Citation
Allelochemicals That Modify Mosquito Host-Seeking Behavior: Gas Chromatography/Mass Spectrometry and High-Field Asymmetric-Waveform Ion Mobility Spectrometry

Material Information

Title:
Allelochemicals That Modify Mosquito Host-Seeking Behavior: Gas Chromatography/Mass Spectrometry and High-Field Asymmetric-Waveform Ion Mobility Spectrometry
Creator:
SANTOS, SAMARET M. OTERO ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Chickens ( jstor )
Desorption ( jstor )
Electric fields ( jstor )
Electric potential ( jstor )
Electrodes ( jstor )
Feathers ( jstor )
Ions ( jstor )
Mass spectrometers ( jstor )
Signals ( jstor )
Skin ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Samaret M. Otero Santos . Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2008
Resource Identifier:
659874730 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

1 ALLELOCHEMICALS THAT MODIFY MOS QUITO HOST-SEEKING BEHAVIOR: GAS CHROMATOGRAPHY/MASS SPECTROMETR Y AND HIGH-FIELD ASYMMETRICWAVEFORM ION MOBILI TY SPECTROMETRY By SAMARET M. OTERO SANTOS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 Copyright 2007 Samaret M. Otero Santos

PAGE 3

3 To my dear mom

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Richar d A. Yost, for giving me the opportunity to join his research group and for hi s support and guidance. He is an excellent advisor who taught me how to be a leader and an independent thinker. It was a pleasure to work with this project and with our collaborators at the U.S Department of Agriculture . A special thanks to Dr. Ulrich Bernier for all of his help in the developm ent of the methodology for the GC/MS experiments and for sharing his knowledge about mosquito attr actants. Thanks to Dr . Sandra Allan, from the USDA, for answering questions rela ted to mosquitoes and West Ni le Virus. Thanks to Erin Vrzal for finding the time to help me with the co llection of samples from chickens. I must thank Dr. Brian Quinn, also from the USDA, for help ing me with the sorbent/solvent extraction methodology and for suggesting ideas for improving the method. I must also thank Dr. Matthew Booth for answering questions about GC/MS, for helping in troubleshooting the GC/MS instruments in our lab, and for letting me use hi s equipment for the thermal desorption of chicken feathers. I cannot forget to thank Dr. David Powell at the ma ss spectrometry facility in the Chemistry Department for letting me use the GC/M S instrument in his lab. I thank Dr. Maria C. Dancel for reserving time for me to use the inst rument in Dr. Powell’s lab and for helping me with the experiments. I am thankful for my friend and ex-colleague Dr. Alisha Mitchell-Roberts from whom I learned not only about science but also about daily things in life, for always editing my written work, and for sharing her knowledge of FAIMS. Sh e always was there for me when I needed her the most. I give special thanks to my colleague Dodge Baluya. I am always going to be grateful to him for being so kind and working his “mag ic” around the instruments even when he was not a user of the instrument. I also would like to thank Dr. Timothy Garrett for taking the time to read through the chapters and giving me great suggestions. I cannot forget Erick Molina, who

PAGE 5

5 has helped me with this work for the past ye ar and suggested a better idea for working with headspace samples. I must also thank Antonjia Macek for being so kind and having the patience to check for grammatical errors throughout this dissertation. Of course I cannot forget my family and frie nds for always believing in me. They have always given me lots of encouragement when I needed it the most. I must thank my friend Michelle Rodriguez for being my family away from home and with whom I shared great memories. Thanks to my wonderful family. I could not imagine going through graduate school without their support. Finally, my greatest thanks are extended to my mother, Carmen M. Santos, for being a hard worker and taking care of my two brothers and me and for guiding us through difficult times. Thanks to her, I am the pe rson that I am today. I will always be grateful to her for encouraging me to obtain more education.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 ABSTRACT....................................................................................................................... ............13 CHAPTER 1 INTRODUCTION..................................................................................................................15 Research Overview.............................................................................................................. ...15 Mosquito Host-Seeking Behavior...........................................................................................16 Sensilla of Mosquitoes....................................................................................................16 Short and Long Range Signals........................................................................................17 Olfactory Cues.................................................................................................................18 Carbon dioxide.........................................................................................................20 1-octen-3-ol..............................................................................................................20 Human emanations...................................................................................................21 Animal emanations other than octenol.....................................................................22 Overview of Analytical Methods............................................................................................22 Gas Chromatography/Mass Spectrometry (GC/MS).......................................................23 Quadrupole Ion Trap Mass Spectrometer........................................................................23 Quadrupole Mass Spectrometer......................................................................................26 Ion Mobility Spectrometry..............................................................................................27 Overview of Dissertation....................................................................................................... .28 2 IDENTIFICATION OF CANDIDATE ATTRACTANT COMPOUNDS FROM AVIAN HOSTS FOR THE MOSQ UITO VECTOR OF WEST NILE VIRUS BY GAS CHROMATOGRAPHY/MASS SPECTROMETRY............................................................35 Introduction................................................................................................................... ..........35 Overview of Analytical Methods............................................................................................36 Thermal Desorption Methods..........................................................................................36 Sorbent/Solvent Extraction Method................................................................................37 Separation..................................................................................................................... ...37 Ionization..................................................................................................................... ....38 Electron ionization (EI)............................................................................................38 Chemical ionization (CI)..........................................................................................39 Characteristic EI ion fragmentation.........................................................................40 Experimental................................................................................................................... ........41 Thermal Desorption of Glass Beads................................................................................41 Sorbent/Solvent Extraction Method................................................................................43

PAGE 7

7 Direct Thermal Desorption of Chicken Feathers Method...............................................43 Results and Discussion......................................................................................................... ..44 Thermal desorption of glass beads following cryofocusing GC/MS..............................44 Sorbent/Solvent Extraction Method................................................................................48 Direct Thermal Desorption of Chicken Feathers.............................................................50 Conclusions.................................................................................................................... .........51 3 OPTIMIZATION OF HIGHFIELD ASYMMETRIC-WAVEFORM ION MOBILITY SPECTROMETRY.................................................................................................................69 High-Field Asymmetric-Waveform Ion Mobility Spectrometry (FAIMS)............................69 Ionalytics Selectra Unit...................................................................................................73 Compensation Voltage (CV)...........................................................................................73 Experimental................................................................................................................... ........74 Results and Discussion......................................................................................................... ..75 Effect of Curtain Gas Flow Rate.....................................................................................75 Effect of Dispersion Voltage (DV)..................................................................................77 Effect of Carrier Gas Composition..................................................................................77 Conclusions.................................................................................................................... .........80 4 CHARACTERIZATION OF ALLOMONE S USING ATMOSPHERIC PRESSURE CHEMICAL IONIZATION/HIGH-FIELD ASYMMETRIC-WAVEFORM ION MOBILITY SPECTROMETR Y/MASS SPECTROMETRY................................................98 Introduction................................................................................................................... ..........98 N,N-diethyl-3-methylbenzamide (DEET).......................................................................99 3,7-dimethylocta-1,6-dien-3-ol (linalool)........................................................................99 Vapor Samples.................................................................................................................99 Atmospheric Pressure Chemical Ionization (APCI)......................................................100 Finnigan LCQ................................................................................................................101 Mass Analysis on the LCQ............................................................................................102 Experimental................................................................................................................... ......103 Standard Solutions.........................................................................................................104 Headspace Analysis.......................................................................................................104 Results and Discussions........................................................................................................105 APCI/FAIMS/MS & APCI/FAIMS/MS/MS................................................................105 Methanolic solutions..............................................................................................105 Headspace analysis.................................................................................................107 Quantitative Studies.......................................................................................................110 Methanolic solutions..............................................................................................110 Headspace...............................................................................................................111 Conclusions.................................................................................................................... .......112 5 CONCLUSION AND FUTURE WORK.............................................................................128 Conclusions.................................................................................................................... .......128 Future Work.................................................................................................................... ......132

PAGE 8

8 LIST OF REFERENCES.............................................................................................................135 BIOGRAPHICAL SKETCH.......................................................................................................142

PAGE 9

9 LIST OF TABLES Table page 2-1 Characteristic EI fragment ation ions for different cl asses of compounds used for identification of components from the GC/MS analyses...................................................53 2-2 Compounds tentatively identified from chicken skin emanations.....................................58 2-3 Compounds found to be pres ent in chicken feathers.........................................................66 3-1 Chemical structures and molecular wei ghts for the two allom ones studied in this work........................................................................................................................... ........87 4-1 Chemical structures and molecular wei ghts for the two allom ones studied in this work........................................................................................................................... ......114

PAGE 10

10 LIST OF FIGURES Figure page 1-1 Quadrupole ion trap (QIT) composed of three electrodes; two end-cap electrodes and a ring electrode............................................................................................................... ....29 1-2 A three dimensional ideal ion trap showing the dimensions r0 and z0, where r0 2 = 2z0 2............................................................................................................................... ......30 1-3 Mathieu stability diagram for a quadrupole ion trap mass spectrometer...........................31 1-4 Quadrupole ma ss spectrometer..........................................................................................32 1-5 Mathieu stability diagram for a quadrupole mass spectrometer........................................33 1-6 A conventional ion mobility spectrometer (IMS)..............................................................34 2-1 The thermal desorption of glass bead s followed cryofocusing GC/MS method used for analyzing chicken skin emanations..............................................................................54 2-2 The sorbent/solvent extraction method us ed for analyzing compounds from chicken feathers by GC/MS............................................................................................................55 2-3 Typical total ion chromatogram (TIC) of compounds thermally desorbed from of 10 glass beads rubbed on chicken skin followed analysis by GC/MS on a Guardian ZB5ms ........................................................................................................................... .........56 2-4 Typical TIC of compounds thermally desorbed from 10 glass beads rubbed on chicken skin followed analysis by GC/MS on a ZB-wax column.....................................57 2-5 Mass spectra corresponding to glycerol; peak detected at 25.31 minutes using the ZB-wax column.................................................................................................................60 2-6 Mass spectra of 1-amino1H-pyrrole-2,5-dione; peak de tected at 25.10 minutes using the ZB-wax column............................................................................................................61 2-7 Mass spectra of oleyl al cohol; peak detected at 2 7.99 minutes using the ZB-wax column......................................................................................................................... .......62 2-8 Mass spectra of 3-isobutylhexahydropyrro lo[1,2-a]pyrazine-1,4-di one; peak detected at 37.05 minutes using the ZB-wax column .....................................................................63 2-9 Mass spectra of 3,5,7-trihydroxy-2H1-be nzopyran-2-one; peak detected at 43.93 minutes using the ZB-wax column....................................................................................64 2-10 The TIC from the EI analysis of com pounds from chicken feathers extracted from Anasorb-747 using carbon disulfide..................................................................................65

PAGE 11

11 2-11 Chromatogram from the EI analysis of compounds thermally desorbed from chicken feathers....................................................................................................................... ........68 3-1 Effect of the electric field strength on the ion drift velocity at low field in conventional IMS...............................................................................................................81 3-2 Effect of the electric field strength, E , on the ion drift velocity at high electric fields in FAIMS....................................................................................................................... ....82 3-3 Plots for the hypothetical dependence of ion mobility on electric field strength for three types of ions............................................................................................................ ..83 3-4 Ion motion between two parallel plates dur ing the application of an electric field in FAIMS.......................................................................................................................... .....83 3-5 The behavior of ions at low a nd high electric fields in FAIMS.........................................84 3-6 Asymmetric waveform used in FAIMS. The peak maximum of the waveform is called the dispersion voltage (DV)....................................................................................85 3-7 FAIMS cell design used in experiments............................................................................86 3-8 Line-of-sight FAIMS cell de sign used in experiments......................................................86 3-9 Total ion chromatogramcompensation voltage (TIC-CV) and ion selectivecompensation voltage (IS-CV) spectra of 1 ppm standard of linalool and DEET.............88 3-10 Effect of the curtain gas flow rate on th e signal intensity for DEET and linalool ions.....89 3-11 Effect of the curtain glass flow rate on the compensation voltage for DEET and linalool ions.................................................................................................................. ......90 3-12 Effect of increasing the dispersion vo ltage on the signal inte nsity of DEET and linalool....................................................................................................................... ........91 3-13 Effect of increasing the dispersion vo ltage on the compensation voltage of DEET.........92 3-14 Effect of increasing the dispersion volta ge on the compensation voltage of linalool.......93 3-15 Compensation voltage as a function of increasing He content in the N2 carrier gas for DEET and linalool.............................................................................................................94 3-16 Compensation voltage as a function of increasing CO2 content in the N2 carrier gas for DEET and linalool........................................................................................................95 3-17 Signal intensity as a function of increasing He content in the N2 carrier gas for DEET and linalool................................................................................................................... ......96

PAGE 12

12 3-18 Signal intensity as a function of increasing CO2 content in the N2 carrier gas for DEET and linalool.............................................................................................................97 4-1 The atmospheric pressure chemical i onization (APCI) proce ss used for LC/MS...........115 4-2 Finnigan LCQ APCI source used in all experiments.......................................................115 4-3 Instrumentation set-up used for th e characterization of the allomones by APCI/FAIMS/MS............................................................................................................116 4-4 The TIC-CV spectrum for a mixture of 1 ppm each of DEET and linalool in methanol....................................................................................................................... ....117 4-5 The APCI/FAIMS/MS spectrum from area under the peak at a CV of -10.6 V..............117 4-6 The IS-CV spectra obtained by selecting exclusively the ions with m/z 192 and 137 of DEET and linalool.......................................................................................................118 4-7 The APCI/FAIMS/MS/MS spectra. A) APCI/FAIMS/MS/MS spectrum of DEET, obtained by setting the CV at -10.5 V, B) APCI/FAIMS/MS/MS spectrum of linalool, obtained by setting the CV at -8.4 V.................................................................119 4-8 The APCI/FAIMS/MS/MS/MS spectrum using as precursor ion the m/z 119...............120 4-9 Comparison of IS-CV spectra (m/z 192, the [M+H]+] ion) of a 10 ppm DEET standard in methanol........................................................................................................121 4-10 The APCI/FAIMS/MS spect rum of peak at -2.1 V.........................................................122 4-11 The APCI/FAIMS/MS spectrum of the peak at -8.8 V...................................................122 4-12 The APCI/FAIMS spectra of the head space of a mixture of 5 ppm DEET and 1000 ppm of linalool................................................................................................................ .123 4-13 The APCI/FAIMS/MS calibration curv es for DEET standard solutions.........................124 4-14 The IS-CV and mass spectrum of a 100 ppb DEET standard in methanol analyzed by APCI/FAIMS/MS............................................................................................................125 4-15 The APCI/FAIMS/MS calibration curves for linalool standard solutions ranging from 100 ppb to 100 ppm................................................................................................126 4-16 Concentration dependence of APCI/FAI MS/MS response for the headspace of DEET standards...................................................................................................................... ....127

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ALLELOCHEMICALS THAT MODIFY MOSQUITO HOST-SEEKING BEHAVIOR: GAS CHROMATOGRAPHY/MASS SPECTROMETR Y AND HIGH-FIELD ASYMMETRIC-WAVEFORM ION MOBILI TY SPECTROMETRY By Samaret M. Otero Santos May 2007 Chair: Richard A. Yost Major: Chemistry Mosquitoes (Diptera: Culicidae) are nuisance pests responsible for the transmission of several human diseases such as yellow fever, malaria, St. Louis encephalitis, Japanese encephalitis and West Nile Virus (WNV). Although the mechanisms that determine host preference are poorly understood, it is known that mosquitoes respond to chemical and physical signals to orient themselves towards a host fr om a distance, known as mosquito host-seeking behavior. Allelochemicals that modify arthropod host-seeking behavior consist of kairomones and allomones. Kairomones are chemical cues that are used for host finding and are thought to play a large role in this process. Allomones c onsist of attraction-inhibito rs and other repellents that interfere with the host-locati on process. The primary goal of this research is to contribute to the understanding of the hos t-seeking behavior in mosquitoes so that new and more effective methods can be employed for controlling mosquito populations. There have been many attempts to identify possible attractants for several species of mosquitoes. It is known that the mosquito from the Culex specie, the main vector of WNV, feeds primarily on birds, but ther e is little information in th e literature about odors emanating from avian hosts. This problem has been approach ed in the first objective of this work by the

PAGE 14

14 identification of candidate-att ractant compounds from chicken skin and feathers, using gas chromatography/mass spectrometry (GC/MS). A thermal desorption method and a sorbent/solvent extraction method were used; when combined with GC/MS, they allowed for the detection of different functiona l groups of compounds such as carboxylic acids, aldehydes, ketones and alcohols. The release of allelochemicals in the field is normally accomplished with surveillance traps baited with attractant lures, or appropriate can dles or aerosol-dispensing devices to release repellents. The downfield distri bution of these compounds in the field and subsequent detection by arthropods is not well understood. The use of high-field asymmetric-waveform ion mobility spectrometry (FAIMS) to characte rize the distribution of these pl umes in the field is being explored. FAIMS separates ions based on the diff erence in their mobility at high electric field relative to their mobility at low electric fiel d. The results from initial studies with APCI/FAIMS/MS in the laboratory show promise fo r FAIMS to be used as a portable instrument that is amenable to the downstream monitoring of the dispersion from targeted compounds from released sources.

PAGE 15

15 CHAPTER 1 INTRODUCTION Research Overview Mosquito-borne viruses are of major concern to the health and safety of society. Several methods are used to control mosquitoes such as insecticides, repell ents and environmental management. Despite all of this, no method has been found to be constant in effectiveness. To develop a better understanding of th e mosquito host-seeking behavior is the main goal of this research, which will contribute in the developm ent of new, safe and effective methods for controlling mosquito populations. There have been many attempts to identify possible attractants for several species of mosquitoes. The interest of the work reported in this dissertation is focused towards the mosquito vect or of West Nile Virus (WNV) because there are not sufficient data on what at tracts this vector to it s hosts. It is known that Culex mosquitoes, the main vector of the WNV, feed primarily on birds, but there is little inform ation in the literature about odors emanating from avian hosts. This probl em has been approached in the first objective of this work by determining compounds emanati ng from chicken skin and feathers using gas chromatography/mass spectrometry (GC/MS). The second objective has been to evaluate highfield asymmetric-waveform ion mobility spectrometr y (FAIMS) as a potential analytical tool for the characterization of odor plumes from field release of compounds that modify mosquito host seeking-behavior. The mapping of odor plumes in the field by FAIMS will give more insight about the downfield distribution of these com pounds in the field once released and most importantly about the subsequent detection of these odor plumes by the mosquitoes. The objectives of this work are addressed in detail throughout the sections and chapters to follow.

PAGE 16

16 Mosquito Host-Seeking Behavior Mosquitoes (Diptera: Culicidae) are responsib le for the transmission of several human diseases such yellow fever, malaria, St. Loui s encephalitis, Japanese encephalitis and WNV.1, 2 Human and animal hosts are in risk of being inf ected when a mosquito carrier of one of these diseases comes into contact and bites them. Even though all animals are a ttractive to mosquitoes, usuall y the mosquito host-seeking behavior is normally associated to the interac tion between mosquitoes and mammals and birds. The female mosquitoes are the ones that require a blood meal in order to derive the protein vitellogen, which is necessary for egg production.3 The host preference will depend on the mosquito species. Some mosquitoes preferentia lly feed on humans, and are therefore known as anthropophilics, others prefer mammals othe r than humans, known as zoophophilics, and/or ornithophilics that feeds mostly on birds.4 Although the mechanisms that determine host preference are poorly understood, it is known that mosquitoes respond to chemical and physical signals to orient themselves toward a host from a distance. This is known as mosquito hostseeking behavior. These signals are detected by sensory receptors on th e antennae and palps of mosquitoes. In the next sections, the present-da y knowledge of the sensilla and cues that play a major role in the host-seeking beha vior of mosquitoes is discussed. Sensilla of Mosquitoes To be able to understand the host-location pr ocess use by mosquitoes , it is necessary to know and understand the basis of mosquito sens ory physiology. The use of electroantennogram recordings (EAGs) and techniques such as scan ning and transmission electron microscopy have been useful in giving insight about the sensory system of blood-sucking arthropods. The mosquito nervous system consists of three comp onents: central nervous sy stem (CNS), visceral nervous system and the peripheral system.5 The CNS is composed of the brain, ventral nerve

PAGE 17

17 cord and ganglia, while the viscer al nervous system consists of both motor and sensory fibers that stimulate activity of parts of the alimentary canal. At last but not least, the peripheral receptor system involves the motor and sensory neurons as well as the different sense organs. Additionally, the differences in responsiveness to physical and ch emical signals are regulated by changes in sensitivities of the neur ons in the peripheral receptor system.6 The sense organs, also known as sensilla, work by transforming a biological stimulus into information that can be transmitted in the form of a nervous impulse. In the mosquito, each sequential response to certain si gnals is mediated by its own se nsory stimulus. Moreover, the receptor systems at the antennae and maxillary palps have been already classified according to the behavior and signal that they perceive and encode. However, some receptors can actually provide input for more than one behavior in a dult mosquitoes such as mating, host location and feeding or finding ovi position sites. Short and Long Range Signals Host-seeking behavior by the mosquito will depend on physical, visual and olfactory cues. It is described by several behavioral steps, be ginning with the activation of a receptive arthropod to a signal produced from a host and ending w ith the landing of the insect on the host. Mosquitoes respond to physical cues in the proxim ity of the host, whereas at long distances from the host they react to visual a nd olfactory cues. The short and long range signals are presented below and olfactory cues are disc ussed in an individual section. At short distances from the host, it has been determined that female mosquitoes respond to body temperature and moisture wh en searching for blood meals.7-9 Even though it is not well known at what distances the mosquito can perceive heat and humidity, it is believed that these physical factors are registered at a distance of 1 m or less. Th ese two physical cues are involved in the orientation and landing behavi or in the vicinity of a host. Temperature is a physical signal

PAGE 18

18 that has been related not only to mosquitoes but also to other vector or ganisms that can detect and orient to it, and therefore use it in host location.6 The neurons sensitive to changes in temperature are found to be in the coeloconica sensilla located at the tip of the mosquito’s antennae. Takken et al. studi ed in a wind tunnel how the re lative humidity affects the hostseeking behavior of the female mosquito Anopheles gambiae .9 It was determined that mosquitoes were more capable to orient them selves towards the ports during periods of high relative humidity than during periods of lower or stable humidity. Vision plays an important role in the mosquito host-location process. It helps the mosquito in the location of mates, hosts, sugar sources an d oviposition sites. Also, it guides the mosquito flight path.10 The eyes have several functions such as discerning light level intensities, movements, colors, shapes and patterns.11-13 Additionally, mosquitoes have a minimal angle of perception between 4 and 8 .14 However, the resolving power of the mosquito’s eye is lower than the human, but the wide aperture offers a better vision during dark ni ghts compared to other vertebrates. Numerous authors ha ve studied the response of mosquitoes to colored visual stimuli by measuring the spectral sensitivities of the eyes using behavioral techni ques. It was discovered that mosquitoes preferred dark colors rather than light and it has been suggested that this preference can be related to the response to surfaces with a low reflectance factor. Olfactory Cues Olfactory cues, known also as allelochemicals, are signals that modify mosquito hostseeking behavior. There are four major differe nt types of allochemicals, the pheromones, synomones, allomones and kairomones.3, 15 At longer distances, mosquitoes are known to respond to these types of signals in order to locate their hosts.

