Application of comprehensive gas chromatography (GCxGC) to measurements of volatile organic species in ambient air

The focus of this thesis was the in-situ application of the new analytical technique "GCxGC" in both the marine and continental boundary layer, as well as in the free troposphere. Biogenic and anthropogenic VOCs were analysed and used to characterise local chemistry at the individual measurement sites. The first part of the thesis work was the characterisation of a new set of columns that was to be used later in the field. To simplify the identification, a time-of-flight mass spectrometer (TOF-MS) detector was coupled to the GCxGC. In the field the TOF-MS was substituted by a more robust and tractable flame ionisation detector (FID), which is more suitable for quantitative measurements. During the process, a variety of volatile organic compounds could be assigned to different environmental sources, e.g. plankton sources, eucalyptus forest or urban centers. In-situ measurements of biogenic and anthropogenic VOCs were conducted at the Meteorological Observatory Hohenpeissenberg (MOHP), Germany, applying a thermodesorption-GCxGC-FID system. The measured VOCs were compared to GC-MS measurements routinely conducted at the MOHP as well as to PTR-MS measurements. Furthermore, a compressed ambient air standard was measured from three different gas chromatographic instruments and the results were compared. With few exceptions, the in-situ, as well as the standard measurements, revealed good agreement between the individual instruments. Diurnal cycles were observed, with differing patterns for the biogenic and the anthropogenic compounds. The variability-lifetime relationship of compounds with atmospheric lifetimes from a few hours to a few days in presence of O3 and OH was examined. It revealed a weak but significant influence of chemistry on these short-lived VOCs at the site. The relationship was also used to estimate the average OH radical concentration during the campaign, which was compared to in-situ OH measurements (1.7 x 10^6 molecules/cm^3, 0.071 ppt) for the first time. The OH concentration ranging from 3.5 to 6.5 x 10^5 molecules/cm^3 (0.015 to 0.027 ppt) obtained with this method represents an approximation of the average OH concentration influencing the discussed VOCs from emission to measurement. Based on these findings, the average concentration of the nighttime NO3 radicals was estimated using the same approach and found to range from 2.2 to 5.0 x 10^8 molecules/cm^3 (9.2 to 21.0 ppt). During the MINATROC field campaign, in-situ ambient air measurements with the GCxGC-FID were conducted at Tenerife, Spain. Although the station is mainly situated in the free troposphere, local influences of anthropogenic and biogenic VOCs were observed. Due to a strong dust event originating from Western Africa it was possible to compare the mixing ratios during normal and elevated dust loading in the atmosphere. The mixing ratios during the dust event were found to be lower. However, this could not be attributed to heterogeneous reactions as there was a change in the wind direction from northwesterly to southeasterly during the dust event.

The formation of OH radicals from Criegee intermediates: a LIF-FAGE study from laboratory to ambient

