Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Laser desorption - laser photoionization ion trap mass spectrometry for the direct analysis of solid… Specht, August Anders 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-902714.pdf [ 11.7MB ]
Metadata
JSON: 831-1.0061200.json
JSON-LD: 831-1.0061200-ld.json
RDF/XML (Pretty): 831-1.0061200-rdf.xml
RDF/JSON: 831-1.0061200-rdf.json
Turtle: 831-1.0061200-turtle.txt
N-Triples: 831-1.0061200-rdf-ntriples.txt
Original Record: 831-1.0061200-source.json
Full Text
831-1.0061200-fulltext.txt
Citation
831-1.0061200.ris

Full Text

Laser Desorption - Laser Photoionization Ion Trap Mass Spectrometry for the Direct Analysis of Solid Samples B y Augus t Ander s Specht B . S c . H . , Queen ' s Unive r s i ty , 1998 A T H E S I S S U B M I T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y In T H E F A C U L T Y O F G R A D U A T E S T U D I E S Department o f Chemis t ry W e accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A December 2003 © Augus t Ander s Specht, 2003 ABSTRACT T h i s thesis describes w o r k associated w i t h the creation o f a nove l ana ly t ica l instrument, a two-laser i o n trap mass spectrometer. T h i s device combines one o f the most sensit ive and selective direct so l id sampl ing techniques, two-laser s o l i d sampl ing , w i t h a very versatile mass spectrometer, the i o n trap. T h i s w o r k demonstrates some o f the potential advantages associated w i t h coup l ing these two techniques. The results f rom this w o r k can be d iv ided into three sections. Sec t ion one shows data associated w i t h the development and characterization o f this new instrument w h i c h couples I R laser desorption fo l lowed by U V laser photo ioniza t ion and analysis us ing an i o n trap mass spectrometer. F o r cal ibrat ion, a new type o f so l id sample preparation method i n v o l v i n g activated charcoal as the so l id substrate was used. It was found that both the I R and U V intensity, and the delay between them, p lay an important role i n both the magnitude and type o f signals observed. A method o f gas phase ion accumula t ion was also examined . F i n a l l y , this section demonstrates the technique 's abi l i ty to provide direct quali tat ive informat ion for P A H content for N . I . S T . S R M 1944 r iver sediment w i t h no sample pre-treatment. The second section o f this thesis shows data relating to the detection o f the pharmaceut ical agent Spiperone direct ly o n a so l id b io log i ca l l y relevant tissue matr ix . T h i s data shows that the two-laser i o n trap method is suitable for applicat ions w i t h compl ica ted matrices w i t h no need for sample pre-treatment. F i n a l l y , the th i rd section o f this thesis describes the addi t ion o f a third, tunable laser to the system. T h i s laser a l lows for opt ical p robing o f the trapped i o n c loud . V i s i b l e absorpt ion spectra for the gas phase P A H cations isomers phenanthrene and anthracene are shown. The th i rd laser also a l lows for the poss ib i l i ty o f wavelength selective photodissocia t ion o f P A H isomers for reso lv ing compl ica ted isomer mixtures. i i TABLE OF CONTENTS A B S T R A C T II T A B L E OF C O N T E N T S I l l LIST OF T A B L E S VI I LIST OF F I G U R E S VIII L IST OF A B B R E V I A T I O N S X V A C K N O W L E D G M E N T S X I X 1 I N T R O D U C T I O N 1 1.1 O V E R V I E W 1 1.2 M A S S SPECTROMETRY 2 1.2.1 Theory of the Quadrupole Ion Trap 4 1.2.2 Mathematical Description of the Ion Trap 7 1.2.3 The 3D Quadrupole Ion Trap as a Mass Spectrometer 15 1.3 S A M P L I N G A N D IONIZATION FOR M A S S SPECTROMETRY . . . . . 2 2 1.3.1 Two-Laser Solid Sampling (L2MS) 24 1.3.2 Laser - Solid Interactions ..26 1.3.3 Laser Photoionization 29 1.4 C O U P L I N G T w o - L A S E R SOLID S A M P L I N G WITH A N ION T R A P 3 3 1.4.1 This Work : ... .-. 36 2 E X P E R I M E N T A L 38 2.1 INTRODUCTION 3 8 i i i 2.2 T H E ION T R A P ELECTRODES A N D V A C U U M M A N I F O L D 4 0 2.3 ELECTRONICS A N D TIMING 4 7 2.3.1 Ion Trap Electronics 47 2.3.2 The Laser Timing 49 2.4 SOFTWARE FOR ION T R A P OPERATION 53 2.5 O V E R A L L S Y S T E M A N D TIMING 59 T W O L A S E R ION T R A P M A S S S P E C T R O M E T R Y F O R T H E A N A L Y S I S OF E N V I R O N M E N T A L S A M P L E S 62 3.1 INTRODUCTION 6 2 3.2 E X P E R I M E N T A L 66 3.3 RESULTS A N D DISCUSSION 67 3.3.1 Test Of Instrument Effectiveness 67 3.3.2 Sample Preparation 74 3.3.3 Two Laser Ion Trap Mass Spectra 78 3.3.4 Effect of IR Power on Observed Signal 81 3.3.5 Effect of UV power on Observed Signal 85 3.3.6 Semi-Quantitative Analysis 88 3.3.7 Selective Ion Accumulation 90 3.3.8 SRM1994 Analysis 94 3.4 CONCLUSIONS 97 D E S O R P T I O N P R O F I L E S OF P A H S F R O M A C T I V A T E D C H A R C O A L 98 4.1 INTRODUCTION 98 i v 4.2 E X P E R I M E N T A L 100 4.3 RESULTS A N D DISCUSSION 100 4.4 . CONCLUSIONS 109 5 D E T E C T I O N O F T H E D R U G S P I P E R O N E O N B I O L O G I C A L M A T R I C E S H I 5.1 INTRODUCTION I l l 5.2 E X P E R I M E N T A L , 114 5.3 RESULTS A N D DISCUSSION 115 5.4 CONCLUSIONS 124 6 O P T I C A L S P E C T R O S C O P Y I N A N I O N T R A P . .. ..' . . . .132 6.1 INTRODUCTION ". . 