microsomal epoxide hydrolase [41]. Some pathways include the oxidation to
9-hydroxymethyl-10-carbamoyl acridan (9-AC), CBZ-2OH, CBZ-3OH and CBZ-10OH [42-
43].
11
N
ONH2
OH N
ONH2
OH
N
ONH2
N
H
O
NH2
CH2OH
N
ONH2
O
N
ONH2
OH OH
(minor)
(major)
(80 % of EP-CBZ)
40 % 25 %
CBZ
CBZ-3OH CBZ-2OH
EP-CBZ
Diol-CBZ
9-AC
Fig. 2.1: Major pathways of the oxidative metabolism of CBZ in human [43]
DCF
The major metabolite from DCF in humans and rats is 4`-hydroxy-DCF (OH-DCF) (40%)
[44]. Additionally, 10-20 % of each 3`-hydroxy-, 5-hydroxy-, and 4`,5-dihydroxy-DCF (Fig.
2.2) and 3`-hydroxy-4`-methoxy-DCF, furthermore an acyl glucuronide species was identified
[45-46].
IBU
The metabolism of IBU in the human body is well known from pharmaco-kinetic studies.
Main excretion products including possible conjugates are IBU (15 %), OH-IBU (26 %),
carboxyibuprofen (CX-IBU) (43 %) and carboxyhydratropic acid (CA-HA) in minor amounts
[47-50] (Fig 2.3).
SFM
SFM is metabolised in the human body and about 50-60 % of the applied dose is excreted as
the inactive metabolite Ac-SFM, 15 % as the conjugate Glucu-SFM and only 15-20 % as the
unchanged compound [51-55] (Fig. 2.4).
12
N
HCl
Cl
HOOC
N
HCl
Cl
HOOC
OH
N
HCl
Cl
HOOC
OH
N
HCl
Cl
HOOC
OH
N
HCl
Cl
HOOC
OH
OMe
N
HCl
Cl
HOOC
OH
OH
III
III IV
VVI
I: Diclofenac
II: 4`-Hydroxy-diclofenac
III: 5-Hydroxy-diclofenac
IV: 3`-Hydroxy-diclofenac
V: 4`, 5-Dihydroxy-diclofenac
VI: 3`-Hydroxy-4`-methoxy-diclofenac
Fig. 2.2: The major oxidative metabolism products of DCF in urine [56]
CH3CH3
COOH
CH3
HOOC
CH3
COOH
CH3CH3
OH
CH3
COOH
CH
CH3
COOH
CH3
COOH
HCH3
COOH
CH3H
COOH
CA-IBU
IBU
OH-IBU
CA-HA
Further degradation
(R)-(-)-IBU (S)-(+)-IBU
inactive pharmacologically active
Fig. 2.3: Major pathways of the oxidative metabolism of IBU in human [57]
13
S
O
O
N
O
NCH3
NH2
H
S
O
O
N
O
NCH3
N
CH3
O
H
H
O
S
O
O
N
O
N
CH3
NH2
OH
OH
COOH
OH
Ac-SFM
SFM
Glucu-SFM
Fig. 2.4: Major pathways of the oxidative metabolism of SFM in human [58]
2.2. Fate of the drugs in aquatic environment
DCF
When DCF entering the water sources more than 90 % are degraded photolytically via
formation of 8-chlorocarbazole-1-aceticacid (8-CCA) and carbazole-1-acetic acid (Fig. 2.5)
[59-60]. The two compounds are not reported as human metabolites [61], however, they were
previously identified as photolysis products of DCF in buffer solutions [62-63]. 8-CCA is
considered to be the initial, cyclized dehydrochlorination product of DCF, and carbazole-1-
acetic acid is a further photoreduction (i.e. dechlorination) product, in which chlorine is
replaced by hydrogen. When the photolysis of DCF was carried out in pure water without a
H-source (i.e. methanol), carbazole-1-acetic acid was not formed. Instead, another major
product is formed, which was identified by mass spectra as hydroxycarbazole-1-acetic acid
[60].
DCF showed in the laboratory experiment that there is only a negligible adsorption onto
sediment particles and so it is not surprising that DCF was not detected in the sediments of
lakes [60].
N
RCOOH
H
Fig. 2.5: The major DCF photodegradation products: (R: -H: carbazole-1-acetic acid,
-Cl: 8-CCA, -OH: 8-hydroxycarbazole-1-acetic acid)
14
IBU
Under environmental conditions, this compound has different transformation kinetics.
Stumpf et al. [64] found that the excretion pattern is hardly changed on the way: Major
changes occurred during biological treatment of the activated sludge in the STP and only
slightly during primary pre-clarification. CX-IBU was almost quantitatively eliminated,
while OH-IBU was the dominant compound in STP effluents and rivers.
This indicates that OH-IBU is the most stable of the three compounds under these conditions
(if it is not continuously formed from IBU or from conjugate cleavage). Zwiener et al. [35]
reported that OH-IBU was formed from IBU under aerobic conditions in activated sludge,
while CA-HA was formed under anaerobic conditions. In both cases, these transformation
products did not add up to more than 10 % of the initial IBU concentration, suggesting that
the major amounts in sewage can be derived from human excretion.
For IBU, the direct phototransformation can be neglected because this compound does not
absorb sunlight [65].
CBZ
CBZ is ubiquitously present in the aquatic environment [6]. However, the principle reason for
the ubiquitous high CBZ concentration is the extremely low removal rate in municipal STP.
Investigation of influent and effluent samples from different municipal STPs have shown that
it is not significantly removed (less than 10 %) during sewage treatment [6]. Different field
studies have shown that CBZ are not attenuated during bank infiltration [6]. This explains
why CBZ has been detected in a number of groundwater samples at a maximum concentration
up to 1.1 µg/L [6] and was also found with a concentration of 30 ng/L in drinking water [66].
However, as revealed by pharmacokinetic data only 1-2 % of CBZ is excreted unmetabolized.
Glucuronide-conjugates can presumably be cleaved in sewage and STP and thus increase the
environmental concentrations.
Because of its low biodegradability and chemical properties, photodecomposition becomes
significant as a pathway of natural elimination. Despite the absorbance of sunlight by CBZ,
Andreozzi et al. [67] observed only a slight elimination through phototransformation. On the
other hand, Frimmel et al. observed that the presence of natural organic matter (HO19)
increased the photochemical degradation rate of CBZ [30].
SFM
It was reported that SFM could be removed up to 60 % during the biological step in a
municipal STP [68]. Heberer et al. [6] observed an efficient removal of various antibiotic and
bacteriostatic drugs during bank filtration. In principal, SFM can be removed by traditional
UV-irradiation as reported by Thiemann et al. [69].
There are only a few information available in literature about chemical behaviour and the fate
of SFM in the aquatic environment [1].
15
3. Results and discussion
3.1. Methodical approach
This study is divided into two main parts:
In the first part pieces of animal intestines were tested as natural membrane for separation and
enrichment of drugs from water. In the second part analytical methods were developed based
on enrichment steps and GC/MS and/or LC-ESI/MS determination. Eventually, the methods
should be applied to monitor the biodegradation of drugs in model plants.
Membrane study
Pieces of pig, sheep and cattle intestines were applied in a homemade permeation chamber
device in order to investigate the basic membrane properties, such as membrane stability
under various conditions and reproducibility of the drugs permeation. Furthermore, surface
structure and pore size distribution were characterized. Moreover, direct influences on the
permeation processes across the membrane wall like analyte concentration, surface area,
stirring velocity, pH, temperature and the presence of other water ingredients such as
inorganic salts, chelators, high molecular compounds and extracting materials are of great
interest.
Additionally, the applicability of these natural membranes was tested with different aquatic
matrices such as surface water, ground water, wastewater and sewage sludge water.
Ultimately, comparative study between the applied natural flat membranes and technical
membranes was carried out.
To fulfill this task at first the long-term stability of the selected analytes was to investigate
and then LC/UV method was to develop for the registration of the permeated compounds.
Analytical method development
The second part of the present work will focus on the development and application of
analytical methods for the simultaneous extraction and determination of acidic, neutral and
basic target pharmaceuticals from environmental water samples. The specific approach will
be done in the following steps:
¾ A pre-concentration step based on solid phase extraction applying different
commercially available materials in order to find out the best material suitable for a
wide polarity range of analytes
¾ Investigation in an optimal derivatization in order to render such compounds more
volatility prior to GC/MS analysis
¾ Development of a GC/MS method based on SIM mode and assured by MS/MS for
qualitative and quantitative analysis of the volatile analytes
¾ Development of a LC-ESI/MS method based on MS/MS technique for qualitative and
quantitative analysis of the non-volatile and thermal labile compounds
¾ Testing the influence of water ingredients on the analytical methods (matrix effects)
¾ Validation of both methods by measuring the linearity range, repeatability, accuracy,
limit of detection and limit of quantitation
16
¾ Examination of the developed methods for its intended purpose by analysing samples
of surface, ground, wastewater and sewage sludge
¾ At least the biodegradation of the target pharmaceuticals and some of their synthesized
metabolites will be followed for particular pilot plants such as batch and biofilm
reactors model systems used in the Institute for Water Research in Schwerte-Geisecke
17
3. 2. Membrane Extraction
In this chapter the focus will turned to the possible applications of cheap animal intestine as
natural membranes. The idea was based on such types of membrane used in the last century as
dialysis membrane in medicine. So, the applying potential as membranes will be evaluated.
The major concern in this work was given to some selected organic drugs as representative
model.
3.2.1. Stability of the drugs stock solutions and monitoring of drug transport
As already mentioned, prior to the membrane studies, the stability of the aqueous drug
solutions was investigated. Furthermore, HPLC-UV methods for the determination of the
drugs and some of their metabolites were developed in order to follow permeation tests, both
in feed and permeate side.
At the beginning of the study the stability of the single analytes in the aqueous stock solutions
at different temperatures and pH-values were tested. For that purpose the aqueous stock
solutions (2 mg/L) were stored in a water bath at (50 ºC) for 70 hours. After this treatment
IBU, CBZ and SFM showed nearly no change in the concentration. Whereas, the
concentration of DCF decreased about 20 %. Moreover, a new stock solutions were stored for
6 weeks in the refrigerator at 4 oC and afterwards further 6 weeks at room temperature at ~ 25
oC. CBZ, IBU and SFM showed no decomposition over a period of three months. In contrast,
a loss of concentration of DCF was observed. In the refrigerator the decrease was about 5 %
within one month, afterwards the concentration was constant until the bottles were stored at
room temperature. From that time the concentration decreased continuously to 70 % (Fig.
3.1).
This loss in concentration could be attributed to a partly re-crystallisation of the analyte due to
its low solubility and to the formation of oligomers (dimers up to hexamers) as has been
detected by LC-ESI/MS. It is known that in some cases the source of electrospray can play an
important role in adduct formation [107]. Many attempts were made to separate those
oligomers by liquid chromatography to confirm oligomers formation in solution. Oligomeres
could not be detected in a fresh solution of the same concentration. In our case, the
oligomerization observed might be an effect of aging favoured by higher concentrations.
Therefore, the concentration of the stock solutions should not be higher than 25 mg/L. Also,
they should be stored in the refrigerator in order to decrease the degradation processes. It is
recommended to treat the solutions with ultrasonic at a bath temperature of 40 ºC for 15
minutes before carrying out membrane studies.
Fig. 3.1: DCF stability in water as function of time and temperature (10 mg/L); 0-43 days:
4 ºC, 44-90 days: 25 ºC
40
50
60
70
80
90
100
0 153045607590
Time [day]
Percentage [%]
40
50
60
70
80
90
100
0 153045607590
Time [day]
Percentage [%]
18
Development of HPLC-UV methods
The transport of the analytes was monitored by HPLC and UV-detection at a wavelength of
225 nm. Aliquots were taken from the liquid phases at intervals by means of a micro-litre
syringe. The developed methods for the selected pharmaceuticals and some of their
metabolites are described as shown in the chromatogram (Fig 3.2) and described in section
5.3 (Method I+II). The external calibration curves were built from 6 concentrations (n=3) in a
concentration range of 0.1 - 5 mg/L. In the membrane tests the starting concentration of the
drugs were 10 mg/L and the pH-value was adjusted to 9.5 at the feed side. The permeation
phase was deionized water and the solutions in both chambers were stirred (500 RPM) in
order to avoid a concentration gradient within the chambers.
The different pH values (3 and 10) did not influence the substances SFM and CBZ. In the
other hand, pH 3 had nearly double the response of DCF and IBU as well. Therefore, the
calibration standards should be at the same sample pH level in order to compensate the higher
response.
A)
B)
Fig. 3.2: HPLC-UV chromatogram for the selected pharmaceuticals and some of their
synthesized metabolites, (A: 1. SFM, 2. CBZ, 3. DCF, 4. IBU, B: 1. Glucu-SFM, 2. Diol-
CBZ, 3. OH-IBU, 4. Ac-SFM), Nucleosil-120 C-18 as analytical column,
acetonitrile/NaH2PO4-buffer mobile phase (225 nm)
3.2.2. Natural flat membranes
History
Animal intestines are available on the market and used since ancient times in food industries
as packaging material [70]. Moreover, they were used as membrane for dialysis in the last
century to discharge the blood from toxic substances [71].
19
The principle has been used to design highly efficient artificial kidneys based on dialysis
[102]. In this process the blood from a patient suffering from acute or chronic kidney failure is
passed into dialyser from a connection to one of his arteries. Low molecular weight toxins in
the blood, such as urea, creatinine and uric acid pass a cross the membrane into a dialysate
solution of such composition that the osmotic pressure is the same as that of the blood; the
rate of transport of certain salts is thus controlled. The blood then returns to the patient
through a connection to one of his veins.
Transport mechanism
The oral absorption of a drug administered in a solid dosage form depends on a series of
events which occurs in the intestinal lumen: the release of the drug from the solid dosage
form; dissolution of the drug in the luminal fluids; drug metabolism and transport of the drug
molecules from intestinal lumen to the blood across the intestinal wall. The rate-limiting step
in the absorption of many orally administered drugs is the transport across the intestinal wall.
The small intestinal wall consists of three main layers: the muscularies mucosae, the lamina
propria, and the epithelium [72]. The rate limiting barrier is the intestinal epithelium. In in
vivo systems, drugs are usually transported across the intestinal epithelium by one or more of
the following routes: 1. the passive transcellular route; 2. the passive paracellular route; 3.
active carrier-mediated transcellular routes; 4. the transcytosis route (Fig.3.3) [19].
Most of the transport occur by the transcellular route since the membrane has a surface area
which is over 1000 times greater than the area of the paracellular spaces [73]. Several
molecular properties, such as size, charge, lipophilicity and conformation, could influence the
passive transcellular transport of drugs across intestinal epithelial cells [74].
Fig. 3.3: Drug absorption routes across the intestinal epithelium [19]
Selection background
The transport mechanism through the natural cell wall of the intestines of living animals could
happen in different routes described above, in contrast the separation process using technical
membranes are mainly a function of molecular size and pore size distribution.
The reasons for the investigation to apply natural organic material are based on the splendid
permeation characteristics and the high enrichment factor achieved with living cells in
medicine. Our experiments should reveal, if similar transport properties would be maintained
under in vitro conditions.
Origin, preparation and enrichment device used
Pieces of animal intestines such as from cattle, sheep, and pig were used as membranes.
Fig. 3.4 shows the technical device constructed to test the natural flat-membranes. The
membranes were fixed in a window between the feed and permeation PTFE-chamber with
20
silicon seals. Both chambers were closed with caps to avoid loss of water by evaporation and
the solutions were stirred on both sides by micro magnetic stirrers. Different setups have been
modulated with two different sizes of 150 and 450 mL. The chamber volume was filled up to
4/5 of the nominal volume.
Before the special part of the cattle appendix, called ‘Goldschlägerhäutchen’, was finally
selected for the further tests, other parts of the intestines of cattle, sheep, and pig had been
studied, as example pig and cattle intestines (Fig 3.5). They showed similar permeation
trends, however, Goldschlägerhäutchen was favourable because large pieces are available in
different dry and wet form and faster equilibration time is remarkable.
Goldschlägerhäutchen, normally used as sausage skins, are available dried or wet under
preservation of salt. For membrane application, the salted product has to be watered in order
to get rid of the salt. Afterwards, it can easily be inserted between the silicon seals of the
membrane device though its surface is a little slippery. The fate layers were already separated
from the delivered intestines. The dried trading products have a surface of about 30 x 40 cm,
the wet ones of about 50 x 70 cm.
1. Feed-chamber
2. Permeation window
3. Fixing plates
4. Natural membrane 5. Permeate-chamber
1. Feed-chamber
2. Permeation window
3. Fixing plates
4. Natural membrane 5. Permeate-chamber
Fig. 3.4: Technical device used for natural flat membrane tests
Fig. 3.5: Comparison of drugs permeation between Goldschlägerhäutchen and small intestine
of pig (15.2 cm2; membrane area); 10 mg/L of each pharmaceutical
0
20
40
60
80
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
pig small intestine Goldschlägerhäutchen
0
20
40
60
80
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
pig small intestine Goldschlägerhäutchen
21
3.2.3. Natural flat membrane characteristics
3.2.3.1. Membrane stability
In order to have knowledge how long this membrane remains working under various
conditions, we tested pieces of the membranes by soaking them up to three months in NaOH
solution (pH 10) and HCl (pH 2). These values were selected as extreme values because in the
normal wastewater the pH-value ranges usually between 8 and 9 and because in water
analysis natural water samples are usually acidified to pH 2. The membranes were stable at
pH 9, whereas, in the dilute acid they were destroyed within 3 weeks.
Then these stressed pieces were compared with fresh wet and dry pieces. As shown in Fig. 3.6
there were no significant differences among the different membranes.
Fig. 3.6: Comparison of pH-stressed membranes with fresh wet and dried membranes,
SM = 3 months in basic solution, WM = new wet membrane, DM = new dried membrane,
membrane surface area (8.1 cm2), F: Feed, P: Permeate, Avg: the mean from the four drugs
We found different membrane stability based on the water source. In ground water the
membranes were stable for more than two weeks. Even in sewage sludge water we observed
similar stabilities. Wastewater and wastewater effected surface water destroyed the
membranes within 1-2 days because of their high bioactivities. By adjusting the water to pH ≈
2 the stability can be extended to more one day.
There are two possibilities to overcome this problem: The application of photolysis or
chemical cross-linking. The water matrix was photolysed for different periods of time to
reduce any present bioactivity by using UV-lamp radiation. But this operation elongated the
stability only for one further day. A second possibility of stabilisation is based on cross-
linking with formaldehyde [75]. For this purpose the membrane was placed in a formaldehyde
solution (1 %) for two hours, then washed and dried. Such prepared membranes could extend
their stability up to 10 days.
3.2.3.2. Reproducibility of permeation properties
The reproducibility of the membranes permeation process was tested by series of tests under
the same conditions. After the first run the same membrane piece was used repeatedly with
fresh feed solutions three times. Then the membrane position was changed in such a way that
the surface which contacted at first the feed side was turned to the permeation phase in a
second series of measurements. The results are presented in (Table 3.1). Furthermore,
different pieces of one intestine batch were investigated under the same conditions. It is
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg-SM(F) Avg-SM(P)
Avg-WM(F) Avg-WM(P)
Avg-DM(F) Avg-DM(P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg-SM(F) Avg-SM(P)
Avg-WM(F) Avg-WM(P)
Avg-DM(F) Avg-DM(P)
22
remarkable that the permeation characteristics are nearly independent on different batches as
well of the position of the surface.
Table 3.1: Reproducibility for piece of membrane on both sides
M = membrane, Fr = front side, B = back side (by definition); n = 4
F = feed side, P = permeate side, SD = Standard deviation
3.2.3.3. Membrane surface structure and pore size
Scanning electron microscope studies have been done to get an impression about the structure
of the intestine surfaces (Fig. 3.7). For all wet membranes we observed a somewhat cloudy
structure with little remarkable details. Against that, the surface of the dried pieces showed a
relative strong structuring, resulting from the protein conglutination of different individual
parts. In Fig. 3.7 a pore structure is not obvious.
Because of the effect that about 10 % of a standard humic substance (HUS; HO13), having a
molecular weight > 1000 Dalton, permeated through the membrane within the first 24 hours,
an attempt was started using polyethylene glycol (PEG) with a molecular weight of 1500 and
3500 in order to characterize the pore size. The PEGs needed about 20 hours to reach
equilibrium on both sides (Fig. 3.8). PEG tests were analysed by HPLC and refractive index
detection (RI) as described in section 5.3 (Method III). While, HUS test was measured by
spectrophotometer at 254 nm.
However, no clear information of pore size distribution can be given by these results.
Presumably, the test substances transport through the membrane wall by diffusion as
described in section 3.2.2. Therefore, the question of correct pore size could not be answered
exactly.
Time
[h]
F
[%]
SD
[%]
P
[%]
SD
[%]
3
80.0
3.2
17.3
1.5
M-Fr
6
72.1
2.1
30.9
1.7
24
51.4
3.9
46.4
1.4
27
48.3
2.7
47.9
1.8
3
78.5
6.8
14.2
3.1
M-B
6
67.0
6.7
27.8
4.3
24
49.5
3.5
45.0
3.2
27
49.3
3.7
45.9
3.1
23
Fig. 3.7: Scan electron microscope of the Goldschlägerhäutchen (A) and Sheep-Appendix (B)
Fig. 3.8: Goldschlägerhäutchen pore size characterization using 1 % PEG, 10 mg/L HUS
(HO13) and membrane surface area (8.1cm2), HUS was measured at 254 nm and PEG using
HPLC/RI (Fractogel TSK HW-40S was used as stationary phase and water as mobile phase)
3.2.4. Influences on permeation characteristics
3.2.4.1. Influence of general parameters on the permeation process
The influence of some basic parameters such as membrane size (Fig. 3.9), stirring speed (Fig.