Das Hydroxyl Radikal ist, auf globalem Maßstab, das bedeutendste Oxidant in der Atmosphäre. Es initiiert den Abbauprozess vieler, teilweise schädlicher, Spurengase und insbesondere den von flüchtigen Kohlenwasserstoffen (VOC). Die OH Konzentration ist somit ein gutes Maß für die augenblickliche Selbstreinigungskapazität der Atmosphäre. Messungen zu nächtlicher Zeit mit LIF-FAGE-Instrumenten (engl.: laser-induced fluorescence - fluorescence assay by gas expansion) haben Konzentrationen des Hydroxylradikals (OH) ergeben, die signifikant höher waren, als sich mit der bekannten Chemie erklären ließ. Um herauszufinden, ob ein solches Signal wirklich atmosphärisches OH ist oder von einer störenden Spezies stammt, die im Messinstrument OH produziert, wurde das LIF-FAGE-Instrument des Max-Planck-Instituts für Chemie (MPIC) im Rahmen dieser Doktorarbeit modifiziert und getestet. Dazu wurde ein so genannter Inlet Pre-Injector (IPI) entwickelt, mit dem in regelmäßigen Abständen ein OH-Fänger in die Umgebungsluft abgegeben werden kann, bevor das OH vom Instrument erfasst wird. Mit dieser Technik ist es möglich, ein Hintergrund-OH (OHbg), d. h. ein im Instrument erzeugtes OH-Signal, vom gemessenen OH-Gesamtsignal (OHtot) zu trennen. Die Differenz zwischen OHtot und OHbg ist die atmosphärische OH-Konzentration (OHatm). Vergleichsmessungen mit der hier entwickelten Technik, dem IPI, in zwei verschiedenen Umgebungen mit Instrumenten basierend auf Massenspektrometrie mit chemischer Ionisation (CIMS, engl.: chemical ionization mass spectrometry) als alternativer Methode des OH-Nachweises, zeigten eine weitgehende Übereinstimmung. Eine umfassende Beschreibung des Systems zur Ermittlung der Ursache des OHbg hat ergeben, dass es weder von einem Artefakt des Instruments noch von hinlänglich bekannten und beschriebenen LIF-FAGE-Interferenzen stammt. Zur Bestimmung der Spezies, die das OHbg-Signal verursacht, wurden verschiedene Laborstudien durchgeführt. Die Arbeit im Rahmen dieser Doktorarbeit hat ergeben, dass das LIF-FAGE-Instrument leicht auf OH reagiert, das beim monomolekularen Zerfall stabilisierter Criegee-Intermediate (SCI) im Niederdruckbereich des Instruments gebildet wird. Criegee-Intermediate oder Carbonyloxide entstehen bei der Ozonolyse ungesättigter flüchtiger Kohlenwasserstoffverbindungen (VOC, engl.: volatile organic compounds) und können daher in der Umgebungsluft vorkommen. Anhand von Tests mit verschiedenen Verweilzeiten der SCI im Niederdruckbereich des Instruments in Verbindung mit einem detaillierten Modell mit der neuesten SCI-Chemie wurde die monomolekulare Zerfallsgeschwindigkeit von 20  10 s-1 für den syn-Acetaldehyd-Oxykonformer bestimmt. Der in Feldkampagnen gemessene OHbg-Wert wurde dahingehend untersucht, ob SCI die Quelle des beobachteten Hintergrund-OH im Feld sein könnten. Das Budget für die SCI-Konzentration, das für die Kampagnen HUMPPA-COPEC 2010 und HOPE 2012 berechnet wurde, ergab eine SCI-Konzentration zwischen ca. 103 und 106 Molekülen pro cm3. In der Kampagne HUMPPA-COPEC 2010 ergab die Schwefelsäurekonzentration, dass die OH-Oxidation von SO2 allein die gemessene H2SO4-Konzentration nicht erklären konnte. In dieser Arbeit konnte gezeigt werden, dass das Hintergrund-OH mit dieser ungeklärten Produktionsrate von H2SO4 korreliert und somit die Oxidation von SO2 durch SCI als mögliche Erklärung in Frage kommt. Ferner korreliert das Hintergrund-OH in der Kampagne HOPE 2012 mit dem Produkt aus Ozon und VOC und konnte mit SO2 als SCI Fänger entfernt werden. Qualitativ zeigen wir somit, dass das in der Umgebungsluft gemessene Hintergrund-OH wahrscheinlich durch den monomolekularen Zerfall von SCI verursacht wird, doch sind weitere Studien notwendig, um die quantitativen Beziehung für diese Spezies und dem Hintergrund-OH in unserem Instrument zu bestimmen.