1 3 2 6.2 E X P E R I M E N T A L 136 6.3 RESULTS A N D DISCUSSION 141 6.3.1 Photodissociation of Trapped Cations 141 6.3.2 Fragmentation Pathways in Anthracene and Phenanthrene 150 6.3.3 Visible Spectra of Phenanthrene and Anthracene cations.... .....153 6.4 CONCLUSIONS 159 7 S E M I - Q U A N T I T A T I V E D E T E R M I N A T I O N O F P A H I S O M E R S D I R E C T L Y F R O M S O L I D M A T R I C E S 162 7.1 INTRODUCTION 162 7.2 E X P E R I M E N T A L 167 7.3 RESULTS A N D DISCUSSION : 168 v 7.4 CONCLUSION 178 8 C O N C L U S I O N S 182 8.1 GENERALITIES 182 8.2 E N V I R O N M E N T A L A N A L Y S I S 183 8.3 BIOLOGICAL A N A L Y S I S 186 8.4 SPECTROSCOPY OF T R A P P E D IONS 188 R E F E R E N C E S 191 v i LIST OF T A B L E S Table 3.1 P A H compounds certified to be contained i n S R M 1944a 96 Table 4.1 Exper imen ta l ly determined ideal delay t imes between laser events 103 Tab le 4.2 Table o f P A H mass and experimental ly determined M a x w e l l - B o l t z m a n n temperature for laser desorption f rom activated charcoal 108 Table 7.1 S u m m a r y o f symbols and reference slopes used i n this chapter 172 Table 7.2 S u m m a r y o f observed data for the " u n k n o w n " sample 179 v n LIST OF FIGURES Figure 1.1 P ic to r i a l d iagram o f the i o n trap mass spectrometer, (a) A cross section through the longi tudina l axis, (b) A three-dimensional representation 6 F igu re 1.2 A graphical representation o f the stable solutions to the M a t h i e u equat ion i n both the radial (r) and ax ia l (z) directions i n terms o f au and qu [13] 12 F igure 1.3 A h enlarged v i e w o f the experimental ly important region o f the M a t h i e u stabil i ty d iagram [13] 13 F igure 1.4 A n enlarged v i e w o f the stabili ty d iagram showing the possible posi t ions o f several ions o f different mass on the stabili ty diagram under var ious R F voltage condi t ions 16 F igure 1.5 Schemat ic cartoon o f two-laser mass spectrometry as it has been p rev ious ly appl ied ... 25 F igure 1.6 S i m p l i f i e d Jab lonski diagram for a number o f possible photon/molecule interactions, (a) S ing le photon ion iza t ion (b) N o ion iza t ion (c) Resonant two photon ion iza t ion (d) N o n Resonant mul t iphoton ion iza t ion (e) T w o photon resonant ion iza t ion 31 F igure 2.1 A conceptual representation o f the two-laser i o n trap system at U B C 39 F igure 2.2 V a c u u m man i fo ld for the two-laser i o n trap system at U B C 41 F igure 2.3 S o l i d sample probe for the two-laser i o n trap system at U B C 43 F igure 2.4 C l o s e up v i e w o f the I R and U V lasers as they pass through the i o n trap 45 F igure 2.5 T i m i n g diagram for the two-laser system used at U B C 52 F igure 2.6 P r imary operating w i n d o w for Ion Trap control software. Wr i t t en i n the L a b V i e w environment 54 v i i i Figure 2.7 F l o w chart for the Ion Trap software....... ". .' 56 F igure 2.8 L o g i c a l b lock diagram showing the flow o f energy and informat ion for the two-laser i o n trap system at U B C 60 F igure 3.1 The s ix p o l y c y c l i c aromatic hydrocarbons ( P A H s ) p r imar i ly used i n this w o r k . ; 63 F igure 3.2 M a s s spectrum o f C C L ion ized by electron impact ion iza t ion , resul t ing i n the format ion o f C C L * ions 68 F igure 3.3 M a s s spectrum o f CCI4 i on ized by electron impact ion iza t ion and ejected f rom the i o n trap by the resonance ejection mode o f operation 70 F igure 3.4 M a s s spectrum o f CCI4 i on ized by electron impact ion iza t ion w i t h a single N B B W c y c l e after ion iza t ion . The N B B W cyc le caused the r emova l o f a l l mass components b e l o w 119 T h and above 120 T h 71 F igure 3.5 M a s s spectrum o f CCI4 i on ized by electron impact ion iza t ion and w i t h removal o f 119 T h by a single frequency ejection after ion iza t ion 73 F igure 3.6 M e c h a n i c a l press and sample cup used i n the creation o f so l id samples for the two-laser system at U B C 77 F igure 3.7 T y p i c a l mass spectrum col lected w i t h the two-laser mode o f ion iza t ion o f a sample o f charcoal sp iked w i t h f ive P A H s ; 79 F igure 3.8 Expanded v i e w o f F igure 3.7 between mass 130-265 T h 80 F igure 3.9 Magn i tude o f the pyrene peak area as a function o f the number o f laser cyc les for a charcoal sample sp iked w i t h pyrene 82 i x Figure 3.10 Effect o f I R power on observed signal for phenanthrene as a funct ion o f the number o f laser shots for three I R laser powers. • Measured I R P o w e r (Left H a n d Uni t s ) . • Integrated Peak Areas (Right H a n d Uni t s ) 83 F i g u r e 3.11 U V power (uJ) vs. integrated peak areas for f ive P A H s . T h e x-ax i s is i n the units o f uJ o f U V energy i n the trap vo lume and the y-axis is i n terms o f integrated peak areas 86 F igure 3.12 Concentra t ion vs. measured s ignal for a series o f pyrene/charcoal standards. ; . 