3.10) and temperature (Fig. 3.11) on the permeation of SFM, CBZ, DCF and IBU were
studied in addition to some of their synthesised metabolites; Diol-CBZ, OH-IBU, Ac-SFM
and Glucu-SFM.. The optimum settings were kept for all succeeding tests. Though the
selected compounds differ in structure and polarity, they represent always the same
permeation behaviour as demonstrated by the result in (Fig. 3.14). Moreover, the synthetic
metabolites showed similar trends of permeation. Therefore, the drugs permeation results
were demonstrated by the average values from the all test analytes as one unit.
As expected, the relationship of chamber volume and membrane size was important and
although the temperature influenced the dynamic of the permeation, the stirring velocity was
of particular importance. When increasing the velocity a decrease of the equilibrium time was
observed. A further increase above an optimal value resulted in turbulent flows without
further enhancement. A velocity of 500 rotations per minute was fixed as the practical
optimum.
The influence of different drug concentrations on the permeation process as presented in (Fig.
3.12) is low.
B (10µm
x 500)
A (10µm
x 500)
0
20
40
60
80
100
01020304050
Time [h]
Percentage [%]
PEG 1500 PEG 3500 HUS (1000-3000)
0
20
40
60
80
100
01020304050
Time [h]
Percentage [%]
PEG 1500 PEG 3500 HUS (1000-3000)
24
Fig. 3.9: Influence of membrane surface area in cm2: M-I 3.84,
M-II 15.2, M-III 25
Fig. 3.10: Influence of stirring rate (3.84 cm2)
Fig. 3.11: Influence of temperature (15.2 cm2)
Fig. 3.12: Influence of concentration (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 1020304050607080
Time [h]
Percentage [%]
Avg (F/P)-M-I Avg (F/P)-M-II Avg (F/P)-M-III
0
10
20
30
40
50
60
70
80
90
100
0 1020304050607080
Time [h]
Percentage [%]
Avg (F/P)-M-I Avg (F/P)-M-II Avg (F/P)-M-III
0
10
20
30
40
50
60
70
80
90
100
0 10203040506070
Time [h]
Percentage [%]
Avg (F/ P) 0 RPM
Avg (F/ P) 300 RPM
Avg (F/ P) 660 RPM
0
10
20
30
40
50
60
70
80
90
100
0 10203040506070
Time [h]
Percentage [%]
Avg (F/ P) 0 RPM
Avg (F/ P) 300 RPM
Avg (F/ P) 660 RPM
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg (F/ P) 5°C Avg (F/ P) 25°C Avg (F/ P) 50°C
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg (F/ P) 5°C Avg (F/ P) 25°C Avg (F/ P) 50°C
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Avg (F/ P) 1mg/L
Avg (F/ P) 5mg/L
Avg (F/ P) 10mg/L
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg (F/ P) 1mg/L
Avg (F/ P) 5mg/L
Avg (F/ P) 10mg/L
25
No significant influence of pH on the analyte transport through the membrane could be
observed (Fig. 3.13). The reason might be found in the pH-balance between both chambers,
which is achieved within a very short time. Throughout the most experiments the feed
samples were adjusted within pH 8 - 10, representing the pH range of wastewater.
Fig. 3.13: Influence of pH (15.2 cm2)
3.2.4.2. Influence of additives on the permeation process
To increase or hinder the permeation process many additives were added to the feed and
permeate solutions such as salts, chelating agents (β-CD, EDTA) and surfactants (ionic and
non-ionic).
The results we got from the addition of salt (Fig. 3.14) and chelating agents (Fig. 3.15-16)
referred to no actual variation in permeation kinetics.
Fig. 3.14: Influence of salt; 1% (w/w) Na2SO4 (15.2 cm2)
Fig. 3.15: Influence of chelators; 0.1 % (w/w) EDTA+ 1 % (w/w) Na2SO4 (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time[h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time[h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg (F/ P) pH 4 Avg (F/ P) pH 7 Avg (F/ P) pH 9
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg (F/ P) pH 4 Avg (F/ P) pH 7 Avg (F/ P) pH 9
26
Fig. 3.16: Influence of chelators; 0.5 % (w/w) β-CD (15.2 cm2)
In the presence of other ingredients, e.g. by addition of sodium dodecylsulfate (SDS; ionic) or
polysorbate (non-ionic), the analyte transport was affected adversely (Fig. 3.17-18).
But that depends on the concentration of the surface-active or complex-forming agent; when
the concentration of the extra is > 0.5 % a tangible retardation is the consequence for CBZ
and DCF. However, in the concentration range known to be present in wastewaters the effect
can be neglected and, besides that, it can be attenuated by relative high salt concentrations
(salting-out effect).
Fig. 3.17: Influence of tenside: 2 % (w/w) SDS added to feed side (15.2 cm2)
Fig. 3.18: Influence of complexation agent: 1 % (v/v) polysorbate
added to feed side (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time[h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time[h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P)
DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P)
DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
27
3.2.4.3. Influence of aqueous matrix on the permeation process
Wastewater contains a high spectrum of substances like salts, HUS and other organic and
inorganic compounds. Therefore, we have investigated the behaviour of the drugs in the
presence of HUS (Fig. 3.19). Two HUS fractions were added in different batches; it was the
standard material Hohlohsee13 (HO13, Black Forest area) and Venner Moor (VM, North-
Rhine-Westphalia district).
Fig. 3.19: The influence of the 10 mg/L hydro-colloids HO13 and VM on
pharmaceutical permeation through the natural membrane (8.1 cm2), Avg = average values
Because of their chemical composition the presence of such high molecular weight hydro-
colloids like HUS can lead to different interactions with the analytes such as hydrogen
binding to polar groups, or covalent binding, or only coordination. This again can have a
direct influence on analyte permeation kinetics. Actually this is the fact for HUS of a younger
genesis such it is the case of bog-water from the Venner Moor Fig. 3.19. The lower curve in
the Fig. 3.19 is caused by analytes with polar structure moieties as the carboxylic groups of
DCF and IBU. But in the same figure it is also demonstrated that no significant influence was
observed in the presence of the HUS (HO13), which had reached its final stage of
humification.
Other investigations have been done using real water samples resulting from sludge and
wastewater (Fig. 3.20-21). It couldn’t be found any influence on permeation by filtrated
sludge and wastewaters.
Fig. 3.20: Influence of the filtrated sludge water on the permeation (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg Avg+HO13 Avg+VM
0
10
20
30
40
50
0 5 10 15 20 25
Time [h]
Percentage [%]
Avg Avg+HO13 Avg+VM
28
Fig. 3.21: Influence of the filtrated wastewater (pH = 3.0) on the permeation (15.2 cm2)
3.2.4.4. Influence of additional extraction steps on the permeation process
It is characteristic for permeation processes in a defined volume that only an equilibrium
concentration can be reached between both sides. To improve the recovery, the concentration
gradient must be forced. This can be achieved by removing the analytes from the permeate
chamber using an additional acceptor phase. The target pharmaceuticals could be extract from
the permeate phase as reported by M. Grote et al. [105] by overlaying with water immiscible
liquid phase. The concept of liquid/liquid extraction was carried out by overlaying the
permeate phase a water-immiscible film of an organic solvent such as n-octanol or a mixture
of n-octanol or n-decanol with n-decan in different ratios (Fig. 3.22-23). Furthermore, we
tested the influence of additives on the extraction such as tris(2-ethylhexyl)-phosphate in
order to enhance the selectivity and the efficiency of the extraction processes (Fig. 3.24).
In the system with octanol CBZ was eliminated with 80 %, DCF and IBU with 60 % from the
feed side, while SFM showed no improvement. The system with the solvent mixture increased
the depletion for CBZ of only about 10 %. Additives did not show any remarkable
improvement on the permeation efficiency.
Fig. 3.22: Influence of an additional liquid phase extraction on the permeation
using n-octanol (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
SFM (F/P) CBZ (F/P) DCF (F/P) IBU (F/P)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
29
Fig. 3.23: Influence of an additional liquid phase extraction on the permeation
using n-decan: n-decanol (80:20) (15.2 cm2)
Fig. 3.24: Influence of an additional liquid phase extraction on the permeation
using n-octanol in addition to 2 ml tris(2-ethylhexyl)-phosphate (15.2 cm2)
Another way for the depletion is an extraction by means of solid adsorbent materials. The
most effective procedure is pumping of the permeation solution through a column filled with
charcoal as effective adsorbent materials (Fig. 3.25). The result showed more than 90 % of all
test compounds could be eliminated from the feed chamber. An immediate use of charcoal
filtration without the membrane technique is inappropriate because a clogging of the filters by
particles and high molecular compounds would deteriorate the sorption efficiency. Hence, the
natural membrane has an important role as retardation reactor to prevent the adsorbent
material from the overloaded processes.
Fig. 3.25: Influence of an additional solid phase extraction on the permeation
using charcoal column (15.2 cm2)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F) CBZ(F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM (F) CBZ(F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20
Time [h]
Percentage [%]
SFM (F) CBZ (F) DCF (F) IBU (F)
30
3.2.5. Comparison between natural flat membranes and technical membranes
In order to get an idea where the natural membranes are positioned in comparison with
technical membranes used in dialysis, cellulose ester membranes with a molecular weight cut
off of 1000 Dalton has been tested. An additional question was how they behave in the
presence of HUS of younger genesis (VM)(Fig. 3.26).
Fig. 3.26: Influence of humic substances on the permeation of pharmaceuticals, using a
technical membrane (Cellulose ester; MWCO =1000 Dalton) (8,1 cm2)
In the Fig. 4.26 it can be noticed that CBZ permeated similarly to the natural membrane but
the other pharmaceuticals permeate retard clearly. While the permeation yield of CBZ was
about 45 %, the one of the other substances were less than 25 %. This loss increased
especially in the presence of HUS, which had an effect to all substances of about 10 %.
Another interesting point, demonstrated in Fig. 4.26, is that CBZ permeated relative quickly
within the first 8 h in comparison to SFM, DCF, and IBU. During that time SFM seems to
have no permeation but afterwards the kinetic increased clearly stronger than IBU and DCF.
The influence of the HUS besides the above mentioned interaction with the analytes may be
caused by adsorption onto the membrane surface or a blocking of the pores by HUS/analyte-
adducts [104].
0
10
20
30
40
50
60
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM CBZ DCF IBU
SFM+VM CBZ+VM DCF+VM IBU+VM
0
10
20
30
40
50
60
0 5 10 15 20 25
Time [h]
Percentage [%]
SFM CBZ DCF IBU
SFM+VM CBZ+VM DCF+VM IBU+VM
31
3.2.6. Conclusion
As shown by the single results presented in this chapter, it is obviously that intestines from
animals can work as dialysis membrane. Different types of intestine parts of cattle, sheep and
pig were applied in the original wet and dried form. The permeation is mainly influenced by
the concentration gradient between the both chambers. Therefore, increasing the stirring
velocity and membrane surface area to optimum leads to a decrease of the equilibrium time to
about six hours. pH values, temperature and concentration in both champers showed nearly no
influence in the total permeation process.
Goldschlägerhäutchen was chosen for further investigations because large pieces are available
in different dry and wet form and faster equilibration time is tangible.
The interferences with water ingredients such as chelating agent and surfactants can interfere
the permeation process but only in unusual high concentrations. Additional water matrices
such as humic substances could retard the permeation process of some analytes depending on
their chemical structure and the composition of the matrix as well. To increase the depletion
of the analytes a re-extraction by solid or liquid phases must be performed. Activated carbon
showed to be a superior combination method. A depletion of more than 90 % of all analytes is
feasible.
The behaviour of the studied analytes and in dependence on the test conditions was
investigated with stocks solutions. Within three months, all stocks solutions showed
remarkable high stabilities except DCF, which had a loss of about 30 % from the original
concentration. The loss could be interpretated to a partly re-crystallization or
photodegradation as well. Therefore, the stock solutions should not be higher than 25 mg/L
and should be store in the refrigerator.
In such samples the natural membranes showed high stabilities and reproducible results. But
the stability is decreased tremendously in presence of a bioactivity in the water samples. In
the presence of bioactivity the lifetime of the membranes are not satisfactorily. To overcome
that problem and to increase the lifetime for some more days, the bioactivity should be
decreased by pre-filtration or radiation with UV-light and the membranes should be treated
with formaldehyde solutions.
Different polyethylene glycols and standard fractions of humic substances were applied to
estimate the pore sizes. No clear answer of the pore size could be given because the
permeation through the membrane presumably occur by diffusion as described in theory of
the membrane in vivo system. Then the intestines were compared to technical membrane. In
the presence of humic substances some of the analytes were highly retarded from cellulose
ester dialysis membrane with fixed molecular weight cut off of 1000 Dalton. This result
revealed that the natural membrane used miscellaneous pathways in permeation processes.
As a consequence, possible applications in water treatment are restricted by membrane
stability. Furthermore, analytical enrichment is not postulated due to slow permeation kinetics
and a second extraction step is necessary after membrane permeation for the permeate side.
32
3.3. Methods development based on LC/MS and GC/MS
Besides improved separation and enrichment procedures reliable analytical techniques based
on an enrichment step, chromatographic separation and mass spectrometric determination
with and without derivatization are fundamental. Therefore, in this chapter the attention is
turned to it in the following units.
Though, advantageously in water treatment natural flat membranes are unsuitable for analysis
purpose for two main reasons as point out in previous conclusion: the first, the permeation
velocity is too slow, hence the total time analysis would be extended unfavorably. The
second, a second extraction step is necessary after membrane permeation for the permeate
side. If necessary, a filtration with conventional filter is without problems, because only
maximal sample volume of 1 L is needed.
3.3.1. Real water sample preparation
Sample clean up and extraction
In order to remove suspended material, the aqueous samples were vacuum filtered through
0.45µm cellulose membrane filter.
One intention of this work was to develop a method for the extraction of acidic, hydrophilic,
neutral and basic pharmaceuticals simultaneously from water samples. Generally, a selective
extraction of pharmaceutical compounds from environmental waters is a critical step but in
this study it was especially difficult, because the analytes cover a broad range of polarity.
Ultimately, the final enrichment method could only be the best compromise.
As pointed out in chapter 1, for the extraction of high polar analytes SPE with polymeric
sorbents often proved to be superior to alkyl-bonded silica (e.g. octadecasilane). A variety of
hyper-crosslinked sorbents are commercially available, differing in the degree of linkage,
porosity and surface area. They are either co-polymerisates of styrene and a polar component
(e.g., methacrylate or N-vinylpyrrolidone) or the functional groups are introduced after
polymerisation (e.g., by sulfonation). This functionalisation results in mainly two effects:
improved wetting characteristics for better mass transfer and additional possibilities for
interactions with functional groups of the analytes and thus a higher degree of retention.
At the preliminary stage in this study, a variety of cartridges were investigated such as
octadecasilane (C-18), ethylvinylbenzene-di-vinylbenzene-copolymer (Lichrolut EN) and
[poly(divinylbenzene-co-N-vinylpyrrolidone)] (Oasis HLB).
To avoid contamination of plasticizers which may be caused by the polypropylene martial,
glass cartridges were used exclusively. These cartridges were self-filled with the SPE
materials. The amounts of filling were listed in Table 3.2. The influence of the adsorbent
amount was investigated successively by the recovery. The yields were calculated from 1 L of
ultrapure water (pH ≈ 2) spiked with 1 µg/L each of the analytes by using a HPLC/UV
method which allowed the separation and quantitation of all selected analytes as described in
section 3.2.1. The results are listed in Table 3.2.
Oasis HLB adsorbent was finally chosen as the best phase for our purpose. The structures of
the compounds studied and the pka values of the investigated drugs are shown in Table 1.1.
33
The samples were acidified to pH 2 as water samples are normally stabilised at that pH to
obtain the undissociated form of the analytes. Several parameters especially the elution
solvent, the drying temperature and time were optimised for conditioning the SPE material.
The following conditions were fixed as represented in Fig. 3.27 and described in the
experimental section as well.
To achieve sufficient sensitivity, or to change the solvent for further analysis, the extracts had
to be concentrated by evaporation of the solvent, occasionally several times within the
complete analytical procedure. The volume reduction could be critical and results in losses of
the more volatile compounds, especially of IBU, had to be considered. Therefore, special care
was taken and the sample was not allowed to come to complete dryness.
Table 3.2: Percentage of drug recoveries and relative standard deviation (RSD) obtained with
various SPE cartridges (concentration: 1 µg/L, 1 L; pH = 2 and n = 3), (LC-UV)
Derivatization for GC-MS
The major purposes of analytical derivatization are the enhancement of the volatility of
analytes by decreasing the polarity, for example of OH-, COOH-, NH2- groups or increasing
the thermal stability. In order to render such compounds more volatile prior to GC analysis,
they are mostly esterified, silanated or acetylated using one of the diverse methods reported.
The selected pharmaceuticals in this study are characterised by a wide polarity range; IBU has
an acidic group, SFM, DCF are characterised by both acidic and basic groups in the same
molecule and for CBZ the basic group is dominant.
The best choice for derivatization of acidic pharmaceuticals containing carboxylic moieties is
given by methylation with trimethylsulphonium hydroxide (TMSH) or diazomethane
(CH2N2). Other possibilities besides methylation are acylation with trifluoroacetic anhydride
(TFAA), benzylation with pentafluorobenzyl bromide (PFBB) and silylation with N,O-
bis(trimethylsilyl)acetamide (BSA) [91]. CBZ contains an amide and SFM a sulfonamide and
a primary amine as functional groups which may be derivatised by alkylation, silylation or
acylation reagents depending on their special structure.
Based on qualitative experiments, many reagents from divers described methods were
investigated for analytical derivatization concerning the functional groups of the analytes,
namely esterification [34,85-86,91], alkylation [82,87], silylation [88-89,103] and acylation
[84]. The qualitative results could be illustrated as the following:
Sorbent C-18
(500 mg)
Lichrolut EN
(200 mg)
Oasis HLB
(100mg)
Analyte Recovery
[%]
RSD
[%]
Recovery
[%]
RSD
[%]
Recovery
[%]
RSD
[%]
SFM 22 22 45 7 71 4
CBZ 86 16 97 0 94 0
DCF 98 14 80 4 100 4
IBU 85 3 83 3 95 2
34
1. Methylation with CH2N2: suitable for DCF, IBU, OH-IBU, SFM and Ac-SFM (Fig. 3.28)
2. Benzylation with PFBB: suitable for DCF, IBU, and SFM (Fig. 3.29)
3. Silylation with BSA: suitable for DCF and IBU (Fig. 3.30). CBZ could be silylated by
MSTFA/TMSI/ DTE 1000:2:2 (v/v/w) only in very low yields (Fig. 3.30c)
4. Acylation: different reagents were tested but without satisfying results
CBZ with nearly neutral properties could generally be analysed without derivatization, but it
was decomposed in the injector forming iminostilbene (IMINO) (Fig. 3.31). The
derivatization of Glucu-SFM with diazomethane was proved by LC-ESI/MS. But due to
elevated polarity the analysis with GC/MS was impossible.
The summarized results are presented as shown in (Table 3.3). Benzylation and methylation
are chosen for further quantification method because they cover nearly the whole analytes.
In order to optimise the method, all parameters having direct influence in the yield of the
reaction such as reaction time, reagent volume, reaction temperature and final solvent were
investigated. The final modify methods are set as described in (section 5.2).
Based on the intention to get little by-products, derivatives of low molecular weight, short
reaction times as well as lower detection limit, derivatization with diazomethane was chosen
as the general method. The quantitative results from diazomethane are discussed in detail in
section 3.3.2.