PAGE 19

19 Pheromones are chemical cues used to send messag es to an individual of the same species. In other words, communication within species occu rs by the use of pheromones. These chemical signals are mostly used to find favorable site s such as food sources, shelter and sites for depositing eggs. Additionally, species use pheromones as a wa rning signal when danger has been encountered and for mating. Sex pheromones in insects are very spec ific and can work not only for the location of the right mate but also fo r the reproductive isolati on of the organism. Another category of allochemicals, the synom ones, refer to semioc hemicals that not only benefit the producer but also the receiver. Good examples of s ynomones are the volatile compounds emanating from plants that serve to attract natural enemies of herbivores. Allomones are catalogued as chemical com pounds that are produ ced or acquired by a species, and when in contact with an organism from another species induce behavioral or physiological response beneficial to the emitter. In the mosquito fiel d, allomones are known as attraction-inhibitors and/or repellents that interfer e with the host-location process. More details about allomones will be given in chapter 4 of this dissertation. The last group of olfactory cues is the kairom ones. By definition, these are chemical cues that are produced by plant, animal or human hosts which are attractive to arthropods that feed on these hosts.15 Kairomones are believed to be the mo st important mechanism use by the mosquito in the host-location process. In particular, bl ood-feeding behavior in female mosquitoes makes use of volatile compounds emanating from a host fo r the purpose of orientat ion and host finding. For many years, researchers have been focuse d on trying to find candi date attractants for many species of mosquitoes in an effort to fi nd better and safer ways for controlling mosquito populations, but little is known a bout the identity of the compounds that cause this kind of behavior in arthropods. It is presumed that the host-seeking beha vior in mosquitoes results from

PAGE 20

20 the attraction to a mixture of volatile co mpounds rather than to a single compound. The following sections briefly describe what is know n about chemical attrac tants to this date. Carbon dioxide As Takken described it, carbon dioxide (CO2) is an universal kairomone which is present in the exhaled breath of all vertebrates.7 Previous studies showed that the removal of CO2 from the exhaled air of human hosts resu lted in a reduction of mosquitoes attracted but it did not affect the number of mosquitoes trying to f eed in the vicini ty of the hosts.16 Later, Gilles explained that CO2 activates and induces upw ind flight of the mosquito towards the host.17 In a review presented by Takken et al. about the role of CO2 in the host-seeking behavior of mosquitoes, it was stipulated that in experiments in the field and laboratory almost all female mosquitoes were attracted to it, with the exception of the female mosquito Anopheles gambiae .18-20 In the field and in mosquito surveillance, CO2 is used in traps with the purpose of attracting mosquitoes. Usually the traps release CO2 from dry ice or gas cylinders. When gas cylinders are used the release rate can be set to deliver CO2 concentrations similar to the ones present in exhaled breath, it has been found that at these concentrations, CO2 can attract mosquitoes at distances of up to 15 m. Furtherm ore, experiments in the field using traps have proven that when CO2 is used in combination with ot her attractants or baits, it acts synergistically. Examples of these attractants and baits are lactic acid, 1-octen-3-ol and chickens.21-23 1-octen-3-ol In 1984, Hall et al. published the discovery of 1octen-3-ol present in bovine breath while they were studying the behavior of tset se flies in the fi eld and laboratory.24, 25 It was demonstrated that this volatile compound serves as an attractant for this type of flies. Later on, Takken and Kline demonstrated that octe nol serves as a mosquito attractant.21, 26 However,

PAGE 21

21 octenol appears to only attract ce rtain mosquito species, and its e ffectiveness in the attraction of the Aedes taeniorhynchus is greatly enhanced by the addition of CO2. Human emanations Most research in the entomology field has been concentrated on the identification of human kairomones due to the greatest inte rest on the mosquito-human interaction. Consequently, human emanations have been st udied in more detail than odors of other vertebrates, a limiting factor in the quest for findi ng attractants for mosquito es that are attracted to avian hosts rather than to human hosts. Acree et al . , in 1968, discovered (L)-lactic acid as the first human-produced mosquito attractant other than CO2 that elicited the attraction of the yellow fever mosquito Aedes aegypti .22 It was obtained from acetone washings of human forearms. Lactic acid can be found in sweat and on the skin, as a product of muscle metabolism.22 Others compounds including butanone, phenols and carboxylic fatty acid s, have shown to attract Ae. aegypti as well.27, 28 Other groups have focused their studies on disc overing olfactory stimulants for the malaria mosquito, An. gambie , by studying human sweat samples. The An. gambie has shown behavioral and electrophysiological respons e to Limburger cheese, carbo xylic acids and human sweat samples.29-31 Park and co-workers found that short-ch ain saturated carboxylic acids present in human sweat extracts have a better electroan tennogram (EAG) respons e than longer-chain saturated carboxylic acids.31 Maijerink and colleagues also id entified olfactory stimulants for An. gambie from human sweat samples by bioassays and EAG.29, 30 They analyzed freshly collected and incubated human sweat samples. The pooled fresh sweat samples did not show behavioral or EAG response. However, incu bated pooled sweat samples showed both. The headspace from pooled fresh and incubated sweat samples was also analyzed by GC/MS and 81 volatile compounds were detected.

PAGE 22

22 EAG and behavioral response studies to chem icals found in human skin emanations have been also performed on the mosquito Culex quinquefasciatus , a major vector of filariasis. Prakash determined from the EAG studies that all carboxylic acids elicit an olfactory response with the exception of tetradecanoic acid and octadecanoic acid. Two al cohols elicit EAG and behavioral response, glycol and benzyl alcoho l, whereas nonanal was the only aldehyde that elicited a high EAG and behavioral res ponses from this type of mosquito.32 Animal emanations other than octenol Even though avian hosts are one of the main re servoirs of diseases that are transmitted by mosquitoes during periods of amplification, litt le is known about the olfactory cues used by mosquitoes to locate these hosts. Many resear ch groups have focused their studies on the analysis of compounds from the ur opygial gland of birds because ot her insects that feed on birds, like mites and flies, have shown to be attracted to odors from this gland. Hunter et al. showed that the Culex pipens and Culex restuans were attracted to the cr ow’s uropygial gland when tested in traps in the field.33 In another study that involved the Culex annulirostris , Smith and his group tested the host preference of this mosquito towards guinea pigs and chickens using host-ba ited traps in the field and in the laboratory. It was determined that Cx. annulirostris from the Riverland region of South Australia preferred guinea pigs and those fo rm the northeast region preferred chickens. In addition, fur and feathers extracts were anal yzed by GC/MS; unfort unately, it was not an exhaustive study in terms of identification of compounds.34 Overview of Analytical Methods The work presented in this dissertation consis ts of studies performed using two analytical techniques, gas chromatography/mass spectrome try (GC/MS) and ion mobility spectrometry (IMS). A brief introduction to ea ch technique is given below.

PAGE 23

23 Gas Chromatography/Mass Spectrometry (GC/MS) GC/MS combines two analytical techniques. It uses GC for separation and MS as a detection analyzer and identification tool. Th e advantages of using GC/MS are that compounds from a complex mixture can be separated so that individual mass spectra of each compound can be acquired for qualitative and qua ntitative purposes. GC/MS is a highly sensitive technique, in which samples of as little as 1 pmol can be anal yzed and a full mass spectrum of the analyte can be obtained. Evaluation of the mass spect rum for fragmentation patterns or “chemical fingerprints” is the basis fo r identification of compounds. The experiments discussed in chapter 2 of this dissertation were completed using commercial GC/MS systems equipped either w ith a quadrupole or a quadrupole ion trap mass analyzer. General quadrupole and quadrupole ion trap theory is addressed in more detail in the following sections. Quadrupole Ion Trap Mass Spectrometer The quadrupole ion trap mass spectrometer (Q ITMS) was invented in 1960 by Paul and Steinwedel.35 It is a very sensitive and specific mass sp ectrometer that also functions as an ion storage device, in which positively or negatively charged gaseous ions can be confined for short periods of time (~1-1000 ms) at a pressure of 1 mTorr and in the absence of solvent molecules.36 The confining of ions in the ion trap is due to the formation of a trapping potential well when appropriate potentials are applied to the electrodes of the ion trap.37 It works as a mass spectrometer when the field within the device is changed, making the trajectories of the stored ions of increasing mass/charge ratios (m/z) unsta ble. Measurement of the m/z ratio of the trapped ions arises as the ions leave the trapping field in order of m/z ratio.38 Once ions are ejected from the ion trap, the ions hit a detect or, providing an out put signal that constitutes a mass spectrum.

PAGE 24

24 Figure 1-1 illustrates the com ponents of the quadrupole ion trap (QIT). The QIT consists of three electrodes: two identical end-cap electrodes and one ring el ectrode located at the center. In ion trap instruments equipped with external ionization sources like the ones used for this research, the GCQ and LCQ (The rmo Finnigan, San Jose, CA), bot h end-cap electrodes have a single perforation. Through the perf oration in one of the end-cap electrode ions are gated into the trap, while through the perforat ion located in the ot her end-cap electrode ions exit the trap and hit the detector. The third electrode, cal led the ring electrode, is placed symmetrically between the two end-cap electr odes. Equation (1-1) demonstr ates the ideal geometrically arrangement of the electrode s in a QIT, in where 2z0 equals the distance between the end-cap electrodes and r0 is the radius of the ring electro de that normally measures 1 cm.37 The arrangement of an ideal trap is shown in figure 1-2. r0 2 = 2z0 2 Eq. (1-1) A major drawback of having the entrance and exit holes in each end-cap, through where ions are injected and ejected, is the crea tion of imperfections that affect the way ions are trapped and scanned out, producing mass spectra that someti mes exhibit incorrect mass assignments. To correct for this, in 1991, the distance between th e end-cap electrodes was increased or stretched in order to create a mass spectrum with correct mass assignments. Therefore, the relationship given by Equation (1-1) is no longe r valid because the value of z0 had been increased typically by 10.6%. Ions are trapped successfully in the ion trap by applying a radio freque ncy potential (RF) to the ring electrode which produces an ideal quadrupole field while the end-cap electrodes are kept grounded. The ion motion within the QIT can be described mathematically by the solutions to the reduced Mathieu equations (Equa tions 1-2 and 1-3), where e is the electronic charge of the

PAGE 25

25 ion, U is the DC potential, V is the amplitude of the RF, is the angular frequency of the applied RF, and r0 and z0 are the internal dimensions of the tr ap. These solutions are in terms of ion stability and instability within the ion trap in the axial and radi al directions that are described by the dimensionless parameters az and qz, respectively. az = -2ar = -16eU/m(r0 2+2z0 2) 2 (Eq. 1-2) qz = -2qr=8eV/m(r0 2+2z0 2) 2 ( Eq. 1-3) The Mathieu stability diagram, shown in figure 1-3, illustrates the regions for ion stability within the ion trap.37 Ions can be trapped in these regions because their trajectories are stable in both the axial (z) and radial (r) dimensions. The point shown in figure 1-3 with value of qz=0.908, and az = 0.0 is referred as the low-mass-cut-o ff (LCMO), which corresponds to the ion of lowest mass/charge ratio that can be stored in the ion trap. One way of ejecting ions from the ion trap is by the method of massselective instability mode developed by Stafford.39 In this mode, ions are first fo cused in the center of the trap by means of momentum-dissipating collisions with helium atoms. Ion ejection from the ion trap occurs by ramping the amplitude of the RF poten tial applied to the ring electrode, which makes the qz of all ions increase. When the qz value for each ion reaches the value of 0.908, then these are ejected axially through the end-cap electrodes in order of increasing m/z charge ratio. The problem with this method of scanning is th at it can produce mass spectra with poor mass resolution. This is due to space charging, particul arly for the low m/z io ns with least kinetic energy found near the center of th e trap as these ions move thr ough the higher m/z ions still in the trap on their way to the exit end-cap. To correct for this problem, the mass-selective instability mode has been modified by the us e of axial modulation or resonant ejection.37 In this way, a fixed frequency is applied across the end-cap electrodes and the ions are made to come

PAGE 26

26 into resonance at this frequency before reaching the LCMO as the RF amplitude is ramped. This causes resonant ejection of the ions from the ion trap, producing a higher mass resolution spectrum. Quadrupole Mass Spectrometer The quadrupole mass spectrometer consists of f our parallel hyperbolic or round electrodes that make the quadrupole field.40 These four electrodes combin e direct-current (DC) and RF fields which create the mass filter.41, 42 Figure 1-4 shows the components of a quadrupole mass spectrometer. Opposite pairs of rods are electrically connected; th e two pairs are then connected to RF and DC voltage sources of opposite polarity. Once ions are formed and accelerated into the quadrupole field, mass analysis is perform by ramping the RF amplitude and DC voltages at constant DC:RF ratio, while the RF frequency is kept constant. For any given set of DC:RF voltages, ions having only a specific m/z value w ill be transmitted through the quadrupole filter avoiding collisions with the rods , while all other ions crash in to the quadrupole rods. A mass spectrum is obtained by scanning the RF and DC voltages from a minimum to a maximum. The ion motion inside the quadrupole mass spectrometer can also be defined mathematically by the Mathieu equations (Equations 1-4 and 1-5), where a is related to the DC, q is related to the RF, e is the electronic char ge of the ion, U is the DC potential, V is the magnitude of the RF, and r0 is the distance from the z axis to each of the quadrupole rods. Solutions to these equations co rrespond to stable tr ajectories of ions in the quadrupole mass filter. That is, the ion will avoid collisions with the quadrupole surfaces , reaching the detector.41 ax = -ay = 8 zeU/m2r0 2 (Eq. 1-4) qx = -qy = 2zeV/m2r0 2 (Eq. 1-5) The stability diagram for a quadrupole mass spectrometer is shown in figure 1-5. This diagram illustrates the region of ion stability within the quadrupole. Ions with stable trajectories

PAGE 27

27 will be those that have m/z values defined by the a and q space coordinates at the tip of the stability diagram.43, 44 Ion Mobility Spectrometry Studies discussed in this work in chapte r 3 and 4 involved the use of ion mobility spectrometry (IMS), specifically the relativel y new technique of hi gh-field asymmetricwaveform ion mobility spectrometry (FAIMS). A brief introduction to IMS and FAIMS is given below. A detailed discussion of FAIMS can be found in chapter 3. IMS is suitable for field portable applicati ons because sample ionization and sample characterization occur at atmospheric pressure.45, 46 It is also the leading technology for on-site detection of chemical warfare agents, explosiv es, illicit drugs and ai r quality monitoring. 47-50 IMS is highly sensitive, has low power consump tion and is already made portable. Figure 1-6 shows a schematic of a conventional IMS drift ce ll. This technique separate gas-phase ions based on their drift velocities in weak electric fields.51 The determination of gas-phase ion mobilities in IMS is based on gating the ions into a drift tube under a consta nt electric field. The drift velocity, vd, of the ion is directly proportion al to the electric field strength, E , at electric fields lower than 500 V/cm, with the proportiona lity constant termed the mobility constant, K = vd/ E . In IMS, the mobility constant ( K ) is independent of the applied electric field.52 Despite all the advances for the past years in trying to make conventional IMS suitable for on-site detection, IMS drift tubes still remain expensive and are re latively large and/or when miniaturized, suffer from losses in detection limits.53 More recently, instruments have been devel oped based on a method of ion filtering using high-field asymmetric waveforms. This tec hnique of high-field asymmetric-waveform ion mobility spectrometry (FAIMS) works similarly to IMS, in which motion of ions is induced by electric fields at atmospheric pressure.54, 55 However, FAIMS separates gas-phase ions based on

PAGE 28

28 the field-dependence of mobility at high electric field (i.e., 10,000 V/cm) to mobility at lower electric fields (i.e., 5000 V/cm). A version of FAIMS, based on voltage applied to two planar plates, has been used already as a detector for GC in the analysis of pheromones and mosquito attractants in an effort to make FAIMS suitable for on-site analysis.56, 57 More details on the FAIMS technique appear in chapter 3. Overview of Dissertation This dissertation focuses in th e study of allelochemicals that modify mosquito host-seeking behavior. Two applications are particularly di scussed, the identification of candidate attractant compounds for the mosquito vector of WNV by GC/MS, and the evaluation of FAIMS for the characterization of allomones. The dissertation consists of five chapters. Th e first chapter reviews th e objectives of this work, the mosquito-host seeking behavior pr ocess and basic entomological fundamentals concerning the mosquito, physical and chemical signals used by the mosquito for host location, and also an overview of the analytical methods employed in this dissertation. Sampling methods used for the identification of co mpounds from chicken skin and feather by GC/MS in an effort to find candidate attractant compounds for the mosquito vector of WNV are di scussed in detail in Chapter 2. Chapter 3 focuses on the evaluati on of FAIMS for the study of allomones odor plumes. FAIMS history and funda mentals are introduced and optimum parameters for best ion transmission through the FAIMS cell are examined. Chapter 4 mainly describes the use of APCI/FAIMS/MS for the characterization of allo mones in which MS/MS was also employed for further characterization of com pounds. Quantitation of allomone s in methanolic solutions and headspace are addressed as well in this chapter. The final chapter of this dissertation, Chapter 5, presents the conclusions and future work.

PAGE 29

29 Figure 1-1. Quadrupole ion trap (QIT) composed of three electrodes; two end-cap electrodes and a ring electrode.

PAGE 30

30 Ring electrode End-cap electrode End-cap electrode Ring electrode End-cap electrode End-cap electrode Figure 1-2. A three dimensional ideal ion trap showing the dimensions r0 and z0, where r0 2 = 2z0 2 . (Adapted from March et al., International Journal of Mass Spectrometry , 2000, 200, 285-312).

PAGE 31

31 Figure 1-3. Mathieu stability diagram for a quadrupol e ion trap mass spectrometer. It illustrates the regions for ion stability within the ion tr ap. Ions can be trapped in these regions because their trajectories are stable in bot h the axial (z) and radial (r) dimensions. The value of qz=0.908 is referred as the low-mass-cutoff (LCMO) which is the ion of lowest mass/charge ratio that can be stored in the ion trap. (Adapted from March et al., International Journal of Mass Spectrometry , 2000, 200, 285-312).

PAGE 32

32 Figure 1-4. Quadrupole mass spectro meter. Shown are the ionizat ion source, the four parallel round rod electrodes that make the mass anal yzer, and the detector. (Adapted from Watson, J. T., Introduction to Mass Spectrometry , 1997, 3ed Edition, LippincottRaven, NY, p. 140)

PAGE 33

33 Figure 1-5. Mathieu stability diagram for a qua drupole mass spectrometer. An operating line is shown here in which if the RF and DC volta ges applied are adjusted so an ion of mass m2 is inside the stability regi on, then heavier ions of mass m3 and lighter ions of mass m1 are outside the stability region and will not be transmitted. (Adapted from Watson, J. T., Introduction to Mass Spectrometry , 1997, 3ed Edition, LippincottRaven, NY, p. 140.)

PAGE 34

34 Ionization regionDrift region To data acquisition Electric field Electric field Faraday plate Aperture grid Ionization source Shutter grid Ionization regionDrift region To data acquisition Electric field Electric field Faraday plate Aperture grid Ionization source Shutter grid Figure 1-6. A conventional ion mobility spectrome ter (IMS). (Adapted from Eiceman et al., Applied Spectroscopy , 1999, 53, 338A.)