8 9 F igure 3.13 M a s s spectrum from a sample containing 6 P A H s (chrysene depleated) [black l ine] . Af te r chrysene gas phase pre concentration [gray l ine] . 92 F igure 3.14 A v e r a g e chrysene peak height normal ized relative to a single laser cyc l e as a funct ion o f the number o f laser cycles 93 F igure 3.15 T w o laser mass spectrum observed f rom a sample o f standard reference mater ial 1944a 95 F igure 4.1 Integrated peak areas vs. I R - U V delay t ime for Acenaphthene, Phenanthrene, Pyrene, Chrysene, and Benzo[a]pyrene 102 F igure 4.2 Chart o f exper imental ly determined M a x w e l l - B o l t z m a n n temperature vs. P A H mass 108 F igure 5.1 Photo o f bra in tissue extracted f rom a male Sprague D a w l e y rat used i n this w o r k 116 F igure 5.2 Photo o f l iver tissue extracted f rom a male Sprague D a w l e y rat used i n this w o r k 116 F igure 5.3 T w o laser mass spectrum o f Spiperone o n a charcoal matr ix 118 Figure 5.4 T w o laser mass spectrum o f Spiperone o n a charcoal matr ix w i t h 1,5, and 10 laser cycles f o l l o w e d by a single N B B W pulse 119 F igure 5.5 T w o laser mass spectrum o f Spiperone on a charcoal matr ix w i t h f ive laser cycles f o l l o w e d by a single N B B W pulse and a c o l l i s i o n induced dissocia t ion ( C I D ) wave fo rm 121 F igure 5.6 T w o laser mass spectrum o f a sl ice o f bra in tissue f rom a male Sprague-D a w l e y rat 122 F igure 5.7 T w o laser mass spectrum o f a sl ice o f l iver tissue f rom a male Sprague-D a w l e y rat 123 F igure 5.8 T w o laser mass spectrum o f a sl ice o f bra in tissue f rom a male Sprague-D a w l e y rat w h i c h had been spiked w i t h a Spiperone solut ion 125 F igure 5.9 T w o laser mass spectrum w i t h 5 laser cycles fo l l owed by a single N B B W pulse o f a s l ice o f bra in tissue from a male Sprague-Dawley rat that had been sp iked w i t h a Spiperone solut ion 126 F igure 5.10 T w o laser mass spectrum w i t h 5 laser cycles fo l lowed by a single N B B W pulse and a single C I D wavefo rm o f a sl ice o f b ra in tissue f rom a male Sprague-D a w l e y rat w h i c h had been spiked w i t h a Spiperone solut ion 127 F igure 5.11 T w o laser mass spectrum o f a sl ice o f l iver tissue f rom a male Sprague-D a w l e y rat w h i c h had been spiked w i t h a Spiperone solut ion 128 F igure 5.12 T w o laser mass spectrum w i t h 5 laser cycles fo l l owed by a single N B B W pulse o f a s l ice o f l i ve r tissue froiri a male Sprague-Dawley rat that had been sp iked w i t h a Spiperone solut ion 129 x i Figure 5.13 T w o laser mass spectrum w i t h 5 laser cycles fo l lowed by a single N B B W pulse and a single C I D waveform o f a sl ice o f l iver tissue f rom a male Sprague-D a w l e y rat w h i c h had been spiked w i t h a Spiperone solut ion 130 F igure 6.1 D i a g r a m o f the three-laser set-up at U B C 138 F i g u r e 6.2 Enhanced photo o f the three-laser set-up at U B C 139 F igure 6.3 D i a g r a m o f the I R desorption, U V photoionizat ion, and v i s ib l e photo fragmentation lasers interacting w i t h the interior o f the i o n trap 140 F igure 6.4 The P A H isomers at 178 T h phenanthrene and anthracene 142 F igure 6.5 T w o laser mass spectrum o f a sample o f phenanthrene o n activated charcoal . v 143 F igure 6.6 T w o laser mass spectrum o f a sample o f phenanthrene on activated charcoal w i t h the addi t ion o f a N B B W isola t ion pulse 145 F igure 6.7 T w o laser mass spectrum o f phenanthrene o n activated charcoal w i t h the addi t ion o f a N B B W isola t ion pulse fo l lowed b y 5. photodissocia t ion laser shots (892 nm) .,_ 146 F igure 6.8 T w o laser mass spectrum o f a sample o f anthracene on activated charcoal w i t h the addi t ion o f a N B B W isola t ion pulse fo l l owed by 5 photodissocia t ion laser shots (682 nm) 147 F igure 6.9 Daughter i o n popula t ion observed over 0-4 photodissocia t ion laser shots (focusing around 152 T h ) o n a sample o f anthracene 148 F igu re 6.10 Daughter i o n popula t ion observed over 0-4 