Table 3.3: Summarized results of the derivatization study
+ = suitable; high product yields
- = not-suitable; nearly no reaction products
0 = not tested
Analyte
Reagent
SFM CBZ IBU DCF Ac-
SFM
Glucu-
SFM
OH-
IBU
Methylation + - + + + + +
Benzylation + - + + 0 0 0
Silylation - - + + 0 0 0
Acylation - - - - 0 0 0
35
Fig. 3.27: Sample clean up and extraction scheme
Filtration (0.45 µm)
Adjust to pH 2
Solid-phase Extraction:
Conditioning: 12 ml MeOH, 3 ml EtOAc, 12 mL water pH 2
Loading: at 5 mL/min
Washing: 3 mL water
Drying: N stream
Elution: 4.5 mL MeOH, 1.5 mL EtOAc
2
Solvent reduction to final volume
(300 µL at 40°C)
Dividing the sample
in 2 aliquots
GC/MS analysis
Evaporate to dryness
Redissolve in 200 µL 10 mM AA
buffer pH 4, (10% MeOH)
LC/MS analysis
Evaporate to dryness
Redissolve in 200 µL EtOAc
Spiking with
internal standard PCB 169
Evaporate to dryness
Derivatization
1 mL CH N
22
36
40 60 80 100 120 140 160 180 200 220 240 260 280 300 320
0
50
100
39 51 59
63 76 89
93 107 125
151
164
179
214
242
250 277
309
N
HCl
Cl
H3COOC
Abundance [%]
m/z
B
40 60 80 100 120 140 160 180 200 220 240 260 280 300 320
0
50
100
39 51 59
63 76 89
93 107 125
151
164
179
214
242
250 277
309
N
HCl
Cl
H3COOC
Abundance [%]
m/z
B
40 60 80 100 120 140 160 180 200 220
0
50
100
41 51 59 65 77
91
105
117
131 145
161
177
220
CH3
CH3CH3
COOCH3
Abundance [%]
m/z
A
40 60 80 100 120 140 160 180 200 220
0
50
100
41 51 59 65 77
91
105
117
131 145
161
177
220
CH3
CH3CH3
COOCH3
Abundance [%]
m/z
A
40 60 80 100 120 140 160 180 200 220 240 260 280
0
50
100
39
43 55
65
80
92
98
108
119
140
156
162
188
203
S
O
O
N
O
NCH3
N
CH3H
H
Abundance [%]
m/z
C
40 60 80 100 120 140 160 180 200 220 240 260 280
0
50
100
39
43 55
65
80
92
98
108
119
140
156
162
188
203
S
O
O
N
O
NCH3
N
CH3H
H
Abundance [%]
m/z
C
37
Fig. 3.28: Electron ionization mass spectra of the diazomethane derivative of the IBU (A),
DCF (B), SFM (C), AC-SFM (D) and OH-IBU (E), the measurement was based on GC/MS
using capillary column (HP-5MS), split/splitless injector and gradient temperature program as
described in chapter 5
40 60 80 100 120 140 160 180 200 220 240 260 280 300
0
50
100
39
43
52
55
65
82
92
98
108
119
134
140150
156
161
188
204 230
245
S
O
O
N
O
NCH3
N
CH3
O
CH3
H
Abundance [%]
m/z
D
40 60 80 100 120 140 160 180 200 220 240 260 280 300
0
50
100
39
43
52
55
65
82
92
98
108
119
134
140150
156
161
188
204 230
245
S
O
O
N
O
NCH3
N
CH3
O
CH3
H
Abundance [%]
m/z
D
40 60 80 100 120 140 160 180 200 220
0
50
100
31 41
43
51
59
65 77
91
105
119
128 145
161
178
221
CH3
CH3CH3
O
COOCH3
CH3
Abundance [%]
m/z
E
40 60 80 100 120 140 160 180 200 220
0
50
100
31 41
43
51
59
65 77
91
105
119
128 145
161
178
221
CH3
CH3CH3
O
COOCH3
CH3
Abundance [%]
m/z
E
41030 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390
0
50
100
41 57 91 105
118
145
161
181
343 386
Abundance [%]
m/z
CH3
CH3CH3
O
O
CH2
F
F
F
F
F
A
41030 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390
0
50
100
41 57 91 105
118
145
161
181
343 386
Abundance [%]
m/z
CH3
CH3CH3
O
O
CH2
F
F
F
F
F
A
38
Fig. 3.29: Electron ionization mass spectra of the pentafluorobenzyl derivative of the IBU
(A), DCF (B) and SFM (C)
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480
0
50
100
75 89
151
161
179
214
242
259
277
294 394
475
m/z
Abundance [%]
N
HCl
Cl
O
OC
H2
FF
F
F F
B
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480
0
50
100
75 89
151
161
179
214
242
259
277
294 394
475
m/z
Abundance [%]
N
HCl
Cl
O
OC
H2
FF
F
F F
B
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390 410 430 450
0
50
100
4352
65
80
92
108
119
131
140
156
162
181
188 206 234 272286 306
326
352369
m/z
Abundance [%]
S
O
O
N
O
NCH3
H
N
F
F
F
F
F
H
C
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390 410 430 450
0
50
100
4352
65
80
92
108
119
131
140
156
162
181
188 206 234 272286 306
326
352369
m/z
Abundance [%]
S
O
O
N
O
NCH3
H
N
F
F
F
F
F
H
C
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350
0
50
100
41
45 57
73
91
117
131145
160
191205220
234 263
278
Abundance [%]
m/z
A
CH3
CH3CH3
O
O
Si CH3
CH3CH3
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350
0
50
100
41
45 57
73
91
117
131145
160
191205220
234 263
278
Abundance [%]
m/z
A
CH3
CH3CH3
O
O
Si CH3
CH3CH3
39
Fig. 3.30: Electron ionization mass spectra of the trimethylsilyl (TMS) derivative of the IBU
(A), DCF (B) and CBZ(C)
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390 410 430
0
50
100
45 75 151165
179
214
242
277
352
367
m/z
Abundance [%]
B
N
HCl
Cl
O
OSi
CH3
CH3
CH3
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330 350 370 390 410 430
0
50
100
45 75 151165
179
214
242
277
352
367
m/z
Abundance [%]
B
N
HCl
Cl
O
OSi
CH3
CH3
CH3
50 80 110 140 170 200 230 260 290 320 350 380 410 440 470 500
0
50
100
8495 139
165
193
211 308
m/z
Abundance [%]
C
N
O N
H
Si
CH3
CH3
CH3
50 80 110 140 170 200 230 260 290 320 350 380 410 440 470 500
0
50
100
8495 139
165
193
211 308
m/z
Abundance [%]
C
N
O N
H
Si
CH3
CH3
CH3
40 60 80 100 120 140 160 180 200 220 240
0
50
100
3944 51 6369 75 82 89 95101 115 126 139 152
165
177
193
236
N
ONH2
Abundance [%]
m/z
A
40 60 80 100 120 140 160 180 200 220 240
0
50
100
3944 51 6369 75 82 89 95101 115 126 139 152
165
177
193
236
N
ONH2
Abundance [%]
m/z
A
40
Fig. 3.31: Electron ionization mass spectra of CBZ (A) and IMINO (B)
3.3.2. GC/MS
Because of the high separation power of capillary columns followed by a specific detection by
means of mass spectrometry the combined GC/MS technique is the most appropriate tool for
volatile and thermal stable compounds.
All single parameters were varied and tested for the investigated pharmaceuticals including
the main metabolites, which were synthesised at the University of Paderborn. The final
working conditions based on methylation with diazomethane as described in chapter 5.
3.3.2.1. Determination of underivatized analytes
Direct analysis without derivatization was only possible for CBZ. Besides the signal for CBZ
a second very sharp one with a shorter retention time could be observed, often more intensive
than that of CBZ. It is caused by IMINO, a decomposition product of CBZ formed in the
injector of the GC.
3.3.2.2. Methylation by diazomethane
As described in section 3.3.1 the analytical determination of the selected drugs was performed
after derivatization with diazomethane as reagent for all analytes. The methylated products of
IBU, OH-IBU, DCF, SFM, and Ac-SFM are shown in a total ion current (TIC) chromatogram
(Fig. 3.32).
40 60 80 100 120 140 160 180 200
0
50
100
39 51 63 69 75
83
89
95
101 115 128 139 152
165
177
193
N
H
Abundance [%]
m/z
B
40 60 80 100 120 140 160 180 200
0
50
100
39 51 63 69 75
83
89
95
101 115 128 139 152
165
177
193
N
H
Abundance [%]
m/z
B
41
Fig. 3.32: GC/MS chromatogram of the methylated analytes (CH2N2), 1. IBU-Me, 2. OH-
IBU-Me, 3. IMINO, 4. DCF-Me, 5. CBZ, 6. SFM-Me, 7. PCB, 8. Ac-SFM-Me, Using fused-
silica capillary column (HP-5MS), split/splitless injector and gradient temperature program as
shown in chapter 5
3.3.2.3. Fragmentation pattern of the methylated analytes
The chemical structure and typical mass spectra of the methylated derivatives are shown in
Fig. 3.28 and the selected ions are listed in Table 3.4.
The mass spectrum of methylated IBU indicates a molecular ion [M]+• at m/z 220, the major
fragment ions appear at m/z 177 [M-C3H7]+ and 161 [M-COOCH3]+. Further fragments
appear at m/z 91, 105 and 119 indicating the structure of an alkyl-substituted aromatic system
[35]. The mass spectrum of methylated DCF indicates a molecular ion [M]+• at m/z 309, the
major fragment ions are m/z 277 [M-CH3OH]+, 242 (further loss Cl), 214 (further loss CO)
and 179 (further loss Cl) [59]. Though the mass spectrum of methylated SFM doesn’t show a
molecular ion [M]+• at m/z 267, the methylated products could be validated by LC/MS as well
as to Ac-SFM. The major fragments in GC/MS have m/z 203 (presumable loss of SO2) and
188 (presumable further loss of CH3). Further fragments are at m/z 156, 108 and 92. The mass
spectrum of methylated Ac-SFM doesn’t show a molecular ion [M]+• at m/z 309, the major
fragment ions appear at m/z 245 (presumable loss of SO2), 230 (presumable further loss of
CH3), 161 (presumable loss of methyl-acetylaminophenyl) and 134 (acetylaminophenyl).
Further fragments are at m/z 156, 108 and 92. The fragmentations pattern for both SFM and
Ac-SFM are in agreement with data reported in the litreture [76].
The mass spectrum of methylated OH-IBU has only a weak molecular ion [M]+• at m/z 236,
the methylated products could be validated by LC/MS too. The major fragment ions are at
m/z 178 [M-58]+•, 177 [M-COOCH3]+ and 118 (further loss of C3H7O). Further fragments
have the m/z 161 and 91 [35, 77].
3412 14 16 18 20 22 24 26 28 30 32
0
20
40
60
80
100
Abundance [%]
Time [min]
1
2
3
4
5
678
TIC
3412 14 16 18 20 22 24 26 28 30 32
0
20
40
60
80
100
Abundance [%]
Time [min]
1
2
3
4
5
678
TIC
42
The underivatized mass spectra of CBZ as shown in Fig. 3.31a indicates a molecular ion
[M]+• at m/z 236, the major fragment at m/z 193 [M-NHCO]+ and a further fragment at m/z
165.
3.3.2.4. Identification by SIM
Selected ion monitoring (SIM) is a technique in which a particular ion or a set of ions are
selected and monitored. SIM experiments are useful for detecting small quantities of a target
compound or to eliminate overlapping by signals of a complex mixture. The precondition is
that the mass spectrum as well as the retention time of the target compound are known. Thus,
SIM is useful for trace analysis for a rapid screening of a large number of samples for known
target compounds.
Depending on the analyzer system of the mass spectrometer SIM can provide lower detection
limits and greater speed of analysis than a full scan MS because only a few ions are
monitored.
Therefore, quantification was performed by SIM using the most abundant and specific ions of
each compound. The retention time - sometimes is an important help in identification - and
the single ion masses of the selected pharmaceuticals are listed in (Table 3.4).
Table 3.4: Specific retention times of the analyte used for quantification in the (EI+, SIM)
a = instrument internal standard
3.3.2.5. LOD and LOQ
Instrument detection limit (IDL) is defined as the minimum concentration of the analyte that
the instrument can detect with a signal to noise ratio 3:1. It is only a value for the efficiency of
the instrument and, consequently, independent from sample preparation. The method
detection limit (MDL) is similar to an IDL, but is based on samples, which passed all single
method steps. A limit of quantitation (LOQ) is normally 6 to 10 times the MDL value and is
considered to be the lowest concentration that can be accurately measured. DLs are actually
determined by analysis of multiple low-level samples in addition to blanks. This information
gives the variation in instrument response at levels near the detection limit.
The IDL and instrument LOQ were determined by measuring standard solutions of the
analytes in the concentration range of 1-10 µg/L and in addition blank samples as well.
Whereas, MDL and method LOQ were determined by spiking the analytes into 1 L ultrapure
Analyte Rt
[min]
Molecular ion
[m/z]
Ion 1 for SIM
[m/z]
Ion 2 for SIM
[m/z]
IBU-Me 13.63 220 161 117
OH-IBU-Me 16.72 236 178 119
IMINO 20.22 193 193 165
DCF-Me 22.39 309 214 242
CBZ 23.65 236 193 165
SFM-Me 24.71 267 156 92
PCB 169a 26.11 360 360 290
Ac-SFM-Me 30.10 309 161 134
43
water sample in the concentration range 1-10 ng/L, leading to an enrichment factors of 2500
after sample preparation steps. In case of SFM and Ac-SFM this range could not be achieved.
The MDL and LOQ investigation gave different results based on the chemical structure
characteristics of the analytes. Acidic and neutral compounds showed MDLs between 1-5
ng/L and LOQs between 3-10 ng/L (Table 3.5), but the polar compounds, e.g. SFM, have
relatively higher detection limits up to 100 ng/L. The reason might be interpreted due to the
efficiency of the enrichment and derivatization procedures. Therefore, gas chromatography is
not the most suitable method for SFM and its metabolites especially in a lower concentration
range. For neutral and acidic analytes, IDL and MDL are 1-5 and 2-12 pg absolute per
injection respectively.
Table 3.5: Regression coefficients, limits of detection and limits of quantification obtained in
SIM mode of GC/MS, * = underivatised
a = 1 L sample (distilled water), b = 0.1 L sample (surface water), I = instrument, M = method
3.3.2.6. Calibration
Stock solutions were prepared by solubility the analytes in ultrapure water. Different
concentrations were prepared by dilution the concentrations solution in ultrapure or real
waters. All stock solutions were stored in the refrigerator to protect them against
photodegradation. They were warmed up to room temperature before use.
Quantification was performed using external calibration. Calibration curves were created from
6 concentration points (n=3) in the concentration range 10-1000 ng/L. The sample volume
was 1.0 L using ultrapure water and 0.1 L using surface water (river Ruhr). GC/MS data
acquiring was done by the SIM technique. The peak areas were corrected due to a background
concentration of the analytes in the real water samples. The peak areas were plotted against
the corresponding concentration of the analytes (Fig. 3.33).
Sample volumes of 100-1000 mL were extracted, leading to enrichment factors ranging
between 250 and 2500 in order to test the influence of the volume. As mentioned before in the
derivatization section CBZ partially formed IMINO as a thermal degradation product caused
by the GC injector. Therefore, the sum of both areas was accumulated as one unit for
calibration.
Based on MDL and LOQ, the methylated form of SFM, Ac-SFM, and Glucu-SFM were
excluded from GC/MS calibration. An alternative procedure using LC-ESI/MS/MS is
described later in section 3.3.3.
Analyte r2
DWa
r2
SWb
IDL
[pg]
I-LOQ
[pg]
MDL
[ng/L]
M-LOQ
[ng/L]
IBU-Me 0.996 0.9982 1 2.5 5 10
OH-IBU-Me 0.9985 0.9998 1 2 1 3
DCF-Me 0.9989 0.9998 1 2.5 3 5
CBZ* 0.9982 0.9987 5 10 5 10
SFM-Me - -
50 100 100 250
Ac-SFM-Me - -
500 750 500 1000
44
The quality of the developed method was confirmed by the linearity of the calibration curves
up to 1 µg/L (Fig. 3.33). No further investigation was done for higher concentrations because
the analytes expected to be present in a relative low concentration range in surface water. In a
case of higher concentration, such as in wastewaters, the samples could be fitted by dilution.
Regression coefficients are > 0.99 in most cases, indicating a good linearity of the calibration
curves (Table 3.5).
In order to have an overview about the GC/MS instrument reproducibility, 3,3',4,4',5,5'-
hexachlorobiphenyl (PCB 169) was spiked as an internal instrument standard after the
derivatization step. From numerous injections within a period of weeks, the relative standard
deviation was found to be lower than 4 %. In GC/MS all analytes except IBU showed high
precise measurements. The reason may be the sample handling that means a loss of an amount
during the drying step in the heating block under a stream of nitrogen.
A)
B)
Fig. 3.33: Calibration curves from GC/MS analysis after SPE and derivatization with
CH2N2 (A = 1 L spiked ultrapure water, B = 0.1 L spiked river water; n=3)
0
10000
20000
30000
40000
50000
60000
70000
0 0.2 0.4 0.6 0.8 1
Concentration [µg/L]
Area [counts]
OH-IBU
IBU
CBZ
DCF
0
10000
20000
30000
40000
50000
60000
70000
0 0.2 0.4 0.6 0.8 1
Concentration [µg/L]
Area [counts]
OH-IBU
IBU
CBZ
DCF
0
10000
20000
30000
40000
50000
60000
70000
80000
0246810
Concentration [µg/l]
Area [counts]
OH-IBU
IBU
CBZ
DCF
0
10000
20000
30000
40000
50000
60000
70000
80000
0246810
Concentration [µg/l]
Area [counts]
OH-IBU
IBU
CBZ
DCF
45
3.3.3. LC/MS
Though the liquid chromatography has a restricted separation power in comparison to
capillary GC and though the combination with mass spectrometry is not the best support for
compound identification because of fragment-poor spectra, this technique is indispensable in
the analysis of polar and thermo-labile analytes. The possibility to apply the MS/MS-
technique to verify aimed fragments increases the relevance of this method immensely.
3.3.3.1. Separation characteristics
To achieve reliable analytical separations in LC different mobile phases and the influence of
the pH were tested. The pH-value can have a great influence in the retention time as it is
demonstrated for DCF and IBU in (Fig. 3.34), while SFM and CBZ have nearly constant
retention time in the whole pH-range. To achieve an optimum in separation power and to
reach high sensitivity a method based on capillary LC-ESI/MS was developed. The
experimental conditions are collocated in chapter 5.
In all cases a compromise has to be made between the improvement of separation and
deterioration of the ionization process. Therefore, many mobile phase compositions with
different pH values were studied to achieve suitable chromatographic separation and a stable
ionization spray. Methanol is a better protic solvent and generally produces more ions than
acetonitrile. Whereas, acetonitrile lead to better HPLC separation and a lower column back
pressure. Initially, both acetonitrile and methanol were tested as organic mobile phases for LC
separation. The measurements were finally carried out with a mixture of both of them. An
adequate mobile phase, as discussed later in section 5.3, led to superior composition for
interface ionization, shorter retention time and better resolution of the analytes.
Fig. 3.35 shows the extracted mass chromatogram of the target pharmaceuticals and their
metabolites after the SPE enrichment and HPLC separation. Under the described conditions
all analytes were resolved chromatographically within a retention time of 23 min.
Fig. 3.34: Influence of the HPLC mobile phase pH on retention times
0
5
10
15
20
25
30
35
40
4.5 5.0 5.5 6.0 6.5 7.0 7.5
pH-Value
Rt [min]
SFM DCF IBU CBZ
0
5
10
15
20
25
30
35
40
4.5 5.0 5.5 6.0 6.5 7.0 7.5
pH-Value
Rt [min]
SFM DCF IBU CBZ
46
Fig. 3.35: Extracted mass chromatograms from conventional LC-ESI/MS analysis of standard
solution 1. Glucu-SFM (ESI-), 2. SFM (ESI+), 3. Ac-SFM (ESI+), 4. OH-IBU (ESI-),
5. CBZ (ESI+), 6. DCF (ESI-), 7. IBU (ESI-), using Aquasil C-18 as analytical column and a
gradient program as shown in chapter 5 from mobile phase A and B
3.3.3.2. Mass spectrometer parameters
All analyses were carried out with an electrospray ionisation (ESI) interface and an ion trap
mass analyzer. Instrument control, data acquisition and evaluation were done with Xcalibur
software.
Several mass spectrometric parameters had to be optimised in order to obtain the highest
possible abundance of the analytes in MS. With ESI most of the ions are already formed in
the liquid phase and thus the eluent composition has a significant role in the ionisation
process. Optimisation of the eluent composition is discussed in the previous section. The first
parameters to be optimised in ESI are those having a crucial effect on the spray formation,
namely the voltage of the capillary and the flow rate of the nebulising gas.
Electrospray operation parameters were optimized by direct infusion of the analytes by means
of a syringe pump at a flow rate of 10 µL/min into the ESI source. Nitrogen was used as
sheath gas at a flow rate of approximately 0.3 L/min, the spray voltage was set to 3.5 kV in
the positive ionization mode and to 3.2 kV in the negative ionization mode, the transfer
capillary was set to 200 ˚C.