PAGE 35

35 CHAPTER 2 IDENTIFICATION OF CANDIDATE ATTR ACTANT COMPOUNDS FROM AVIAN HOSTS FOR THE MOSQUITO VECTOR OF WEST NILE VIRUS BY GAS CHROMATOGRAPHY/MASS SPECTROMETRY Introduction West Nile virus (WNV) is a mosquito-borne ar bovirus transmitted in nature through a birdmosquito-bird cycle, birds being th e primary reservoir for the virus.58 WNV is a member of the Flaviviridea family of viruses, which include s Japanese encephalitis, St. Louis encephalitis, Kunjin and Murray Valley encephalitis.59-62 In 1937 this virus was first isolated from the blood of a woman in the West Nile province of Ug anda. During the 1950s the virus caused outbreaks in Israel, South Africa, France, Romania and Russia. The West Nile virus was first discovered in the Western Hemisphere in 1999 in New York City.63 One particular deta il of the outbreak in New York was the mortality of avian species, a pattern not seen previ ously in the Old World.58, 61 For the past forty years, a la rge group of researchers have demonstrated the importance of chemical compounds as candidate attractants fo r arthropods. Kairomone s, already known to elicit mosquito attraction are ca rbon dioxide, lactic acid, acetone , 1-octen-3-ol and short-chain carboxylic acids.7 Unfortunately, most of these studies focused on determining chemical cues for mosquitoes that are more attracted to humans ra ther than other vertebrates. As a result, little is known about odors emanating from other vertebra tes hosts, particularly avian hosts which are known to carry many arboviruses such as WNV. Animal-baited traps have been used for samp ling adult mosquitoes. In Japan, Buescher et al. used baited traps with different type s of birds for studying Japanese encephalitis.64 This resulted in the collection of mostly Culex mosquitoes, which are known to be vectors of WNV. Also, Vickery et al. showed that chickens, when combined with CO2, provided the best bait for trapping the mosquito vector of St. Louis encephalitis, Culex nigripalpus .23 In addition, Allan

PAGE 36

36 evaluated avian odors for mosqu ito attraction and found that Culex mosquitoes were more attracted to a chicken than to a human arm when tested in an olfactometer.65 Furthermore, it was demonstrated that chicken f eathers elicit attraction of Culex mosquitoes which are likely to contain numerous volatile compounds that can play a role in mosquito attraction. However, chemical identities of these co mpounds were not described. The purpose of this work is to report com pounds tentatively identified from chicken skin emanations and feathers that can serve as potential attractants for the Culex mosquitoes. Thermal desorption of glass bead s rubbed on chicken skin, thermal desorption of feathers and the use of sorbent/solvent extraction to collect or ganic compounds from feathers followed GC/MS were the three methods used and the results are discussed here. Overview of Analytical Methods Thermal Desorption Methods Thermal desorption is a widely used techni que for extracting and isolating volatile and semivolatile compounds from various matrices . Two thermal desorption methods were employed in this work for the identification of candidate attractant compounds for mosquitoes vector of WNV: thermal desorption of glass bead s previously rubbed on ch icken skin, and direct thermal desorption of compounds from chicken feat hers. Thermal desorption of glass beads was chosen because it allows for the analysis of a volatilized sample in the gas phase, a sample detection method similar to that which ar thropods encounter, as proven by Bernier et al.66, 67 For that reason, the thermal desorption of glass beads method applied in experiments for studying chicken skin emanations was ta ken from methodology employe d by Bernier et al. for the analysis of human skin emanations.67 The analysis of human skin emanations by th ermal desorption of cr yofocused glass beads using GC/MS allowed Bernier et al. to identi fy polar and non-polar compounds of moderate

PAGE 37

37 volatility.67 The development and employment of th is method for the analysis of candidate attractant compounds from human skin emanations for the mosquito vector of yellow fever was led by the discovery of Schreck et al.68 These researchers found that substances from skin were transferable to a glass surface wh ich then became attractive to mo squitoes when tested in an olfactometer.68-70 Sorbent/Solvent Extraction Method Complimentary information about avian odors obtained from the thermal desorption of glass beads method was desired. Therefore, an alternative method for collecting compounds from chicken feathers was employed. The beaded active carbon resin, Anasorb-747, used normally in solvent desorption, was chosen because it is an effective coll ector of non-polar and polar organics compounds, and with CS2 as the extraction solven t a very high desorption efficiency (> 90% ) can be obtained.71 Separation Due to the complex nature of the samples co llected, a separation technique was needed. Gas chromatography (GC) was chosen as the separation technique because of the interest on detecting volatiles and semivolatile compounds from sa mples. In the initial stages of this work, a non-polar GC column with 5% phe nyl content in the stationary phase was used. Later, a polar column was employed to improve peak resolutio n of polar components (i .e carboxylic acids). In the thermal desorption me thods, cryofocusing was empl oyed to narrow sample component bands prior to separation on the colu mn due the wide desorption profiles from thermal desorption of compounds. A portion of the column from the inlet end was immersed into liquid nitrogen (LN2), preventing migration of compone nt bands beyond this point and focusing them into a narrow band prior to the separation step. The problem observed and also

PAGE 38

38 reported by Bernier et al. with cryofocusing this way is the in troduction of contaminant peaks, probably coming from the septum or from co lumn bleed from the portion immersed in LN2. Ionization MS is a technique for analyzi ng ions in the gas phase. There are different types of ionization that can be used in MS. However, for the GC/MS anal yses discussed in this work, electron ionization (EI) was mostly employed; pos itive chemical ionization (PCI) was also used in some cases to help with the identification pr ocess. A brief descrip tion and the historical significance of both EI and CI mode s are presented in this chapter. For CI, emphasis is given to the discussion of PCI. Electron ionization (EI) Dempster et al. first used an EI source in 1921 and later it was further developed by Nier et al. in 1943.72 The most important feature of EI is that it provides structural information of the analyte molecule; this is the main reason why EI was chosen as a pr imary ionization mode. In EI, ionization and fragmentation of the analyte molecules occurs when these come into contact with electrons emitted from a hot filament having kinetic energies of ~ 70 eV. In nearly all mass spectrometers equipped with an EI source, the hot filament is usually a thin ribbon or filament of some metal that is heated to a temperature that can em it electrons. These free electrons are then attracted to an anode that is situated on the opposite side of the ionization chamber from the cathode. A small magnetic field and an aperture near the filament collimate the electron beam causing an increase in the probab ility of interaction with analyte molecules. As the analyte molecule interacts with the high energy electron (70 eV), it absorbs some of the electron energy which causes the ej ection of one of the electrons from the molecule to form a radical molecular ion, M+•, and one residual energetic elec tron, as shown in Scheme 2-1. M + eM+ • + 2 eScheme 2-1

PAGE 39

39 Chemical ionization (CI) The PCI mode on the GC/MS system Finni gan DSQ was used to help with the identification of some compounds detected from the thermal desorption of glass beads rubbed on chicken skin. CI is a soft i onization technique that forms a mo lecular type ion (typically a protonated molecule, [M+H]+) from which molecula r weight information can be obtained. In this way, CI offers complimentary information to the structural information obtained by EI. CI was developed from studies of ion-molecu le reactions in simple hydrocarbons gases by Munson and Field around 1966.73, 74 The CI process is started by EI of the reagent gas; in this case methane (CH4) was used. Due to the high pressure in the ionization chamber (~ 1 torr), the reagent gas molecules outnumber the analyte molecu les and thus ions are formed from EI of the reagent gas. The reagent ions ionize the analyt e molecules via ion-molecu le reactions. EI of methane produces the methane molecular ion, as shown in scheme 2-2:36, 74 CH4 + eCH4 + + 2eScheme 2-2 Some of these molecular ions have enough internal energy to decompose into fragment ions, as shown on Schemes 2-3 and 2-4. These molecular and fragment ions of CH4 react further with neutral molecules of CH4 to produce protonated ions which are formed through the following ion-molecule reactions (Scheme 2-5 to 2-8): CH4 + CH3 + + H Scheme 2-3 CH4 + CH2 + + H2 Scheme 2-4 CH4 + + CH4 CH5 + + CH3 Scheme 2-5 CH3 + + CH4 C2H5 ++ H2 Scheme 2-6 CH2 + + CH4 C2H3 + + H2 + H• Scheme 2-7 C2H3 + + CH4 C3H5 + + H2 Scheme 2-8

PAGE 40

40 Because the CH5 +, C2H5 +, and C3H5 + do not react further with CH4, they become the dominant ions and serve as reagent ions to ionize the analyte molecules. The CI spectra include product ions formed by three possi ble reactions: These reactions are presented below, with the reagent ions typically involved and M referri ng to analyte molecules (Scheme 2-9 to 2-11): CH5 + + M MH+ + CH4 proton transfer Scheme 2-9 C3H5 + + M [M-H]+ + C3H6 hydride ion abstraction Scheme 2-10 C2H5 + + M [M+C2H5]+ adduct formation Scheme 2-11 In the proton transfer reaction, a proton is transferred to the an alyte molecule if the analyte molecule has a higher proton affinity than CH4 +, the conjugate base of CH5 +. On the other hand, for hydride ion abstraction reaction to happen the hydride anion affinity of C3H5 + most be greater than the hydride anion affi nity of the analyte [M-H] + ion. Finally, adduct formation with the C2H5 + and the C3H5 + ions typically occurs when the proton affinity of the analyte molecule is less than their conjugate bases (C2H4 and C3H4). Characteristic EI ion fragmentation The national institute of sta ndards and technology (NIST) ma ss spectra database was used to examine the EI mass spectra acquired in this part of the dissertation. When the library searches failed to give a reasonable match, CI wa s used in some cases to further restrict the search in the database; if this still failed, the mass spectra were manually examined. Manual examination involved the inspecti on for distinctive EI compound class fragmentation patterns; this information was then used in a new library s earch. When available, standards were used to confirm identity of compounds given by the library. Compounds that could not be identified are not listed in the tables that ar e shown throughout this chapter, how ever, tentative identifications are listed. Table 2-1 summarizes the characteris tic EI ion fragmentati on patterns for different classes of compounds that were taken into account when unknown spectra were evaluated.

PAGE 41

41 Experimental Samples were collected at the United States Department of Agriculture (USDA), Center for Medical, Agricultural and Veterinary Entomo logy in Gainesville, Florida. Samples were taken from skin and feathers of a brown leghorn chicken, Gallus domesticus. Thermal Desorption of Glass Beads A schematic illustration of the thermal desorp tion of glass beads rubbed on chicken skin followed by cryofocusing GC/MS method is presente d in figure 2-1. Experiments involved the use of 2.9 mm diameter glass beads to collect chicken skin emana tions. Glass beads were rubbed on chicken skin prior to loading into a modified GC injector port liner. The person handling the beads and the person holding the chicken used gloves (100% nitrile) at all times to avoid contamination from human skin emanations. Once the beads were rubbed, th ey were stored in a 3.0 mL glass vial that was placed in an ice bath and then take n to the laboratory for GC/MS analysis. Ten glass beads were then transfe rred into the GC injector port liner that was previously modified in order to prevent th e beads from dropping down onto the GC column entrance that extends up into the injector insert. After the glass beads were loaded into the liner, it was placed in the injection port that was held at 35 C to minimize evaporation of the sample. Helium head pressure was set to 0.0 psig during the changing and placement of the injector port liner to prevent migration of th e volatiles past the portion of the column that was cryofocused. Before starting the thermal desorption process on the injector port, a styrofoam cup (12oz) filled with LN2 was placed in the oven such that appr oximately 8 cm of column could be looped in the container around 15 cm below the point wher e the column passes from the injector port to the oven. After placing the cup, the helium carri er gas velocity was set to 40.0 cm/s and the initial desorption was started by increasi ng the injector port temperature from 35 C to 250 C.

PAGE 42

42 Throughout the cryofocusing part of the met hod, the GC oven temperature was set to 35 C. As mentioned already, the purpose of this process was to collect the volatile compounds and focused them into a narrow band before initiation of separation. After the cryofocusing phase, the styrofoam cup was removed and the GC oven temperature program was started. Two types of columns were used for separation, a polar and a non-polar. A 30.0 meter ZB-wax column (Phenom enex, Torrance, CA) with an i.d of 0.25 mm and a film thickness of 0.25 m was used with a temperature program that consisted of a 3.0 minute hold at 35 C, followed by a ramp rate of 8.0 C /min with a hold at 250 C for 5.0 minutes. The 30.0 meter Guardin ZB-5ms column (Phenom enex) with an i.d of 0.25mm and a film thickness of 0.25 m was used with a temperature program that consisted of a 3.0 minute hold at 35 C, followed by a ramp rate of 17.0 C /min with a hold at 220 C for 24.0 minutes. The helium carrier gas was run at a constant veloc ity of 40.0 cm/s and the injector was used in splitless mode. Analyses were performed on a Finniga nMAT GCQ ion trap mass spectrometer (ThermoFinnigan, San Jose, CA) and on a DSQ (T hermoElectron, San Jose, CA). Full scan mass spectra (mass range 34-450 amu) were obtained using EI. The transfer line and the ionization source temperatures were 200 C and 180 C, respectively. Positive CI was performed on the DSQ. The reagent gas used wa s methane at a flow rate of 0.3 mL/min. The temperature of the CI source was set to 150 C. The instrument was tuned for EI and CI using perfluorotributylamine (PFTBA) prior to each analysis. In addition, blanks consisting of 10 glass beads without sample were run.

PAGE 43

43 Sorbent/Solvent Extraction Method The use of sorbents was also explored in th is study to collect chemical compounds from chicken feathers that can function as chemical attractants. Figure 2-2 shows a schematic of the sorbent/solvent extraction method. Fresh chicke n feathers taken from a brown leghorn chicken were weighed out. Compounds from 12.5 g of chic ken feathers were colle cted onto 500.0 mg of Anasorb-747 (SKC, Eighty Four, PA ); the sorbent was stored in a sealed glass container and collection times were evaluated. Compounds trapped in the sorben t were then extracted using 2.0 mL of carbon disulfide (CS2) (Sigma-Aldrich, MO). 1.0-L volume of a blank and a sample were injected separately in the GC/MS. The blank was composed of the CS2 extraction of Anasorb-747. Separation was performed using a 30.0 meter Guardian ZB-5ms column with an i.d of 0.25mm and a film thickness of 0.25 m (Phenomenex, Torrance, CA). The oven temperature ramp consisted of a 3.0 minute hold at 35 C, followed by a ramp at 8.0 C /min with a hold at 250 C for 6.0 minutes. The carrier gas linear ve locity was set to 40.0 cm/s and splitless mode of injection was used. Full-scan mass spectra (mass range 34-450 amu) were obtained using EI on a Finnigan MAT GCQ ion trap mass spectrometer (ThermoFinnigan, San Jose, CA). The transfer line temperature was set at 200 C and the ionization source temperature was set at 180 C. The MS was set with a filament delay of 7.0 minutes after GC injection. Anasorb-747 was conditioned prior to use. It was rinsed three times with CS2 and then baked-out at 150 C in thermal desorption tubes for 30 minutes. Direct Thermal Desorption of Chicken Feathers Method A thermal desorption tube was packed with 60.0 mg of freshly collected chicken feathers. Compounds were desorbed on a co mmercial desorber unit at 200 C for 4.0 min and then cryotrapped using LN2, at -160 C for 6.0 min. The cryo-trap was th en heated to purge volatiles onto

PAGE 44

44 the head of a 30.0 meter DB-WAX etr column (J & W Scientific, CA) with an i.d of 0.25mm and a film thickness of 0.25 m. Blanks were composed of the thermal desorption tube by itself. The GC analysis ramp consisted of a 6.0 minute hold at 35 C, and then a ramp to 200 C at 6 C/min. The carrier gas flow rate was set at 1.2 mL/min and the inj ector was operated in splitless mode. Analysis was performed on a Finnigan Trace GC/MS quadrupole mass spectrometer (ThermoFinnigan, San Jose, CA). Full-scan mass spectra (mass range 34-450 amu) were obtained using EI. Results and Discussion Thermal desorption of glass be ads following cryofocusing GC/MS The glass beads method was chosen as the init ial method for the collection of chicken skin emanations because it was proven to be a good sa mpling method for the collection of oily/waxy volatile material from the skin.66, 67 Bernier et al. demonstrated th at the primary benefit of using glass beads to analyze skin emanations is the ability to perform a solventless injection of volatiles onto a GC column without excessive introduction of water.66 Glass beads were rubbed on the skin of a chicken around the wing, neck or thigh areas; each bead was rubbed for a period of 2 minutes. A white residue was left on the rubbed beads, indicative of waxy material transf erred from the skin. The rubbed beads were placed in a closed vial and this vial was kept in a dry ice bath until sample collection pha se was completed. The samples were then taken to the laboratory for GC/MS analyses. A downside of this method was that reproducibility between sample runs was quite poor, most likely due to the loss or evaporation of the more volatile components befo re GC/MS analysis. Consequently, this made the identification process more challenging.

PAGE 45

45 Thermal desorption of compounds from the glass beads followed by cryofocusing GC/MS was performed first using a ZB-5ms column. Initial experiments conducted using 10 glass beads with this non-polar column s howed poor peak shape in the ch romatograms for some of the eluting components, presumably due to the pola r nature of the compounds. This problem was alleviated by changing to a polar column. A set of 10 beads were analyzed by the method described earlier and the ZB-wax column was used for separation of the components. The problem of peak fronting was eliminated when using a polar column indicative of the polar nature of the eluting components. Typical total ion chromatogram (TIC) traces from the separation of thermally desorbed compounds from 10 glass beads followed by cryofocusing GC/MS using the ZB-5ms and the ZB-wax column s appear in figure 2-3 and 2-4, respectively. As can be seen from the GC/MS traces in figure 2-3 through 2-4, several components were detected. The mass spectrum from the area under each peak was examined and the spectra were exported to the NIST library database for identif ication purposes. As mentioned earlier, PCI mass spectral data were also acquired. Many co mponents in these analys es provided some level of difficulty in terms of identification, presum ably due to difference between the EI spectra obtained using the GCQ ITMS and the EI mass spect ra from the NIST library database, typically acquired on a quadrupole MS. As a result, the Th ermo DSQ quadrupole MS was used as well, in order to have closer matches with the databa se and ease the identification process. Only compounds present and suspected to be present on chicken skin eman ations are listed in table 22. A discussion of the identification of some of these compounds follows. The major peaks found to be present on the TI C of these figures are free fatty acids (FFAs), as might be anticipated due to the source of the sample, the chicken’s skin. Previous work by Bernier et al. also showed that major components observed from the analysis of human

PAGE 46

46 skin emanations by thermal desorption of glass beads followed cryofocusing GC/MS were FFAs. In addition, these were the components observe d to have poor peak shape when a non-polar column was used. It was found in these analyses that the most abundant FFAs in the TIC were hexadecanoic acid, heptadecanoic acid and octadecanoi c acid. The chicken skin is composed of lipids that can be separated into six major gr oups: phospholipids, free sterols, triglycerides, methyl esters, sterol esters and lastly FFAs.75-77 Wertz et al. reported that FFAs present in chicken skin lipids are composed mainly of aliphatic carboxylic acids ranging from 14 to 20 carbons (C14-C20). Moreover, it has been reported that similar chain distributions are found in chicken liver tryglycerides and abdominal fat.77 FFAs ranging from C30-C40 are also found to be present on the chicken skin,77 however, these longer chains of FFAs were not observed by the method employed in this dissertation, pr esumably due to their low volatility. Another component detected was glycerol , which is an important component of tryglycerides (i.e., fats and oils) and of phospholipids. At first, glycerol could not be identified due to the absence of a molecular ion in the EI mass spectrum; thus, PCI was employed to aid in the identification. Figure 2-5 shows the EI and PCI mass spectra obtained for glycerol on the DSQ. From inspection of the PCI spectrum, a mo lecular weight of 92 was determined due to the abundant [M+H]+ at m/z 93. Taking this into account , a new library search restricted to compounds with molecular weight of 92 was made, which ended up matching the compound with 1,2,3-propanetriol (glycerol). Several other compounds were also identified with the aid of PCI. Th e identification of 1amino-1H-pyrrole-2,5-dione is shown in figures 2-6 A and B. The abundance of the [M+H]+ ion at m/z 113, shown in figure 2-6B, revealed a mol ecular weight of 112, which then was used as reference when making a new library search. In addition, by inspecting the PCI spectrum, it was

PAGE 47

47 determined that the EI spectrum, appearing in figure 2-6A, already contained the molecular ion at m/z 112 for 1-amino-1H-pyrrole-2,5-dione. Another compound identified this way was oleyl alcohol. The EI and PCI mass spectra for oleyl alcohol appear in figure 2-7 A and B, respectively. The PCI sp ectrum shows the [M+H]+ and the [M+29]+ ions at m/z 269 and 297, respectively. The [M+29]+ resulted from the adduct formation with C2H5 +. The EI spectrum obtained in the anal ysis for oleyl alc ohol does not show the molecular ion for this compound making it difficu lt to match with library entries. However, the PCI data helped in determining a molecular weight of 268 which made the identification of oleyl alcohol possible. The presence of 3-isobutylhexahydropyrrolo[ 1,2-a]pyrazine-1,4-dione was detected after using a longer oven temperature ramp (250 C hold for 15.0 minutes) and by examining the EI and CI mass spectra (Figure 2-8 A and B). Co mparison of the EI spectrum (Figure 2-8A) with the NIST database entries did not produce a cl ose match, by examining the PCI spectrum (Figure 2-8B), however, it was possible to assign a mol ecular weight of 210 to this component, which then was used as a reference when studying matches from the library. The appearance in the PCI spectrum of both the [M+H]+ ion from proton transfer and the [M+C2H5]+ ion from adduct formation was the key in the identification of 3-isobutylhe xahydropyrrolo[1,2-a]pyrazine-1,4dione. Otherwise, identification of this compound using solely EI spectrum data could not have been possible due to the ab sence of a molecular ion. A peak at 43.93 minutes was observed after using a longer oven temperature ramp (250 C hold for 15.0 minutes). The presence of the [M+H]+, [M+C2H5]+ and [M+C3H5]+ ions in the PCI mass spectra on figure 2-9B was useful in de termining the chemical identity of 3,5,7-tri hydroxyl-2H1-benzopyran-2-one. The [M+H]+, [M+29]+ and [M+41]+ ions in the PCI spectrum

PAGE 48

48 appear at m/z 195, 223 and 235, respectively. Even though the EI spectrum, figure 2-9A, shows the presence of the molecular i on at m/z 194, the use of the PCI data confirmed that 194 was the molecular ion and not a fragment ion. The group of amides found to be present in chic ken skin is listed in table 2-2. These were identified by looking for the char acteristic ions at m/z 59 and 72 in the EI spectrum. With the exception of acetamide, the long chain amides (C12 C15) were detected when using the ZB-5ms column and were the last to el ute during the chromatographic run. Sorbent/Solvent Extraction Method Previous work reported in th e literature and performed by USDA personnel showed that hexane extracts of chicken feathers elicited attract ion of mosquitoes and thus are likely to contain volatile compounds that can serv e as candidate attractants.65 Unfortunately, the chemical identity of those compounds was not reported. Because the feathers were likely to contain volatile compounds, the use of Anasorb-747 to collect compounds of chicken feathers was explored. The analysis by GC/MS of the CS2 extraction of compounds retained by the sorbent provided some chemical identities of possible compounds present in chicken feathers. This offers complimentary information about avian odors if comp ared to the data obtained from the analyses of chicken skin emanations previously discussed. The sorbent Anasorb-747 was used because it efficiently collects non-polar and polar organic compounds, and with CS2 as the extraction solvent, a high desorption efficiency is obtained.71 Although different collection times were evaluated when using Anasorb-747, results from this method were finally obtained after two months of collecti ng compounds from chicken feathers onto the resin. Chromatograms for the collection times tw o weeks and one month, showed no discernible GC peaks above the solvent tail from the CS2. Thus, a longer collection

PAGE 49

49 time (two months) was used along with a slow er GC temperature ramp to allow the CS2 solvent peak to elute before most analytes. Figure 2-10 shows the TIC of volatiles tr apped in the Anasorb-747 for two months, followed by extraction with CS2. A non-polar column (ZB-5m s) was used for separation purposes. The mass spectrometer filament was set with a solvent delay of 7.0 minutes in order to minimize ion formation from the CS2 solvent. The EI-MS spectrum for each peak in the chromatogram in figure 2-10 was studied and iden tified using the NIST library database. The results from this study are listed in table 2-3, which contains compounds found to be present in chicken feathers after the two-month collection time. The use of this method resulted in the de tection of compounds from two major classes, esters and aldehydes. The esters dominate d the TIC (figure 2-10) obtained from the Aasorb747/CS2 extraction method. These esters range d from a six carbon chain to eighteen carbon chain (C6-C18). The presence of esters in the sa mple is likely probably a product of bacterial degradation of lipids caused by the two-month collection time. It has been reported that lipids are also produced by the uropy gial gland of most birds, and are known to be essential for the maintenance of good plumage condition.78 Aldehydes were the second group of compounds to dominate the GC/MS data. From the aldehydes detected, emphasis in this discussion is given to no nanal. The reason this compound is discussed here is because pr evious studies have shown that nonanal elicited EAG response in the mosquitoes Cx. quinquefasciatus and Cx. tarsalis , two vectors of WNV.34, 79, 80 The abundance of the nonanal peak, with retention time of 16.72 minutes in the TIC that appears in figure 2-10, is greater than that of any other aldehyde. The dete ction of nonanal is consistent

PAGE 50

50 with previous reports in the literature,34 where it was determined to be a compound constituent of chicken feather odor. It is worth mentioning that several background p eaks were detected as well. Some of the background components are Anasorb-747/CS2 related and others are due to column bleed. The peaks at 13.82, 22.92, 25.95 minutes are CS2 related, and these were present in the blank which was composed of an inje ction of 1.0 L of the CS2 extraction from the Anasorb-747 alone. The peaks at 13.82 and 25.92 minutes have been identified as O,S-dimethylester carbonodithioic acid and di(methoxythiocarbonyl)disulfide, respectivel y. Di(methoxythiocarbonyl)disulfide resulted in the most intense peak observed in Figure 2-10. Direct Thermal Desorption of Chicken Feathers A limited study of compounds thermally desorbed from chicken feathers was conducted. The direct thermal desorption method allowed fo r the detection of compounds from chicken feathers. The results obtained from this met hod are catalogued as preliminary due to the limited accessibility of the instrumental set-up used. The TIC trace obtained from the direct thermal desorption of chicken feathers is shown in Figure 2-11, and the compounds tentatively identified are la beled with numbers. The identification process relied on EI-MS library searches. A few aldehydes, ketones and alcohols were observed using this method. Octanal, nonanal and benzaldehyde were obs erved. Nonanal was the most abundant component peak in the TIC with retention time of 14.82 minutes. Two alcohols were tentatively identified using this method, octanol and nonanol. It has been reported that the presence of alcohols in feathers play an im portant role in the maintenance of good plumage in birds and to act as fungicidal, bactericidal or in other hygienic processes.78

PAGE 51

51 Conclusions The main goal of this work was to identify compounds from avian hosts that can serve as potential attractants for the mosquito vector of WNV. Compounds from an avian host were studied due to the preference of the main vector of WNV, Culex mosquitoes, for this type of host. The analytical technique of GC/MS was empl oyed for these analyses due to the interest in analyzing volatile compounds. Three different methods were used to analyze chicken odors by GC/MS: thermal desorption of glass beads rubbed on chicken skin, direct thermal desorption of compounds from chicken feathers, and sorbent/ solvent extraction of compounds from chicken feathers. The thermal desorption of glass b eads following cryofocusing GC/MS allowed for the detection of polar and non-polar compounds of moderate volatility from chicken skin emanations. However, due to the complex nature of the sample and the time it took to collect, transport and analyze the samples, the reproducibility between sample runs was quite poor due to the loss of more volatile components. Volatile compounds from chic ken feathers were studied as well, which provided complimentary information about avian odors to that obtained by thermal desorption of glass beads. The sorbent/solvent ex traction method made possible the identification presumably of esters and aldehydes. The presen ce of the esters in this anal ysis is thought to come from bacterial degradation of lipids over the long collection time used to desorb compounds off the feathers and trapped them into the sorbent. Ad ditionally, few compounds from the direct thermal desorption of chicken feathers were tentat ively identified by EI-MS searches, including aldehydes, ketones and alcohols. The author hopes that this work gives th e foundation for future bioassays in which mosquito attraction to a single or mixture of co mpounds could be tested. However, since several

PAGE 52

52 compounds remained unidentified at this time, th is could mean that the important compounds or potential attractants for the mosquito vect or of WNV have not been yet detected.