photodissocia t ion laser shots (focusing around 178 T h ) o n a sample o f anthracene 148 x n Figure 6.11 N o r m a l i z e d areas for the ratio o f 152/178 vs. the number o f photodissocia t ion laser shots on a sample o f anthracene 149 F igure 6.12 Fragmentat ion eff iciency (ratio o f 152/178 Th) vs. energy at two different photodissocia t ion laser wavelengths (720 n m and 740 nm) for a sample o f anthracene 155 F igu re 6.13 T o p spectrum- Photodissocia t ion spectra o f the anthracene cat ion (ratio o f 152/178 T h vs. wavelength). B o t t o m spectrum - Anthracene cat ion spectrum acquired i n a frozen argon matr ix at 12 K [196] 156 F igure 6.14 T o p spectrum- Photodissocia t ion spectrum o f the phenanthrene cat ion (ratio o f 152/178 T h vs. wavelength). B o t t o m spectra- Phenanthrene cat ion spectrum acquired i n a frozen neon matr ix at 4.2 K [165] 158 F igure 7.1 Anthracene and phenanthrene photofragmentation spectra acquired i n an i o n trap by the R E M P D method 166 F igure 7.2 T w o laser mass spectrum o f a sample o f phenanthrene and pyrene o n activated charcoal w i t h the addi t ion o f a N B B W laser pulse. The gray l ine is w i t h the addi t ion o f 20 photofragmentation laser shots at 892 n m 174 F igure 7.3 T w o laser spectrum o f a sample o f anthracene and pyrene o n activated charcoal w i t h the addi t ion o f a swift laser pulse. The gray l ine is w i t h the addi t ion o f 20 photofragmentation laser shots at 892 n m 175 F igure 7.4 Ca l ib ra t i on curve for anthracene and phenanthrene measured relative to pyrene w i t h and wi thout the addi t ion o f the photofragmentation laser 176 F igure 7.5 T w o laser mass spectrum o f an " u n k n o w n " sample o f anthracene, phenanthrene, and pyrene on activated charcoal w i t h the addi t ion o f a N B B W laser x i i i pulse. The gray l ine is w i t h the: addi t ion o f 20 photofragmentation laser shots at 892 n m . : : • 177 F igure 8.1 T w o laser mass spectrum o f Buckminsterful lerene 190 x i v LIST OF ABBREVIATIONS [A] Anthracene concentration [P] phenanthrene concentrat ion [Py] pyrene concentration 0-p zero-to-peak a acceleration (m-s" ) A C Al te rna t ing Current A D C A n a l o g - t o - D i g i t a l Conve r s ion a m u atomic mass unit a u M a t h i e u parameter (dimensionless) C Constant used i n M - B equation C I C h e m i c a l Ionizat ion C I D C o l l i s i o n Induced Dissoc ia t ion c m Cent imeter d distance between desorption and ion iza t ion events i n M - B equation D a D a l t o n D A C D i g i t a l - t o - A n a l o g Conve r s ion D C Direc t Current D I B Diffuse Interstellar Bands E E lec t r i c f ie ld (V-m" 1 ) e elementary charge (1 .6021-10" 1 9 C) E I E lec t ron Impact Ioniza t ion E S I Electrospray Ionizat ion e V electron V o l t (1.602189 x 10- | 9 J ) f frequency (cycles/second) . F Force (kg -n r s" ) F A B Fast A t o m Bombardment f r agA fragmentation eff ic iency o f anthracene induced by v i s i b l e laser fragP fragmentation efficiency o f phenanthrene induced by v i s i b l e x v laser fragPy fragmentation efficiency o f pyrene induced by v i s ib l e laser F T - I C R • Four ie r Transform Ion C y c l o t r o n Resonance F W H M F u l l W i d t h at H a l f M a x i m u m g(t) the observed s ignal i n the M - B as a funct ion o f t ime G C Gas Chromatography G P I B Genera l Purpose Interface B u s I /O Input/Output I A observed peak intensity o f anthracene I E E E Institute o f E lec t r i ca l and E lec t ron ic Engineers I P observed peak intensity o f phenanthrene Ip y observed peak intensity o f pyrene I R Infrared K K e l v i n k B o l t z m a n n Constant (1.3807 x 10" 2 3 J / K ) L 2 M S laser desorption laser photo ioniza t ion mass spectrometry L C L i q u i d Chromatography L I T D Laser Induced The rma l Desorp t ion m mass (kg) m / z mass/charge M A L D I M a t r i x Ass i s t ed Laser Desorpt ion Ioniza t ion M - B M a x w e l l - B o l t z m a n n M H z mega H z (10 6 cycles/second) M - n mass peak at parent mass ( M ) minus a who le number (n) M P I M u l t i Pho ton Ionizat ion M S M a s s Spectrometry M S N Tandem mass spectrometry to the N - l level M W M o l e c u l a r Weigh t N B B W N o t c h e d B r o a d B a n d W a v e f o r m N d : Y A G N e o d y m i u m : Y t t r i u m A l u m i n u m Garnet Laser P A H P o l y c y c l i c A r o m a t i c Hydroca rbon x v i P C I Personal Computer Interface B o a r d P D Photo Dis soc ia t ion PI Photo Ionizat ion q u M a t h i e u parameter (dimensionless) r radial coordinate i n the trap vo lume R E M P D Resonance Enhanced M u l t i p h o t o n D i s soc i a t i on R E M P I Resonance Enhanced M u l t i p h o t o n Ioniza t ion R F R a d i o Frequency R L O Ra t io (wi th the ) Laser O N R N L Ra t io (with) N o (vis ible) Laser r 0 internal radius o f r ing electrode s second S I M S Secondary Ion M a s s Spectrometry t t ime (seconds) - . . • T O F T i m e o f F l igh t mass spectrometry U D C component o f <P0 U B C Unive r s i ty o f B r i t i s h C o l u m b i a U S E P A U n i t e d States Envi ronmenta l Protect ion A g e n c y U V Ul t rav io le t : ^ V A C component o f O 0 v s stream veloc i ty vs. versus :.