Successive work was done to optimize the spray voltage and transfer capillary; generally, low
voltage leads to a smaller detector response, because the formed droplets carry less charge, a
too high value causes arcs in the source. To obtain the best sensitivity for the detection,
different capillary temperatures were investigated in order to assist the dissolution of the ions
produced by ESI. 200 ˚C was the optimum value because an increasing temperature showed
0 2 4 6 8 10 12 14 16 18 20 22
0
50
100
50
100
50
100
50
100
50
100
50
100
50
100 NL: 1,90E5
m/z= 251,5-252,5 F: - c ESI
sid=5,00 Full ms2 428,00@26,00
[ 115,00-500,00]
NL: 3,95E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
254,00@45,00 [ 65,00-300,00]
NL: 2,65E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
296,00@40,00 [ 80,00-300,00]
NL: 7,08E6
m/z= 220,5-221,5 F: - c ESI
sid=5,00 Full ms [ 200,00-300,00]
NL: 5,84E7
m/z= 193,5-194,5 F: + c ESI
sid=5,00 Full ms2 237,00@33,00
[ 65,00-300,00]
NL: 1,33E6
m/z= 249,5-250,5 F: - c ESI
sid=5,00 Full ms2 294,00@25,00
[ 80,00-300,00]
NL: 1,03E6
m/z= 158,5-159,5 F: - c ESI
sid=5,00 Full ms2 205,00@35,00
[ 55,00-300,00]
Time [min]
Abundance [%]
1
2
3
4
5
6
7
0 2 4 6 8 10 12 14 16 18 20 22
0
50
100
50
100
50
100
50
100
50
100
50
100
50
100 NL: 1,90E5
m/z= 251,5-252,5 F: - c ESI
sid=5,00 Full ms2 428,00@26,00
[ 115,00-500,00]
NL: 1,90E5
m/z= 251,5-252,5 F: - c ESI
sid=5,00 Full ms2 428,00@26,00
[ 115,00-500,00]
NL: 3,95E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
254,00@45,00 [ 65,00-300,00]
NL: 3,95E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
254,00@45,00 [ 65,00-300,00]
NL: 2,65E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
296,00@40,00 [ 80,00-300,00]
NL: 2,65E6
m/z= 187,5-188,5 F: + c ESI
sid=10,00 Full ms2
296,00@40,00 [ 80,00-300,00]
NL: 7,08E6
m/z= 220,5-221,5 F: - c ESI
sid=5,00 Full ms [ 200,00-300,00]
NL: 7,08E6
m/z= 220,5-221,5 F: - c ESI
sid=5,00 Full ms [ 200,00-300,00]
NL: 5,84E7
m/z= 193,5-194,5 F: + c ESI
sid=5,00 Full ms2 237,00@33,00
[ 65,00-300,00]
NL: 5,84E7
m/z= 193,5-194,5 F: + c ESI
sid=5,00 Full ms2 237,00@33,00
[ 65,00-300,00]
NL: 1,33E6
m/z= 249,5-250,5 F: - c ESI
sid=5,00 Full ms2 294,00@25,00
[ 80,00-300,00]
NL: 1,33E6
m/z= 249,5-250,5 F: - c ESI
sid=5,00 Full ms2 294,00@25,00
[ 80,00-300,00]
NL: 1,03E6
m/z= 158,5-159,5 F: - c ESI
sid=5,00 Full ms2 205,00@35,00
[ 55,00-300,00]
NL: 1,03E6
m/z= 158,5-159,5 F: - c ESI
sid=5,00 Full ms2 205,00@35,00
[ 55,00-300,00]
Time [min]
Abundance [%]
1
2
3
4
5
6
7
47
degradation of DCF. Isopropanol/water (5:1) was used as sheath liquid at a flow rate of 4
µL/min in order to stabilize and enhance the ESI signal.
Following the optimization of capillary voltage, ion optics parameters were optimized for
each polarity mode based on mass peak intensity. After the method development was finished,
the formation of ammonium adducts in the positive mode and acetate adducts in the negative
mode were observed. Moreover, oligomers were formed at a higher concentration in the both
ionization mode. It is often observed that adduct ions hamper qualitative and quantitative
results, especially when unexpected. These are formed by either association between species
present in the mobile phase system that are preserved due to the soft ionization of the
electrospray process or to the presence of gas-phase collisions in the spray chamber prior to
sampling by the mass spectrometer. Thereby, the most important parameter is the declustering
potential, which should be tuned in order to reduce the formation of adducts and to increase
the abundance of the parent ions. To dissociate as much as possible from those adducts
without any dissociation of the parent ions, an ion source collision energy was applied in the
skimmer region. The optimal collision energy for those compounds was between 5 and 10 V
in negative and positive ion mode (Table 3.6). The analytes were detected using a time-
segments mode with positive and negative voltage switching.
The properties of the functional groups define the ionisation mode; DCF, IBU and OH-IBU
contain carboxylic acid, so they will favor the negative mode while CBZ, SFM, Ac-SFM and
Glucu-SFM prefer the positive mode because of their amine functionality. Due to their
structural ambivalence of SFM and its metabolites, the detection was possible in negative
ionization mode as well and for Glucu-SFM it was also favorably.
Considering Glucu-SFM, ESI is specially favourable because it operate without heat input in
the spray-ionization step, thus allowing the polar glucuronide to be ionized without thermal
degradation [78].
3.3.3.3. Confirmation by MS/MS
For the analysis of complex mixtures, the MS/MS feature - also called ‘selected reaction
monitoring’ (SRM) - provides a high degree of selectivity, specificity, and sometimes a better
limit of detection than full scan MS. In SRM scan mode one species of ions, called parent
ions, are selected and stored in the mass analyzer. Then these ions are excited by energy so
that they collide with the background gas atoms. By that, the parent ions fragment to one or
more product ions, which are recorded in an SRM product ion mass spectrum. This procedure
can be repeated several fold. Like SIM, SRM is of advantage for a rapid analysis of trace
components in complex mixtures.
At first, the optimum parameters of the MS/MS acquisition had to be determined for each
individual compound, in order to characterize the typical fragmentation pattern and to
maximise abundances of fragment ions for each compound. It was performed using an online
injection valve; 20 µL (1 mg/L) standard solution containing the analytes was injected into the
mobile phase effluent of the HPLC; 50 % of buffer B at a flow rate of 13µL/min. After the
selection of the parent ions by mass analyser, the collisions of the parent ions caused the
fragmentation producing the product ions. After determination of the most significant product
ions, optimisation of the collision energy was carried out to produce the maximum abundance
of product ions. Parent and product ion masses of the individual compounds are given in
(Table 3.6).
48
Table 3.6: Optimized LC-ESI-MS/MS parameters for the analysis of the pharmaceuticals and
some their metabolites
3.3.3.4. Analyte fragmentation pattern
All analytes were measured in full scan mode and under MS/MS conditions.
The mass spectrum of CBZ contains an abundant [M+H]+ ion at m/z 237 and only a small
fragment ion at m/z 194 (Fig. 3.36A). Fig 3.36B shows the MS/MS product ion spectra of the
[M+H]+ ion; as expected the collision results in the structural fragment at m/z 194 which
corresponds to the loss of HNCO. The mass spectrum of SFM contains an abundant [M+H]+
ion at m/z 254 and a weak ammonium adduct [M+NH4]+ (Fig. 3.37A). Fig 3.37B shows the
MS/MS product ion spectra of the [M+H]+ ion with m/z 188 which corresponds to
rearrangement by losing H2SO2 [79]. Other minor fragments appear at m/z 156, 108 and 92;
the interpretation for these fragmentations is shown in (Fig. 3.38) and affirmed by literature
[80]. The mass spectrum of Ac-SFM contains an abundant [M+H]+ ion at m/z 296 and a weak
ammonium adduct [M+NH4]+ (Fig. 5.39A). Fig 5.39B shows the MS/MS product ion spectra
of the [M+H]+ ion with m/z 188 which corresponds to the loss (CH3CO+HSO2). Other
fragments appear at m/z 230 (loss of H2SO2), 198 and 136; the interpretation of these
fragmentations follows the fragmentation of the SFM pattern as shown in (Fig 3.38) [23,80].
The mass spectrum of Glucu-SFM contains an abundant [M-H]- ion at m/z 428, the other ion
at m/z 252 corresponds to deprotonated SFM formed by the cleavage of the glycoside bond
(Fig. 3.40A). Fig. 3.40B shows the MS/MS product ion spectra of the [M-H]- ion with m/z
252 which corresponds to the loss of the glucuronide moiety. The mass spectrum of DCF
contains an abundant [M-H]- ion at m/z 294 and only a small fragment ion with m/z 250 (Fig.
3.41A). Fig 3.41B shows the MS/MS product ion spectra of the [M-H]- ion of DCF with m/z
250, corresponding to the expulsion of CO2 [81]. The mass spectrum of IBU contains an
abundant [M-H]- ion at m/z 205 in addition to the acetate adduct [M-H+CH3COOH]- and the
dimer [2M-H]- (Fig. 3.42A). Fig. 3.42B shows the MS/MS product ion spectra of the [M-H]-
ion of IBU with m/z 159 which corresponds to the loss of (CO+H2O). The mass spectrum of
OH-IBU contains an abundant [M-H]- ion at m/z 221 in addition to the acetate adduct [M-
H+CH3COOH]- and the dimer [2M-H]- (Fig 3.43). The MS/MS experiment for OH-IBU
showed unstable product ions at m/z 177, 159 and 133.
Rt SRM Fragment ion
Analyte
[Min]
Mode
Collision
energy
[V] [m/z]
Normalized
collision energy
[%] [m/z]
Glucu-SFM 4.6 -ve 5 428 26 252/175
SFM 7.6 +ve 10 254 45 188/156
Ac-SFM 9.2 +ve 10 296 40 236/188
OH-IBU 10.7 -ve 5 221 0 -
CBZ 14.8 +ve 5 237 33 194
DCF 19.1 -ve 5 294 25 250
IBU 20.7 -ve 5 205 35 159
49
Fig. 3.36: LC-ESI/MS spectrum of CBZ (A) and the MS/MS product ion spectrum of the
[M+H]+ ion at m/z 237 (B)
190 195 200 205 210 215 220 225 230 235 240 245 250
m/z
50
100 237,2
238,1
194,4
Abundance [%]
[M+H]+
A
190 195 200 205 210 215 220 225 230 235 240 245 250
m/z
50
100 237,2
238,1
194,4
Abundance [%]
[M+H]+
A
80 100 120 140 160 180 200 220 240 260 280 300
m/z
50
100 194,3
219,9
192,3
Abundance [%]
B
80 100 120 140 160 180 200 220 240 260 280 300
m/z
50
100 194,3
219,9
192,3
Abundance [%]
B
220 225 230 235 240 245 250 255 260 265 270 275 280 285 290 295
m/z
50
100 254,1
255,1
256,1 270,8
Abundance [%]
[M+H]+
[M+NH4]+
A
220 225 230 235 240 245 250 255 260 265 270 275 280 285 290 295
m/z
50
100 254,1
255,1
256,1 270,8
Abundance [%]
[M+H]+
[M+NH4]+
A
50
Fig. 3.37: LC-ESI/MS spectrum of SFM (A) and the MS/MS product ion spectrum of the
[M+H]+ ion at m/z 254 (B)
S
O
O
NN
O
NCH3
H
R
H
SO2
N
R
H
ON
R
H
N
NH
R = H (m/z 108)
m/z 92
R = H (m/z 156)
R = Ac (m/z 198)
+Or +
+
+
Fig. 3.38: Common CID fragmentation reaction for SFM and Ac-SFM
90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260
m/z
50
100 188,0
156,0
147,2 190,2
194,1
235,8
148,2160,2 254,0
146,2 235,1
108,1
Abundance [%]
B
90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260
m/z
50
100 188,0
156,0
147,2 190,2
194,1
235,8
148,2160,2 254,0
146,2 235,1
108,1
Abundance [%]
B
280 285 290 295 300 305 310 315 320 325
m/z
50
100 296,1
297,1
312,7
298,1
Abundance [%]
[M+H]+
[M+NH4]+
A
280 285 290 295 300 305 310 315 320 325
m/z
50
100 296,1
297,1
312,7
298,1
Abundance [%]
[M+H]+
[M+NH4]+
A
51
Fig. 3.39: LC-ESI/MS spectrum of Ac-SFM (A) and the MS/MS product ion spectrum of
the [M+H]+ ion at m/z 296 (B)
Fig. 3.40: LC-ESI/MS spectrum of Glucu-SFM (A) and the MS/MS product ion spectrum of
the [M-H]- ion at m/z 428 (B)
80 100 120 140 160 180 200 220 240 260 280 300
m/z
50
100 188,1 236,1
136,1 197,9
194,1 230,0
214,2 278,0
202,2 254,0
160,9 218,1 277,1
134,1 219,0
94,1 156,0 296,0
102,9 146,2
Abundance [%]
B
80 100 120 140 160 180 200 220 240 260 280 300
m/z
50
100 188,1 236,1
136,1 197,9
194,1 230,0
214,2 278,0
202,2 254,0
160,9 218,1 277,1
134,1 219,0
94,1 156,0 296,0
102,9 146,2
Abundance [%]
B
250 260 270 280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430
m/z
100
Abundance [%]
428,1
429,1
430,1
252,4
50
[M-H]-
A
250 260 270 280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430
m/z
100
Abundance [%]
428,1
429,1
430,1
252,4
50
[M-H]-
A
160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
50
100 252,0
174,9 427,9156,1
Abundance [%]
m/z
[M-Glucu]-
B
160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
50
100 252,0
174,9 427,9156,1
Abundance [%]
m/z
[M-Glucu]-
B
52
Fig. 3.41: LC-ESI/MS spectrum of DCF (A) and the MS/MS product ion spectrum of the
[M-H]- ion at m/z 294 (B)
240 245 250 255 260 265 270 275 280 285 290 295 300 305 310 315 320
m/z
50
100 294,1
296,0
297,9
250,5
252,4
Abundance [%]
[M-H]-
A
240 245 250 255 260 265 270 275 280 285 290 295 300 305 310 315 320
m/z
50
100 294,1
296,0
297,9
250,5
252,4
Abundance [%]
[M-H]-
A
140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290
m/z
50
100 250,1
Abundance [%]
B
140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290
m/z
50
100 250,1
Abundance [%]
B
160 180 200 220 240 260 280 300 320 340 360 380 400 420
m/z
50
100 205,1
410,9
265,0
206,1
411,9
159,4
Abundance [%]
[M-H]-
[M-H+CH3COOH]-
[2M-H]-
A
160 180 200 220 240 260 280 300 320 340 360 380 400 420
m/z
50
100 205,1
410,9
265,0
206,1
411,9
159,4
Abundance [%]
[M-H]-
[M-H+CH3COOH]-
[2M-H]-
A
53
Fig. 3.42: LC-ESI/MS spectrum of IBU (A) and the MS/MS product ion spectrum of the
[M-H]- ion at m/z 205 (B)
Fig. 3.43: LC-ESI/MS spectrum of OH-IBU
3.3.3.5. LOD and LOQ
The particular definitions are given in the GC/MS chapter (3.3.2.5). The quantification limits
are a function of chromatographic conditions, mass spectrometry parameters and chemical
characteristics of the analyte, and in LC especially the eluent composition because of the
mobile phase modifiers. Moreover, analyte features (i.e. hydrophobicity, lacking polarity) can
improve the ionization efficiency; thereby, instrument DLs can decrease significantly. MLD
and quantification follow the same criteria.
Instrument DLs varies between 20 and 140 pg absolute for all analytes. LOQs are distributed
between 60 and 200 pg absolute (Table 3.7). The analyte showed MDLs between 3-5 ng/L
and LOQ between 5 –10 ng/L.
160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
m/z
50
100 221,2
443,0
281,0
222,2 444,0
282,1
Abundance [%]
[M-H]-
[M-H+CH3COOH]-
[2M-H]-
160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500
m/z
50
100 221,2
443,0
281,0
222,2 444,0
282,1
Abundance [%]
[M-H]-
[M-H+CH3COOH]-
[2M-H]-
145 150 155 160 165 170 175 180 185 190 195 200 205
m/z
50
100 159,1
161,1
149,1 205,0
175,1
147,9
Abundance [%]
B
145 150 155 160 165 170 175 180 185 190 195 200 205
m/z
50
100 159,1
161,1
149,1 205,0
175,1
147,9
Abundance [%]
B
54
Table 3.7: Regression coefficients, rounded limits of detection and limits of quantification
obtained in LC-ESI/MS method
a = 1 L sample (distilled water), b = 0.1 L sample (surface water), I = instrument, M = method
3.3.3.6. Calibration
LC-ESI/MS quantification was performed using external calibration. The calibration curves
were built from 6 concentrations (n=3) in a measurement range of 10-1000 ng/L by spiking
1.0 L ultrapure water and 0.1 L river water samples with the analytes; the enrichment factors
were ranging between 250 and 2500 after sample preparation steps.
When the analyte concentration is too high for the capillary column a shift in the retention
time can be observed. In order to solve this problem the injection volume was decreased from
20 µl to 10 µl and additionally the column equilibration time was extended.
The acquisition by LC-ESI/MS method was based on MS/MS experiments except for OH-
IBU. In this case the full scan mode was used because the fragmentation product ions are
instable. For the real water samples the analyte peak areas were background corrected. The
high linearity of this method could be demonstrated for the selected calibration range (Fig.
3.44). No further investigation was done with higher concentrations because the selected
analytes are present in a relative low concentration range in surface water. In the case of
higher concentration, the sample could be diluted. The regression coefficient was > 0.97 in
most cases, indicating a good linearity of the calibration curves (Table 3.7).
LC-ESI/MS instrument reproducibility was studied by successive injections of the same
sample, the RSD was found to fluctuate a little bit from day to day, which is quite normal in
LC-ESI/MS and probably a result of the instability of the spray. To solve this problem, an
internal standard (e.g. isotope label compounds) should be used in order to compensate the
uncontrolled random errors caused by other components in the system or the instrument itself.
Isotopes labeled compounds of our selected drugs are not available. Therefore, no internal
standards were included in this investigation.
Analyte r2
DWa
r2
SWb
IDL
[pg]
I-LOQ
[pg]
MDL
[ng/L]
M-LOQ
[ng/L]
Glucu-SFM 0.9994 0.9774 100 200 5 10
SFM 0.9987 0.9996 20 60 5 10
Ac-SFM 0.9988 0.998 80 120 5 10
OH-IBU 0.9986 0.9996 60 120 5 10
CBZ 0.9878 0.998 60 120 3 5
DCF 0.9987 0.9998 140 200 5 10
IBU 0.9991 0.997 100 200 5 10
55
A)
B)
Fig. 3.44: Calibration curves from LC-ESI/MS analysis after SPE (A: 1 L spiked ultrapure
water, B: 0.1 L spiked river water; n=3)
0
500
1000
1500
2000
2500
0.0 0.2 0.4 0.6 0.8 1.0
Concentration [µg/L]
Area [counts x105]
CBZ
OH-IBU
SFM
Ac-SFM
DCF
IBU
Glucu-SFM
0
500
1000
1500
2000
2500
0.0 0.2 0.4 0.6 0.8 1.0
Concentration [µg/L]
Area [counts x105]
CBZ
OH-IBU
SFM
Ac-SFM
DCF
IBU
Glucu-SFM
0
500
1000
1500
2000
2500
0246810
Concentration [µg/L]
Area [counts x105]
CBZ
OH-IBU
SFM
Ac-SFM
DCF
IBU
Glucu-SFM
0
500
1000
1500
2000
2500
0246810
Concentration [µg/L]
Area [counts x105]
CBZ
OH-IBU
SFM
Ac-SFM
DCF
IBU
Glucu-SFM
56
3.3.4. Validation of the developed methods
To prove if the developed methods are robust against interferences, parameters having direct
influence in the extraction recoveries and signal response, e.g. sample volume, sample
concentration and matrix effect were studied.
Three sample volumes (250, 500 and 1000 mL) were chosen to study the enrichment recovery
at constant final concentration of a 625 µg/L. To study concentration effects 50, 250 and 1000
ng/L were used at a constant sample volume of 1 L. To ensure the robustness of the developed
method, surface water from river Ruhr was used as model for the matrix effect. Detectable
concentrations of the selected drugs might be found in real water samples. Therefore, blank
values were considered. A control matrix without our analytes as a standard solution was not
available. The percentage recovery of the analytes spiked into each type of tested sample was
calculated as shown in the (equation 1).
100
tan
covRe ×
−
=dardsenrichednonofArea
sampleunspikedofAreasamplespikedofArea
ery (Equation 1)
As shown in (Table 3.8 – 3.9), we can recognise that neither an increasing sample volume nor
the sample concentration had a noticeable influence in extraction recoveries for both methods.
However, the response factors obtained from test solutions and matrix loaded samples differ
significantly for both methods especially in LC-ESI/MS analysis for some analytes (Table
3.10). The recoveries of the spiked analytes from different volumes (250-1000 mL) and
concentrations (50-1000 ng/L) by means of GC/MS were ranged between 91-123 % and 71-
113 % respectively, while in LC-ESI/MS were 57-90 % and 47-94 %. For surface water the
recoveries of the spiked analytes from different concentrations (2.5-10 µg/L) were between
91-139 % based on GC/MS and 44-81 % for LC-ESI/MS, but for Glucu-SFM the recovery
was only about 10 % (Table 3.10). At lower concentration (0.5 µg/L), the analytes showed
lower recoveries values for most of the analytes as shown in Table 3.10. In contrast to the
standard solution in ultrapure water a loaded matrix had a significant influence on Glucu-
SFM. The reason is that the glucuronide moiety hydrolysed in the matrix as proved in chapter
3.4.
In GC/MS, in some cases the extraction recovery values were 10-20 % higher than the non-
enriched standard solutions. A feasible interpretation may be that these compounds are
present in the original surface water in very low concentrations, which could not be detected
before spiking as reported by Sacher F. et al. [82]. In fact, the matrix did not show any
influence and that might be attributed to the high molecular compounds which are retarded in
the GC injector. Moreover, the ionization technique used prevents the formation of any
artifact during the ionization process in comparison with other techniques.
In LC-ESI/MS method, the matrix showed significant influence in the recovery values (Table
3.10). These losses were caused by matrix impurities, which either reduced the sorption
efficiencies or led to signal suppression in the ESI interface, which is assumed to result from
the competition of the analyte ions and matrix components for access to the droplet surface
for the gas-phase emission [83].
The comparison of the recovery values of enriched standard solutions using both methods
showed higher recovery values in GC/MS especially for IBU and OH-IBU as demonstrated in
Table 3.8 – 3.9. Theses results prove that the ionisation in the ESI is responsible for the loss in
the recovery and not to the sorption efficiencies of the SPE material.