Development, characterization, and application of flowing atmospheric-pressure afterglow ionization for mass spectrometric analysis of ambient organic aerosols

Addressing current limitations of state-of-the-art instrumentation in aerosol research, the aim of this work was to explore and assess the applicability of a novel soft ionization technique, namely flowing atmospheric-pressure afterglow (FAPA), for the mass spectrometric analysis of airborne particulate organic matter. Among other soft ionization methods, the FAPA ionization technique was developed in the last decade during the advent of ambient desorption/ionization mass spectrometry (ADI–MS). Based on a helium glow discharge plasma at atmospheric-pressure, excited helium species and primary reagent ions are generated which exit the discharge region through a capillary electrode, forming the so-called afterglow region where desorption and ionization of the analytes occurs. Commonly, fragmentation of the analytes during ionization is reported to occur only to a minimum extent, predominantly resulting in the formation of quasimolecular ions, i.e. [M+H]+ and [M–H]– in the positive and the negative ion mode, respectively. Thus, identification and detection of signals and their corresponding compounds is facilitated in the acquired mass spectra. The focus of the first part of this study lies on the application, characterization and assessment of FAPA–MS in the offline mode, i.e. desorption and ionization of the analytes from surfaces. Experiments in both positive and negative ion mode revealed ionization patterns for a variety of compound classes comprising alkanes, alcohols, aldehydes, ketones, carboxylic acids, organic peroxides, and alkaloids. Besides the always emphasized detection of quasimolecular ions, a broad range of signals for adducts and losses was found. Additionally, the capabilities and limitations of the technique were studied in three proof-of-principle applications. In general, the method showed to be best suited for polar analytes with high volatilities and low molecular weights, ideally containing nitrogen- and/or oxygen functionalities. However, for compounds with low vapor pressures, containing long carbon chains and/or high molecular weights, desorption and ionization is in direct competition with oxidation of the analytes, leading to the formation of adducts and oxidation products which impede a clear signal assignment in the acquired mass spectra. Nonetheless, FAPA–MS showed to be capable of detecting and identifying common limonene oxidation products in secondary OA (SOA) particles on a filter sample and, thus, is considered a suitable method for offline analysis of OA particles. In the second as well as the subsequent parts, FAPA–MS was applied online, i.e. for real time analysis of OA particles suspended in air. Therefore, the acronym AeroFAPA–MS (i.e. Aerosol FAPA–MS) was chosen to refer to this method. After optimization and characterization, the method was used to measure a range of model compounds and to evaluate typical ionization patterns in the positive and the negative ion mode. In addition, results from laboratory studies as well as from a field campaign in Central Europe (F–BEACh 2014) are presented and discussed. During the F–BEACh campaign AeroFAPA–MS was used in combination with complementary MS techniques, giving a comprehensive characterization of the sampled OA particles. For example, several common SOA marker compounds were identified in real time by MSn experiments, indicating that photochemically aged SOA particles were present during the campaign period. Moreover, AeroFAPA–MS was capable of detecting highly oxidized sulfur-containing compounds in the particle phase, presenting the first real-time measurements of this compound class. Further comparisons with data from other aerosol and gas-phase measurements suggest that both particulate sulfate as well as highly oxidized peroxyradicals in the gas phase might play a role during formation of these species. Besides applying AeroFAPA–MS for the analysis of aerosol particles, desorption processes of particles in the afterglow region were investigated in order to gain a more detailed understanding of the method. While during the previous measurements aerosol particles were pre-evaporated prior to AeroFAPA–MS analysis, in this part no external heat source was applied. Particle size distribution measurements before and after the AeroFAPA source revealed that only an interfacial layer of OA particles is desorbed and, thus, chemically characterized. For particles with initial diameters of 112 nm, desorption radii of 2.5–36.6 nm were found at discharge currents of 15–55 mA from these measurements. In addition, the method was applied for the analysis of laboratory-generated core-shell particles in a proof-of-principle study. As expected, predominantly compounds residing in the shell of the particles were desorbed and ionized with increasing probing depths, suggesting that AeroFAPA–MS might represent a promising technique for depth profiling of OA particles in future studies.