PAGE 53

53 Table 2-1. Characteristic EI fr agmentation ions for different classes of compounds used for identification of components from the GC/MS analyses. Compound Class Characteristic EI Fragmentation Ions Aliphatic acids m/z 60, m/z 73, losse s: [M-18]+, [M-28]+, [M-45]+ Amines m/z 58, loss: [M-27]+, m/z series: 44,58,72,86... Esters m/z 74 methyl substituted m/z 88 ethyl substituted Sulfur-containing compounds loss 34 Da, thiols m/z series:47,61,75,89 Aldehydes m/z 44, losses:[M-1]+, [M-18]+, [M-28]+ m/z series: 57,71,85,99 Ketones m/z series: 57,71,85,99 m/z 58 methyl ketones m/z 72 ethyl ketones Aliphatics m/z series: 57,71,85,99 Alcohols C>5 losses:[M-18]+, [M-46]+ Aromatics m/z 77, cluster of ions within 2 mass units Amides m/z series: 59, 72, 86, 98, 128

PAGE 54

54 EI/MS GC Host Sample Collectionglass beads Thermal Desorptionand Cryofocusing (liquid N2) EI/MS GC Host Sample Collectionglass beads Thermal Desorptionand Cryofocusing (liquid N2) Figure 2-1. The thermal desorption of glass beads followed cryofocusing GC/MS method used for analyzing chicken skin emanations.

PAGE 55

55 EI-MS Sample CollectionAnasorb747 (500 mg) 2.0 mLcarbon disulfide Extraction GC Feathers EI-MS Sample CollectionAnasorb747 (500 mg) 2.0 mLcarbon disulfide Extraction GC Feathers Figure 2-2. The sorbent/solvent extraction met hod used for analyzing compounds from chicken feathers by GC/MS.

PAGE 56

56 0 5 10 15 20 25 30 35 Time (min) 0 10 20 30 40 50 60 70 80 90 100Relative Abundance Non-polar column 0 5 10 15 20 25 30 35 Time (min) 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 0 5 10 15 20 25 30 35 Time (min) 0 10 20 30 40 50 60 70 80 90 100Relative Abundance Non-polar column Figure 2-3. Typical total ion ch romatogram (TIC) of compounds thermally desorbed from of 10 glass beads rubbed on chicken skin followed analysis by GC/MS on a Guardian ZB5ms .

PAGE 57

57 5 10 15 20 25 30 35 Time (min) 0 4 8 12 16 20 24Relative Abundance Polar column 5 10 15 20 25 30 35 Time (min) 0 4 8 12 16 20 24Relative Abundance Polar column Figure 2-4. Typical TIC of compounds thermally desorbed from 10 glass beads rubbed on chicken skin followed analysis by GC/MS on a ZB-wax column.

PAGE 58

58 Table 2-2. Compounds tentatively identified from chicken skin emanations. List of compounds was derived from analyses of thermal deso rption of glass beads previously rubbed on chicken skin followed by cryofocusing GC/M S. EI-MS library searches and PCI were used for tentative iden tification of compounds. Rt refers to retention time in non-polar (ZB-5ms) and polar (ZB-wax) columns. Compound Formula Molecular Weight Rt ZB-5ms (min) Rt ZB-wax (min) Carboxylic Acids Acetic acid C2H4O2 60 12.82 Nonanoic acid C9H18O2 158 16.74 Dodecanoic acid C12H24O2 200 21.72 27.12 Tetradecanoic acid C14H28O2 228 24.47 29.24 Pentadecanoic acid C15H30O2 242 25.81 Hexadecanoic acid C16H32O2 256 27.34 31.46 Heptadecanoic acid C17H34O2 270 28.47 32.84 Octadecanoic acid C18H36O2 284 29.80 34.60 Nonadecanoic acid C19H38O2 298 30.81 Eicosanoic acid C20H40O2 312 32.24 Amides/ Nitorgen Realated Acetamide C2H5NO 59 17.64 Dodecanamide C12H25NO 199 32.66 Tridecanamide C13H27NO 213 34.40 Tetradecanamide C14H29NO 227 36.73 Pentadecanamide C15H31NO 241 39.59 1-isocyanatopropane C4H7NO 85 2.78 Alcohols Glycerol C3H8O3 92 25.31 Oley alcohol C18H36O 268 27.99 Aliphatics Undecane C11H24 156 5.96 Dodecane C12H26 170 8.22 Tetradecane C14H30 198 11.95

PAGE 59

59 Table 2-2. Continued Compound Formula Molecular Weight Rt ZB-5ms (min) Rt ZB-wax (min) Heterocyclics/ Aromatics Ethylene sulfide C2H4S 60 1.66 2.74 Cyclobutanone C4H6O 70 2.14 1.70 4-pyridinamine C5H6N2 94 10.23 2,4-dimethylpyrrole C6H9N 95 15.16 3,5,7-trihydroxy-2H-1-ben zopyran-2-one C9H6O5 194 43.93 1-amino-1H-pyrrole-2,5-dione C4H4N2O2 112 25.10 3-Isobutylhexahydropyrrolo[1,2a]pyrazine-1,4-dione C11H18N2O2 210 37.05 Chlorophene C13H11ClO 218 26.52 Aldehydes/Ketones Ethanal C2H4O 44 18.37 3-pentanone C5H10O 86 3.07 Sulfides/ Thio/ Thioethers Methanethiol CH4S 48 1.04 1.31 Carbon disulfide CS2 76 2.84 1.44 Dimethyltrisulfide C2H6S3 126 11.63 Compounds from Background Dimethylamine C2H7N 45 0.97 1.02 Methylene chloride CH2Cl2 84 2.40

PAGE 60

60 45 50 55 60 65 70 75 80 85 90 95 100 m/z 0 20 40 60 80 100Relative Abundance 61.04 60.05 62.06 55.04 A) EI 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 m/z 0 20 40 60 80 100Relative Abundance 61.0975.09 93.03 62.09 76.07 94.11B) PCI [M+H]+ -H20-CH3OH 45 50 55 60 65 70 75 80 85 90 95 100 m/z 0 20 40 60 80 100Relative Abundance 61.04 60.05 62.06 55.04 A) EI 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 100 m/z 0 20 40 60 80 100Relative Abundance 61.0975.09 93.03 62.09 76.07 94.11B) PCI [M+H]+ -H20-CH3OH Figure 2-5. Mass spectra correspon ding to glycerol; peak detect ed at 25.31 minutes using the ZB-wax column. A) EI-MS spectrum, a nd B) PCI-MS spectrum. Spectra obtained on a Thermo DSQ.

PAGE 61

61 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 m/z 0 20 40 60 80 100Relative Abundance 113.03 72.08 85.07 B) PCI [M+H]+ 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 m/z 0 20 40 60 80 100Relative Abundance 112.05 69.08 54.03 84.06A) EIM+• 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 m/z 0 20 40 60 80 100Relative Abundance 113.03 72.08 85.07 B) PCI [M+H]+ 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 m/z 0 20 40 60 80 100Relative Abundance 112.05 69.08 54.03 84.06A) EIM+• Figure 2-6. Mass spectra of 1-amino-1H-pyrro le-2,5-dione; peak dete cted at 25.10 minutes using the ZB-wax column. A) EI-MS spect rum, and B) PCI-MS spectrum. Spectra obtained on a Thermo DSQ.

PAGE 62

62 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 m/z 0 20 40 60 80 100Relative Abundance 55.05 82.05 96.08 69.09 110.11 123.11152.08221.14 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 m/z 0 20 40 60 80 100Relative Abundance 83.10 111.14 97.10 269.23 69.18 250.24 139.13 151.38 180.18297.02 [M+H]+[M+C2H5]+B) PCI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 m/z 0 20 40 60 80 100Relative Abundance 55.05 82.05 96.08 69.09 110.11 123.11152.08221.14 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 m/z 0 20 40 60 80 100Relative Abundance 83.10 111.14 97.10 269.23 69.18 250.24 139.13 151.38 180.18297.02 [M+H]+[M+C2H5]+B) PCI Figure 2-7. Mass spectra of oley l alcohol; peak detected at 27.99 minutes using the ZB-wax column. A) EI-MS spectrum, and B) PC I-MS spectrum. Spectra obtained on a Thermo DSQ.

PAGE 63

63 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 154.03 70.09 86.04 125.08 55.04 139.11 108.08 96.05 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 211.08 239.08 154.12 70.19 86.04[M+H]+[M+C2H5]+B) PCI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 154.03 70.09 86.04 125.08 55.04 139.11 108.08 96.05 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 154.03 70.09 86.04 125.08 55.04 139.11 108.08 96.05 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 211.08 239.08 154.12 70.19 86.04[M+H]+[M+C2H5]+B) PCI Figure 2-8. Mass spectra of 3-isobutylhexahydro pyrrolo[1,2-a]pyrazine-1,4dione; peak detected at 37.05 minutes using the ZB-wax column . A) EI-MS spectrum, and B) PCI-MS spectrum. Spectra obtained on a Thermo DSQ.

PAGE 64

64 60 80 100 120 140 160 180 200 220 240 260 280 300m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 70.09 194.09 96.07 110.08 138.11 124.05 166.09 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 195.08 223.03 70.16 235.15 96.08[M+H]+[M+C2H3]+[M+C3H5]+B) PCIM+• 60 80 100 120 140 160 180 200 220 240 260 280 300m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 70.09 194.09 96.07 110.08 138.11 124.05 166.09 A) EI 60 80 100 120 140 160 180 200 220 240 260 280 300m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 195.08 223.03 70.16 235.15 96.08[M+H]+[M+C2H3]+[M+C3H5]+B) PCIM+• Figure 2-9. Mass spectra of 3,5,7-trihydroxy-2H 1-benzopyran-2-one; peak detected at 43.93 minutes using the ZB-wax column. A) EI-MS spectrum, and B) PCI-MS spectrum. Spectra obtained on a Thermo DSQ.

PAGE 65

65 8 10 12 14 16 18 20 22 24 26 28 30 32 34 Time (min) 0 10 20 30 40 50 60 70 80 90 100Relative Abundance * * * 8 10 12 14 16 18 20 22 24 26 28 30 32 34 Time (min) 0 10 20 30 40 50 60 70 80 90 100Relative Abundance * * * Figure 2-10. The TIC from the EI analysis of compounds from chicken feathers extracted from Anasorb-747 using carbon disulfide. Se paration was performed using a ZB-5ms column. * Refers to CS2 related background.

PAGE 66

66 Table 2-3. Compounds found to be present in chic ken feathers. List of compounds derived from the Anasorb747/CS2 extraction and GC/MS an alysis. Tentative identification was made using EI-MS library searches. Rt refers to retention time in the polar column (ZB-wax). Compound Formula Molecular Weight Rt ZB-wax (min) Esters Butanoic cid methyl ester C5H10O2 102 8.19 Pentanoic acid methyl ester C6H12O2 106 10.50 Hexanoic acid methyl ester C7H14O2 130 12.79 Heptanoic acid methyl ester C8H16O2 144 14.98 Octanoic acid methyl ester C9H18O2 158 17.05 Nonanoic acid methyl ester C10H20O2 172 18.97 Decanoic acid methyl ester C11H22O2 186 20.77 Undecanoic acid methyl ester C12H24O2 200 22.46 Dodecanoic acid methyl ester C13H26O2 214 24.06 Tridecanoic acid methyl ester C14H28O2 228 25.58 Tetradecanoic acid methyl ester C15H30O2 242 27.01 Pentadecanoic acid methyl ester C16H32O2 256 28.36 Hexadecanoic acid ,methyl ester C17H34O2 270 29.66 Heptadecanoic acid methyl ester C18H36O2 284 30.97 Octadecanoic acid methyl ester C19H38O2 298 32.42 (Z)-9-hexadecenoic acid methyl ester C17H32O2 268 29.43 (Z)-9-octadecenoic acid methyl ester C19H36O2 296 32.07 Sulfurous acid methyl ester C2H6O3S 110 8.44 Aldehydes Pentanal C5H10O 86 7.70 Hexanal C6H12O 100 9.96 Heptanal C7H14O 114 12.33 Octanal C8H16O 128 14.59 Nonanal C9H18O 142 16.72 2-nonenal C9H16O 140 17.86

PAGE 67

67 Table 2-3. Continued Compound Formula Molecular Weight Rt ZB-wax (min) Ethers/Ketones 2,2-dimethoxybutane C6H14O2 118 8.99 3,3-dimethoxy-2-butanone C6H12O3 132 10.55 Background Compounds O,O-dimethylmonothiocarbonate C3H6O2S 106 9.36 O,O-dimethyl ester carbonodithioic acid C3H6OS2 122 13.82 Methoxy carbonyl methoxythioxomethyl disulfide C4H6O3S3 198 19.63 Di(methoxythiocarbonyl) disulfide C4H6O2S4 214 25.92 Hexamethyldisiloxane C6H18OSi2 162 7.30 Hexamethylcyclotrisiloxane C6H18O3Si3222 10.25

PAGE 68

68 4 6 8 10 12 14 16 18 20 22 24 26 Time (min) 10 20 30 40 50 60 70 80 90 100Relative Abundance 10 9 1 2 11 4 3 56 7 8 4 6 8 10 12 14 16 18 20 22 24 26 Time (min) 10 20 30 40 50 60 70 80 90 100Relative Abundance 10 9 1 2 11 4 3 56 7 8 Figure 2-11. Chromatogram from the EI anal ysis of compounds thermally desorbed from chicken feathers. A DB-WAX etr was used for the GC separation. Numbered peaks correspond to the following tentatively iden tified compounds: 1) heptanone, 2) 2pentylfuran, 3) octanone, 4) octanal, 5) tetradecane, 6) 6-methyl-5-hepten-2-one, 7) dimethyl trisulfide, 8) nonanal, 9) be nzaldehyde, 10) octanol, and 11) nonanol.

PAGE 69

69 CHAPTER 3 OPTIMIZATION OF HIGHFIELD ASYMMETRIC-WAVEFORM ION MOBILITY SPECTROMETRY High-Field Asymmetric-Waveform Io n Mobility Spectrometry (FAIMS) FAIMS was first reported in 1993, but the origin of this technique da tes back to the 1980s with work by Gorshkov et al. in Russia.81 It was introduced in the United States, by the name of field ion mobility (FIS), as a pot ential portable analytical tool fo r field detection of explosives and drugs of contraband by the Mine Safety Appliances Company (MSA).82 Since then, the technology has been described by different names such as ion drift non-linearit y spectrometry83, ion mobility increment spectrometry84, radio frequency ion mobility spectrometry57, 85, and differential mobility spectrometry.86, 87 Initial FAIMS experiments employed the flat parallel plate geometry that was first described by Buryakov et al.,88this design was improved by repl acing the flat plates with concentric cylinders, as described by the work of Carnahan et al.89 FAIMS is different from conventional ion mobility spectrometry (IMS) in th at the electric fields applied use a high-field asymmetric-waveform rather than DC voltages, and the separation of ions is based on the differences in their mobility at high electric fields, Kh, relative to their mobility at low electric fields, K . In conventional IMS, the ion drift velocity, vd, is proportional to the applied electric field (<500 V/cm) and thus, K the proportionality constant in vd= KE , is independent of the electric field, E , applied.49, 90 The linear relationship between the ion drift velocity and electric field strength is illustrated in figure 3-1, where K is obtained fr om the slope of the plot of vd vs. E . Contrary to IMS, at high elec tric fields (> 5000V/cm) in FAIMS, the ion drift velocity starts to deviate from what is expected and is no longer proportional to the applied electric field (figure 3-2). As a result, Kh becomes dependent upon the applied fiel d, which is the basis for operation in FAIMS. The deviation observed in FAIMS from the linear relationship at high electric fields

PAGE 70

70 is thought to be related to the ion size, its struct ural rigidity and its in teraction with the bath gas.54 The plots A and C in figure 3-2 describe two different types of ion be havior in FAIMS. The mobility of an ion under the influence of an electric field is described by Equation 3-1, where f ( E ) refers to the functional dependence that the ion mobility has on the electric field and K is the coefficient of ion mobility at low electric field.81 Figure 3-3 illustrates the behavior observed in FAIMS for three types of changes in ion mobility as a function of the electric field. For a type A ion, the mobility increases with an increase in the electr ic field strength. The mobility of a type C ion decrease s at high electric fields. Fina lly, for a type B ion, mobility initially increases but then decr eases at higher electric fields . Typically, type A ions are considered to be of low mass (< 300 amu), whereas type C ions are observed to be higher mass ions (> 300 amu). Kh ( E ) = K [ 1 + f ( E )] (Eq. 3-1) To ilustrate an ion’s motion in the FAIMS cell, two parallel plates can be used, as shown in figure 3-4. In this figure, a type A ion, as the ones studied in this wor k, is shown being carried by a gas stream between the two parallel plates. While the lo wer plate is kept at ground potential, a simplified asymmetric waveform, shown as V ( t ), is applied to the upper plate to produce the required electric fiel d. This asymmetric waveform is composed of a high-positive voltage component, V1, lasting for a shorter period of time, t1, and a low-negative voltage component, V2, lasting for a longer period of time, t2. The waveform is synthesized such that the integrated voltage-time product a nd also the field-time product appl ied to the upper plate is zero during a complete cycle of the waveform, as shown in equation 3-2; V1 t1 + V2 t2 = 0 (Eq. 3-2)

PAGE 71

71 During the application of the high-positive vol tage component of the waveform, the ion will move with a velocity equal to v1 which is the product of the mobility constant at high electric field, Kh, and the applied high electric field, E1 (Equation 3-3). Thus the distance traveled by the ion will be d1 that is represented by equation 3-4. v1 = Kh E1 (Eq. 3-3) d1 = v1 t1 = Kh E1 t1 (Eq. 3-4) The application of this portion of the waveform will cause the ion to drift toward the lower plate as a positive voltage is applied to the upper plate. However, when the waveform applied to the upper plate is switched to the low-negative vo ltage component, the ion will start to migrate back toward the upper plate. The ion will move with a velocity, v2, described by equation 3-5, in where K is the ion mobility constant at low-field and E2 the applied low electric field. The distance traveled during this portion of the waveform is d2, as given by equation 3-6. v2 = K E2 (Eq. 3-5) d2 = v2t2 = KE2t2 (Eq. 3-6) Because the integrated voltage-time products are equal to zero, then the field-products, E1 t1 and E2 t2 are equal in absolute magnitude. If Kh and K are the same, d1 and d2 are equal; therefore the ion will experience no net displacemen t, and will return to its original position relative to the plates. This is exactly what happens when both components of the waveform are of low voltage. However, if Kh > K , as is true for type A ions at high electric field (> 5000 V/cm), then the ion will travel further during the application of the high-positive component of the waveform, and it will move toward the lower pl ate. For type C ions at high electric field, Kh < K , thus this ion will migrate further toward the upper plate during the application of the negative portion of the waveform. These three beha viors are illustrated in figure 3-5. As for the

PAGE 72

72 positive type A ion, to compensate for the ion drif t toward the lower plate, a small negative DC voltage, called the compensation voltage (CV), is applied to the upper plate to help balance the ion’s drift. The CV is characteristic of the i on species. The applicati on of a CV prevents the ions from migrating toward either plate, and as result the ion can pass through the FAIMS cell into the MS. Therefore at a certain CV, only ions th at have a stable trajectory at that CV will be transmitted, while the others will strike either of the plates and be lost.91 This allows FAIMS to act as an ion filter device capable of transmitting onl y ions of interest with an appropriate ratio of Kh/ K . Figure 3-6 illustrates the asymmetric waveform s that are commonly used in FAIMS. The waveform described as B its the reverse polar ity of waveform A. These waveforms are a combination of a sine wave and its first harm onic (twice the frequency), as given by equation 38, where D is the peak voltage of the high portion of the waveform (DV), is the waveform frequency in radians/s and is the phase shift of /2 radians (90 ). The waveform has a frequency of approximately 1 MHz. V ( t ) = (2/3) D sin (t ) + (1/3) D sin (2t ) (Eq. 3-8) The selection of which waveform to apply to the inner electrode of the FAIMS cell will depend on the ionic species to be analyzed. Th e dispersion voltage (DV) refers to the peak maximum of the waveform. The waveform repr esented as A has positive DV and waveform B has a negative DV (figure 3-6). The work discus sed in this chapter and in chapter 4 focused on positive ions of type A. To separate such ions, the waveform shown in figure 3-6B was used, with a negative DV.