57
Table 3.8: Influence of sample volumes in the extraction recoveries (R) of selected analytes
(625 µg/L; final concentration, 1 L sample volume, n = 2)
Table 3.9: Influence of concentrations in the extraction recoveries (R) of selected analytes
(625 µg/L; final concentration, 1 L sample volume, n = 2)
Table 3.10: Influence of surface water matrix from the river Ruhr in the extraction recoveries
(R) of target analytes (100 mL sample volume, enrichment factors 250 and n=3)
250 mL 500 mL 1000 mL
GC/MS LC/MS GC/MS LC/MS GC/MS LC/MS
Analyte R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
Glucu-SFM - - 65 11 - - 60 4 - - 58 3
SFM - - 73 10 - - 71 10 - - 72 8
Ac-SFM - - 90 6 - - 69 13 - - 76 4
OH-IBU 119 2 88 4 110 7 87 2 111 3 70 7
CBZ 123 3 81 4 115 8 72 11 114 2 82 1
DCF 96 2 82 1 91 9 81 1 92 0 82 7
IBU 116 3 66 0 93 0 66 2 100 9 57 6
50 ng/L 250 ng/L 1000 ng/L
GC/MS LC/MS GC/MS LC/MS GC/MS LC/MS
Analyte R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
Glucu-SFM - - 47 6 - - 58 3 - - 68 9
SFM - - 60 8 - - 72 8 - - 69 13
Ac-SFM - - 77 4 - - 76 4 - - 81 5
OH-IBU 94 6 71 8 111 3 69 7 111 3 64 10
CBZ 95 3 80 5 113 2 81 1 106 3 85 8
DCF 75 2 94 4 90 0 82 7 88 3 84 7
IBU 71 25 63 4 99 9 57 3 79 0 60 17
500 ng/L 2500 ng/L 10000 ng/L
GC/MS LC/MS GC/MS LC/MS GC/MS LC/MS
Analyte R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
R
[%]
RSD
[%]
Glucu-SFM - - 9 19 - - 8 7 - - 11 9
SFM - - 39 3 - - 44 3 - - 46 3
Ac-SFM - - 63 9 - - 74 5 - - 76 2
OH-IBU 132 2 41 11 139 2 57 6 129 4 61 2
CBZ 132 3 51 10 120 3 54 2 109 5 78 8
DCF 98 2 37 19 110 7 81 13 97 4 78 3
IBU 118 6 21 2 123 5 50 4 91 3 58 5
58
3.3.5. Comparison between the methods developed
GC-MS and LC-ESI/MS techniques were compared both to find out the most suited method
concerning chromatographic (peak shape, resolution, etc.), mass spectrometric (efficiency,
sensitivity, etc.) properties and to explore the potential for a multiresidual method.
GC/MS method separated the target analytes with high resolution. In LC-ESI/MS, the drugs
and their metabolites can be separated in a single run. But because of the wide pKa range of
the analytes the pH-value of the LC eluent and the sample has a great influence on retention
times and separation characteristics and claims much attention.
The MS/MS technique in LC-ESI/MS method provides better sensitivities and gives
additional information. As well in GC/MS, SIM gives high sensitivities.
When comparing the recovery values from the both methods, we could discriminate the
influence of the matrix on the ionization efficiency in LC-ESI/MS.
The GC/MS measurements reproducibility was represented with lower RSD’s. On the other
hand, acceptable reproducibility was achieved for LC-ESI/MS due to the absence of internal
standard correction.
GC/MS method provides a better accuracy for all investigated analytes except SFM and its
metabolites.
Both methods showed very good linearity (r2> 0.97) in most cases and have a LOQs from 3 to
10 ng/L for all compounds, except for SFM and its metabolites in GC method.
3.3.6. Applications of the developed methods for a real surface water sample
The developed methods were tested for applicability to real water samples by measuring
surface water samples from river Ruhr. Concentrations of the selected drugs and some of their
metabolites are listed in (Table 3.11). The table demonstrates that GC/MS is best for the
determination of IBU, OH-IBU, DCF, while LC-ESI/MS is best for SFM and Ac-SFM; CBZ
can be determined with both methods. Quantification on both methods was based on a
calibration curves build with the same matrix as described in the previous sections.
Four drugs and some of their main metabolites could be identified in relative high
concentrations. That is not surprisingly when considering that the effluent of a hospital enters
the river about to the place of sampling. Using GC/MS the concentrations in river water
ranged DCF, IBU, OH-IBU and CBZ from 107 to 249 ng/L. The results for the two SFM
analytes - measured by LC-ESI/MS – were 229 ng/L for SFM and 316 ng/L for Ac-SFM,
while the concentration of CBZ was 290 ng/L.
In comparison with diverse studies for drugs concentrations in German rivers [2], these
magnitudes are mentioned for drugs detected at similar concentrations. Glucu-SFM was not
detected due to possible cleavage of the glucuronides in surface water.
59
Table 3.11: Concentration (ng/L) of the selected pharmaceuticals and some of their
metabolites in surface water (Ruhr river, Germany on 04.06.2004)
- : method non-suitable
n.d. : not detected
Analyte GC/MS method
[ng/L]
RSD
[%] LC-ESI/MS method
[ng/L]
RSD
[%]
Glucu-SFM - - n.d. -
SFM - - 229 51
Ac-SFM - - 316 46
OH-IBU 249 10 - -
CBZ 241 13 290 39
DCF 107 25 - -
IBU 169 31 - -
60
3.3.7. Conclusion
Two analytical methods based on a pre-concentration step by solid phase extraction followed
by GC/MS and LC-ESI/MS were developed for simultaneous determination of traces of target
drugs and their main metabolites. Derivatization step was necessary prior to GC/MS analysis.
After examination a varieties of sorbent materials, Oasis HLB showed the best extraction
recoveries for all target analytes. The analytes concentrations and sample volumes showed
nearly no influence in the extraction efficiency.
Due to the elevated polarity of structure moieties of the analytes, the extract has to be
derivatize prior to the measurement for GC/MS. Among many reagents, diazomethane was
the best reagent for most analytes except CBZ and Diol-CBZ. But that doesn’t matter,
because CBZ and Diol-CBZ could be analysed without derivatization. The advantage of
diazomethane is that there are a few by-products and the molecular weight of the derivatives
increased slightly.
Regarding the chromatography conditions, the HPLC mobile phase composition and it pH
value had tremendous influence on the retention time of acid functionality and ions formation
in ESI interface. Therefore, the final method was the best compromise between these
parameters.
For more selectivity and specificity, GC/MS method was based on selected ion monitoring,
while selected reaction monitoring was used in LC-ESI/MS.
For ultrapure water samples applying the final procedures the recoveries of the spiked
analytes at constant sample volume (1 L) and concentration (1 µg/L) were 79-111 % by
means of GC/MS and 60-85 % using LC-ESI/MS. For surface water the recoveries were 91-
129 % based on GC/MS and with LC-ESI/MS 46-78 % except for Glucu-SFM. The method
detection limits were 1-5 ng/L in GC/MS except the SFM and its metabolites and 3-5 ng/L in
LC-ESI/MS for all analytes.
The GC/MS method showed high linearity, high resolution, good precision, low DLs and
nearly no matrix influence. The only disadvantage of this method is that SFM, Ac-SFM and
Glucu-SFM are not suitable for quantifications. On the other hand also the LC-ESI/MS
method showed high linearity, good resolution, acceptable precision, low DLs and tangible
matrix dependence as results of the signal suppression, caused by high amounts of organic
and inorganic ions in the sample.
As a consequence, the determination of DCF, IBU and OH-IBU were preferred by GC/MS,
whereas, SFM, Ac-SFM and Glucu-SFM were preferred by LC-ESI/MS. CBZ could be
determined by the both methods.
The developed methods were verified by analysis of surface water samples from the river
Ruhr. All the target analytes were detected except Glucu-SFM. Using GC/MS the
concentration of DFC, IBU, OH-IBU and CBZ ranged from 107 to 249 ng/L, while in LC-
ESI/MS the concentration of SFM, Ac-SFM and CBZ ranged from 229 to 316 ng/L.
61
3.4. Investigation of the biodegradation of drugs in different model systems
There is still a lack of data on the fate, elimination mechanisms and efficiencies in sewage and
drinking water treatment especially for the metabolites and transformation products. In
addition, the understanding of physical, biological and chemical behavior in the aquatic
system such as adsorption, degradation and hydrolysis is of great importance as pathways for
natural treatment. The fate of the most selected drugs is not well understood yet in the aquatic
environment. Up to now only a few data’s are available about the fate of the human
metabolites of CBZ, DCF and SFM in the aquatic environment.
The analytical developed methods were tested to study the behavior fate of our target analytes
in batch and column experiments. Such experiments give an important knowledge about the
degradation behaviour of the pharmaceuticals in the dependence of the amount of the
dissolved oxygen concentration, pH-values or microbial populations.
The experiments are carried out with relatively higher concentration (1 mg/L) to simplify the
detection of any possible transformation products. The described LC-UV developed methods
in section 3.2.1 were used in order to monitor the behaviour of the analytes concentration
during the period of test time. Besides that, the attempt was started to identify the single
degradation products by means of LC-ESI/MS and GC/MS. For that purpose an enrichment
procedure as described in section 3.3.1 was applied. These results are discussed in section
3.4.4. The aim from the further mass spectrometry investigation was qualitative aspects.
3.4.1. Purity investigation of the standard drug materials
Besides that the analysis is complicated by a lot of compounds similar to the drugs, which
have their origin in the drug preparation and are still present in the applied products.
Therefore, the purity of the commercially available standard compounds and synthesised
metabolites were justified in the preliminary stage of this study in order to discriminate
between the presence of trace products in the original starting materials and the products
formed during the biodegradation processes.
Using GC/MS, traces of the OH-IBU could be identified in IBU (Fig. 3.45) and the
synthesized OH-IBU showed traces from IBU, IBU-Me, and CA-HA (Rt: 16.05) (Fig. 3.46).
The LC/MS method verified these results (Fig. 3.47).
By the analysis of GC/MS, CBZ showed traces of Diol-CBZ (Rt: 21.08), CBZ-ME (Rt:
21.47), and a signal at Rt: 18.29 can probably be assessed to CBZ-OH and at Rt: 19.88 to EP-
CBZ (Fig. 3.48). Both substances cannot be clearly identified. LC/MS verified the presence of
Diol-CBZ (Rt: 8.7), CBZ-OH and/or EP-CBZ (Rt: 11.04) and CBZ-Me (Rt: 11.47) (Fig. 3.49).
Using GC/MS, DCF showed traces of OH-DCF (Rt: 26.03), 8-CCA (Rt: 23.27) and indole
(Rt: 22.0) (Fig. 3.50). LC/MS verified the presence of traces from OH-DCF (Rt: 14.81) and 8-
CCA (Rt: 17.44) (Fig. 3.51).
Using GC/MS, SFM showed traces of Acetamide-N-Phenyl (Rt: 10.73), 4-Aminobenzene-
sulfonamide (Rt: 20.99) and methylated SFM (Rt: 24.75). LC/MS verified the results (Fig.
3.52). Using GC/MS, the synthesized Ac-SFM showed traces from Acetamide-N-Phenyl, 4-
Aminobenzensulfonamide, SFM and methylated Ac-SFM. LC/MS verified the presence of
methylated Ac-SFM (Fig. 3.53). The LC/MS chromatogram from synthesized Glucu-SFM
showed the presence of mono-, di- and tri-methylated Glucu-SFM as synthesised products
(Fig. 3.54).
62
The source of the methylated products could be by-products of synthesis or contaminants of
the starting compounds, is not yet clear. It could be recognised based on LC/MS and GC/MS
analysis of the standard solutions presence of additional products from synthesis, which must
be considered in the further study.
Fig. 3.45: Extracted mass chromatograms from LC/MS analysis of the standard IBU solution
Fig. 3.46: Electron ionization mass spectra of the diazomethane derivative of the CA-HA
Fig. 3.47: Extracted mass chromatograms from LC/MS analysis of the
synthesised OH-IBU solution
0246810 12 14 16 18 20 22 24 26 28
Time [min]
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Abundance [%]
10,40
20,03
IBU
OH-IBU
0246810 12 14 16 18 20 22 24 26 28
Time [min]
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
Abundance [%]
10,40
20,03
IBU
OH-IBU
0246810 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100
7,31
8,44
10,13
20,28
10,11 10,29
12,02
CA-HA
IBU-Me
OH-IBU
IBU
0246810 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100
7,31
8,44
10,13
20,28
10,11 10,29
12,02
CA-HA
IBU-Me
OH-IBU
IBU
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
40 51
59
77 91
103
117
131
148
163
177
191 222
CH3
O
OCH3
O
O
CH3
Abundance [%]
m/z
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
40 51
59
77 91
103
117
131
148
163
177
191 222
CH3
O
OCH3
O
O
CH3
Abundance [%]
m/z
63
Fig. 3.48: Electron ionization mass spectra of the CBZ-Me (A), Diol-CBZ (B),
CBZ-OH (C) and EP-CBZ(D)
30 50 70 90 110 130 150 170 190 210 230 250 270
0
50
100
31
43
51
58
69
74
87
95
115 129
143
152
165
178
193
206
218
235 250
Abundance [%]
m/z
N
ON
Me
H
A
30 50 70 90 110 130 150 170 190 210 230 250 270
0
50
100
31
43
51
58
69
74
87
95
115 129
143
152
165
178
193
206
218
235 250
Abundance [%]
m/z
N
ON
Me
H
A
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330
0
50
100
32 51 63
76 89
103
113 126
151
164
179
193
207
Abundance [%]
m/z
N
ONH2
OH OH B
30 50 70 90 110 130 150 170 190 210 230 250 270 290 310 330
0
50
100
32 51 63
76 89
103
113 126
151
164
179
193
207
Abundance [%]
m/z
N
ONH2
OH OH B
30 60 90 120 150 180 210 240 270 300 330 360
0
50
100
32 45 63 76
89
96 113 126 151
179
192 205 267 325
Abundance [%]
m/z
N
ONH2
OH
C
30 60 90 120 150 180 210 240 270 300 330 360
0
50
100
32 45 63 76
89
96 113 126 151
179
192 205 267 325
Abundance [%]
m/z
N
ONH2
OH
N
ONH2
OH
C
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450
0
50
100
41 55
63
69
85
111
122
131
165
192
207
223
243 281 305322 341 381 429 446
Abundance [%]
m/z
N
ONH2
OD
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450
0
50
100
41 55
63
69
85
111
122
131
165
192
207
223
243 281 305322 341 381 429 446
Abundance [%]
m/z
N
ONH2
O
N
ONH2
OD
64
Fig. 3.49: Extracted mass chromatograms from LC/MS analysis of the standard CBZ solution
Fig. 3.50: Electron ionization mass spectra of the diazomethane derivative of the
OH-DCF (A), 8-CCA (B) and Indole (C)
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
50
100
0
50
100
0
50
100
Abundance [%]
0
50
100
0
50
100
8,70
11,04
11,04
11,47
14,21
CBZ
CBZ-Me
CBZ-OH
EP-CBZ
Diol-CBZ
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
50
100
0
50
100
0
50
100
Abundance [%]
0
50
100
0
50
100
8,70
11,04
11,04
11,47
14,21
CBZ
CBZ-Me
CBZ-OH
EP-CBZ
Diol-CBZ
Abundance [%]
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
39
45
51 63 77 89 107 125
151
164
179
214
228
242
277
294
339
m/z
N
HCl
Cl OMe
MeOOC
A
Abundance [%]
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
39
45
51 63 77 89 107 125
151
164
179
214
228
242
277
294
339
m/z
N
HCl
Cl OMe
MeOOC
A
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
44 63
75 89
99
107
125
151 178
213
241
273
N
H
Cl COOMe
m/z
Abundance [%]
B
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
44 63
75 89
99
107
125
151 178
213
241
273
N
H
Cl COOMe
m/z
Abundance [%]
B
Abundance [%]
m/z
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
39 51 63 78
89 107
125
151
179
214
242
277
N
Cl
Cl
OC
Abundance [%]
m/z
30 60 90 120 150 180 210 240 270 300 330 360 390
0
50
100
39 51 63 78
89 107
125
151
179
214
242
277
N
Cl
Cl
OC
65
Fig. 3.51: Extracted mass chromatograms from LC/MS analysis of the standard DCF solution
Fig. 3.52: Extracted mass chromatograms from LC/MS analysis of the standard SFM solution
Fig. 3.53: Extracted mass chromatograms from LC/MS analysis of the
synthesised Ac-SFM solution
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100 18,16
14,81
17,38
Abundance [%]
8-CCA
OH-DCF
DCF
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100 18,16
14,81
17,38
Abundance [%]
8-CCA
OH-DCF
DCF
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100 7,42
10,74
14,99
SFM
SFM-Me
4-Aminobenzensulfonamide
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100 7,42
10,74
14,99
SFM
SFM-Me
4-Aminobenzensulfonamide
6 7 8 9 10 11 12 13 14 15 16 17 18
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100 7,27
8,79
14,18
SFM
Ac-SFM
Ac-SFM-Me
6 7 8 9 10 11 12 13 14 15 16 17 18
Time [min]
0
20
40
60
80
100
0
20
40
60
80
100
Abundance [%]
0
20
40
60
80
100 7,27
8,79
14,18
SFM
Ac-SFM
Ac-SFM-Me
66
Fig. 3.54: Extracted mass chromatograms from LC/MS analysis of the
synthesised Glucu-SFM solution
3.4.2. Batch experiments
The MITI-tests, following the OECD method (described in chapter 5.4), examined the
degradation of the model substances CBZ, DCF, IBU and SFM in ground water (GW) and
surface water from the river Ruhr affected by the run off of a wastewater sewage plant, but
further on, to simplify matters, only called surface water mix (SWM). The experiments were
prepared parallel in sterile and unsterile water in order to be able to differentiate between
microbiological and chemical conversions.
The composition of the batch solution contains 1 mg/L of the 4 analytes each, dissolved in a
matrix of culture medium, which contains 16 % ground water or the surface water mix. The
culture medium composition was prepared as described in chapter 5.4.
3.4.2.1. The behaviour of the target drugs
Sterile waters
The experiments under sterile conditions showed no decrease in concentration of the active
substances within a month for both ground water and surface water mix (Fig. 3.55). Because
of the distinct results with the target compounds the experiments in sterile waters were
dropped.
Unsterile waters
A strong loss in concentration could be observed for SFM in both batches (Ground water,
surface water mix) after one month. Besides that, in the surface water mix occurred a dramatic
decrease of concentration of IBU after 25 days as shown in the (Fig. 3.56).
Two months later the four tests were measured again to get additional information. In surface
water mix IBU and SFM showed a complete degradation process, while the concentration of
SFM in ground water batch decreased to 26 %, and within the next 2 weeks it was degraded
totally. The concentration of IBU in the ground batch was constant; the degradation of IBU
started after 6 months, the concentration decreased up to 33 % within 35 days. The other
analytes showed constant concentrations in both batches as shown in Fig. 3.56.
In both Fig. 3.55 and Fig. 3.56, there is little gap in the graph after 30 days. The reason is a
necessary exchange of the analytical column. Repeating the test under unsterile surface water
3 4 5 6 7 8 9 10
Time (min)
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100 4,45
6,94
4,87
8,30
Abundance [%]
Glucu-SFM
Glucu-SFM-Me
Glucu-SFM-(Me)2
Glucu-SFM-(Me)3
3 4 5 6 7 8 9 10
Time (min)
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100 4,45
6,94
4,87
8,30
Abundance [%]
Glucu-SFM
Glucu-SFM-Me
Glucu-SFM-(Me)2
Glucu-SFM-(Me)3
67
mix conditions, the result validated the degradation of IBU, which was significantly
dismantled within two weeks. On the other hand the degradation of SFM started two months
later and was completed after two and a half month (Fig. 3.57). The other analytes showed
similar behaviour as shown in the previous tests.
The difference between the two tests may be caused by the differences of water ingredients of
the surface water, taken at different times from the river. In the conventional water treatment
plants the wastewater storage period normally is very short. Therefore, the time isn’t long
enough to degrading these substances significantly. An exception is IBU, which was degraded
depending on the seasonal water ingredient.
3.4.2.2. The behaviour of the synthesised metabolites
Applying the same approach of batch experiments, the synthesised metabolites Glucu-SFM,
Ac-SFM, OH-IBU and Diol-CBZ were investigated in order to compare the behaviour of
degradation with the parent analytes in both water types.
Surface water mix
The results showed that in this water the metabolites Glucu-SFM, OH-IBU, and Ac-SFM
degraded significantly under batch conditions with a different speed (Fig. 3.58); the signal of
Ac-SFM has disappeared completely after 2 weeks, Glucu-SFM after 5 weeks and OH-IBU
after 8 weeks. Only Diol-CBZ kept its concentration value.
Ground water
All the investigated metabolites showed high stability under such water batch conditions (Fig.
3.58).