„•• ] . ' V U V V a c u u m U l t r a V i o l e t W Watt z charge z ax ia l coordinate i n the trap vo lume z 0 XA o f the endcap electrode spacing CXA constant for desorption/ionization/trapping efficiencies for anthracene ap constant for desorption/ionization/trapping efficiencies for phenantherene x v i i constant for desorption/ionization/trapping efficiencies for pyrene p i o n trajectory stabili ty value (dimensionless) y weight ing constant for z X weight ing constant for x u m micrometers $ M a t h i e u equation parameter a weight ing constant for y angular frequency o f the A C potential (radians/second) COu,n nt h order frequency o f a stable i o n trajectory o 0 appl ied electric potential X V l l l ACKNOWLEDGMENTS The comple t ion o f a doctoral thesis represents a turning point i n ones l i fe . H o w e v e r , this is i n no way an achievement o f a single person. Instead, it is the result o f the interactions o f many people w h o have contributed their t ime and energy to m a k i n g this possible . I have been fortunate throughout m y life that these contributions have a lways arr ived i n a posi t ive and t ime ly manner. A technical thesis o f this type can on ly be accompl i shed w i t h the dedicat ion, sk i l l s , and talents o f an entire team o f workers . I w o u l d l ike to thank the gentlemen o f the M e c h a n i c a l and Elec t ron ics shops at U B C for their advice and qual i ty workmansh ip . E s p e c i a l l y ; D a v e B a i n s , M a r t i n Car l i s l e , M i l a n C o s c h i z z a , B r i a n D i t chburn , Jason G o z j o l k o , B r i a n Greene, K e n L o v e , R o n M a r w i c k , B r i a n Snapkauskas, and D a v e T o n k i n . I w o u l d also l ike to thank m y many academic colleagues and mentors w h o m prov ided m u c h support and encouragement over the years. Thanks are especia l ly due to m y supervisor M i k e Blades , for a l l o w i n g ample intel lectual freedom w h i l e p r o v i d i n g ready support and encouragement when needed. I w o u l d also l ike to thank John H e p b u r n and D o n Doug la s for a l l o w i n g me to be a guest i n their respective groups. Thanks are also extended to the many other analyt ical members o f the department for there help and advice , i nc lud ing D a v i d Chen , A l a n Ber t ram, Bruce T o d d , the Doug la s G r o u p , the C h e n Group , the Hepburn Group , and o f course m y f e l l o w Blades group members . I w o u l d especial ly l ike to thank Chr i s Barbosa , D a v e M c L a r e n , Den i s R o l l a n d , and K e n W r i g h t for many useful discussions, and even more not so useful ones. I also w o u l d l ike to thank m y fami ly and friends both i n Canada, the U S , and abroad. I feel t ruly blessed to have them i n m y life. Espec i a l ly to m y mother and father w h o never doubted m y decisions or m y methods. Specia l thanks are also extended to Joanna K i r k e m y friend and companion . I w o u l d also l ike to acknowledge the f inancial support I received f rom the U n i v e r s i t y o f B r i t i s h C o l u m b i a and the Natura l Science and Engineer ing Research C o u n c i l o f Canada. F i n a l l y , I w o u l d l ike to dedicate this thesis to G l e n Nagano , Grade 7 teacher at L o r d Strathcona Elementary . I often wonder h o w m y life w o u l d have turned out i f I were not so fortunate as to have been i n your class. x i x C H A P T E R 1 INTRODUCTION 1.1 Overview T h i s thesis describes the construct ion and characterizat ion o f a n e w analy t ica l instrument that combines one o f the most useful and versatile mass spectrometers, the quadrupole i o n trap, w i t h a very selective and sensitive method for direct s o l i d sample analysis , two laser sampl ing and ioniza t ion . Th i s instrument w i l l be shown to have many advantages over the tradit ional means o f performing these techniques. T o appreciate the advantages o f c o m b i n i n g these two methods, the remainder o f this chapter w i l l describe the i o n trap mass spectrometer and the typ ica l modes o f its operation. It w i l l also provide an introduct ion to laser sampl ing theory and the combina t ion o f laser desorption - laser photoioniza t ion as a methodology for chemica l analysis. F i n a l l y , this chapter w i l l conclude w i t