PAGE 73

73 Ionalytics Selectra Unit All experiments discussed in this chapter and in chapter 4 were performed using a Ionalytics Selectra FAIMS unit (Beta model) from the former company Ionalytics (Ottawa, Canada), now a subsidiary of Thermo Electron. A picture of the FAIMS cell used with the Finnigan LCQ (San Jose, CA) for the following studi es is presented in figure 3-7. The FAIMS unit consisted of the FAIMS cell (electrodes housed in a PEEK body), the asymmetric waveform generator which provides the waveform to the i nner electrode, and gas pu rification cartridges and gas flow controllers to ensure a regulated fl ow of dry carrier gas into the FAIMS cell. The Ionalytic FAIMS unit uses concentric cyli nders which provide higher sensitivity than the flat plate configurations.88 The cylindrical geometry creates a non-uniform electric field between electrodes which gives rise to higher ion transmission efficiency and atmospheric pressure ion focusing.92-94 Figure 3-8 shows a schematic of the line-of-sight FAIMS cell design used in these experiments, showing the inner and outer electrodes. Both the asymmetric waveform and the CV are applied to the inner el ectrode. In this FAIMS configuration, the gap between the inner and outer electr ode is 2.0 mm; this gap corresponds to the space that the ions travel through the cell. A curtain plate is locate d at the front of the FAIMS cell; it is made of stainless steel, and has an opening hole in the center to allow ions to pass into the space between the electrodes. A positive or negative voltage of 1kV is applied to this curtain plate depending on the polarity of the ions being analyzed. Compensation Voltage (CV) The CV is an important parameter in FAIMS because the CV needed to prevent the ion drift toward either plate at high electric fields is i on-dependent. As stated before, this is the additional voltage applied to the in ner electrode of FAIMS to balan ce the drift of the ion so it can pass through the cell to the mass spect rometer. All ions of a given polarity and ion type (A or C)

PAGE 74

74 can be analyzed by scanning the CV, resulting in a CV spectrum, called the total ion current-cv spectrum, TIC-CV. Also, an ion-selective CV spectrum, IS-CV, can be obtained for each individual ion by plotting the intensity of ions of that sp ecific m/z vs. CV. Experimental The use of FAIMS requires the op timization of several paramete rs in order to obtain the best ion transmission through the device. For the purpose of this work, three parameters were optimized; including the curtain ga s flow rate, DV, and carrier ga s composition. The effects that these parameters have on the compensation volt age and the signal intensity were monitored for two kinds of allomones. The two allomone s chosen were N,N-diethyl-3-methylbenzamide (DEET) and 3,7-dimethylocta-1,6-dien -3-ol (linalool) because of the interest of using FAIMS as a potential tool for characteri zing these types of compounds in the field. Table 3-1 shows the chemical structures and molecular weights of DEET and linalool. More information about these two compounds and their characterization using FAIMS is given in chapter 4. The optimization was conducted using the line-o f-sight FAIMS cell attached to the heated capillary of a Finnigan LCQ ion trap mass spectrometer (San Jose, CA) and positive ion atmospheric pressure chemical ionization (APCI) as the ionizati on technique. The APCI source is held in place at a distance of approximately 1 cm away from the curtain plate of the FAIMS cell by the help of two cust om-made brass extensions. Solutions of 1 ppm of DEET (MW = 191.4 g/mo l) and linalool (MW = 154.25 g/mol) in methanol were used in the following studies. For the APCI source, the vaporizer temperature was set to 300 C. The heated capillary temperature and voltage were set to 150 C and 7.0 V, respectively. The discharge poten tial was set to +2.5 kV and the t ube lens offset was set to -5.0 V. The sheath gas was set to 30.0 (arbitrary units) and the injection flow ra te of the analyte was

PAGE 75

75 maintained at 10.0 L/min. The CVs and signal in tensities were monitored for the m/z ions at 192 and 137 that correspond to the [M+H]+ ion for DEET and the [M+H-H20]+ ion for linalool, respectively. Figure 3-9 shows how TIC-CV and IS-CV spectra are obtained in order to determine CV values. CVs for each ion are obtain ed by first scanning the CV (i.e., 0 to -20 V), from which a TIC-CV spectrum of all the ions is constructed, and th en IS-CV spectra are obtained for each ion by selecting the m/z each individual ion. Values for the CV and signal intensity are then taken from the maximum peak height in the IS-CV spectra. The mean of the peak height for four replicates was reported in these studies. Results and Discussion Effect of Curtain Gas Flow Rate Proper functioning of FAIMS requires that the cell be extremely dry. In FAIMS, optimum performance is achieved by the intr oduction of a clean, dry gas into the FAIMS device. This gas is referred as the curtain gas and is introduced through the port adjacent to the curtain plate region, as shown in figures 3-7 and 3-8. In these experiments, nitrogen (N2) was used as the curtain gas. The curtain gas ha s two functions; 1) to help wi th the desolvation of ionized analytes and 2) to carry the desolvated ions through the FAIMS cell into the MS. The portion entering the FAIMS cell is called the carrier gas. The curtain gas flow rate was optimized experimentally in order to have the best ion transmission through the FAIMS device. For this purpose, 1 ppm solutions of DEET and linalool in me thanol were used. Th e data were acquired with the DV at -4000 V and the curt ain plate at +1000 V. The eff ect that the curtain gas flow rate has on the signal intensity and the compen sation voltage was studied . Experiments were performed by scanning the CV from 0.0 V to -20.0 V in 2 minutes intervals. CV values were then determined as the voltage for maximum transmission of ions at m/z of 192 and 137 for DEET and linalool, respectively. Once CVs were determined for DEET and linalool, the signal

PAGE 76

76 intensity values were taken from the maximum p eak height on the IS-CV spectra. The curtain gas flow rate was varied from 2.0 L/min to 4.3 L/min, which is the maximum that can be achieved by the gas flow controlle rs in the FAIMS Selectra Beta unit. Figure 3-10 shows the effect that the curtain gas flow rate has on the signal intensities for D EET and linalool ions. The results showed that the transmission of these ions, according to the individual plots of both DEET and linalool, was best achieved at the ma ximum curtain gas flow rate of 4.3 L/min. Operation of FAIMS at very low gas flow rates af fects the signal intensity of the monitored ion due to poor desolvation of the ionized ions caused by the low fl ow of gas exiting through the orifice in the curtain plate. At such low flow rates, solvent molecules introduced into the APCI may enter through the orifice in the curtain plat e, causing ion-molecule reactions in the FAIMS cell. On the contrary, if the flow rate is too high it may cause a decrease in intensity due to a decreased number of ions entering the FAIMS cell against the high flow ra te exiting the orifice in the curtain plate. However, this was not obs erved at the highest flow rate used in these studies, as seen from the plots in figure 3-10. The effect of gas flow rate on compensation vo ltage value was also studied. As previously described, the APCI source and the FAIMS condi tions used were kept the same for these experiments. Figure 3-11 illustrates the effect of increasing the gas flow rate from 2.0 L/min to 4.3 L/min on the compensation voltage for 1 ppm solu tions of DEET and linalool in methanol. It was observed from the plot of linalool that the compensation voltage remained relatively constant when increasing the N2 gas flow rate, whereas for DEET the CV increases with increasing gas flow rate. The behavior showed by DEET is not typically seen in FAIMS, in which the gas flow rate does not affect the comp ensation voltage. The sh ifting in the CV values

PAGE 77

77 for DEET is probably due to moisture present in the nitrogen gas line, which can undergo ionmolecule reactions with the DEET ions, sh iftting the CVs to more negative values. Effect of Dispersion Voltage (DV) The dispersion voltage was another parameter that was optimized for FAIMS. Under the same ionization source and FAIMS conditions, the effect of the dispersi on voltage on the signal intensity and on the compensati on voltage was studied for D EET and linalool. The maximum dispersion voltage that the Selectra Beta unit wa veform generator can supply is 4000 V. The effect of the dispersion voltage on signal intens ity is shown in figure 312 for DEET and linalool. Data shown in figure 3-12 were collected using 400 V increments from a dispersion voltage of 2400 to -4000 V. The trend of these plots shows that the signal intens ity, hence sensitivity, increases with increasing DV. At a low DV of 2400, the ion intensity for both ions is low which is attributed to diffusional losses to the electrode walls.55 The increase in sensitivity is caused by the ion focusing mechanism at atmospheric pressu re, which results from a decrease in the ion current lost to the electrode walls due to ion focusing in the FAIMS analyzer region.55 Increasing the dispersion voltage from -2400 to -4000 V also affected the compensation voltage of DEET and linalool. Plots of the comp ensation voltage as a f unction of the dispersion voltage are given in figure 313 for DEET and figure 3-14 for linal ool. These two trends showed that the magnitude of the CV increases as the DV increases. It was observed that for DEET the CV shifted to more negative values and for linalool to more positive values. The dependence of the CV on the dispersion voltage is a consequence of the change in high-field mobility relative to the low-field mobility. Effect of Carrier Gas Composition The carrier gas composition can also be varied in FAIMS. Several gases and mixtures of gases (i.e., N2, CO2, He, NO2, O2) can be used with FAIMS in order to improve sensitivity and

PAGE 78

78 the degree of separation between compounds.95 Here, the effect of mixing helium (He) with the N2 carrier gas or mixing carbon dioxide (CO2) with N2 on both compensation voltage and signal intensity for the DEET and linalool ions was studied. The He or CO2 content in the N2 carrier gas was varied from 0% to 50 %. DV was maintain ed at -4000 V and the total gas flow rate was maintained at 4.3 L/min. The magnitude of the compensation vo ltage changes when mixtures of N2/He or N2/CO2 are used. The experiments performed demonstrat e this effect on the CVs for DEET and linalool when He or CO2 were used up to 50%. The use of the N2/He mixture had the most dramatic effect on the CV. Figure 3-15 shows how the CV necessary to balance th e ion’s drift for DEET and linalool is shifted to more ne gative values by up to 15 V as th e He content is increased from 0% to 50% in the carrier gas composition. The CV increases because the mobility at high field for both ions deviates increasingly from their mob ility at low electric fiel d. This effect can be explained as follows; an ion drif ting through a high electric fiel d has an average energy that consists of a thermal component and a kinetic component dependent on th e velocity of the ion through the bath gas.95 Therefore, because He atoms are light, small, and have low polarizability, the mobility of an ion will increase when mixing He with the N2 carrier gas because the ions will have more energetic collisions with He atoms than with N2 molecules, resulting in higher CVs. Thes e collisions are dominated by th e short-range repulsive forces rather than long-range attractive interactions.96 However, when using the CO2/N2 mixture as carrier gas, the opposite effect is observed, as shown in figure 3-16. Increasing the CO2 content shifted the CV for both DEET and linalool to more positive values which reflects a decrease in the mobility of the ions. This behavior obser ved might be indicative that the ions are now behaving as type C ions, which would require the application of opposite polarity of CV to

PAGE 79

79 transmit them through the cell. This can be a resu lt of energy lost by the ion through interacting collisions with CO2 molecules as compared to N2 molecules and He atoms. Additionally, Guevremont et al. have suggested that decreases in ion mobility at high fields can be related to the formation of new orbitals led by the interactio n of the outer electron sh ell of the ion and the gas.82 Higher intensities were seen for DEET when adding He to the N2 carrier gas. This same pattern was also observed for linalool. The plot s showing the effect of increasing He content on signal intensity for DEET and linalool appear in figure 3-17. This kind of behavior has been previously described for type A ions, in which an increase in ion signal resulted from increasing the He content up to 60%.82, 96, 97 In the case of DEET and linal ool, increasing the He content from 0% to 50% increases the ion signal by a fa ctor of approximately 5x and 8x, respectively. The electrical breakdown of the carrier gas in the FAIMS cell limits the use of He to 60 %. The opposite trend in intensitie s was observed when mixing CO2 with N2. Increments of CO2 produced ~9x lower signal inte nsities for DEET and linalool, as illustrated in figure 3-18. Indeed, the ion signal of linalool wa s lost upon addition of a 50% of CO2, whereas DEET was still detected at 50:50% N2/CO2. At this moment, it is not certain why i on signal is affected by varying carrier gas composition but presumably is related to the i on focusing mechanism inside the FAIMS cell at higher electric fields. For instan ce, the boost in ion signal with N2/He gas mixture may be attributed to the higher DC compensation voltage needed to balance the ion drift, which will cause an increase of the electric field insi de the FAIMS cell, resulting in a higher ion transmission because of the focusing of the ions in the FAIMS analyzer region.

PAGE 80

80 Conclusions FAIMS is an ion mobility-based technique that has recently emerged and has gained scientific interest for many a pplications. However, there is much to understand about the chemistry and physics of how ions behave in side the FAIMS cell when certain operating parameters are changed, such as electric field st rength and bath gas composition. Also, is still unknown the behavior and characteristics of type B ions because interest has been focused on the more common type A and C ions. The use of FAIMS as a potential tool for char acterizing allomones odor plumes is currently being evaluated. Optimization of several parameters is required in order to achieve the best ion transmission through the FAIMS cell. In this work, the curtain gas flow rate, dispersion voltage and the carrier gas composition were optimized. The effect of these parameters on the signal intensity and compensation voltage was studied for two allomones of interest, DEET and linalool. Experiments showed that a higher sensitivity was achieved when the maximum available gas flow rate of 4.3 L/min was used; which indicates that better transmission of ions in FAIMS is obtained when the conditions inside th e cell are dry. Furthermore, an increase in sensitivity was observed at higher DV due to the ion focusing mechanism at atmospheric pressure in the cylindrical electrodes. Finally, it can be concluded that the degree of separation between compounds in a mixture and sensitivity are also affected in FAIMS by changing the carrier gas composition. Further stud ies of the effect of using othe r gases, such as dry air which is use in portable FAIMS units, is of importa nce because it will contribute to the understanding of how ions behave in FAIMS when using different gases. This would be beneficial for different types of applications.

PAGE 81

81 Ion Velocity vd vs Electric Field E0.0 0.2 0.4 0.6 0.8 1.0 1.2 050100150200250 Electric Field E (V/cm)Ion Velocity vd Ion Velocity vd vs Electric Field E0.0 0.2 0.4 0.6 0.8 1.0 1.2 050100150200250 Electric Field E (V/cm)Ion Velocity vd Figure 3-1. Effect of the electr ic field strength on the ion drift velocity at low field in conventional IMS. K , slope of the plot, is independent of the applied electric field.

PAGE 82

82 Ion Velocity vd vs Electric Field E0.0 0.2 0.4 0.6 0.8 1.0 1.2 020004000600080001000012000 Electric Field E (V/cm)Ion Velocity vd A C Ion Velocity vd vs Electric Field E0.0 0.2 0.4 0.6 0.8 1.0 1.2 020004000600080001000012000 Electric Field E (V/cm)Ion Velocity vd A C Figure 3-2. Effect of the electric field strength, E , on the ion drift velocity at high electric fields in FAIMS. Ion drift velocity starts to de viate from what is expected and is no longer proportional to the applied electric field. Kh becomes dependent on the applied field. Traces A and C represent two different ion mobility behavior in FAIMS.

PAGE 83

83 Figure 3-3. Plots for the hypotheti cal dependence of ion mobility on electric field strength for three types of ions. (Ada pted from Purves et al. Analytical Chemistry , 1999, 71 (13), 2346-2357) Gas Flow + 0 t1t2V(t) V1V1V2V2 DV DVV (t) Gas Flow Gas Flow + + 0 t1t2V(t) V1V1V2V2 DV DVV (t) Figure 3-4. Ion motion between two parallel plates during the appli cation of an electric field in FAIMS. V ( t ) is the simplified asymmetric waveform applied to the upper plate. (Adapted from Purves et al . Analytical Chemistry , 1999, 71 (13), 2346.)

PAGE 84

84 At low electric field (<500 V/cm) At low electric field (<500 V/cm) At high electric field (>5000 V/cm) At high electric field (>5000 V/cm) • • A A – – low low mass ions mass ions • • C C – – high high mass ions mass ions To compensate for the drift To compensate for the drift towards an electrode, add a towards an electrode, add a DC compensation voltage DC compensation voltage (CV) (CV) At low electric field (<500 V/cm) At low electric field (<500 V/cm) At high electric field (>5000 V/cm) At high electric field (>5000 V/cm) • • A A – – low low mass ions mass ions • • C C – – high high mass ions mass ions To compensate for the drift To compensate for the drift towards an electrode, add a towards an electrode, add a DC compensation voltage DC compensation voltage (CV) (CV) Figure 3-5. The behavior of ions at low and high electric fields in FAIMS. At low field (i.e., 500 V/cm), the ion will experience no net di splacement relative to the plates because Kh and K are identical. However, at hi gher electric field (>5000 V/cm) Kh deviates from K . If Kh > K , then the ion will behave as a type A ion and it will move toward the lower plate during the application of the high positive portion of the waveform. In contrast, if Kh < K , then the ion will migrate toward the upper plate during the negative-voltage portion of the waveform, which is the behavi or of a type C ion. As shown in the bottom center figure, a DC co mpensation voltage (CV) can be used to compensate for the drifting of either type A or C ions.

PAGE 85

85 A) B) A) B) Figure 3-6. Asymmetric waveform used in FAIM S. The peak maximum of the waveform is called the dispersion voltage (DV). ( Adapted from Purves et al., Review of Scientific Instruments , 1999, 70 (2), 1370.)

PAGE 86

86 Gas flow in Asymmetric waveform lead Outer bias voltage lead Curtain plate voltage lead Curtain plate PEEK housing Gas flow in Asymmetric waveform lead Outer bias voltage lead Curtain plate voltage lead Curtain plate PEEK housing Figure 3-7. FAIMS cell design used in experiments. Asymmetric waveform and CV are applied to the inner electrode. Figure 3-8. Line-of-sight FAIMS cell design used in experiments. Asymmetric waveform and CV are applied to the inner electrode.

PAGE 87

87 Table 3-1. Chemical structures and molecular weights for the two allomones studied in this work. Molecular Weight (MW) Structure IUPAC Name Common Name Molecular Weight (MW) Structure IUPAC Name Common Name DEET linalool N,N N,N dimethyl dimethyl m m toluamide toluamide 3,7 3,7 dimethyl dimethyl 1,6 1,6 octadien octadien 3 3 ol ol 191 154 Molecular Weight (MW) Structure IUPAC Name Common Name Molecular Weight (MW) Structure IUPAC Name Common Name DEET linalool N,N N,N dimethyl dimethyl m m toluamide toluamide 3,7 3,7 dimethyl dimethyl 1,6 1,6 octadien octadien 3 3 ol ol 191 154

PAGE 88

88 0 1 2 3 4 5 6 7 8Time (min) 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 0 20 40 60 80 100 3.05 5.07 1.05 7.10 0.79 4.86 2.81 6.83 3.05 5.07 1.04 7.08 IS-CV m/z192TIC-CVIS-CV m/z137 CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V 0 1 2 3 4 5 6 7 8Time (min) 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 0 20 40 60 80 100 3.05 5.07 1.05 7.10 0.79 4.86 2.81 6.83 3.05 5.07 1.04 7.08 IS-CV m/z192TIC-CVIS-CV m/z137 0 1 2 3 4 5 6 7 8Time (min) 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 0 20 40 60 80 100 3.05 5.07 1.05 7.10 0.79 4.86 2.81 6.83 3.05 5.07 1.04 7.08 IS-CV m/z192TIC-CVIS-CV m/z137 CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V CV: 0 to -20 V Figure 3-9. Total ion chromatogramcompensa tion voltage (TIC-CV) and ion selectivecompensation voltage (IS-CV) spectra of 1 ppm standard of linalool and DEET. Spectra shows how CVs are obtained in FAIMS. First, the CV is scanned (i.e., 0 to 20V) and a TIC-CV spectrum containing all i ons in the mixture is generated. IS-CV spectra are obtained by selecting the m/z for each individual ion in the mixture. The CV is then obtained from the maximum peak height and the mean of the replicates is reported as the CV value.

PAGE 89

89 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 2.02.53.03.54.04.5Gas Flow Rate (L/min)Intensity (cps x 105 ) DEET linalool 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 2.02.53.03.54.04.5Gas Flow Rate (L/min)Intensity (cps x 105 ) DEET linalool Figure 3-10. Effect of the curtain gas flow ra te on the signal intensity for DEET and linalool ions. The values are the mean of three measurements the standard deviation of the mean.

PAGE 90

90 -6 -5 -4 -3 -2 -1 0 2.02.53.03.54.04.5Gas flow rate (L/min)Compensation Voltage (V) linalool DEET -6 -5 -4 -3 -2 -1 0 2.02.53.03.54.04.5Gas flow rate (L/min)Compensation Voltage (V) linalool DEET Figure 3-11. Effect of the cu rtain glass flow rate on the co mpensation voltage for DEET and linalool ions. The values are the mean of three measurements the standard deviation of the mean.

PAGE 91

91 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Intensity (cps x 105) DEET linalool 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Intensity (cps x 105) DEET linalool Figure 3-12. Effect of increas ing the dispersion voltage on th e signal intensity of DEET and linalool. Values are the mean of three m easurements standard deviation of the mean.

PAGE 92

92 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 -4200 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Compensation Voltage (V) DEET -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 -4200 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Compensation Voltage (V) DEET Figure 3-13. Effect of increas ing the dispersion voltage on the compensation voltage of DEET. Values are the mean of three measuremen ts standard deviation of the mean

PAGE 93

93 0 0.5 1 1.5 2 2.5 3 3.5 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Compensation Volatge(V) linalool 0 0.5 1 1.5 2 2.5 3 3.5 -4000 -3800 -3600 -3400 -3200 -3000 -2800 -2600 -2400Dispersion Voltage (V)Compensation Volatge(V) linalool Figure 3-14. Effect of increasi ng the dispersion voltage on the co mpensation voltage of linalool. Values are the mean of three measuremen ts standard deviation of the mean.

PAGE 94

94 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 0102030405060He percent (%)Compensation Voltage (V) linalool DEET -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 0102030405060He percent (%)Compensation Voltage (V) linalool DEET Figure 3-15. Compensation vol tage as a function of increasing He content in the N2 carrier gas for DEET and linalool. Values are the mean of four measurements standard deviation of the mean.