Fig. 3.55: The influence of sterile batch conditions on the target pharmaceuticals
(GW: ground water, SWM: surface water mix)
0
20
40
60
80
100
120
140
0 40 80 120 160 200
Days
Percentage [%]
SFM-GW CBZ-GW DCF-GW IBU-GW
SFM-SWM CBZ-SWM DCF-SWM IBU-SWM
0
20
40
60
80
100
120
140
0 40 80 120 160 200
Days
Percentage [%]
SFM-GW CBZ-GW DCF-GW IBU-GW
SFM-SWM CBZ-SWM DCF-SWM IBU-SWM
68
Fig. 3.56: The influence of unsterile batch conditions on the target pharmaceuticals
(GW: ground water, SWM: surface water mix)
Fig. 3.57: The influence of unsterile SWM batch condition on the target pharmaceuticals
(trial 2), (GW: ground water, SWM: surface water mix)
0
20
40
60
80
100
120
140
0 20406080100
Days
Percentage [%]
SFM CBZ DCF IBU
0
20
40
60
80
100
120
140
0 20406080100
Days
Percentage [%]
SFM CBZ DCF IBU
0
20
40
60
80
100
120
140
0 40 80 120 160 200
Days
Percentage [%]
SFM-GW CBZ-GW DCF-GW IBU-GW
SFM-SWM CBZ-SWM DCF-SWM IBU-SWM
0
20
40
60
80
100
120
140
0 40 80 120 160 200
Days
Percentage [%]
SFM-GW CBZ-GW DCF-GW IBU-GW
SFM-SWM CBZ-SWM DCF-SWM IBU-SWM
69
Fig. 3.58: The behaviour of the investigated metabolites under unsterile batch conditions
(GW: ground water, SWM: surface water mix)
3.4.3. Column experiments
The column experiments were performed at the pilot plant station by the Institute for Water
Research in Schwerte-Geisecke. The columns set up are shown in (Fig. 3.59) and the
technical dates are given in section 5.4. The interpretation of the tests samples were done at
the Institute For Analytical Sciences in Dortmund.
The goal of the column experiments was the semi-natural simulation of the rinsing water
through the sediment passage. The columns were filled with Rhine sand on which a biofilm
was grown by feeding the columns with ground or surface waters from the river Ruhr.
The columns were operated at the following conditions:
¾ Column I: operated with natural aerobic ground water, which is relative poor in
microorganisms
¾ Column II: operated by conditioning with surface water from river Ruhr
¾ Column III: operated by conditioning with surface water from river Ruhr in addition to
potassium chloride (400 mg/L)
¾ Column IV: repetition of the column II under the same condition (flow path)
¾ Column V: repetition of the column II, but the test solution was cycled in a closed
circuit (closed circuit)
0
20
40
60
80
100
120
140
0 14284256
Days
Percentage [%]
Glucu-SFM-GW Diol-CBZ-GW OH-IBU-GW Ac-SFM-GW
Glucu-SFM-SWM Diol-CBZ-SWM OH-IBU-SWM Ac-SFM-SWM
0
20
40
60
80
100
120
140
0 14284256
Days
Percentage [%]
Glucu-SFM-GW
0
20
40
60
80
100
120
140
0 14284256
Days
Percentage [%]
Glucu-SFM-GW Diol-CBZ-GW OH-IBU-GW Ac-SFM-GW
Glucu-SFM-SWM Diol-CBZ-SWM OH-IBU-SWM Ac-SFM-SWM
70
The content of the microorganisms in surface water was characterised by a comparison with
ground water (chapter 5.4). After stabilization of the hydraulic, chemical and biological
conditions, surface and ground waters were spiked with 100 µg/L each of the four analytes
and sucked through the columns at a flow rate of 1.6 L/h. The period of the test took four
weeks continuously.
3.4.3.1. The behaviour of the drugs
The results of this investigation (column I - III) showed that only the concentration of IBU
decreased significantly in the columns II and III. The other drugs showed high stability under
the adapted conditions. In column I the concentration kept constant.
The sample measurements from column IV showed constant concentration levels during the
test time (Fig. 3.60A). Against that, after the adaptation phase of the first week, column V
showed a retarded decrease in concentration of SFM, DCF and IBU within the first 7 weeks
followed then by a strong degradation (Fig. 3.60B). CBZ nearly kept a constant value over the
test period.
This is a known effect in biodegradation. The microbes first consume carbon source easy to
digest resulting an increasing population. Later on, they attack other carbon sources such as
the drugs.
Fig. 3.59: The set-up of the column experiment, carried out at the
‘Institute for Water Research’ in Schwerte-Geisecke
71
Fig. 3.60: The target pharmaceuticals behaviour under biodegradable conditions
A: Column IV (flow path), B: Column V (closed circuit)
0
50
100
150
200
250
0 14284256
Days
Percentage [%]
SFM CBZ DCF IBU
A
0
50
100
150
200
250
0 14284256
Days
Percentage [%]
SFM CBZ DCF IBU
A
0
20
40
60
80
100
120
140
160
0 1428425670
Days
Percentage [%]
SFM CBZ DCF IBU
B
0
20
40
60
80
100
120
140
160
0 1428425670
Days
Percentage [%]
SFM CBZ DCF IBU
0
20
40
60
80
100
120
140
160
0 1428425670
Days
Percentage [%]
SFM CBZ DCF IBU
B
72
3.4.4. Attempts to identify of degradation products
In order to understand the fate of the target drugs and their synthesised metabolites, samples
from the batch and column tests were successively collected and analysed with LC- and
GC/MS according to the methods developed in chapter 3.3.
3.4.4.1. Degradation products of the target drugs under batch conditions
Because no degradation could be observed in sterile waters, the investigation was
concentrated only on unsterile waters.
Under such batch conditions (ground water, surface water mix) the concentration of IBU and
SFM decreased significantly as mentioned in section 3.4.2.
For IBU the results were very odd; while in ground water the formation of OH-IBU (Fig.
3.28E) and CA-IBU (Fig. 3.61) were detected, in surface water mix no single IBU metabolite
could be detected though the signal from IBU disappeared totally. Against that, SFM showed
the same behaviour in both waters. SFM is partially transformed to Ac-SFM which itself is
then degraded completely (Fig. 3.71). DCF and CBZ didn’t show any formations of
degradation products under these conditions.
3.4.4.2. Degradation products of the target drugs under column conditions
In the samples of the first three columns no IBU metabolites could be observed. Though
column IV ran under the same conditions as Column II now the formation of OH-IBU, CA-
IBU and CA-HA could be detected as shown in (Fig. 3.62). The reason might be a much
greater biological activity caused by the flood water situation in that time. Column V showed
OH-IBU and a very weak signal of CA-IBU. The concentration of OH-IBU in column IV and
V increased with time, but on a very different level (Fig. 3.63). The other metabolite
concentrations showed a gradual decrease. The behaviour of IBU metabolites is in agreement
with the data mentioned in the literature so far [11,35,90].
In contrast to column IV, where no significant degradation of SFM could be detected, Column
V showed a significant decrease in the concentration of SFM and an adequate increase in the
concentration of Ac-SFM (Fig. 3.64). Finally a similar fate was observed for SFM and Ac-
SFM in the column V as we have seen in the batch experiment.
No metabolites were detected for DCF in column IV. On the other hand, the formation of an
unknown DCF-artifact in column V could be found at (Rt: 29.0) in the GC/MS chromatogram
(Fig. 3.65). Its concentration increased during the test phase (Fig. 3.66). Under the applied
conditions CBZ didn’t show the formation of any possible metabolites.
3.4.4.3. Degradation products of the synthesised metabolites under batch condition
Samples from the metabolites batch tests were collected after 5 weeks for further verification
of the degradation products by GC/MS and LC/MS.
In contrast to surface water mix where the degradation process needed 10 weeks, OH-IBU
showed a more retarded degradation in ground water (Fig. 3.67). But none of the expected
transformation compounds could be identified.
73
The investigation of Ac-SFM behaviour in ground water showed a significant decrease in
concentration after 12 weeks (Fig. 3.68); at the same time SFM was detected in a very low
concentration. This concentration was always present in the synthesised Ac-SFM as a residual
contaminant as mentioned in section 3.4.1. In surface water mix Ac-SFM degraded strongly
within 2 weeks as showed in Fig. 3.58. An unknown product from surface water mix batch
was detected in the GC/MS chromatogram at (Rt: 25.94) (Fig. 3.69), but a further
identification is too difficult to predict the resulted product. A control measurement of the
derivatized sample by LC/MS confirmed the formation of an unknown compound at (Rt:
16.88) (Fig. 3.70) It seemed to be a dimethylated SFM, as result from the hydrolysis product
of Ac-SFM (Fig. 3.71).
When applying Glucu-SFM in the ground water batch a signal for SFM appeared after 60
days with increasing tendency (Fig. 3.72). That might be caused by hydrolysis of the
glucuronide moiety. In the surface water mix batch, SFM was also formed from Glucu-SFM
within one and a half week (Fig. 3.71). The formation of SFM was verified with GC/MS.
Later on, the formed SFM decreased gradually and degraded totally after 8 weeks in surface
water mix. Similar to Ac-SFM, the same unknown peak was detected in the GC/MS
chromatogram at (Rt: 25.94) in addition to a new unknown product at (Rt: 22.61) (Fig. 3.73)
in both batch tests. Both unknowns showed dismantling fate in surface water mix. The
derivatized samples were confirmed by LC/MS, the result confirmed the formation of
methylated Ac-SFM, mono-, di- and trimethylated SFM (Fig. 3.74). Possibly these unknown
compounds in GC/MS are derivative products as a result from derivatization process.
Diol-CBZ showed constant concentration levels in both tests with ground water and surface
water mix.
Fig. 3.61: Electron ionization mass spectra of the diazomethane derivative of the IBU
metabolite (CA-IBU) formed under batch conditions
30 60 90 120 150 180 210 240 270 300 330 360
0
50
100
31
43
51
59 77
91
105
117
131
145
155
177 205
230
264
CH3CH3
O
OCH3
O
O
CH3
Abundance [%]
m/z
30 60 90 120 150 180 210 240 270 300 330 360
0
50
100
31
43
51
59 77
91
105
117
131
145
155
177 205
230
264
CH3CH3
O
OCH3
O
O
CH3
Abundance [%]
m/z
74
Fig. 3.62: The formation level of IBU metabolites during the flow path column test (GC/MS)
Fig. 3.63: The formation level of IBU-metabolites during the column tests (GC/MS)
0736 43 57
CA-HA
CA-IBU
OH-IBU
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Days
CA-HA
CA-IBU
OH-IBU
Metabolites
Area [counts]
0736 43 57
CA-HA
CA-IBU
OH-IBU
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Days
CA-HA
CA-IBU
OH-IBU
CA-HA
CA-IBU
OH-IBU
Metabolites
Area [counts]Area [counts]
0
50
100
150
200
250
300
350
01428425670
Days
Area [counts x 103]
OH-IBU - closed circuit column OH-IBU - flow path column
0
50
100
150
200
250
300
350
01428425670
Days
Area [counts x 103]Area [counts x 103]
OH-IBU - closed circuit column OH-IBU - flow path column
75
A)
B)
Fig. 3.64: The transformation behaviour of SFM (A) and the formation level of Ac-SFM
(B) during the column tests (LC/MS)
Fig. 3.65: Electron ionization mass spectra of the diazomethane derivative of an
unknown DCF-artifact (?) formed during closed circuit column test
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450
0
50
100
39 51
59
77 89
103
120
138
151
165
178
193
213
228
241 258
272
300
336
367
Abundance [%]
m/z
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450
0
50
100
39 51
59
77 89
103
120
138
151
165
178
193
213
228
241 258
272
300
336
367
Abundance [%]
m/z
0
5
10
15
20
25
0 1428425670
Days
Area [counts x 107]
flow path column
closed circuit column
0
5
10
15
20
25
0 1428425670
Days
Area [counts x 107]Area [counts x 107]
flow path column
closed circuit column
flow path column
closed circuit columnclosed circuit column
0
5
10
15
20
25
30
0 1428425670
Days
Area [counts x 109]
flow path column
closed circuit column
0
5
10
15
20
25
30
0 1428425670
Days
Area [counts x 109]
flow path column
closed circuit column
76
Fig. 3.66: Formation of an unknown DCF-artifact (?) during the
closed circuit column test (GC/MS)
Fig. 3.67: OH-IBU metabolite behaviour under batch conditions (GC/MS)
(GW: ground water, SWM: surface water mix)
0
5
10
15
20
25
35 42 49 56 63 70 77
Days
Area [counts x10
5]
OH-IBU-GW OH-IBU-SWM
0
5
10
15
20
25
35 42 49 56 63 70 77
Days
Area [counts x10
5]
OH-IBU-GW OH-IBU-SWM
0
500
1000
1500
2000
2500
3000
0 14284256
Days
Area [counts]
0
500
1000
1500
2000
2500
3000
0 14284256
Days
Area [counts]
77
Fig. 3.68: Ac-SFM metabolite behaviour and the formation of SFM under batch conditions
(GC/MS), (GW: ground water, SWM: surface water mix)
Fig. 3.69: Electron ionization mass spectra of the diazomethane derivative of an unknown
Ac-SFM-artifact (?) formed under batch condition in surface water mix
Fig. 3.70: LC-ESI+/MS mass spectra of an unknown methylated Ac-SFM-artifact (?)
formed during batch tests
160 180 200 220 240 260 280 300 320 340
m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Abundance [%]
282,11
283,11
218,39 253,30 284,10
175,28 219,29 254,25 304,13
160 180 200 220 240 260 280 300 320 340
m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Abundance [%]
282,11
283,11
218,39 253,30 284,10
175,28 219,29 254,25 304,13
0
10
20
30
40
50
60
70
40 50 60 70 80
Days
Area [counts x10
4]
Ac-SFM-GW SFM produced in GW
Ac-SFM-SWM SFM produced in SWM
0
10
20
30
40
50
60
70
40 50 60 70 80
Days
Area [counts x10
4]
Ac-SFM-GW SFM produced in GW
Ac-SFM-SWM SFM produced in SWM
30 50 70 90 110 130 150 170 190 210 230 250 270 290
0
50
100
39 55
65
77
90
106
122
133
154
170
176
202
217 281
Abundance [%]
m/z
30 50 70 90 110 130 150 170 190 210 230 250 270 290
0
50
100
39 55
65
77
90
106
122
133
154
170
176
202
217 281
Abundance [%]
m/z
78
Fig. 3.71: The transformation of Glucu-SFM, Ac-SFM and the formation of SFM under batch
condition in surface water mix (LC/UV)
Fig. 3.72: The formation of SFM from Glucu-SFM metabolite under
batch conditions (GC/MS)
Fig. 3.73: Electron ionization mass spectra of the diazomethane derivative of an unknown
Glucu-SFM-artifact (?) formed under batch conditions
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0 14284256708498
Days
Concentration [mg/L]
Glucu-SFM SFM produced from Glucu-SFM
Ac-SFM SFM produced from Ac-SFM
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0 14284256708498
Days
Concentration [mg/L]
Glucu-SFM SFM produced from Glucu-SFM
Ac-SFM SFM produced from Ac-SFM
0
5
10
15
20
25
30
35
40
45
50
35 49 63 77
Days
Area [counts x104]
SFM produced from Glucu-SFM in GW SFM produced from Glucu-SFM in SWM
0
5
10
15
20
25
30
35
40
45
50
35 49 63 77
Days
Area [counts x104]
SFM produced from Glucu-SFM in GW SFM produced from Glucu-SFM in SWM
30 50 70 90 110 130 150 170 190 210 230 250
0
50
100
43
51
55 66
77
91
104
120
131
152
159 168
174
187
200 215
Abundance [%]
m/z
30 50 70 90 110 130 150 170 190 210 230 250
0
50
100
43
51
55 66
77
91
104
120
131
152
159 168
174
187
200 215
Abundance [%]
m/z
79
Fig. 3.74: Extracted mass chromatograms of an LC/MS run, as a result of a
derivatization of a Glucu-SFM sample in surface water mix batch test
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time [min]
0
50
100
0
50
100
0
50
100
0
50
100
0
50
100 7,69
13,36
15,00
16,88
20,02
Abundance [%]
SFM-Me
Ac-SFM-Me
SFM-(Me)2
SFM-(Me)3
Glucu-SFM-Me
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Time [min]
0
50
100
0
50
100
0
50
100
0
50
100
0
50
100 7,69
13,36
15,00
16,88
20,02
Abundance [%]
SFM-Me
Ac-SFM-Me
SFM-(Me)2
SFM-(Me)3
Glucu-SFM-Me
80
3.4.5. Conclusion
In this chapter, the biodegradation was carried out in long-term batch and biofilm reactor
experiments for the target drugs; CBZ, DCF, IBU and SFM. The batches were run in present
of culture medium under ground water (GW) and surface water affected by the run off of a
wastewater sewage plant (surface water mix; SWM) from the river Ruhr, while the biofilm
reactors were run under ground water and non loaded surface water (SW) from the same river.
Moreover, some of their synthesised metabolites; Diol-CBZ, Ac-SFM, Glucu-SFM and OH-
IBU were investigated under unsterile batch conditions in GW and SWM. The batch
experiments were carried out both in sterile and unsterile for the selected drugs.
In order to discriminate between the products from the biodegradation tests and the present
contaminants in the applied analytes, the purity of the commercial standards and of the
synthesised metabolites were studied first by GC/MS and LC-ESI/MS.
The sterile batches revealed no degradation processes for all analytes during the tests time due
to the absence of the bioactivity in the water matrix. Whereas, IBU and SFM were degraded
significantly under unsterile SWM batch condition within 4-14 weeks depending on the
bioactivity in the water matrix. In contrast to that, DCF and CBZ are not readily
biodegradable under the applied operation conditions. Moreover, the synthesised metabolites
are degraded at the same condition significantly in different rates within 2-9 weeks except
Diol-CBZ, which was stable under the applied conditions.
Concerning the GW batches, SFM showed degradation as similar as in SWM but IBU showed
long-term stability. It decreased after 6 months dramatically. Under similar conditions, DCF
and CBZ and the synthesised metabolites showed no degradation.
In the biofilm reactors with GW and SW no degradation of the analytes was observed within
4 weeks. On the other hand, when cycling the test solution in surface water within a closed
circuit column, a significant degradation within 7-10 weeks for all the analytes except CBZ
could be observed.
In order to understand more about the degradation mechanisms, further verification based on
LC- and GC/MS was adapted.
OH-IBU was identified as the major metabolites of IBU biodegradation in the unsterile
ground water batch and surface water biofilm test. The other conditions showed degradation
processes to unknown fates but no IBU metabolites could be identified.
In term of SFM biodegradation, low yield from Ac-SFM in unsterile GW, unsterile SWM and
closed circuit column was observed as transformation product. Eventually, degradation leads
to unknown products. CBZ and DCF didn’t show any transformation products in all batches,
whereas DCF showed the formation of an unknown-artifact in closed circuit column as
detected by GC/MS.
OH-IBU as synthesised metabolites significantly degraded to unknown products in both
batches. Whereas, Ac-SFM showed the formation of SFM in SWM batch. Additionally, in
SWM batch an unknown signal was detected by GC/MS. Ultimately, the whole components
degraded to unknown fate. Glucu-SFM degraded in a similar way as Ac-SFM. In GW and
SWM batches SFM was detected as a transformed product from Glucu-SFM, in addition to
two unknown products were detected by GC/MS.
81
4. Summary and conclusions
In this work the attention was turned to investigate the possibility of an application of
particular animal intestines as natural membranes for the depletion of target drugs in water
treatment and for sample preparation prior to specific laboratory analysis. Also, analytical
methods for the simultaneous qualitative and quantitative determination of the selected
analytes and further identification of unknown products should be developed based on liquid
and gas chromatography and mass spectrometry. Moreover, long-term biodegradation of the
analytes at particular conditions in batches and biofilm reactors should be investigated to
estimate their fate in the aquatic environments.
The target drugs were carbamazepine (CBZ), diclofenac (DCF), ibuprofen (IBU) and
sulfamethoxazole (SFM), furthermore, 10,11-dihydro-10,11-dihydroxycarbamazepine (Diol-
CBZ), N-1-glucuronidesulfamethoxazole (Glucu-SFM), N-4-acetylsulfamethoxazole (Ac-
SFM) and 2-hydroxyibuprofen (OH-IBU) as some of their main metabolites. The analytes
were chosen as leading drugs for the present work according to the amount consumed in
medicine, high concentrations occurrence in the aquatic environment and the varieties of
physical and chemical characters.
Within the first part of the presented work it could be shown that membranes of natural
intestines behave like dialysis membranes. A very favourable intestine is the so-called
Goldschlägerhäutchen, a special part of the cattle appendix, which has optimal permeation
characteristics and besides that, it is commercially available in large sizes in different formats
and with low costs. Permeation processes by means of Goldschlägerhäutchen are mainly
influenced by the concentration gradient, surface area and stirring velocity. Water ingredients
such as surfactants can interfere the permeation process but only at unusual high
concentrations. Humic substances interfere the permeation process of some analytes when its
in young genesis. The depletion of the drugs can reach > 90 % by combination of the
permeate solution with additional solid or liquid phase extraction.
When comparing the animal intestines to technical membranes, no clogging effects could be
observed during the permeation through the natural membrane, even if the water sample
contained a complex mixture of other water ingredients.
The stability of the natural membranes was not sufficient for all water matrices depending on
the bioactivity. Particularly wastewater promoted instability for this membrane type. So, they
might only be applied to technical application in combination with disinfection or filtration
steps. A modifying treatment of the intestine with formaldehyde could extend the stability up
to 10 days. But the best way would be to reduce the bioactivity of water.
Due to slow permeation kinetics the natural membranes cannot be recommended for
analytical purpose, such as clean-up procedures.
In the second part, GC/MS and LC-ESI/MS quantification methods were developed based on
a pre-concentration step by solid phase extraction and analytical derivatization prior to the
GC/MS analysis.
For sample preparation many attempts were done to find out the most proper adsorbent and
derivatization agent. Oasis HLB as polymeric sorbent showed superior extraction recoveries
for the target drugs and diazomethane proved to be the best reagent for derivatization.