h a descript ion o f the advantages and disadvantages that result f rom combin ing this mode o f sampl ing and ion iza t ion w i t h an i o n trap mass spectrometer. C H A P T E R 2 w i l l describe the design, development, construction, and in i t i a l characterizat ion o f the instrument. It w i l l provide specific details o f the components and parts o f the instrument and general details o n its methods o f operation. The remain ing chapters w i l l describe experiments that were performed to characterize the potential uses o f this instrument. A l l o f these remain ing "data" chapters w i l l have their o w n int roduct ion and experimental sections i n order to place the w o r k i n context relative to 1 the literature and describe the specific operating condit ions under w h i c h the experiments were performed. C H A P T E R 3 w i l l show typica l results obtained i n the routine operat ion o f this instrument w i t h a variety o f analytes and samples o f environmental interest. It w i l l also demonstrate some typ ica l modes o f operation o f the instrument. C H A P T E R 4 describes the laser desorption process for a particular sample type w h i c h was examined i n detai l . In C H A P T E R 5, the results o f experiments concerning b io log i ca l samples are presented. Spec i f i ca l ly , these experiments were concerned w i t h the examinat ion o f drug molecules direct ly o n b i o l o g i c a l matrices. C H A P T E R S 6 and 7 describe the addi t ion o f a th i rd (tunable) laser to the apparatus. T h i s laser provides a means o f probing ions that have been contained i n the i o n trap. C H A P T E R 6 shows results that demonstrate this method 's ab i l i ty to co l lec t gas phase v i s ib le spectra o f trapped ions by the method o f Resonance Enhanced M u l t i p h o t o n D i s s o c i a t i o n ( R E M D ) . C H A P T E R 7 demonstrates a pract ical appl ica t ion o f the R E M P D technique to discr iminate between isomers o f P o l y c y c l i c A r o m a t i c Hydrocarbons ( P A H s ) . F i n a l l y , C H A P T E R 8, the "conc lus ions" chapter w i l l summarize the key results f rom this course o f w o r k and suggest methods to improve this instrument i n the future. 1.2 Mass Spectrometry A l l mass spectrometry ( M S ) devices have three distinct features i n c o m m o n . Sample atoms or molecules are first converted to gas phase ions - the ion iza t ion step. The ions are then separated or analyzed based on their mass to charge ratio (m/z) and then the ions are detected i n a systematic way . The results are almost a lways converted 2 into a plot (the mass spectrum) o f i o n count (intensity) vs. the mass to charge ratio (m/z) . B y careful analysis o f the mass spectrum, one can gain a weal th o f in format ion (quantitative and qualitative) about the molecules or atoms contained i n the o r ig ina l sample. A l l forms o f mass spectrometry can be traced back to the w o r k o f S i r J .J . T h o m s o n o f Cavend i sh Laboratory at the Un ive r s i t y o f Cambr idge [1]. H i s work , carr ied out i n the latter part o f the 19th century, focused o n examin ing electr ical discharges and led to the d i scovery o f the electron i n 1897. D u r i n g the next decade o f research, T h o m s o n rea l ized that a beam o f ions cou ld be caused to change their trajectories i n the presence o f electric o r magnetic fields. B y 1910 T h o m s o n had created the essential components o f a mass spectrometer [2]. Th i s rudimentary device consisted o f a s imple i o n source (a discharge tube), a dispersion element ( two magnet ic poles) , and a detector (photographic plates). W h i l e the f ie ld o f mass spectrometry has g r o w n immense ly since this t ime, the under ly ing pr inciples have remained the same. D u r i n g the last 100 years o f research, the f ie ld o f mass spectrometry has e v o l v e d f rom the development phase, where the instrument was the experiment, to the d ivers i f ica t ion phase where n o w mass spectrometers are used rout inely i n c o m m e r c i a l labs a l l over the w o r l d for sample analysis and w o r k l ike that presented i n this thesis is done to f ind ways o f m a k i n g the instrument even more useful for the case o f specif ic samples. The f ie ld o f mass spectrometry has evo lved to a point n o w where there are currently f ive dis t inct ly different types o f mass spectrometer commerc i a l l y avai lable . These are the: magnetic/electric sector mass spectrometer [3], R F linear quadrupole mass spectrometer [4], 3 D quadrupole i o n trap mass spectrometer [5], Four ie r t ransform Ion 3 C y c l o t r o n Resonance ( F T - I C R ) mass spectrometer [6, 7] , and the T i m e o f F l i g h t ( T O F ) mass