PAGE 95

95 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 0102030405060CO2percent (%)Compensation Voltage (V) CV Linalool CV DEET -2.0 0.0 2.0 4.0 6.0 8.0 10.0 0102030405060CO2percent (%)Compensation Voltage (V) CV Linalool CV DEET Figure 3-16. Compensation voltage as a function of increasing CO2 content in the N2 carrier gas for DEET and linalool. Values are the mean of four measurements standard deviation of the mean.

PAGE 96

96 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0102030405060 He percent (%)Intensity (cps x 106) DEET linalool 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0102030405060 He percent (%)Intensity (cps x 106) DEET linalool Figure 3-17. Signal intensity as a function of increas ing He content in the N2 carrier gas for DEET and linalool. Values are the mean of three measurements standard deviation of the mean.

PAGE 97

97 0.0 0.5 1.0 1.5 2.0 2.5 0102030405060 CO2percent (%)Intensity (cps x 104) DEET linalool 0.0 0.5 1.0 1.5 2.0 2.5 0102030405060 CO2percent (%)Intensity (cps x 104) DEET linalool Figure 3-18. Signal intensity as a function of increasing CO2 content in the N2 carrier gas for DEET and linalool. Values are the mean of three measurements standard deviation of the mean.

PAGE 98

98 CHAPTER 4 CHARACTERIZATION OF ALLOMONE S USING ATMOSPHERIC PRESSURE CHEMICAL IONIZATION/HIGH-FIELD ASYMMETRIC-WAVEFORM ION MOBILITY SPECTROMETRY/MASS SPECTROMETRY Introduction This chapter focuses on the characterizati on of two allomones, DEET and linalool, in methanolic solutions and headspace by at mospheric pressure chemical ionization/highfield asymmetric-waveform ion mobi lity spectrometry/mass spectrometry (APCI/FAIMS/MS). Allomones consist of attr action-inhibitors and other repellents that interfere with the mosquito host-location process. The release of allomones in the field is normally accomplished by using appropriate candles or aerosol-dispensing devices. The downfield distribution of these compounds in the field, and subsequent detection by arthropods, is not well understood due in part to the fact that ento mologists lack of a portable analytical tool amenable to ma p these odor plumes once released in the environment. Therefore, the use of FAIMS as a potential analytical tool to characterize allomone odor plumes is being explored. A repellent is a substance that blocks a mos quito’s sensors so that it is not able to follow attractant plumes in air cu rrent given off by warm-blooded hosts.98 Several types of repellents are commercially available and their efficacy varies. In the contrary, attraction inhibitors reduce the number of mosquitoes that take flight in the presence of an attractant by masking the attraction at a cellular level, which means that a mosquito finds it difficult to orient to and locate the source of the attractant.99 These are also known as spatial repellents; but a 100% effec tive inhibitor has not yet been discovered.

PAGE 99

99 99 N,N-diethyl-3-methylbenzamide (DEET) Millions of people have used DEET si nce it was developed in 1946 by the U.S. Department of Agriculture for the U.S Army.100 It is widely recognized as the most effective and long-lasting repe llent available against mosqu itoes, biting flies, fleas and ticks. Sensory biological studies have show n that DEET inhibits the attraction of the Aedes aegypti towards lactic acid.6 Furthermore, it has been demonstrated that DEET inhibits the response of two neurons of the antennal grooved pegs on Ae. Aegypti that are responsible for the detection of skin acids and essential oils.101 Also, DEET has been found to inhibit cells in arthr opods that are not related in the detection of attractants.102 3,7-dimethylocta-1,6-dien-3-ol (linalool) 3,7-dimethylocta-1,6-dien-3-ol (linalool) is a naturally-occurring terpene alcohol with many commercial applications, the major ity of which are based on its pleasant scent. It is found in many flowers and spice plants, su ch as coriander seeds. Previous work has proven that linalool attracts mosquito es to a trap, but when used with CO2, a reduction was observed in the number of mosquitoes coll ected. This suggested that linalool can act as an attractant and as an at traction-inhibitor. Kline et al. demonstrated through studies of spatial repellency responses of Ae. Aegypti that linalool was the most attractive compound and also exhibited spatial repellency.103 Vapor Samples Because the dispersion of allomones in the field will be monitored from vaporreleasing dispensers, it is of interest to study the response of FAIMS to headspace vapors. In headspace analysis, the solvent for the analyte is a gas (i.e., air, nitrogen, helium, etc.) and the analyte must have a partial pre ssure high enough for easy distribution between

PAGE 100

100 100 the gas and liquid phases.104 This implies that the partia l pressure of the analyte should exceed its water solubility. The planar FAIMS configuration has been used more frequently for the detection of chemical vapors.105 Miller et al. was able to test the concentration dependence of the planar FAIMS, called rf-IMS, to increasing co ncentrations of toluene sample vapors. The concentration dependence resulted in an estimated limit of de tection (LOD) of 10 ppb. Moreover, planar FAIMS have been us ed to monitor volatiles in ambient air.53 With the cylindrical configuration, there is lack of information in th e literature of the use of this design to study chemical vapors othe r than the introducti on of gas-phase vapors coming from the ionization process. Atmospheric Pressure Chemical Ionization DEET and linalool were ioni zed by atmospheric pressu re chemical ionization (APCI) because this method of ionization resembles the one s used on portable FAIMS devices, in which chemical ionization occurs at atmospheric pre ssure. Here, a brief description is given about the APCI process and its origin. APCI is a soft ionization technique used in mass spectrometry. This ionization technique dates back to the ear ly 1960s when Kebarle et al. studied ion-proton affinities at high pressure.106 Then in 1966, studies of ion-clus tering at atmospheric pressure was made possible by the introduction of the co rona discharge as an ionization source.107, 108 During the 1970s, published work by Horning et al., using 63Ni and a needle-to-plane corona discharge sources provided informa tion about the ion-molecule chemistry at atmospheric pressure in nitrogen.109 Figure 4-1 illustrates the ionization process in APCI. After the sample is vaporized, ionization occurs at atmospheric pressure.110 The ionization process is

PAGE 101

101 101 initiated by low-energy electrons emitted from a corona discharge needle. These low energy electrons ionize the reagent gas (i.e., nitrogen, H20, etc.) that thro ugh a series of reactions efficiently produces positive and negative ions of the sample analyte. Because H2O has the highest proton affinity of those gases present at atmospheric conditions, it captures protons and can undergo ion molecule reactions to form small clusters of H3O+(H2O)n, where n = 0,1,2,3 etc. Once a molecule with a proton affinity higher than H2O enters the source region, the proton is tr ansferred to it, to form the protonated monomer ion, [M+H]+, as shown in scheme 4-1. Th is new protonated specie can undergo new ion molecule reactions to form ne w series of clusters such as the protonbound dimer, M2H+(H20)n. H3O+ (H2O)n + M MH+(H2O)n (Scheme 4-1) Finnigan LCQ The benchtop Finnigan LCQ quadrupole i on trap mass (QIT) spectrometer (San Jose, CA) was used for the allomone analyses by FAIMS. The LCQ can be operated with two different atmospheric pressure ionizati on sources, electrospray ionization (ESI) and APCI, both of which are design to ionize co mpounds in liquid solutions as in liquid chromatography/mass spectrometry (LC/MS). Once sample molecules are ionized, these ions are passed through a heated capillary from atmosphere to vacuum that helps in the desolvation of the ions. Al so, the use of a sheath gas a nd auxiliary gas can also be employed to help desolvate and cool the ions . After the heated capillary, the ions are transmitted through the skimmer cone using a tube lens that acts as a gate. Sampled ions are collected by the first RF octopole and then transmitted to the second octopole by means of an interoctopole lens. The second octopole is placed insi de the entrance endcap of the ion trap so ions can be transmitte d directly into the ion trap chamber. The

PAGE 102

102 102 octopole guides help with the transmission of ions through a high pressure region where ion scattering would normally occur.111 Moreover, these guides allow for efficient differential pumping. Mass Analysis on the LCQ Mass analysis can be performed on the LCQ using three mass ranges: low mass (m/z 50-200), normal mass (m/z 50-2000) a nd high mass (m/z 100-4000). The LCQ was operated in the normal mass range for the purpos e of this work. Additionally, the three scan modes offered by the LCQ were employed in the experiments described in this chapter. The MS scan mode refers to one stage of mass analysis in single full-scan experiment. Tandem mass spectrometry (MS/MS) was employed in this part of the work to further characterize the ions formed from the compounds of interest, DEET and linalool. MS/MS scan mode employs two consecutive stages of mass spectrometry.112 Early MS/MS studies were performed in multiple sector instruments and on triple quadrupole mass spectrometers. MS/MS in these instruments was termed tandem-inspace because ions move through different stages of the mass spectrometer.113 The quadrupole ion trap (QIT) also allows MS/MS to be perfor med but here ions remain within the trap and this is described as tandem-in-time. This makes MS/MS in QIT instruments more efficient because the stages of MS are performed sequentially with the ions staying in one place, mi nimizing ion losses. Therefore this can lead to achieve greater sensitivity over tandem-in-space instruments. In the QIT, an MS/MS spectrum is gene rated by filling the ion trap and then scanning the ions out of the trap using same process described in chapter 1 for generating a mass spectrum. However, for tandem MS tw o additional steps are used. The first step involves having the ion trap filled. Then, the ion of the m/z of inte rest, called the parent

PAGE 103

103 103 ion, is isolated from the rest of the ions in the trap. This is done by applying a broadband waveform that ejects all ions from the trap except for the parent ion.111 After isolation of the parent ion, the sec ond step in generating an MS/MS spectrum requires the fragmentation of this ion by collision-induced dissociation (CID).114 In CID, the parent ion is excited by the app lication of a dipolar AC signal, 180 out of phase and at the ion’s secular frequency ( z), across the end-cap electrode s. This method is called resonant excitation and the frequency z corresponds to a specific qz value. Consequently, the parent ion is moved to the qz of excitation before the application of the resonant excitation waveform. The parent ion gains kinetic energy when is excited at its secular frequency which increases the amplitude the ion’s oscillations within the trap but not enough to be ejected. Due to the increa sed flight path and higher kinetic energy, the parent ion undergoes a series of energetic collisions with the helium buffer gas atoms. The internal energy gained by the parent ion from these collisions can cause fragmentation of this ion. The fragment i ons (called daughter ions ) are then scanned out of the ion trap and detected by the multiplier to generate the MS/MS spectrum. More than two stages of mass analysis can be obtained as well, and this is called MSn, where n = 3-10. Only MS3 was used in these studies. In MS3, MS/MS/MS, in the third stage of the analysis the daughter ion from the second stage is now the parent ion and is isolated and induced to fragment into its daughter ions. Experimental The two allomones chosen for these e xperiments were DEET and linalool. DEET was chosen because it is a common topical repellent and linalool because it is an

PAGE 104

104 104 attraction-inhibitor that exhibits spatial re pellency in mosquitoes. The structure and molecular weight of these two compounds can be found in table 4-1. Standard Solutions Standard solutions in methanol and th e headspace of DEET and linalool were characterized using a line-of-sight FAIMS ce ll attached onto the heated capillary of a Finnigan LCQ ion trap mass spectrometer (San Jose, CA); atmospheric pressure chemical ionization (APCI) was used as the ionization source. This commercial APCI source is presented in figure 4-2. The AP CI source is held in place at a distance of approximately 1 cm away from the curtain plate of th e FAIMS cell with two custom-made brass extensions. Figure 4-3 illust rates the instrumental setup used for all experiments discussed throughout this chapter. The APCI source was operated with the vaporizer temperature set to 300 C. The capillary temperature and voltage were set to 150 C and +7.0 V, respectively. The discharge current was set to +2.5 kV and the tube lens offset was set to -5.0 V. The sheath gas was operated at a flow rate of 30.0 (arbitrary units) and the injection flow rate was maintained at 10.0 L/min. FAIMS was operated at a dispersion volta ge of -4000 V and the curtain plate voltage was kept at +1000 V. The gas flow rate was set to 4.3 L/min and nitrogen was used as the curtain gas, although some experiments made use of a 60:40 % N2/He mixture. Headspace Analysis The headspace of DEET and linalool standard solutions prepared in methanol was studied by APCI/FAIMS/MS. 500.0 L of the DEET standard was added to a glass vial

PAGE 105

105 105 tube. The vial was then connected to a Sw agelok T-fitting (1/8” stainless steel). Through one end of the T-fitting the nitrogen auxiliar y gas line was attached and through the other end sample vapors were introduced into the APCI source. The gas auxiliary line was operated to deliver N2 at a flow rate of 5 (arbitrary un its). The rest of the APCI source and FAIMS parameters were kept at the sa me values used for the analyses of the methanolic solutions. Results and Discussions APCI/FAIMS/MS & APCI/FAIMS/MS/MS Methanolic solutions Characterization of DEET and linalool by FA IMS involved the acquisition of data using mass spectrometry in order to certainl y identified CV peaks for each compound. A TIC-CV spectrum, shown in figure 4-4, was obtained for a mixture of 1ppm each of DEET and linalool in methanol. The data wa s acquired using a curt ain gas mixture of 60:40 % of N2/He to achieved greater sensitivity. The CV was scanned from 0 to -20 V over 2 minutes. The scanning of the CV was done for 4 cycles for a total period of 8 minutes. The CV values were defined as the average of the CVs of the four peak maxima. The four replicate peaks in the TIC-CV spectra in figure 4-4 occur at an average CV of -10.6 V. The mass spectrum acr oss the first peak s howed a mixture of DEET and linalool ions (Figur e 4-5). The ion at m/z 192 corresponds to protonated DEET, [M+H]+. The ion at m/z 137 corresponds to the [M+H-H20]+ fragment ion, C10H17 +, of linalool. Ion selective-CV spectra we re obtained by plotting the in tensity of the individual m/z ion for each compound versus scan time. For linalool the ion at m/z 137 was chosen and for DEET the ion at m/z 192. The IS-CV sp ectra showed two peaks at different CVs,

PAGE 106

106 106 as illustrated in figure 4-6. The top IS-CV spectrum corresponds to the ion at m/z 192 of DEET and the peak obtained appeared at an average CV of -10.5V. The IS-CV spectrum at the bottom of figure 4-6 has a different p eak at an average CV of -8.4 V that was obtained by selecting only the m/z 137 of linal ool. This figure shows the capability of FAIMS to separate ions in a mixture. Th erefore, knowing the specific value for the CV at which the ion of interest is transm itted through the FAIMS cell, makes FAIMS selective because it acts as an ion filter. Furthermore, by setting the CV to a sp ecific value, compound characterization by MS/MS is possible because of the tran smission of only the ion at that CV. APCI/FAIMS/MS/MS spectra for DEET and linal ool were obtained by setting the CV at -10.5V for DEET (figure 4-7A) and at -8.4V fo r linalool (figure 4-7B ) and isolating the ions at m/z 192 and 137, respectively. A CID of 25% was used for this MS/MS experiment. An average of 200 scans was recorded for each spectrum. The MS/MS spectrum for DEET contains the parent MH+ ion at m/z 192 and a daughter ion at m/z 119 corresponding to C8H7O+. The MS/MS of linalool showed two daughter ions additionally to the parent ion at m/z 137. As shown in figure 4-7B, the two daughter ions for linalool showed up at m/z 95 and 81, which were determined to be for the ions C7H11 + and C6H9 +, respectively. The daughter ion of m/z 192 for DEET, at m/z 119, was isolated and a MS3 spectrum was obtained. The APCI/FAIMS/MS3 spectrum, shown in figure 4-8, was obtained from the average of 200 scans. Th e CV was also kept at -10.5V and a CID energy of 25% was used. It can be observ ed from the spectrum the presence of two daughter ions. The m/z 91 ion is a daughter ion of 119 formed from the lost of CO that

PAGE 107

107 107 corresponds to the C7H7 + ion (tropylium ion). The ion at m/z 109 is presumed to come from an ion-molecule reaction inside the ion trap, since a neutral loss of 10 amu from the m/z 119 ([C8H7O]+) is not possible. The most likely rationale is that the mass-isolated 119 ion reacted with neutral molecules in th e trap (the most likely candidate being methanol and water from the APCI source), and the resulting adduct ion fragmented to form the m/z 109 ion. Note that the m/z 119 ion fragmented by neutral loss of 28 (CO). If the adduct also fragmented by loss of CO to form m/z 109, that would corresponds to an adduct of m/z 137, corresponding to [C8H7O + H2O]+. Headspace analysis The dispersion of allomones in the fiel d will be monitored from vapor-releasing dispensers, which requires an analytical tool capable of detecting and mapping gas-phase compound plumes. The initial characterizati on of gas-phase allomone odor plumes was performed in the laboratory using APCI/FAI MS/MS. For this purposes, a simple set-up was designed for introducing sample vapors in to the APCI region, in where sampling of the solution headspace was achieved by placing th e standard solution into a glass tube vial screwed to a Swagelok T-fitting connected through one of the ends to the auxiliary inlet at the APCI source. In this way, va pors from the headspace were introduced into the APCI source by the aid of N2 auxiliary gas diverted through the T-fitting on its way to the source (see Figure 4-3). The preliminary results using APCI/FAIMS for studying the headspace of methanolic solutions of DEET a nd linalool showed that FAIMS can be used for the analysis of gas-phase vapors. Figur e 4-9 compares the IS -CV spectrum of DEET in a 10 ppm standard solution with the IS -CV spectrum of the headspace of the same solution. The TIC-CV spectra was acquired by scanning the CV from 0 to -20 V and using N2 curtain gas at a flow rate of 4.3 L/ min. The IS-CV spectrum in figure 4-9A

PAGE 108

108 108 contains two peaks; one at a CV of -2.1 V and a less intense p eak at -8.8V. In contrast, the appearance of only one peak at a CV of -2.2 V was observed in the IS-CV spectrum for the headspace in figure 4-9B. The peaks were further characterized by MS, in where the CV was set at either -2.2 V or -8.8 V to transmit only the ions with stab le trajectories at these CV values. An average of 200 scans was acquired for the mass spectra. These mass spectra could offer information on the identity of each peak in the IS-CV spectra. The mass spectrum for the peak at -2.2 V is shown in figure 4-10. Th e FAIMS peak at a CV of -2.2 V, thus corresponds to protonated DEET, that is the MH+, judging by the presence of the ion at m/z 192 amu and the fragment ion at m/z 119 amu. The mass spectrum for the peak at -8.8 V can be found in figure 4-11. This mass spectrum exhibited the same [M+H]+ ion, m/z at 192 amu and an additional lower intensity ion at m/z 383 amu. This ion co rresponds to the proton-bound dimer of DEET, M2H+. This ion was presumably formed from the clustering of protonated DEET with DEET molecules in the APCI source. The [M+H]+ and [M2+H]+ ions pass through the FAIMS cell at different CVs, a nd yield somewhat different spect ra in the ion trap. Note that transport through the 150 C heated capillary interface fr om atmosphere to vacuum and passage through the decl ustering (skimmer/tube le ns) region leads to some fragmentation of the [M+H]+ ion and [M2+H]+ ions in figure 4-10 and 4-11, respectively. The use of FAIMS to analyze sample head space vapors is possi ble as confirmed by the preliminary data discussed above. Note th at DEET is detected in the headspace of a 10 ppm solution at the same CV, although with an ion signal ~100x less than the standard solution (Figure 4-9). The 100x lower abundanc e of neutrals entering the APCI source

PAGE 109

109 109 from the headspace sample, results in no significant dimmer formation, so the second FAIMS peak at a CV of -8.8 V is not obser ved. Linalool could be observed in the headspace as well, but only at a hundred -fold higher concentration of 1000 ppm in solution. This may be due to its higher boiling point of 192 C if compared to DEET (50 C), resulting in a lower vapor pressure in the room temperature headspace vial. The capability of FAIMS to separate co mpounds in the headspace of a mixture was also tested. The headspace of a mixtur e containing 5 ppm of DEET and 1000 ppm of linalool in methanol was studied by APCI/FAI MS/MS; the TIC-CV spectrum is shown in figure 4-12a. The CV was scanned from +5 to -15 V. By selecting the individual m/z for DEET and linalool, an IS-CV spectrum for each compound was obtained. Figure 4-12b shows the IS-CV spectrum for linalool and it can be seen that linalool was transmitted at a CV of +0.3 V. DEET was transmitted at a CV of -2.5 V, as shown in the IS-CV spectrum in figure 4-12c. The change of +0.4 V for the CV of DEET compared to CV in figure 4-10, may be attributed to the use of a different source of nitrogen carrier gas with fewer impurities. Additionally, the peak appe aring at ~ -5.0 V in the TIC-CV (Figure 412a) may resulted from phthalates originating fro m the plastic cap of the glass vial used to hold the headspace sample, which did not have a Teflon covered cap, or from ionmolecule reactions occurring in the FAIMS ce ll. This matter is still currently under investigation. These figures demonstrate that FAIMS is also useful for the separation of compounds present in a mixture in the headsp ace of a solution providing that compounds have enough vapor pressure.