82
A direct comparison of GC/MS and LC-ESI/MS displayed that the latter may have an
advantage for the analysis of the extreme polar analytes due to incomplete derivatization of
polar functional groups. Whereas, the LC-ESI/MS method is especially matrix dependent
caused by high amounts of organic and inorganic ions in the sample, which lead to signal
suppression.
In term of linearity, reproducibility and accuracy, both methods are applicable to the intended
aims.
Applying the final procedures, the recoveries of the spiked analytes at constant sample
volume (1 L) and concentration (1 µg/L) were 79-111 % for ultrapure water samples by
means of GC/MS and 60-85 % using LC-ESI/MS. For surface water the recoveries were 91-
129 % based on GC/MS and with LC-ESI/MS 46-78 %. The method detection limits were 1-5
ng/L for GC/MS (except SFM and its metabolites) and 3-5 ng/L in LC-ESI/MS for all
analytes.
The developed methods were applied to real aqueous samples from the river Ruhr. All the
target analytes could be detected in the surface water in a concentration range from 100 to 320
ng/L, except Glucu-SFM.
Thirdly, long-term biodegradation experiments in batch and in biofilm reactors were carried
out in pilot scales at particular conditions for the selected drugs and some of their main
metabolites.
The drugs in different batches showed different behaviour depending on the type of matrices
and bioactivity as well. IBU and SFM were degradated significantly but in different rates.
CBZ and DCF were resistant to the degradation processes under the applied conditions. The
main metabolites were also degradated in a similar way to the parent drugs.
Except for CBZ, the biofilm reactor experiments significantly revealed a decline in the
concentration of all target drugs at different rates.
Further samples were collected successively in order to follow the degradation of the target
drugs and possible formation of metabolites. Different batches revealed the formation of OH-
IBU as a major metabolite from IBU. Also Ac-SFM was identified as a metabolite from SFM
under the applied conditions. Whereas, the human main metabolites did not show any
remarkable degradation products under the batch conditions except the formation of SFM
resulting from Ac-SFM and Glucu-SFM as well.
As a consequence, in long-term batch experiments IBU and SFM could be degradated under
certain conditions. However, IBU, SFM and DCF were significantly degradated under
specific biofilm reactor conditions. In contrast, CBZ was resistant to all applied conditions.
Additionally, in batch experiments the degradation of the main metabolites showed similar
behaviour compared to parent drug.
The batch and biofilm reactors have been proved to be suitable model systems especially for
ground and surface waters in order to investigate the pathways of the biological degradation
of active drugs as well as their main metabolites under particular conditions.
83
5. Experimental
5.1. Chemicals and materials
All chemicals and solvents were of purity high grades and used without further purification as
described in (Table 5.1). Ultrapure water was prepared with a Seral-Pur Delta UV apparatus
(USF Seral, Ransbach, Germany).
Different types of animal intestines were applied in the original wet form or dried as
membrane.
• Pig : (small intestine; A91M, mast intestine with internal fat layer)
• Sheep : (small intestine; Nova ESS Kal 55/30, Nova ESS Sheets 50x50)
• Cattle : (special appendix part; Goldschlägerhäutchen)
The preferred membranes were special parts of the cattle appendix distributed under the trade
name ‘Goldschlägerhäutchen’ (Jürging GmbH, Versmold, Germany). The intestine was stored
under sodium chloride and was washed before use.
5.2. Sample preparation
The stock solutions of the selected drugs (25 mg/L) were prepared in ultrapure water.
Dissolution of the analytes was always done in the ultrasonic bath at 40 ºC for half an hour.
a) Sample clean up and extraction
The suspended particles in the aqueous samples were vacuum filtered through cellulose
membrane filter (0.45µm).
A 100 mg Oasis HLB cartridge was conditioned first with 6 mL MeOH, 6 mL MeOH-
EtOAC (50:50 v/v), and 3 mL MeOH, followed by 12 mL acidic water (pH ≈ 2). The
filtered water samples were sucked through the cartridge with a flow-rate of 5
mL/min. followed by washing with 3mL fresh water. Then the cartridges were dried
with vacuum under a gentle stream of nitrogen. The analytes were eluted from the
cartridge first by 3 mL MeOH under ground gravity, and followed by 3mL MeOH:
EtOAC (50:50 v/v). The elutes were collected in a 10 mL test tube and concentrated to
about 300 µL by using a heating block at 40 °C under a gentle stream of nitrogen.
Then the sample was divided in 2 halves in order to analyse the same sample with GC
and LC methods.
For LC/MS analysis the sample was dried and the residue redissolved to a final volume of 200
µL with buffer A (10 mM Ammonium acetate (pH 4)) containing 10 % MeOH. For GC
analysis the sample was dried and the following derivatization step was carried out in an
adequate organic solvent.
b) Derivatization procedures applied
1) Diazomethane (CH2N2):
84
Diazomethane is a yellow gas but is used in the form of an ethereal solution. Its reacts
with an organic acid in the following manner,
R
O
OH H+R
OH
OH CH2N N
-
- N2ROH
OCH3-H+R
OMe
O
++
+
1 mL diazomethane solution was added to the analytes in a 1.5 mL vial. It was capped tightly
and then the reaction was allowed to occur within 60 minutes in the refrigerator. Later on the
solution was evaporated under a light stream of nitrogen at room temperature. The residue
was dissolved with EtOAC for GC/MS analysis.
The hint for security must be followed because diazomethane is carcinogenic and can be
extremely unstable with a risk of explosion. All reactions should be carried out in a fume
hood and any stored solutions of diazomethane in diethyl ether should be restricted to a
maximum volume of 100 mL and kept in a refrigerator [84].
Diazomethane preparation: [85]
4.2 g Diazald (N–methyl-N–nitrous-p-toluene sulfonamide) was dissolved in 50 mL
diethyl ether in a separator funnel. 10 g NaOH was dissolved in 40 mL H2O and 30
mL MeOH in a round-bottom flask. The special distillation apparatus with clear seals
was connected tightly in a fume hood. The round bottom flask was warmed up to 40-
45 °C by the heating mantle. The diazald solution was dropped from the separators
funnel into the sodium hydroxide solution very slowly. The released diazomethane gas
was collected in the ice bath as a diethyl ether fraction.
2) Trimethylsulfonium hydroxide (TMSH): [86]
The analytes were redissolved with EtOAC and then 5 µL TMSH was added; the solution
became a little bit turbid, therefore, 1 µL (10 % acetic acid in EtOAC) was added to consume
the excess of TMSH. It’s suitable to measure it directly with GC/MS.
3) Pentafluorobenzyl bromide (PFBB): [87]
The analytes were redissolved with 1 mL acetone, then they were treated with 10 µL PFBB,
30 µL 0.1 mg/mL dicyclohexyl-18-crown-6) and 10 mg powdered potassium carbonate
(K2CO3). The vial was capped tightly and the reaction was allowed to occur within 30
minutes at 60 °C. After that the solution was isolated and dried under nitrogen. The residue
was dissolved with EtOAC for GC/MS analysis.
4) Silylation reagent (I): [88]
Mixture I: 95 % N,O-bis(trimethylsilyl)acetamide (BSA) and 5 % trimethyl-
chlorosilane (TMCS)
The residue was redissolved with 100 µL from mixture I and then the reaction solution was
heated for 2 h (120 °C). Later on it was dried under nitrogen and the rest was redissolved with
EtOAC for GC/MS analysis.
85
5) Silylation reagent (II): [89]
Mixture II: N-methyl-N- (trimethylsilyl) trifluoroacetamide (MSTFA)/trimethyl-
silylimidazol (TMSI)/dithioerytrite (DTE), 1000 µL/2µL/2mg (30 min at
80 °C).
The residue was redissolved with 50 µL of mixture II and then it was heated for 30 min in a
water bath (60 °C). Later on it was dried under nitrogen and the rest was redissolved with n-
hexane for GC/MS analysis.
5.3. Instrumentation parameters
a) UV-VIS-NIR
The absorbance was measured by using a Cary 5 G UV-VIS-NIR spectrophotometer (Varian,
Darmstadt) at varied wavelengths, for the HUS a wavelength of 254 nm was applied.
b) LC-UV
Method I
An analytical method based on HPLC (Sykam, Gilching) and UV-detection (Spectra-Physics,
Darmstadt) has been developed.
Wavelength : 225 nm
Analytical column : Nucleosil-120 C-18 (5 µm, 125 x 4 mm i.d., Knauer, Berlin)
Pre-column : Nucleosil-120 C-18 (5 µm, 10 x 4 mm i.d., Knauer, Berlin)
Mobile phase : Acetonitrile/NaH2PO4-buffer (55/45 v/v)
Flow rate : 0.8 mL/min (isocratic elution)
Injection volume : 5 or 20 µL
Method II
Using the same conditions (Method I) with the exception of the mobile phase
Mobile Phase :10 mM NH4CH3COO: ACN 75:25 (v/v) (pH: 6.0).
Method III
The analytical method used to analyse the PEG tests was based on using HPLC (Merck,
Germany) and RI detection (Water, USA).
Stationary phase : (300 x 16 mm i.d.) column was packed with Fractogel TSK
HW-40S (particle size 20-40 µm, pore size 50 Å and separation
range PEG: 100-3000; Grom GmbH)
Mobile phase : Ultrapure water
Flow rate : 1 mL/min
86
c) LC-ESI/MS
The LC/MS system was a LCQ Deca (Thermo, USA) coupled with a spectrasystem gradient
pump (P4000; Thermo) and a spectrasystem autosampler (AS 3000; Thermo).
Analytical column : Aquasil C-18 (5 µm, 100 Ǻ, 150 x 0.32 mm i.d.,
ThermoHypersil)
Mobile phases : Mobile A: 10 mM Ammonium acetate (pH = 4.13) adjusted
with concentrated acetic acid
Mobile B: 10 mM Ammonium acetate in 98 % (ACN: MeOH
40:60)
Gradient : Eluent B: 25% to 60% for 10 min, increased to 90 % within 4
min, held for 5 min at 90 %, then ramped back to 25 % in 5
min and held finally for 6 min at 25 %
Flow rate : 13 µL/min; Split 200 µL/min, the mobile phases were degassed
with an online spectrasystem degasser (SCM 1000) using
helium
Injection volume : 10 or 20 µL
Spray voltage : 3.5 kV (+ mode) and 3.2 kV (- mode)
Capillary temperature : 200 °C
Sheath gas flow rate : 0.3 L/min N2
Sheath liquid : 4 µl/min (isopropanol/water (5:1))
d) GC/MS
Separation and detection of the analyte was carried out with the GC/MS system HP G1800A
GCD with autosampler 7673A (Hewlett Packard, Ratingen).
Capillary column : A fused-silica (HP-5MS; 30 m; 0.25 mm ID; 0.25 µm FD)
Injector : Split/splitless, splitless time of 2 minutes
Injector temperature : 250 ˚C
njection volume : 1 µL
Carrier gas : Helium
Flow rate : 1 mL/min
Temperature program : Oven = 60 °C (1.5 min), first ramp: 20 ˚C/min to 120 ˚C,
second ramp 4˚C/min to 160 ˚C, third ramp 12 ˚C/min to 250
˚C and then hold 12 min at 250 ˚C
Analysis time : 34 min
Interface temperature : 280 ˚C
Scan mode : Full scan range (m/z = 30-450)
Ionization mode : EI+/ 70 eV
87
5.4. The conditions of the batches and biofilm reactors
The batch tests followed the guidelines of the ‘Organization for Economic Cooperation and
Development’ (OECD) published in 1993 [108]. The guidelines represent a standardized
method, specially the MITI-Test, to test the biodegradation. The principle of this test is the
determination of the biodegradation potential of chemicals under favourable conditions.
Simply, the pharmaceuticals provide as carbon source. They are dissolved in culture medium
and are addited to the water matrix containing the natural bioactivity.
The test was carried out with the model substances CBZ, DCF, IBU and SFM in ground water
and from the river Ruhr surface water affected by the run off of a wastewater sewage plant.
The experiments were prepared parallel in sterile and unsterile water.
Culture medium:
65.25 mg K2HPO4, 25.5 mg KH2PO4, 133.8 mg Na2HPO4.2H2O, 5.1 mg NH4Cl, 67.5
mg MgSO4.7H2O, 82.5 mg CaCl2 and 0.75 mg FeCl3.6H2O; the final volume: 1 L.
Batch solution:
1 mg/L of the 4 analytes each were dissolved in the culture medium containing 16 %
real water.
The column experiments were performed at the pilot plant station by the Institute for Water
research in Schwerte-Geisecke. The columns set up are shown in (Fig. 3.59).
The technical data for the biofilm reactor were as follows:
Column height : 100 cm
Column diameter : 22 cm
Filter height : 80 cm
Flow rate : 1.6 L/h
Backing material : Rhine sand (0.2-2 mm particle size; River Rhine, Germany)
The columns were filled with Rhine sand as supported material on which the biofilm was
grown by feeding the columns with ground or surface waters.
The content of the microorganisms in surface water was characterised by a comparison with
ground water as the following:
1 mL water sample is given on a plate with culture medium DEV (Merck Nr.
1.11471.5000). One plate of each probe is incubated at a temperature of 20+/-2°C and
one plate at 36+/-1°C. After 44+/-4 h incubation time the colony number of each plate
is counted (visible colonies at a 6fold to 8fold loupe amplification).
After the hydraulic, chemical and biological conditions had been stabilized the real waters
were spiked with 100 µg/L of the analytes each and sucked through the columns.
88
Table 5.1: Chemicals and materials used in the present work
Chemical Supplier
Acetic acid Aldrich
Ammonium acetate Aldrich
Ammonium chloride Merck
Ammonium hydroxide Merck
ß-Cyclodextrine Fluka
Calcium chloride Merck
Diazald Aldrich
Dicyclohexyl-18-crown-6 Merck
Ethylenediaminetetraacetic acid Merck
Ferric chloride- 6-hydrate Merck
Hydrochloric acid Baker
Magnesiumsulfate-7-hydrate Merck
Polyethylene glycol Fluka
Polysorbate (TWEEN 80) Sigma
Potassium carbonate Merck
Potassiumdihydrogenphosphate Merck
Potassiumhydrogenphosphate Merck
Sodium dodecylsulfate Merck
Sodium hydroxide Fluka
Sodium sulfate Fluka
Sodiumdihydrogenphosphate Merck
Sodiumhydrogenphosphate-2-hydrate Merck
Sulfuric acid Merck
Tris(2-ethylhexyl)-phosphate Merck
Solvent
Acetone Suprasolv Merck
Acetonitrile Rotisolv HPLC Roth
Decan puriss. Fluka
Dichloromethane Suprasolv Merck
Diethylether Seccosolv Merck
Ethyl acetate Pestilyse Roth
Formaldehyde purum Merck
Methanol Pestilyse Roth
n-Decanol purum Fluka
n-Hexane Suprasolv Merck
Octanol purum Merck
Reference Compound
Carbamazepine Ehrenstorfer
Diclofenac Ehrenstorfer
Ibuprofen Ehrenstorfer
Sulfamethoxazole Ehrenstorfer
89
Reference Compound Supplier
3,3',4,4',5,5'-hexachlorobiphenyl (PCB 169) Ehrenstorfer
N-4-Acetyl-sulfamethoxazole Synthesised at University of Paderborn
10,11-dihydro-10,11-dihyroxycarbamazepine Synthesised at University of Paderborn
2-Hydroxyibuprofen Synthesised at University of Paderborn
N-1-Glucuronide sulfamethoxazole Synthesised at University of Paderborn
Hohlohsee 13 Natural material from Hohloh Lake (Germany)
Venner Moor Natural material from Arnsberger Wald (Germany)
Reagent
Dithioerytrite Fluka
N,O-bis(trimethylsilyl)acetamide Fluka
N-methyl-N- (trimethylsilyl) trifluoroacetamide Fluka
Pentafluorobenzyl bromide Aldrich
Trimethylchlorosilane Merck
Trimethylsilylimidazol Analyt
Trimethylsulfonium hydroxide Machery&Nagel
Material
Cellulose ester dialysis membrane Reichelt Chemietechnik GmbH
Goldschlägerhäutchen Jürging GmbH
Membrane filter (0.45µm) Schleicher-Schüll
SPE Material
Charcoal Merck
Lichrolute EN Merck
Oasis HLB Waters
Octadecasilane (C-18) Restek GmbH
Equipment
Dialyse Chamber Self-production
Heating and drying block Techne DRI-Block DB-20
Micro magnetic stirrers H+P Labortechnik AG
90
6. References
1. K. Kümmerer, Pharmaceuticals in the environment, Springer Verlag: Berlin, 2004,
ISBN 3-540-21342-2.
2. T. Ternes, Occurrence of drugs in German sewage treatment plants and rivers, Water
Research, 1998, Vol. 32, No. 11, pp. 3245-3260.
3. C. Daughton, T. Ternes, Pharmaceuticals and personal products in the environment:
agents of subtle change?, Environmental Health Perspectives, December; 1999, Vol. 107,
Supplement 6, pp. 907–938.
4. D. Kolpin, E. Furlong, M. Meyer, E. Thurman, S. Zaugg, L. Barber, H. Buxton,
Pharmaceuticals, hormones and other organic wastewater contaminants in U.S.
Stream1990-2000, Environ. Sci. Technol., 2002, 36, pp. 1202-1211.
5. S. Richardson, Water analysis: Emerging contaminants and current issues, Anal. Chem.,
2003, 75, pp. 2831-2857.
6. T. Heberer, Occurrence, Fate, and removal of pharmaceutical residues in the aquatic
environment: A review of recent research data, Toxicology letters, 2002 131, pp. 5-17.
7. B. Halling-Sørensen, S. Nielsen, P. Lanzky, F. Ingerslev, H. Lützhøft, S. Jørgensen,
Occurrence, fate and effects of pharmaceuticals substance in the environment,
Chemosphere, 1998, 36, No.2, pp. 357-393.
8. T.A. Ternes, R.W. Hirsch, M. Stumpf, T. Eggert, B.F. Schuppert, K. Haberer: Nachweis
und Screening von Arzneimittelrückständen, Diagnostika und Antiseptika in der
aquatischen Umwelt, Abschlußbericht zum BMBF-Forschungsvorhaben 02WU9667/3,
ESWE-Institut für Wasserforschung und Wassertechnologie GmbH, Mainz, 1999.
9. I. Buerge, T. Poiger, M.D. Müller, H.R. Buser: Caffeine, an anthropogenic marker for
wastewater contamination of surface waters, Environ. Sci. Technol., 2003, 37, pp. 691-
700.
10. T.A. Ternes, R. Hirsch: Occurrence and behavior of X-ray contrast media in sewage
facilities and the aquatic environment, Environ. Sci. Technol., 2000, 34, pp. 2741-2748.
11. C. Zwiener, F.H. Frimmel, Short-term tests with a pilot sewage plant and biofilm
reactors for the biological degradation of the pharmaceutical compounds clofibric
acid, ibuprofen and diclofenac, The Science of the Total Environment, 2003, 309, pp.
201-211.
12. H. Stan, T. Heberer, Analusis Mag., 1997, 27, pp. 20-23.
13. K. Kümmerer, Drugs in the environment: emission of drugs, diagnostic aids and
disinfectants into wastewater by hospital in relation to other sources- a review,
Chemosphere, 2001, 45, pp. 957-969.
14. D. Kolpin, E. Furlong, M. Meyer, E. Thurman, S. Zaugg, H. Buxton, Environ. Sci.
Technol., 2002, 36, pp. 1202-1211.
91
15. T. Ternes, Pharmaceuticals: Occurrence in rivers, groundwater and drinking
water, In: International Seminar: Pharmaceuticals in the Environment,
Technological Institute, Section on Environmental Technology, Brussels, 2000.
16. J. Pawliszyn, Sampling and Sample Preparation for Field and Laboratory, Elsevier
Science B.V., Amsterdam, 2002, ISBN 0-444-50510-5.
17. J. Jönssen, L. Mathiasson, Membrane extraction in analytical chemistry, J. Sep. Sci.,
2001, 24, pp. 495–507.
18. Sedlak, J. Gray, K. Pinkston, Understanding microcontamants in recycled water,
Environ. Sci. Technol., 2000, 34, pp. 508A-515A.
19. S. D. Raeissi, Drug Transport and Metabolism in In Vitro Models of Human Intestine
(Dissertation). Acta Universitatis Upsaliensis, Uppsala, 1998, ISBN 91-554-4341-9.
20. T. Vree, Y. Hekster, Clinical pharmacokinetics of sulfonamides and their metabolites,
Karger: Basel, 1987.
21. C. Lin, C. Chang, W. Lin, J. of chromatography A, 1997, 768, pp. 105-112.
22. C. Metcalfe, B. Koenig, D. Bennin, M. Servos, T. Ternes, R. Hirsch, Occurrence neutral
and acidic drugs in the effluents of Candian sewage treatment plants, Environ. Toxicol.
Chem., 2003, Vol. 22, No. 12, pp. 2872-2880.
23. A. Göbel, C. McArdll, M. Suter, and W. Giger, Trace determination of macrolide and
sulfonamide antimicrobials, a human sulfonamide metabolite, and trimethoprim in
wastewater using liquid chromatography coupled to electrospray tandem mass
spectrometry, Anal. Chem., 2004, 76, pp. 4756-4764.
24. X. Miao, C. Metcalfe, Determination of Carbamazepine and its metabolites in aqueous
samples using liquid chromatography-Electrospray mass spectrometry, Anal. Chem.,
2003, 75, pp. 3731-3738.