spectrometer [8]. A l l totaled, mass spectrometry is a b i l l i o n dol lar business w i t h analyt ica l applicat ions i n such diverse fields as geology, b io logy , chemistry, phys ics , and environmental science. The most recent addi t ion to the fami ly o f commerc i a l l y avai lable mass spectrometers occurred i n 1983 when the F i n n i g a n M A T Corpora t ion o f San Jose found a method o f turning the i o n trap into a useful mass spectrometer [9, 10]. Before this, the i o n trap had been p r imar i l y used by physicists as a means o f containing gas phase ions or as an i o n source for other types o f mass analyzers [11]. Since this t ime, i o n trap mass spectrometers have become one o f the most c o m m o n mass analyzers i n use because they have very h i g h scan speeds, are sensitive, and are inherently capable o f per forming M S N (tandem mass spectrometry) experiments [12]. 1.2.1 Theory of the Quadrupole Ion Trap F o r a comprehensive, detailed account o f the development, theory, and operat ion o f the quadrupole i o n trap the reader is referred to the preeminent text o n the matter: "Quadrupole Storage M a s s Spectrometry" by M a r c h and Hughes [13]. Al te rna t ive ly , more concise versions have appeared elsewhere [5, 14]. The role o f this section is to provide the reader w i t h an understanding o f i o n trap operation sufficient to appreciate its appl ica t ion to the w o r k described i n this thesis. The quadrupole i o n trap had its first pub l ic disclosure i n a patent f i l ed i n 1953 by P a u l and Ste inwedel o f the Un ive r s i t y o f B o n n [15]. A t this t ime, it was l isted as " s t i l l another electrode arrangement" mentioned along w i t h a number o f other geometries that were be ing described to guide and mass select ions. Indeed, it was P a u l ' s group that 4 realized that strongly focusing fields could be utilized to provide mass analysis, and the first detailed account of the operation of the ion trap appeared in the thesis of Berkling in 1956 [16]. The ion trap is, in principle, a simple device consisting o f three electrodes: two parabolic end cap electrodes and a single central ring electrode. A schematic diagram of the ion trap is given in Figure 1.1. The geometry of the electrodes in an idealized mathematical state can be described by the following equations. r .2 2 End Caps: 2 2 R i n g : — 2 _ ± _ = \ Equ. 1.1 r 2 ~2 _ £ _ = _ i E q ° - 1-2 r z o o Where, r 0 is the radius of the ring electrode and z 0 is the closest distance from the center o f the trap to an end cap with r and z denoting the surface of the electrodes. For the quadrupole ion trap used in this thesis, r Q and z a are related by: rl = 2zl Equ. 1.3 Therefore, to define a specific ion trap, one only needs to know the r 0 value; typically, ions traps are built with a r 0 value between l-25mm. The ion trap used in this thesis has an internal radius r 0 = 10.0 mm. The ion trap is a dynamic and complicated device, however, to a first approximation, it can be described simply. Consider a single positive charge placed in the center of the ion trap. If a direct current (DC) potential were applied between the end caps and the ring electrode, so that the ring electrode was negatively charged and the end caps appeared positive, then the ion would find itself confined in the z direction (between Figure 1.1 Pictorial diagram of the ion trap mass spectrometer, (a) A cross section through the longitudinal axis, (b) A three-dimensional representation. end caps). I f the charge made any movement towards one o f the end caps it w o u l d i t se l f experience a restoring force proport ional to the displacement distance. Converse ly , i n the radial d i rect ion, any smal l movement away f rom the center o f the trap w o u l d result i n the charge be ing accelerated towards the r i n g electrode, as the charge w o u l d experience an attractive potential . S i m i l a r l y , i f the r ing were set posi t ive and end caps negative, then the i o n w o u l d be stable i n the radial (r) d i rect ion and unstable i n the ax ia l (z) d i rect ion. Consequent ly , an i o n cannot be confined i n a static electric f ie ld (this is a restatement o f the Ea rnshaw ' s Theorem). A s a result a l l quadrupolar t rapping devices must w o r k on the p r inc ip le o f dynamica l l y changing electric fields. S i m p l y , the potential o n the end caps and r ing must be constantly changing so that the charged particle w i l l never have a chance to impact one part icular electrode. The forces an i o n experiences at one instant w i t h i n the i o n trap can be descr ibed s i m p l y by: F = ma = Ee Equ . 