PAGE 110

110 110 Quantitative Studies Because the APCI/FAIMS/MS benchtop system used for the experiments will not be the system used in the field, the estimati on of the limit of detection (LOD) of this instrumentation was not one of the main goals of this work. However, the response of FAIMS to increasing concentrations of DEET and linalool in methanol and the headspace was recorded in order to estimate the lineari ty of the system towards these compounds. Quantitation signal was based on the FAIMS peak heights for m/z 192 and 137 for DEET and linalool, respectively. The LOD is defined as the lowest concentration of an analyte in a sample that can be detected, but not quantitated. It is defi ned as the amount of analyte necessary to give a si gnal-to-noise ratio (S/N) of 3. LOD were estimated using the Xcalibur software provided with the Finni gan LCQ, with S/N ratios calculated by the program. Serial dilutions were made of DEET and linalool in methanol to give a series of solutions ranging in concentration from 100 parts-per-billion ( ppb) to 10 parts-permillion (ppm). Seven replicates were taken for each standard and calibration curves were constructed for both linalool and DEET usi ng APCI of both the solutions and the headspaces. Methanolic solutions Figure 4-13 shows the FAIMS/MS respons e to solutions of increasing DEET concentration obtained using an APCI ionizati on source. The FAIMS peak height for the m/z 192 ion is plotted as a function of concentr ation of DEET. From the graph it is clear that the concentration dependen ce is not linear over the range of concentrations from 100 ppb to 10 ppm. The curvature shown after 2.5 ppm is a result of th e formation of the DEET proton-bound dimer (m/z 383) which decreas es the intensity of the FAIMS peak for the DEET protonated monomer. If only the first five points (inclu ding the blank) of

PAGE 111

111 111 the calibration curve are plotted, a linear relatio nship is obtained, as seen in the inset figure that appears in figure 4-13B. Becau se the lowest concentration tested was 100 ppb, the LOD for DEET in methanol was extrapolated from the S/N obtained from the Xcalibur program for this concentrati on. Figure 4-14 shows that at 100 ppb, the [M+H]+ion at m/z 192 can be observed with a S/ N of approximately 400. This provides an estimated LOD at a concentration appr oximately hundred times lower, ~1 ppb, which would be expected to give a S/N 3. A calibration curve was also generated for linalool in methanol using APCI/FAIMS/MS, as shown in figure 4-15. This plot shows that the concentration dependence is relatively linea r over 100 ppb to 10 ppm, with little or curvature at higher concentrations in contrast to what was obs erved for DEET. The improved linearity may resulted from the lack of dimerization of linalool at higher concentrations. Again, the LOD was determined from the S/N of th e 100 ppb standard, which was the lowest concentration studied. The S/N for the 100 ppb standard, as given by the Xcalibur software, is around 180; thus, the LOD was estimated to be ~2 ppb. Headspace The response of APCI/FAIMS/MS to the headspace vapors of increasing solution concentrations of DEET was also evaluated. The headspace of standard solutions, with concentrations ranging from 100 ppb to 10 ppm of DEET in methanol, was studied. The calibration curve is shown in figure 4-16; it can be observed from the plot that is not a linear relationship. A headspace equilibration time of 5 minutes was used for each standard before analysis by APCI/FAIMS/MS; lack of linearity may resulted from an inadequate equilibration time between the gas and liquid phases. The evaluation of

PAGE 112

112 112 different equilibration times for headspace analysis may help solve this linearity issue. The use of a more sophisticated headspace se t-up design to introduce sample vapors from the headspace of a solution, and incorporati on of temperature control could also help produce better quantitative resu lts. Examination of the calibration curve (figure 4-16) indicates that linearity can be established in two different regions of the plot, from 100 ppm to 1 ppm, and from 1 ppm to the 10 ppm le vel. One reason for this could have been that the blank still containe d traces of background contam ination of DEET, which could explain the 104 signal for the blank. The LOD was estimated to be around 2 ppb (S/N =3) from the S/N value estimated by Xcalibur fo r the 100 ppb standard. However, this LOD value may be higher due to the DEET cont amination in the blank. One source of contamination could arise from DEET sticking to the walls of the sampling tube used to introduce vapors into the APCI source, particul arly due to the high viscosity of DEET. Solvent washes of the sampling tube redu ced this problem, but DEET could not be completely removed from the trace blank. Conclusions The use of FAIMS as a potential analytic al tool to characterize allomone odor plumes has been explored. Characteriza tion of DEET and linalool by FAIMS involved the acquisition of data using mass spectrometry to help in the identification of CV peaks for each compound. Initial studies demonstrat ed the capability of FAIMS to separate these ions in a mixture. Knowing the speci fic value for the CV at which the ion of interest is transmitted through the FAIMS ce ll makes it possible to employ FAIMS as an ion filter. Monitoring the dispersion of allomones in the field from vapor-releasing dispensers requires an analytical tool capable of de tecting and mapping gasphase compound plumes

PAGE 113

113 113 at trace levels. Preliminary results using APCI/FAIMS for studying the headspace of methanolic solutions of DEET and linalool ha ve shown that FAIMS can be used for the analysis of gas-phase vapors. However, th ese studies showed that APCI/FAIMS/MS was more sensitive to DEET than linalool. Li nalool in headspace was not detected below solution concentrations of 1000 ppm, presum ably due to the low vapor pressure of linalool. The use of heat to warm up the st andard solutions could help in increase the vapor pressure of the compounds so these ca n be detected in th e headspace at lower solution concentrations. Although quantitation of these two compounds was not the main purpose of this work, calibration curves were constructe d in order to test linearity of the APCI/FAIMS/MS system. Linearity was only found over a limited range of concentrations, perhaps due to inadequate equilibration tim e for the distribution between the gas an liquid phase. In addition, as s hown for DEET, increased formation of the proton-bound dimer at higher con centrations (> 2.5 ppm) can affect linearity by lowering signal response of the protonated monomer at higher concentrations . The incorporation of an internal standard may pr ovide more reliable quantitation.

PAGE 114

114 114 Table 4-1. Chemical structur es and molecular weights for the two allomones studied in this work. Molecular Weight (MW) Structure IUPAC Name Common Name Molecular Weight (MW) Structure IUPAC Name Common Name DEET linalool N,N N,N dimethyl dimethyl m m toluamide toluamide 3,7 3,7 dimethyl dimethyl 1,6 1,6 octadien octadien 3 3 ol ol 191 154 Molecular Weight (MW) Structure IUPAC Name Common Name Molecular Weight (MW) Structure IUPAC Name Common Name DEET linalool N,N N,N dimethyl dimethyl m m toluamide toluamide 3,7 3,7 dimethyl dimethyl 1,6 1,6 octadien octadien 3 3 ol ol 191 154

PAGE 115

115 115 Corona discharge needle Corona discharge needle Corona discharge needle Corona discharge needle Figure 4-1. The atmospheric pressure chem ical ionization (APCI) process used for LC/MS. As effluent from the LC is vaporized, ionization is initiated at atmospheric pressure by low-energy electrons emitted from a corona discharge needle. Low-energy electrons ionize a reagent gas that through a series of ion-molecule reactions effi ciently produces positive and negative ions of the sample analyte. Figure 4-2. Finnigan LCQ APCI source used in all experiments. (Adapted from the Finnigan LCQ manual, pg 2-13.)

PAGE 116

116 116 A B C Figure 4-3. Instrumentation set-up used fo r the characterization of the allomones by APCI/FAIMS/MS. A) APCI source on th e Finnigan LCQ (San Jose, CA), B) Open APCI source showing the FAIMS cell (Ionalytics) attached to the heated capillary of the Finnigan LCQ Deca Cl assic, C) Close-up picture of the FAIMS cell attached to the heated capillary.

PAGE 117

117 117 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 0 50 100 5.08 1.05 7.09 NL: 1.10E7 TIC-CV 3.03Relative AbundanceTotal Ion Current Compensation Voltage Spectrum (TIC Total Ion Current Compensation Voltage Spectrum (TIC CV) CV) 0 -20 CV (V)Time (min) 0 -20 CV (V) 0 -20 CV (V) 0 -20 CV (V) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 0 50 100 5.08 1.05 7.09 NL: 1.10E7 TIC-CV 3.03Relative AbundanceTotal Ion Current Compensation Voltage Spectrum (TIC Total Ion Current Compensation Voltage Spectrum (TIC CV) CV) 0 -20 CV (V) 0 -20 CV (V)Time (min) 0 -20 CV (V) 0 -20 CV (V) 0 -20 CV (V) 0 -20 CV (V) 0 -20 CV (V) 0 -20 CV (V) Figure 4-4. The TIC-CV spectru m for a mixture of 1 ppm each of DEET and linalool in methanol. The CV was scanned from 0 to -20 V over a period of 2 minutes, and this was repeated four times over 8 minutes. In this spectrum the peak maxima occur at 1.05, 3.03, 5.08, and 7.09 minutes; given the 2 minutes CV scan from 0 t -20 V, these four ma xima correspond to -10.5 V, -10.3 V, -10.8 V, and -10.9 V, with an average of -10.7 V. NL:1.37E5 60 80 100 120 140 160 180 200 220 240 260 280 300 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 137.00 Linalool [C10H17]+DEET MH+ NL:1.37E5 60 80 100 120 140 160 180 200 220 240 260 280 300 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 137.00 Linalool [C10H17]+DEET MH+ NL:1.37E5NL:1.37E5 60 80 100 120 140 160 180 200 220 240 260 280 300 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 137.00 Linalool [C10H17]+DEET MH+ Figure 4-5. The APCI/FAIMS/MS spectrum from area under the peak at a CV of -10.6 V. A mixture of 1 ppm each of DEET and linalool in methanol was used.

PAGE 118

118 118 Relative AbundanceTime (min) 0 50 100 0 50 100 NL: 8.43E6 IS-CV NL: 1.47E5 IC-CV CV: -10.5V CV: -8.4V 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Ion Selective Compensation Voltage Spectrum (IS-CV) m/z:192 m/z:137 Relative AbundanceTime (min) 0 50 100 0 50 100 NL: 8.43E6 IS-CV NL: 1.47E5 IC-CV CV: -10.5V CV: -8.4V 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Ion Selective Compensation Voltage Spectrum (IS-CV) m/z:192 m/z:137 Relative AbundanceTime (min) 0 50 100 0 50 100 NL: 8.43E6 IS-CV NL: 1.47E5 IC-CV CV: -10.5V CV: -8.4V 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Ion Selective Compensation Voltage Spectrum (IS-CV) m/z:192 m/z:137 Figure 4-6. The IS-CV spectra obtained by se lecting exclusively the ions with m/z 192 and 137 of DEET and linalool. Peaks at CVs -10.5 V and -8.4 V correspond to DEET and linalool, respectively.

PAGE 119

119 119 NL: 9.68E4 100 136.87m/z 0 10 20 30 40 50 60 70 80 90 100Relative AbundanceNL:4.91E5 60 80 100 120 140 160 180 200 192.00 118.93 A [C8H7O]+MH+ DEET CV= -10.5 Vm/zRelative Abundance 60 80 100 120 140 160 180 200 0 10 20 30 40 50 60 70 80 90 80.93 94.93 B[C10H17]+[C6H9]+[C7H11]+Linalool CV= -8.4V NL: 9.68E4 100 136.87m/z 0 10 20 30 40 50 60 70 80 90 100Relative AbundanceNL:4.91E5 60 80 100 120 140 160 180 200 192.00 118.93 A [C8H7O]+MH+ DEET CV= -10.5 Vm/zRelative Abundance 60 80 100 120 140 160 180 200 0 10 20 30 40 50 60 70 80 90 80.93 94.93 B[C10H17]+[C6H9]+[C7H11]+Linalool CV= -8.4V NL: 9.68E4 100 136.87m/z 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100Relative AbundanceNL:4.91E5 60 80 100 120 140 160 180 200 192.00 118.93 A [C8H7O]+MH+ DEET CV= -10.5 Vm/zRelative Abundance 60 80 100 120 140 160 180 200 0 10 20 30 40 50 60 70 80 90 80.93 94.93 B[C10H17]+[C6H9]+[C7H11]+Linalool CV= -8.4V Figure 4-7. The APCI/FAIMS/MS/MS spectra . A) APCI/FAIMS/MS/MS spectrum of DEET, obtained by setting the CV at -10.5 V, B) APCI/FAIMS/MS/MS spectrum of linalool, obtained by setting the CV at -8.4 V.

PAGE 120

120 120 60 80 100 120 140 160 180 200m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 91.00 108.87 119.07 [C7H7]+[C8H70]+MS MS2 MS3192 119 91 60 80 100 120 140 160 180 200m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 91.00 108.87 119.07 [C7H7]+[C8H70]+ 60 80 100 120 140 160 180 200m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 91.00 108.87 119.07 [C7H7]+[C8H70]+MS MS2 MS3192 119 91 Figure 4-8. The APCI/FAIMS/MS/MS/MS spectr um using as precursor ion the m/z 119. Spectrum obtained by setting the CV to 10.5V and with CID of 25 %. Ion at m/z 91 corresponds to the tropylium ion, and ion at m/z 109 is presumed to come from an ion-molecule reaction inside the ion trap, since a neutral loss of 10 amu from the m/z 119 ([C8H7O]+) is not possible. The most likely rationale is that the mass-isolated 119 ion reacted with neutral molecules in the trap ( the most likely candidate be ing methanol and water from the APCI source), and the resulting adduct ion fr agmented to form the m/z 109 ion. Note that the m/z 119 ion fragmented by neutral loss of 28 (CO). If the adduct also fragmented by loss of CO to form m/z 109, that would corresponds to an adduct of m/z 137, corresponding to [C8H7O + H2O]+.

PAGE 121

121 121 Relative Abundance0 20 40 60 80 100 0 20 40 60 80 100 0-5-10-15-20 Compensation Voltage (V) NL: 1.00E5 (B) Headspace 0 -5 -10 -15 -20 NL: 2.04E6 (A) Solution DEET, MH+DEET, MH+ DEET Proton-bound Dimer, M2H+ Relative Abundance0 20 40 60 80 100 0 20 40 60 80 100 0-5-10-15-20 Compensation Voltage (V) NL: 1.00E5 (B) Headspace 0 -5 -10 -15 -20 NL: 1.26E7 (A) Solution DEET, MH+DEET, MH+ DEET Proton-bound Dimer, M2H+ Relative Abundance0 20 40 60 80 100 0 20 40 60 80 100 0-5-10-15-20 Compensation Voltage (V) NL: 1.00E5 (B) Headspace 0 -5 -10 -15 -20 NL: 2.04E6 (A) Solution DEET, MH+DEET, MH+ DEET Proton-bound Dimer, M2H+ Relative Abundance0 20 40 60 80 100 0 20 40 60 80 100 0-5-10-15-20 Compensation Voltage (V) NL: 1.00E5 (B) Headspace 0 -5 -10 -15 -20 NL: 1.26E7 (A) Solution DEET, MH+DEET, MH+ DEET Proton-bound Dimer, M2H+ Figure 4-9. Comparison of IS-C V spectra (m/z 192, the [M+H]+] ion) of a 10 ppm DEET standard in methanol. A) Standard solution of DEET, B) Headspace of a standard solution of DEET.

PAGE 122

122 122 NL:1.83E4 50 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 119.07 MH+ [C8H7O]+ NL:1.18E6 NL: 50 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 119.07 MH+ [C8H7O]+ NL:1.83E4 NL:1.83E4 50 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 119.07 MH+ [C8H7O]+ NL:1.18E6 NL: 50 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 119.07 MH+ [C8H7O]+ Figure 4-10. The APCI/FAIMS/MS spectrum of peak at -2.1 V. Corresponds to protonated DEET with [M+H]+ ion at m/z 192 and the [M+H-NH(C2H5)2]+ ion at m/z 119. NL: 4.41E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67 M2H+ MH+ NL: 4.41E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67 M2H+ MH+ NL: 6.33E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67MH+ NL: 4.41E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67 M2H+ MH+ NL: 4.41E5 NL: 4.41E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67 M2H+ MH+ NL: 4.41E5 NL: 4.41E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67 M2H+ MH+ NL: 6.33E5 100 150 200 250 300 350 400 450 500 m/z 0 10 20 30 40 50 60 70 80 90 100Relative Abundance 192.00 382.67MH+ M2H+ Figure 4-11. The APCI/FAIMS/MS spectrum of the peak at -8.8 V. Peak corresponds to the proton-bound dimer of DEET, [M2+H]+ at m/z 383.

PAGE 123

123 123 5 0 -5 -10 -15Compensation Voltage (V) 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 0 20 40 60 80 100 a) TIC-CV b) linalool m/z= 137 c) DEET m/z=192 5 0 -5 -10 -15Compensation Voltage (V) 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 0 20 40 60 80 100 a) TIC-CV b) linalool m/z= 137 c) DEET m/z=192 Figure 4-12. The APCI/FAIMS spectra of the headspace of a mixture of 5 ppm DEET and 1000 ppm of linalool. A) TIC-CV spectrum of the mixture, b) IS-CV spectrum of linalool, CV set at +0.3 V, and c) IS-CV spect rum of DEET with CV set at -2.5 V.

PAGE 124

124 124 y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 A) B) y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 A) B)y=0.7135x + 0.0383 R2=0.9981 y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 A) B) y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET y = 0.5207x + 0.1786 R2= 0.9924 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) DEET 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 0.0 0.4 0.8 1.2 1.6 2.0 0.00.51.01.52.02.53.0 A) B)y=0.7135x + 0.0383 R2=0.9981 Figure 4-13. The APCI/FAIMS/MS calibrati on curves for DEET standard solutions. Values are the mean of seven measurements the standard deviation of the mean. A) Calibration curve using seve n standard solutions ranging from 100 ppb to 10 ppm, B) Calibration curve using first five data points, shown with red circles, to demonstrate linearity in that region.

PAGE 125

125 125 IS-CV 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Time (min) 0 20 40 60 80 100Relative Abundance RT: 0.55 AH: 82384 SN: 409 m/z= 192 NL: 7.87E3 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z 0 20 40 60 80 100Relative Abundance 192.07 118.87 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 RT: 0.55 AH: 82384 SN: 409 = 192 NL: 7.87E3 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 0 20 40 60 80 100 C8H7O+[M+H]+IS-CV 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Time (min) 0 20 40 60 80 100Relative Abundance RT: 0.55 AH: 82384 SN: 409 m/z= 192 NL: 7.87E3 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z 0 20 40 60 80 100Relative Abundance 192.07 118.87 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 RT: 0.55 AH: 82384 SN: 409 = 192 NL: 7.87E3 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 0 20 40 60 80 100 C8H7O+[M+H]+ Figure 4-14. The IS-CV and mass spectrum of a 100 ppb DEET standard in methanol analyzed by APCI/FAIMS/MS. The Xca libur software used to operate the LCQ reports these values for each FAIMS or chromatographic peak: RT refers to the retention time, AH means average height and SN corresponds to signal-to-noise ratio. CV was scanned fr om +5 to -15 V in 2 minutes, thus, an RT of 0.55 corresponds to a CV of -0.5 V. Ion with m/z 192 corresponds to the [M+H]+ of DEET and the ion at m/z 119 corresponds to the [M+HNH(C2H5)2]+ fragment ion of DEET.

PAGE 126

126 126 y = 0.5404x + 0.071 R 2 = 0.99850.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) linalool y = 0.5404x + 0.071 R 2 = 0.99850.0 1.0 2.0 3.0 4.0 5.0 6.0 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (cps x 105) linalool Figure 4-15. The APCI/FAIMS/MS calibration curves for linalool standard solutions ranging from 100 ppb to 100 ppm. Values are the mean of seven measurements the standard deviation of the mean.

PAGE 127

127 127 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (counts x 105) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.02.04.06.08.010.012.0Concentration (ppm)Intensity (counts x 105) Figure 4-16. Concentration dependence of APCI/FAIMS/MS response for the headspace of DEET standards. Values are th e mean of seven measurements the standard deviation of the mean.

PAGE 128

128 CHAPTER 5 CONCLUSION AND FUTURE WORK Conclusions Mosquitoes are known for their role in transm itting many arbovirus that can put at risk the health and safety of human soci ety. Although the mechanisms that determine host preference are poorly understood, it is known that mosquitoes respond to chemical and physical signals to orient themselves towards a host from a distance . The primary goal of this research was to contribute to the knowledge of the mosquito host-seeking behavior in order to make possible in the near future the development of new, safe and effective methods for controlling mosquito populations. There have been many efforts in trying to iden tify possible attractants for several species of mosquitoes. It is known that the mosquito from the Culex specie, the main vector of WNV, feeds primarily on birds but ther e is little information in th e literature about odors emanating from avian hosts. This problem has been approach ed in the first objective of this work by the identification of candidate attrac tant compounds from chicken skin and feathers using GC/MS. The analytical technique of GC /MS was employed for these anal yses due to the interest of analyzing volatile compounds. Three different met hods were used in order to analyzed chicken odors by GC/MS, including thermal desorption of glass beads rubbed on chicken skin, direct thermal desorption of compounds from chicken feathers, and sorbent/solvent extraction of compounds from chicken feathers. The criteria for choosing these methods was to sample in a manner which biased the compounds to be detect ed while providing some discrimination against compounds that are not thought to play a role in attraction, i.e., non-volatiles. The thermal desorption of glass beads followed by cryofocusing GC/MS allowed for the detection of polar and non-polar compounds of moderate volatility from chicken skin

PAGE 129

129 emanations, as demonstrated by previous work for the analysis of human skin emanations. From these analyses, 39 compounds were identified by using standards or tentatively identified by using EI-MS library searches and PCI spectra. The most abundant compounds detected by this method were free fatty acid s (FFAs) ranging from C12-C20, with hexadecanoic acid and octadecanoic acid being the dominant components in these analyses. A series of long-chain amides (C12C15) were also found to be present in chic ken skin using this method. Due to the complex nature of the sample and the time it took to collect, transport and analyze the samples, several compounds still remain unidentified. Reproducibility between sample runs was quite troublesome due to the loss or evaporation of more volatile components. This made identification of components more challenging. Volatile compounds from chic ken feathers were studied which provided complimentary information about avian odors obtained by th ermal desorption of glass beads. The sorbent/solvent extraction method made possibl e the identification of mostly esters and aldehydes. The presence of the esters in this analysis is thought to come from bacterial degradation of lipids. However, the detection of compounds from the feathers was possible after using a longer collection time (2 months) for co mpounds to be absorbed into the sorbent. A few compounds from the direct thermal desorp tion of chicken feathe rs were tentatively identified by EI-MS library sear ches. The results obtained from this method are considered as preliminary due to the limited acces sibility of the instrumental set-up used. Therefore, chemical identities of only a few compounds in the TIC tr ace were reported, including aldehydes, ketones and alcohols. The author hopes that this work will serve as the foundation for future bioassays in which mosquito attraction to a single or mixture of co mpounds could be tested. However, since several

PAGE 130

130 compounds remain unidentified at this time, this could mean that the important compounds or potential attractants for the mosquito vect or of WNV have not been yet identified. In the latter part of the disse rtation, the evaluation of FAIMS as a potential analytical tool for the characterization of allelochemicals was di scussed. The release of allelochemicals in the field is normally accomplished by using surveillance traps baited with attractant lures, or appropriate candles or aerosol-di spensing devices to release repell ents. The problem is that the downfield distribution of these compounds in the field and subsequent de tection by arthropods is not well understood due to the lack of a field-monitoring analyzer. FAIMS is an ion mobility-based technique that has recently gained a lot of scientific interest for many applications. FAIMS separates ions based on the difference in their mobility at high electric field relative to their mobility at low electric field. There is still much to learn about the chemistry and physics of why ions be have the way they do inside the FAIMS cell when operating parameters such as electric fiel d strength and bath ga s composition are changed. The use of FAIMS as a potential tool for char acterizing allomones odor plumes is currently being evaluated. Optimization of several parameters is required in order to achieve the best ion transmission through the FAIMS cell. In this work, the curtain gas flow rate, dispersion voltage and the carrier gas composition were optimized. The effect of these parameters on the signal intensity and compensation voltage was studied for two allomones of interest, DEET and linalool. These two allomones were chosen b ecause DEET is a common topical repellent and linalool is an attraction-inhi bitor that exhibits spatial repellency in mosquitoes. Experiments showed that a higher sensitivity was obtained when a gas flow rate of 4.3 L/min was used. This supports the hypothesis that be tter transmission of ions in FAIMS is achieved when the conditions in side the cell are dry. Althoug h it is not expected for the