25. C. Metcalfe, X Miao, B Koenig, J. Struger, Distribution of acidic and neutral drugs in
surface waters near sewage treatment plants in the lower Great Lakes and Canada,
Environ. Toxicol. Chem., 2003, 22, pp. 2881-2889.
26. X. Miao, C. Metcalfe, Pharmaceuticals analysis in aqueous using positive and negative
voltage switching microbore LC-ESI-MS/MS, J. mass spectrum., 2003, 38, pp. 27-34.
27. X. Miao, C. Metcalfe, Determination of cholesterol-lowering statin drugs in aqueous
samples using liquid chromatography-electrospray ionization mass spectrometry, J. of
chromatography A, 2003, 998, pp. 133-41.
28. W. Hua, X. Miao, E. Bennett, C. Metcalfe, R. Letcher, Neutral and acidic
pharmaceuticals and major triazine herbicides in wastewater effluent and surface waters
from the upper Detroit River (to be published), 2004.
29. W. Hua, S. Jasim, E. Bennett, S. Mazloum, R. Letcher, The effect of ozone versus
conventional treatment processes on neutral and acidic pharmaceuticals and atrazine on
92
Detroit River drinking water for the city of Windsor, Canada, (to be published), 2004.
30. T. Doll, F. Frimmel, Fate of pharmaceuticals–photodegradation by simulated solar UV-
light, Chemosphere, 2003, 52, pp. 1757- 1769.
31. M. Stumpf, T. Ternes, K. Haberer, P. Seel, W. Baumann, Determination of
pharmaceuticals in sewage plants and river water, Vom Wasser, 1996, 86, pp. 291-203.
32. M. Richardson, J. Bowron, The fate of pharmaceuticals in the aquatic environment, J.
Pharm. pharmacol., 1985, 37, pp. 1-12.
33. S. Öllers, H. Singer, P. Fässler, S. Müller, Simultaneous quantification of neutral and
acidic pharmaceuticals and pesticides at the low ng/l in surface and wastewater, J. of
Chromatography A, 2001, 911, pp. 225-234.
34. U. Jux, R. Baginski, Hans Arnold, M. Krönke, P. Seng, Detection of pharmaceuticals
contaminations of river, pond, and tap water from cologne (Germany) and surroundings,
Int. J. Hyg. Environ. Health, 2002, 205, pp. 393-398.
35. C. Zwiener, S. Seeger, T. Glauner, F. Frimmel, Metabolites from the biodegradation of
pharmaceutical residues of ibuprofen in biofilm reactors and batch experiments, F.
Frimmel, Analytical And Bio analytical chemistry, 2002, 372, pp. 569-575.
36. N. Clarke, D. Rindgen, W. Korfmacher, K.Cox, Systematic LC/MS Metabolite
Identification in Drug Discovery A four-step strategy to characterize metabolites by
LC/MS techniques early in the pharmaceutical discovery process, Anal. Chem.,
August 1, 2001, pp. 430 A-439 A.
37. P. Glue, R. Clement, Cell Mol. Neurobiol., 1999, 19, pp. 309-323.
38. D. Clarke, B. Burchell, The uridine diphosphate glucuronosyltransferase multigene
family: Function and regulation, in Conjugation-deconjugation reactions in drug
metabolism and toxicity, F. C. Kauffman (Ed.), Springer-Verlag, Berlin-Heidelberg
(1994), pp. 3-43.
39. S. Caccia: Metabolism of the newer antidepressants - An overview of the
pharmacological and pharmacokinetic implications, Clin. Pharmacokinet., 1998, 34,
pp. 281-302.
40. K. Lertratanangkoon, M. Horning, Drug Metab. Dispos., 1982, 10, pp. 1-10.
41. N. Kitteringham, C. Davis, N. Howard, M. Pirmohamed, B. Park, J. Pharmacol. Exp.
Ther., 1996, 278, pp. 1018-1027.
42. P. Myllynen, P. Pienimäki, H. Raunio, K. Vähäkangas, Hum. Exp. Toxicol., 1998, 12,
pp. 668-676.
43. O. Pelkonen, P. Myllynen, P. Taavitsainen, Carbamazepine: a blind assessment
of CYP- associated metabolism and interactions in human liver-derived in vitro
systems, Xenobiotica, 2001, Vol. 31, No. 6, pp. 321-343.
93
44. Ciba-Geigy (Hrsg.) 1994: Voltaren® (Diclofenac)-twenty years of clinical experience-an
update, Basel.
45. J. Faigle, I. Böttcher, J. Godbillon, H. Kriemler, E. Schlumpf, W. Schneider, A.
Schweizer, H. Stierlin, T. Winkler, A new metabolites of diclofenac sodium in human
plasma, Xenobiotica, 1998, 18 (10), pp. 1191-1197.
46. J. Kenny, J. Maggs, X. Meng, D. Sinnott, S. Clarke, B. Park, A. Stachulski, Syntheses
and characterization of the acyl Glucuronide and Hydroxy metabolites of diclofenac, J.
Med. Chem., 2004, 47, pp. 2816-2825.
47. R. Mills, S.Adams , E. Cliffe, W. Dickinson, J. Nicholson, Xenobiotica, 1973, 3, pp. 589-
598.
48. M. Spraul, M. Hofman, P. Dvortsak, J. Nicholson, I. Wilson, Anal. Chem., 1993, 65, pp.
327-330.
49. D. Kepp, U. Sidelmann, J. Tjørnelund, S. Hansen, J. Chromatography B, 1997, 696, pp.
235-241.
50. F. von Bruchhausen, G. Dannhardt, S. Ebel, A.W. Frahm, E. Hackenthal, R. Hänsel, U.
Holzgrabe, K. Keller, E. Nürnberg, H. Rimpler, G. Schneider, P. Surmann, H.U. Wolf,
G. Wurm: Hagers Handbuch der pharmazeutischen Praxis, Band 8, Springer Verlag.
51. DAD 9-Kommentar (1986): Deutsches Arzneibuch, 9. Ausgabe 1986 mit
wissenschaftlichen Erläuterungen, Hartke, K. Mutschler, E. (Hrsg.), Wissenschftliche
Verlagsgesellschaft GmbH Stuttgart, Govi-Verlag GmbH Frankfurt.
52. Fachinformation Cotrim-ratiopharm® (September 1997).
53. Fachinformation Kepinol® (July 1997).
54. R. Pfleger (Hrsg.) 1993: Kepinol®/ Kepinol® Fortährte Die bewährte antimikrobielle
Chemotherapeutika-kombination in praxis und Klinik. Eine wissenschaftliche Information
für den Arzt, Bamberg.
55. T. Vree, Y. Hekster, pharmacokinetics of sulfonamides revisited, Karger: Basel, 1985.
56. Degen, W. Dieterle, W. Schneider, W. Theobald, U. Sinterhauf, Pharmacokinetics of
diclofenac and five metabolites after single dose in healthy volunteers and after repeated
doses in patients, Xenobiotica, 1988, 18(12), pp. 1449-1455.
57. Hutt, J. Caldwell, J. Pharm. Pharmacol., 1983, 35, pp. 693-704.
58. Rückstände von Arzneimitteln in Wasserproben Befunde und deren Bewertung
aus Sicht der Trinkwasserversorgung, DVGW-Schriftenreihe Wasser Nr. 94.
59. H. Buser, T. Pioger, M. Müller, Occurrence and fate of the pharmaceutical drug
diclofenac in surface water: Rapid photo degradation in a lake, Environ. Sci.
Technol., 1998, 32, No. 22, pp. 3449-3456.
94
60. T. Poiger, H. Buser, M. Müller, Photodegradation of the pharmaceutical during
diclofenac in a lake: Pathway, field measurements, and mathematical modeling,
Environ. Toxicol. Chem., 2001, 20, No. 2, pp. 256-263.
61. W. Riess, H. Stierlin, J. Faigle, U. Geiger, A. Gerardin, F. Schmid, J. Wagner, W.
Theobald, The pharmacokinetics of diclofenac in animals and man, 1975. In FJ.
Wagenhäuser, ed, A. Voltaren, New Non-Steroid Antirheumatic Agent (Diclofenac). Hans
Huber Publishers, Bern, Switzerland, pp. 19-28.
62. D. Moore, S. Robert-Thomson, D. Zhen, C. Duke, Photochemical studies on the anti-
inflammatory drug diclofenac, Photochem. Photobiol., 1990, 52, pp. 685-690.
63. S. Encinas, F. Bosca, M. Miranda, Phototoxicity associated with diclofenac: A
photophysical, photochemical, and photobiolohical study on the drug and its
photoproducts, Chem. Res. Toxicol., 1998, 11, pp. 946-952.
64. M. Stumpf, T.A. Ternes, K. Haberer, W. Baumann: Isolierung von Ibuprofen-
Metaboliten und deren Bedeutung als Kontaminanten der aquatischen Umwelt, Vom
Wasser, 1998, 91, pp. 291-303.
65. C. Tixier, H. Singer, S. Oellers, S. Müller, Occurrence and fate of Carbamazepine,
clofibric acid, Diclofenac, ibuprofen, ketoprofen, and naproxen in surface water, Environ.
Sci. Technol., 2003, 37, No. 6, pp. 1061-1068.
66. M. Farre, I. Ferrer, A. Ginebreda, M. Figueras, L. Olivella, L. Tirapu, M. Vilanova, D.
Barcelo, Determination of drugs in surface water and wastewater sapmles by liquid
chromatography-mass spectrometry: methods and preliminary results including toxicity
studies with vibrio fisheri, J. of Chromatography A, 2001, 938, pp. 187-197.
67. R. Andreozzi, R. Marotta, G. Pinto, A. Pollio, Carbamazepine in water: persistence in the
enivronment, Ozonation treatment and preliminary assessment on algal toxicity, Water
Research, 2002, 36, pp. 2869-2877.
68. M. Carballa, F. Omil, J. M. Lema, M. Klompart, C. Carcia-Jares, I. Rodriguez, M. Gomez,
T. Ternes, Behavior of pharmaceuticals, cosmetics and hormones in a sewage treatment
plant, Water Research, 2004, 38, pp. 2918-2926.
69. V. Suling, V. Wohlers, M. Reinhard, W. Thiemann, Photooxidation by UV irradiation
and treatment with ionised air of selected antibiotics, Vom Wasser, 2002, 98, pp. 145-
158.
70. http://www.litwak.de and http://www.nfzpronat.de.
71. R. Hofheinz , in www.aerztezeitung.de/docs/1998/11/25/215a2201.asp (1998).
72. L. Madara, J. Trier, Functional morphology of the mucosa of the small intestine. In:
Johnson, L.R. (Eds.), Physiology of the Gastrointestinal Tract. Raven Press, New York,
1994, pp. 1577-1622.
73. J. Pappenheimer, K. Reiss, Contribution of solvent drag through intercellular junctions to
absorption of nutrients by the small intestine of the rat, J. Membr. Biol., 1987, 100, pp.
95
123-136.
74. P. Butron, R. Conradi, R. Hilgers, Mechanism of peptide and protein absorotion. 2.
Transcellular mechanism of peptide and protein absorption: passive aspects, Drug Deliv.
Res., 1991, pp. 365-386.
75. J. D. Guthrie, A. L. Bullock, Ind. Eng. Chem., 1960, 52, pp. 935.
76. Y. Zhang, K. Agarwal, M. Beylot, M. Soloviev, F. David, M. Reider, K. Tserng,
H. Brunengraber, Assay of the Acetyl CoA Probe Acetyl Sulfamethoxazole and
of sulfamethoazole by gas chromatography-mass spectrometry, Analytical
Biochemistry, 1993, 212, pp. 481-486.
77. H. Buser, T. Poiger, M. Müller, Occurrence and Environmental behavior of the chiral
pharmaceuticals drug, Ibuprofen in surface water and in wastewater, Buser, Environ. Sci.
Technol. 1999, 33, pp. 2529-2535.
78. G. Whitehouse, R. Dreyer, M. Yamashita, J. Fenn, Electrospray interface for liquid
chromatographs and mass spectrometers, Anal. Chem., 1985, 57, 1985, pp. 675.
79. K. Dost, D. Jones, G. Davidson, Determination of sulfonamides by packed column
supercritical fluid chromatography with atmospheric pressure chemical ionization mass
spectrometric detection, Analyst, 2000, 125, pp. 1243-1247.
80. C. Hartig, T. Storm, M. Jekel, Detection and identification of sulphonamide drugs in
municipal waste water by liquid chromatography coupled with electrospray ionisation
tandem mass spectrometry, Journal of Chromatography A, 1999, Vol. 854, pp. 163-173.
81. X. Miao, B. Koenig, C. Metcalfe, Analysis of acidic drugs in the effluents of sewage
treatment plants using liquid chromatography-electrospray ionisation tandem mass
spectrometry, J. of Chromatography A, 2002, 952, pp. 139-147.
82. F. Sacher, F. Lange, H. Brauch, I. Blankenhorn, Pharmaceuticals in ground waters
Analytical methods and results of a monitoring program in Baden-Württemberg, Germany,
J.of Chromatography A, 2001, Vol. 938, pp. 199-210.
83. R. Pascoe, J. Foley, A. Gusev, Anal. Chem., 2001, 73, pp. 6014-6023.
84. P.W. Scott, http://www.laboratorytalk.com/books/chem/chrom/rs_2/rs_2_73.html.
85. J. Nolte, B. Jonke, Determination of nitrophenols in water by GC/MS after Solid-Phase
Extraction, Vom Wasser, 2000, 94, pp. 191-201.
86. C. Zwiener, F. Frimmel, Oxidation treatment of pharmaceuticals in water, Water
Research, 2000, Vol. 34, No. 6, pp. 1881-1885.
87. C. Chien, M. Charles, K. Sexton, H. Jeffries, Analysis of Airborne carboxylic acid and
phenol as their Pentafluorobenzyl derivative: Gas chromatography / ion trap
spectrometry with a novel chemical ionization reagent, PFBOH, Environ. Sci. Technol.,
1998, 32, pp. 299-309.
96
88. T. Ternes, M. Meisenheimer, D. Mcdowell, F. Sacher, H. Brauch, B. Gulde, G. Preuss, U.
Wilme, N. Seibert, Removal of pharmaceuticals during drinking water treatment,
Environ. Sci. Technol., 2002, 36, No. 17, pp. 3855-3863.
89. T. Ternes, Method for analysis the Antiepileptics Carbamazepine and Primidone in water
using GC/MS after derivatization, Vom Wasser, 2000, 94, pp. 203-212.
90. C. Zwiener, F. Frimmel, Biodegradation of pharmaceutical residues investigated by SPE-
GC/ITD-MS and online derivatization, J. high resol. Chromatogr., 2000, 23(7/8), pp. 474-
478.
91. T. Ternes, Analytical method for determination of pharmaceuticals in aqueous
environmental sample, Trend in analytical chemistry , 2001, Vol. 20, No. 8, pp. 410-434
2001.
92. S. Snyder, B. Vanderford, R. Pearson, O. Quinones, D. Rexing, Endocrine disrupters and
pharmaceuticals analysis using direct injection LC/MS/MS, proceeding of the AWWA
Water Quality Technology Conference, 02-04 November 2003, Philadelphia, PA.
93. R. Loos, R. Niessner: Analysis of aromatic sulfonates in water by solid-phase
extraction and capillary electrophoresis, J. Chromatography A , 1998, 822, pp. 291-303.
94. T.A. Ternes, M. Stumpf, B. Schuppert, K. Haberer: Simultanous determination of
antiseptics and acidic drugs in sewage and river water, Vom Wasser 1998, 90, pp. 295-
309.
95. T. Renner, D. Baumgarten, K.K. Unger: Analysis of organic pollutants at trace levels
using fully automated solid-phase extraction coupled to high performance liquid
chromatography, Chromatographia, 1997, 45, pp. 199-207.
96. O. Fiehn, M. Jekel: Comparison of sorbents using semipolar to highly hydrophilic
compounds for a sequential solid-phase extraction procedure of industrial wastewaters,
Anal. Chem., 1996, 68, pp. 3083-3089.
97. A. Avdeef, C. Berger, C. Brownell, Pharm. Res., 2000, 17, pp. 85-89.
98. ACD-Physico-Chemical Laboratory, available from <http://www.acdlabs.com>.
99. R. W. Baker, Membrane Technology and applications, McGraw-Hill, 2000, ISBN
0-07-135440-9.
100. International Report “Water Reuse”, IWSA World Congress, 1997, Blackwell Science
Ltd.
101. D. Sedlak, K. Pinkston, Factors Affecting the Concentrations of Pharmaceuticals
Released to the Aquatic Environment, Water Resources Update, Universities Council on
Water Resources, Issue No. 120: September, 2001, pp. 56.
102. B. L. Karger, L. R. Snyder, C. Horvath, An introduction to separation science, John
Wiley & Sons, 1973, pp. 469-495, ISBN 0-471-45860-0.
97
103. O. Jones, N. Voulvoulis, J. Lester, Analytical method development for the simultaneous
determination of five human pharmaceuticals in water and wastewater sample by gas
chromatography-mass spectrometry, Chromatographia, 2003, 58, October No. 7/8, pp.
471-477.
104. P. Burba, H. Geltenpoth, J. Nolte, Ultra filtration behaviour of selected pharmaceuticals
on natural and synthetic membranes in the presence of humic-Rich hydrocolloids,
Current pharmaceuticals analysis, (to be submitted), 2004.
105. M. Grote, B. Haciosmanoglu, M. Bataineh, J. Nolte, Separation of drug traces from
water with particular membrane systems, J. Environ. Sci. Health Part A, 2004,Vol. A39,
No. 4, pp. 1035-1049.
106. M. Grote, A. Vockel, D. Schwarze, A. Mehlich, M. Freitag, Fate of antibiotics in food
chain and environment originating from pig fattening (part 1), Fresenius
Environmental Bulletin, 2004, Vol. 13, No. 11b, pp. 1216-1224.
107. C. Richard, B. Cole, Electrospray ionization mass spectrometry, John Wiley & Sons,
1997, ISBN 0-471-14564-5.
108. OECD (Organization for Economic Co-operation and Development) (1993): OECD
Guidelines for the testing of chemicals, Vol. 2, Part 3, 302C, Paris.
98
7. Publications and presentations of the present work
I) Publications
1. M. Grote, B. Haciosmanoglu, M. Bataineh, J. Nolte, Separation of drug traces from
water with particular membrane systems, J. Environ. Sci. Health Part A, Vol. A39, No.
4, pp. 1035 - 1049, 2004.
2. J. Nolte, M. Bataineh, B. Haciosmanoglu, M. Grote, Membrane systems developed for the
separation of pharmaceutical residues from water, in: M. Cox: Ion Exchange Technology
for Today and Tomorrow, SCI-Verlag, London, 2004, pp. 259-266, ISBN 0-901001-85-6.
3. B. Haciosmanoglu, M. Grote, J. Nolte, M. Bataineh, Separation of drug traces from
contaminated water with particular membrane systems, Proceedings of the ‘ISWA World
Environment Congress & Exhibition’ Istanbul, 8 - 12.7.2002 Eds.: G. Kocasoy, T.
Atabarut, I. Uholgu; In: Appropriate Environmental and Solid Waste Management and
Technologies for Developing Countries 3, 2002, pp. 1801-1808, ISBN 975-518-179-2.
II) Oral presentations
1. M. Bataineh, M. Grote, B. Haciosmanoglu, J. Nolte, Extraction of pharmaceuticals
from water by use of natural flat- und liquid-membrane systems, Trends in sample
preparation, 29.6 - 4.7.2002, Graz, Austria.
2. B. Haciosmanoglu, M. Grote, J. Nolte, M. Bataineh, Separation of drug traces from
contaminated water with particular membrane systems, ISWA 2002 World Environmental
Congress and Exhibition, 8 - 12.7.2002, Istanbul, Turkey.
3. M. Bataineh, J. Nolte, B. Haciosmanoglu, M. Grote, Membrane systems developed for
the separation of pharmaceutical residues from water, IEX 2004, 4 - 7.7.2004,
Cambridge, UK.
III) Poster presentations
1. J. Nolte, M. Bataineh, U. Marggraf, H. Geltenpoth, M. Grote, Development of a method
for the enrichment of pharmaceuticals from waters using selected natural membranes,
Euroanalysis-12, 8 - 13.9.2002, Dortmund, Germany.
2. M. Bataineh, U. Marggraf, H. Geltenpoth, M. Grote, J. Nolte, Nachweis von Diclofenax -
Oligromeren mittels ESI-MS, Jahrestreffen der Deutschen Gesellschaft für
Massenspektrometrie (DGMS 2003), 10 - 12.03.2003, Münster, Germany.
3. M. Bataineh, U. Marggraf, H. Geltenpoth, M. Grote, J. Nolte, Nachweis vom Diclofenac-
Oligomeren mittels ESI-MS, Kooperationsformum Innovation der wwi NRW 30.3.2004,
Mülheim a. d. Ruhr.
4. M. Bataineh, B. Haciosmanoglu, H. Geltenpoth, M. Grote, J. Nolte, Separation of
Selected Pharmaceuticals from Water by Natural Solid and Liquid Membrane
Systems, Euromembrane 2004, 27.9 - 1.10.2004, Hamburg, Germany.
5. M. Bataineh, M. Grote, W. Nigge, J. Nolte, Comparative study on the determination of
selected drugs and their metabolites using LC/ESI-MS and GC/MS, 21st LC/MS
Montreux symposium, 10 - 12.11.2004, Montreux, Switzerland.