PAGE 131

131 compensation voltage to change with increasing gas flow rates, it was observed that for DEET the CV was shifted to more negative values, prob ably due to moisture or other contaminants present in the N2 curtain gas. Furthermore, an increase in sensitivity was observed at higher DV due to the ion focusing mechanism at atmospheric pr essure in the cylindrical electrodes. This is caused by a decrease in the ion current lost to the walls of the FAIMS analyzer due to ion focusing in the FAIMS analyzer region. The change in CV is due to deviations from the ion mobility at high electric field related to the mob ility at low field. Finally, it can be concluded that the degree of separation betw een compounds in a mixture and sens itivity are also affected in FAIMS by changing the carrier ga s composition. The use of a N2/He gas mixture provided with an increase in the ion signal, whereas th e opposite was observed when a mixture of N2/CO2 was used. Even though it is still unknown why sensitiv ity is affected when varying carrier gas composition, the polarizability of the gas is thought to play an important role in this kind of behavior. The results from initial studies with APCI/FAIMS using mass spectrometry to further characterize two allomones, DEET and linalool, demonstrated the capability of FAIMS to separate ions in a solution or h eadspace. Knowing the specific va lue for the CV at which the ion of interest is transmitted through the FAIMS ce ll permits FAIMS to act as an ion filter. Preliminary results using APCI/FAIMS for st udying the headspace of methanolic solutions of DEET and linalool have shown that FAIMS can be used for the analysis of gas-phase vapors. However, these studies showed that APCI/F AIMS/MS was more sensitive for DEET than linalool. Linalool in headspace was not detected below solution concentrations of 1000 ppm, presumably due to the low vapor pressure of linal ool. The use of heat to warm up the standard

PAGE 132

132 solutions could help in increase the vapor pressure of the compounds so they can be detected in the headspace at lower solution concentrations. Although quantitation of these two compounds was not the main purpose of this work, calibration curves were constructed in order to test linearity of the APCI/FAIMS/MS system. Linearity was found over a limited range of c oncentrations, perhaps due to inadequate equilibration time for the distribution between th e gas and liquid phase. In addition, as shown for DEET, increased formation of the proton-bo und dimer at higher concentrations (> 2.5 ppm) can affect linearity by lowering signal res ponse of the protonated monomer at higher concentrations. The incorporation of an in ternal standard may provide more reliable quantitation. Future Work Because there are compounds that remain uni dentified, the use of different sample collection methods, such as microscale purge an d trap and solid-phase micro extraction (SPME), may help provide a more complete knowledge of chicken odors. For example, the use of microscale purge and trap for analyzing chicken feathers may help in the study of the trace components that were masked by the high concentr ation of acids. Microscale purge and trap discriminates against acids, and allows for the detection of highly volatile compounds. If thermal desorption equipment is available in the future, it will be interesting to continue the direct thermal desorption of compounds from chicken feathers, since it is a faster way of analyzing volatiles compounds compared with th e longer collection time required when using the Anasorb-747 method. Studies in chapter 2 yielded several iden tified compounds from PCI and EI alone. Continued use of PCI is necessary for compliment ary information obtained from EI. The use of positive and negative ion CI and MS/MS could help in getting more information from

PAGE 133

133 components that failed to give useful fragmentation patterns by EI or PCI. MS/MS could be employed to obtain daughter spectr a of selected molecular ions or protonated molecular species. The use of GC/Electroantennogram (EAG) resp onse may be very useful in determining only those compounds that elicit a certain type of response in the mosquito. In this way, bioassays could be conducted to test the response of Culex mosquitoes to compounds already known to elicit some kind of response. A singl e compound or a mixture of compounds could be then spiked onto a glass petri di sh, since it has been demonstrated that volatile compounds can be transferred to glass, and this petri dish can be sampled in an olfactometer where mosquito response can be tested. Additional experiments using FAIMS could incl ude the evaluation of different gases or mixture of gases (i.e., dry air). It would be advantageous to know how ions behave in different type of gases in order to know which one to use when more sensitivity is required for the application of interest. Furthermore, the pot ential of a commercial portable FAIMS device, called the microDMx (Sionex) and the evaluation of different ionization sources on this device, will be beneficial for the analysis of other compounds of environmental interest. These compounds could include ultra low volume (ULV) insecticides and pyrethroids which provide a spatial repellency against insects, including mos quitoes. This way, FAIM S could be applied to other type of applications including those of military relevance. The microDMx could be tested as a means to map the ULV treatment in different environments (desert, ju ngle, etc.) to increase our knowledge about how best to use ULV adultic ides. In addition, the use of highly volatile pyrethroids can provide a spatial repellent effect to prevent insects from entering tents that contain military personnel. Therefore, the use of microDMx to indicate the concentration level of spatial repellent and thereby assess the thre shold concentrate at which the efficacy of the

PAGE 134

134 spatial repellent fails could be evaluated. Al so, the microDMx could be used to assess the remaining permethrin treatment to a combat utili ty uniform. Because the residual permethrin in a garment determines its overall bite protection, the microDMx could provide the U.S. military a less expensive route to measuring remaining perm ethrin in a garment, and thus whether the uniform can still protect the soldier from insect bites.

PAGE 135

135 LIST OF REFERENCES (1) James, A. A. Science 1992 , 257 , 37-38. (2) Gillett, J. D. Mosquito News 1979 , 39 , 221-229. (3) Takken, W. Invertebrate Reproduction & Development 1999 , 36 , 67-71. (4) Mboera, L. E. G.; Takken, W. Entomologia Experimentalis et Applicata 1999 , 92 , 83-88. (5) McIver, S. B. Journal of Medical Entomology 1982 , 19 , 489-535. (6) Davis, E. E.; Bowen, M. F. Journal of the American Mosquito Control Association 1994 , 10 , 316-325. (7) Takken, W.; Knols, B. G. J. Annual Review of Entomology 1999 , 44 , 131-157. (8) Knols, B. G. J.; Takken, W.; Cork, A.; DeJong, R. Annals of Tropical Medicine and Parasitology 1997 , 91 , S117-S118. (9) Takken, W.; Knols, B. G. J.; Otten, H. Annals of Tropical Medicine and Parasitology 1997 , 91 , S119-S120. (10) Bidlingmayer, W. L. Journal of the American Mosquito Control Association 1994 , 10 , 272-279. (11) Bowen, M. F. Annual Review of Entomology 1991 , 36 , 139-158. (12) Brown, A. W. A. Bulletin of Entomological Research 1951 , 42 , 575. (13) Browne, S. M.; Bennett, G. F. Journal of Medical Entomology 1981 , 18 , 505-521. (14) Bidlingmayer, W. L.; Hem, D. G. Bulletin of Entomological Research 1980, 70 , 321-342. (15) Brown, W. L.; Eisner, T.; Whittake R. Bioscience 1970 , 20 , 21. (16) Snow, W. F. Bulletin of Entomological Research 1970 , 60 , 43. (17) Gillies, M. T. Bulletin of Entomological Research 1980 , 70 , 525-532. (18) Takken, W.; Dekker, T.; Wijnholds, Y. G. Journal of Insect Behavior 1997 , 10 , 395-407. (19) Mboera, L. E. G.; Knols, B. G. J.; Takken, W.; dellaTorre, A. Bulletin of Entomological Research 1997 , 87 , 173-178. (20) Mboera, L. E. G.; Takken, W.; Sambu, E. Z. Bulletin of Entomological Research 2000 , 90 , 155-159.

PAGE 136

136 (21) Takken, W.; Kline, D. L. Journal of the American Mosquito Control Association 1989 , 5 , 311-316. (22) Acree, F.; Turner, R. B.; Gouck, H. K.; Beroza, M.; Smith, N. Science 1968 , 161 , 1346. (23) Vickery, R. K. Mosquito News 1966 , 507-508. (24) Hall, D. R.; Beevor, P. S.; Cork, A.; Nesbitt, B. F.; Vale, G. A. Insect Science and its Application 1984 , 5 , 335-339. (25) Vale, G. A.; Hall, D. R. Bulletin of Entomological Research 1985 , 75 , 209-217. (26) Kline, D. L. Journal of the American Mosquito Control Association 1994 , 10 , 280-287. (27) Kline, D. L.; Takken, W.; Wood, J. R.; Carlson, D. A. Medical and Veterinary Entomology 1990 , 4 , 383-391. (28) Knols, B. G. J.; vanLoon, J. J. A.; Co rk, A.; Robinson, R. D.; Adam, W.; Meijerink, J.; DeJong, R.; Takken, W. Bulletin of Entomological Research 1997 , 87 , 151-159. (29) Braks, M. A. H.; Meijerink, J.; Takken, W. Physiological Entomology 2001 , 26 , 142-148. (30) Meijerink, J.; Braks, M. A. H.; Brack, A. A.; Adam, W.; Dekker, T.; Posthumus, M. A.; Van Beek, T. A.; Van Loon, J. J. A. Journal of Chemical Ecology 2000 , 26 , 1367-1382. (31) Cork, A.; Park, K. C. Medical and Veterinary Entomology 1996 , 10 , 269-276. (32) Puri, S. N.; Mendki, M. J.; Sukumaran, D.; Ganesan, K.; Prakash, S.; Sekhar, K. Journal of Medical Entomology 2006 , 43 , 207-213. (33) Russell, C. B.; Hunter, F. F. Journal of Medical Entomology 2005 , 42 , 301-305. (34) Williams, C. R.; Kokkinn, M. J.; Smith, B. P. Journal of Chemical Ecology 2003 , 29 , 1889-1903. (35) Paul, W. [US Pa tent 2 939 952]. 1960. (36) March, R. E., Ed. Quadrupole Ion Trap Mass Spectromete r.Encyclopedia of Analytical Chemistry ; 2000, John Wiley & Sons Ltd.: New York, NY. (37) March, R. E. Journal of Mass Spectrometry 1997 , 32 , 351-369. (38) March, R. E. International Journal of Mass Spectrometry 2000 , 200 , 285-312. (39) Stafford, G. C., Kelley, P. E. and Stephens, D. R. [U.S. Patent 45400884]. 1985.

PAGE 137

137 (40) Du, Z. H.; Douglas, D. J.; Konenkov, N. Journal of Analytical Atomic Spectrometry 1999 , 14 , 1111-1119. (41) Dawson, P. H. Mass Spectrometry Reviews 1986 , 5 , 1-37. (42) Miller, P. E.; Denton, M. B. Journal of Chemical Education 1986 , 63 , 617-622. (43) Dawson, P. H.; Yu, B. Q. International Journal of Mass Spectrometry and Ion Processes 1984 , 56 , 25-39. (44) Dawson, P. H.; Yu, B. Q. International Journal of Mass Spectrometry and Ion Processes 1984 , 56 , 41-50. (45) Cohen, M. J.; Karasek, F. W. Journal of Chromatographic Science 1970 , 8 , 330. (46) Karasek, F. W. Research-Development 1970 , 21 , 34. (47) Eiceman, G. A. Z. K. Ion Mobility Spectrometry ; CRC Press: Baton Rouge, 1994. (48) Turner, R. B.; Brokenshire, J. L. Trac-Trends in Analytical Chemistry 1994 , 13 , 275-280. (49) Baumbach, J. I.; Eiceman, G. A. Applied Spectroscopy 1999 , 53 , 338A-355A. (50) Eiceman, G. A.; Ewing, G. J. Talanta 2001 , 54 , 425-426. (51) Eiceman, G. A.; Nazarov, E. G.; Stone, J. A. Analytica Chimica Acta 2003 , 493 , 185-194. (52) Borsdorf, H.; Stone, J. A.; Eiceman, G. A. International Journal of Mass Spectrometry 2005 , 246 , 19-28. (53) Eiceman, G. A.; Nazarov, E. G.; Tadjikov, B.; Miller, R. A. Field Analytical Chemistry and Technology 2000 , 4 , 297-308. (54) Purves, R. W.; Guevremont, R. Analytical Chemistry 1999 , 71 , 2346-2357. (55) Guevremont, R.; Purves, R. W. Review of Scientific Instruments 1999 , 70 , 1370-1383. (56) Eiceman, G. A.; Nazarov, E. G.; Mill er, R. A.; Krylov, E. V.; Zapata, A. M. Analyst 2002 , 127 , 466-471. (57) Eiceman, G. A.; Tadjikov, B.; Krylov, E. ; Nazarov, E. G.; Miller, R. A.; Westbrook, J.; Funk, P. Journal of Chromatography A 2001 , 917 , 205-217. (58) Rappole, J. H.; Hubalek, Z. Journal of Applied Microbiology 2003 , 94 , 47S-58S. (59) Petersen, L. R.; Roehrig, J. T. Emerging Infectious Diseases 2001 , 7 , 611-614.

PAGE 138

138 (60) Goddard, L. B.; Roth, A. E.; Reisen, W. K.; Scott, T. W. Emerging Infectious Diseases 2002 , 8 , 1385-1391. (61) Nasci, R. S.; Savage, H. M.; White, D. J.; Miller, J. R.; Cropp, B. C.; Godsey, M. S.; Kerst, A. J.; Bennett, P.; Gottfried, K.; Lanciotti, R. S. Emerging Infectious Diseases 2001 , 7 , 742-744. (62) Miller, B. R.; Nasci, R. S.; Godsey, M. S. ; Savage, H. M.; Lutwama, J. J.; Lanciotti, R. S.; Peters, C. J. American Journal of Tropi cal Medicine and Hygiene 2000 , 62 , 240-246. (63) Marfin, A. A.; Petersen, L. R.; Eidson, M.; Miller, J.; Hadler, J.; Farello, C.; Werner, B.; Campbell, G. L.; Layton, M.; Smith, P.; Bresnitz , E.; Cartter, M.; Scaletta, J.; Obiri, G.; Bunning, M.; Craven, R. C.; Roehrig, J. T.; Ju lian, K. G.; Hinten, S. R.; Gubler, D. J. Emerging Infectious Diseases 2001 , 7 , 730-735. (64) Scherer, W. F.; Buescher, E. L.; Flemings, M. B.; Noguchi, A.; Scanlon, J. American Journal of Tropical Medicine and Hygiene 1959 , 8 , 665-677. (65) Allan, S. A.; Bernier, U. R.; Kline, D. L. Journal of Medical Entomology 2006 , 43 , 225231. (66) Bernier, U. R.; Booth, M. M.; Yost, R. A. Analytical Chemistry 1999 , 71 , 1-7. (67) Bernier, U. R.; Kline, D. L.; Barnard, D. R.; Schreck, C. E.; Yost, R. A. Analytical Chemistry 2000 , 72 , 747-756. (68) Schreck, C. E.; Smith, N.; Carlson, D. A. ; Price, G. D.; Haile, D.; Godwin, D. R. Journal of Chemical Ecology 1982 , 8 , 429-438. (69) Schreck, C. E.; Gouck, H. K.; Smith, N. Journal of Economic Entomology 1967 , 60 , 1188. (70) Schreck, C. E.; Kline, D. L.; Carlson, D. A. Journal of the Americ an Mosquito Control Association 1990 , 6 , 406-410. (71) Gjolstad, M.; Bergemalm-Rynell, K. ; Ljungkvist, G.; Thorud, S.; Molander, P. Journal of Separation Science 2004 , 27 , 1531-1539. (72) Chapman, J. R. Practical Organic Mass Spectrometry ; 1985, John Wiley & Sons: Chichester, England. (73) Munson, M. S. B.; Field, F. H. Journal of the American Chemical Society 1966 , 88 , 2621. (74) Karasek, F. W., Hutzinger, O. & Safe, S. Mass Spectrometry in Environmental Sciences ; 1985, Plenum Press: New York, NY.

PAGE 139

139 (75) Nicolaid N. Science 1974 , 186 , 19-26. (76) Zibulewsky, J.; Fr ied, B.; Bacha, W. J. Journal of Parasitology 1982 , 68 , 905-908. (77) Wertz, P. W.; Stover, P. M.; Abraham, W.; Downing, D. T. Journal of Lipid Research 1986 , 27 , 427-435. (78) Burger, B. V.; Reiter, B.; Borzyk, O.; Du Plessis, M. A. Journal of Chemical Ecology 2004 , 30 , 1603-1611. (79) Guerenstein, P. G.; Guerin, P. M. Journal of Experimental Biology 2001 , 204 , 585-597. (80) Du, Y. J.; Millar, J. G. Journal of Medical Entomology 1999 , 36 , 158-166. (81) Buryakov, I. A.; Krylov, E. V.; Nazarov, E. G.; Rasulev, U. K. International Journal of Mass Spectrometry and Ion Processes 1993 , 128 , 143-148. (82) Guevremont, R. Journal of Chromatography A 2004 , 1058 , 3-19. (83) Buryakov, I. A.; Kolomiets, Y. N.; Luppu, B. V. Journal of Analytical Chemistry 2001 , 56 , 336-340. (84) Buryakov, I. A. Talanta 2003 , 61 , 369-375. (85) Spangler, G. E. Field Analytical Chemistry and Technology 2000 , 4 , 255-267. (86) Eiceman, G. A.; Krylov, E. V.; Nazarov, E. G.; Miller, R. A. Analytical Chemistry 2004 , 76 , 4937-4944. (87) Eiceman, G. A.; Krylov, E. V.; Tadjikov, B. ; Ewing, R. G.; Nazarov, E. G.; Miller, R. A. Analyst 2004 , 129 , 297-304. (88) Purves, R. W.; Guevremont, R.; Day, S.; Pipich, C. W.; Matyjaszczyk, M. S. Review of Scientific Instruments 1998 , 69 , 4094-4105. (89) Carnahan, B. L. and Tarrasov, A. S. [US Patent 5420424]. 1995. (90) Creaser, C. S.; Griffiths, J. R.; Bramwell, C. J.; Noreen, S.; Hill, C. A.; Thomas, C. L. P. Analyst 2004 , 129 , 984-994. (91) Miller, R. A.; Eiceman, G. A.; Nazarov, E. G.; King, A. T. Sensors and Actuators BChemical 2000 , 67 , 300-306. (92) Krylov, E. V. Technical Physics 1999 , 44 , 113-116.

PAGE 140

140 (93) Guevremont, R.; Purves, R. W.; Barnett, D. A.; Ding, L. Y. International Journal of Mass Spectrometry 1999 , 193 , 45-56. (94) Guevremont, R.; Ding, L. Y.; Ells, B.; Barnett, D. A.; Purves, R. W. Journal of The American Society for Mass Spectrometry 2001 , 12 , 1320-1330. (95) Barnett, D. A.; Ells, B.; Guevremont, R. Journal of the American Society for Mass Spectrometry 2000 , 11 , 1125-1133. (96) Shvartsburg, A. A.; Tang, K. Q.; Smith, R. D. Analytical Chemistry 2004 , 76 , 7366-7374. (97) McCooeye, M. A.; Ells, B.; Barnett, D. A.; Purves, R. W.; Guevremont, R. Journal of Analytical Toxicology 2001 , 25 , 81-87. (98) Wright, R. H. Scientific American 1975 , 233 , 104-111. (99) Bernier, U. R.; Furman, K. D.; Kline, D. L.; Allan, S. A.; Barnard, D. R. Journal of Medical Entomology 2005 , 42 , 306-311. (100) McCabe, E. T.; Barthel, W. F.; Gertler, S. I.; Hall, S. A. Journal of Organic Chemistry 1954 , 19 , 493-498. (101) Sutcliffe, J. F. Journal of the American Mosquito Control Association 1994 , 10 , 309-315. (102) Boeckh, J.; Breer, H.; Geier, M.; Hoever, F. P.; Kruger, B. W.; Nentwig, G.; Sass, H. Pesticide Science 1996 , 48 , 359-373. (103) Kline, D. L.; Bernier, U. R.; Posey, K. H.; Barnard, D. R. Journal of Medical Entomology 2003 , 40 , 463-467. (104) Grob, R. L. Modern Practice of Gas Chromatography , 1985, 2nd ed.; John Wiley & Sons, Inc.: New York, NY. (105) Miller, R. A.; Nazarov, E. G.; Eiceman, G. A.; King, A. T. Sensors and Actuators APhysical 2001 , 91 , 301-312. (106) Kebarle, P.; Godbole, E. W. Journal of Chemical Physics 1963 , 39 , 1131. (107) Shahin, M. M. Journal of Chemical Physics 1966 , 45 , 2600. (108) Shahin, M. M. Journal of Chemical Physics 1967 , 47 , 4392. (109) Griffin, G. W.; Dzidic, I.; Carroll, D. I.; Stillwel R., Horning, E. C. Analytical Chemistry 1973 , 45 , 1204-1209.

PAGE 141

141 (110) Watson, J. T. Introduction to Mass Spectrometry , 1997, 3rd ed.; Lippincott-Raven: Philadelphia, PA. (111) Cole, R. B. Electrospray Ionization Mass Spectrome try. Fundamentals, Instrumentation and Applications ; 1997, John Wiley & Sons , Ltd.: New York, NY. (112) Yost, R. A.; Fetterolf, D. D. Mass Spectrometry Reviews 1983 , 2 , 1-45. (113) Yost, R. A.; Boyd, R. K. Methods in Enzymology 1990 , 193 , 154-200. (114) Louris, J. N.; Cooks, R. G.; Syka, J. E. P.; Kelley, P. E.; Stafford, G. C.; Todd, J. F. J. Analytical Chemistry 1987 , 59 , 1677-1685.

PAGE 142

142 BIOGRAPHICAL SKETCH Samaret M. Otero Santos, daught er of Jos J. Otero and Carmen M. Santos, was born on December 23, 1978, in San Juan, Puerto Rico. Her interest in the area of sciences started to develop in middle school, where she competed in science fairs until attending high school. She graduated from a specialized school in mathematics and sciences, University Gardens High School in May, 1996. In August, 1996, she began her bachelor degree studies in the field of chemistry at the University of Puerto Ric o, Rio Piedras Campus. During her undergraduate years, she became a member of the American Chemical Society and started undergraduate research under the supervision of Dr. Edwi n Quiones and received in 1998 to 2000 the Research Initiative for Scient ific Enhancement Award. In May, 2000, she graduated from the University of Puerto Rico. Two months later, she began graduate studies in the field of physical chemistry. As part of the chemistry graduate school at this univ ersity, she taught the undergraduate level physical chemistry laboratory until December, 2000. While at the graduate program in Puerto Rico, encouragement from fa mily and friends brought her to apply to the chemistry graduate program at the University of Florida, in Gaines ville. In August 2001 upon acceptance to the analytical chemistry graduate pr ogram at UF, she left the graduate school in Puerto Rico, and left her family and moved to the United States to purs ue a doctoral degree in this field. Under the supervision of Dr. Richar d A. Yost, she received her PhD in analytical chemistry in May, 2007.