1.4 Where F is the force, m is the mass, a is the acceleration, E is the electric f i e ld and e is the charge on a particle. W h i l e the forces an i o n experiences m a y be s i m p l y stated, the m o t i o n that results f rom the cont inuously changing electric potentials is very complex . 1.2.2 Mathematical Description of the Ion Trap A single i o n i n an ideal quadrupolar field, w i l l experience an electric potential , ()>, at any point (r, z) w i t h i n the field expressed by the equation [4, 13]: 7 «* = - % ( r 2 - 2 z 2 ) + ^ Equ. 1.5 2rn 2 Where : <p0 appl ied potential radial d imens ion z ax ia l d imens ion r0 radius o f the device I f the appl ied potential at any t ime (t), consists o f a direct current (U) component and an alternating current (V) component, then: <p„=U-VcosQt E q u 1 - 6 Where : <p0 appl ied potential U ampli tude o f D C V ampli tude o f A C (measured 0-peak) Q angular frequency o f the A C potential (radians/second) t t ime (seconds) The appl ied A C potential is l isted i n radians/second and is equal to 27$ w h e r e / i s the frequency i n H z . M o s t analyt ica l ly useful devices operate w i t h an A C frequency i n the 1 M H z range (since 1 M H z is i n the radio-frequency range, the A C potential w i l l sometimes be referred to as the R F potential). Note on the applied potential: for this work, the end caps are earthed and the potential applied to the ring electrode only. The RF potential used here is measured peak to ground. 8 W e m a y n o w rewrite Equ.1.4 i n a differential fo rm, recogniz ing that force FN i n the N d i rec t ion is a funct ion o f acceleration, a, (wh ich is equal to the second derivat ive o f pos i t ion w i t h respect to t ime), and o f the electric f ie ld E ( w h i c h is equal to the negative o f the first derivat ive o f the electric potential w i t h respect to posi t ion) . r, d2z dd> Eau 1 7 F7=ma=m—T = -e— c y u . i.# r d2r d(l> Equ. 1.8 F=ma=m—^- = - e — M dt2 dr Substi tut ing the potential ^ f r o m Equ.1.6 into Equ.1.5 y ie lds : (u - v cos at) 2 _ 2 z 2 (u - v cos at) E q u . 1 - 9 2rl 2 Different ia t ing Equ.1.9 w i t h respect to r and z y ie lds the potential gradients: ^ = _ ^ ( f 7 _ F c o s Q 0 Equ. 1.10 dz rn . ^ = 4 ( t / - F c o s Q r ) Equ. 1.11 d r r<>. . • . Subst i tut ing equations Equ.1.10 and Equ. 1.11 into Equ. 1.7 and Equ.1.8 respect ively gives: Fz=ma = m~^^(U-VcosClt) E(*u- 1 - 1 2 dt2 rl Rearranging Equ.1.12 and Equ.1.13 gives the equation o f m o t i o n o f a single charged i o n i n a quadrupolar t ime va ry ing electric f ie ld : 9 d2z 2e (U-VcosQt)z = 0 Equ. 1.14 dt1 r2m + -^-(U-VcosQt)r = 0 Equ. 1.15 In order for an i o n w i t h mass, m, and charge, e, to be trapped i n a device w i t h a radius r0, the values o f U, V, and i 2 m u s t be chosen i n a specific way so that the trajectory o f the i o n does not increase beyond the dimensions o f the trap as t ime increases. The so lu t ion to these second order differential equations is not t r i v i a l ; fortunately, there exists i n the literature solutions to problems o f this form. It turns out that a second order l inear differential equation, k n o w n as the M a t h i e u Equat ion , w h i c h was developed over 120 years ago to describe bound vibrat ing membranes has a fo rm w h i c h is adaptable to the prob lems i n Equ.1.14 and Equ.1.15 [17, 18]. The canonical fo rm o f the M a t h i e u E q u a t i o n can be wri t ten as: Where , u is the coordinate axis. Solut ions to Equ.1.16 are w e l l k n o w n , so by in t roducing a few parameters, Equ.1.14 and Equ.1.15 can be solved. The three parameters that must be defined are: d2u 7 + K - 2 9 „ c o s 2 c > = 0 Equ. 1.16 d%2 Equ. 1.17 a. -%eU mr2Q2 Equ. 1.18 -AeV mr2Q2 Equ. 1.19 10 The solutions to the M a t h i e u equation are determined i n terms o f the parameters au and qu. F o r every pair o f values au and qu there w i l l be a solut ion where the pos i t ion value, u, as a funct ion o f £ is either stable or unstable. Stable solutions are those for w h i c h the pos i t ion , u, o f the i o n is per iodic and finite as t ime increases. U n d e r these condi t ions , we say the i o n is trapped. Converse ly , unstable solutions are those where the i o n pos i t ion increases without l imi t s over t ime. Ions under these condi t ions are not trapped. Solut ions to the Mathieu.equat ion arebest understood w h e n v i e w e d graphica l ly o n a so-ca l led "s tabi l i ty diagram". Figure 1.2 shows a graphical representation o f the stable solutions to the M a t h i e u equations plotted i n terms o f au and qu [13]. Th i s figure is symmetr ic about the au axis , so on ly one h a l f o f the d iagram is shown. A s ment ioned earlier, for an i o n to be effect ively trapped, it must be stable s imultaneously i n both the axia l (z) and radial (r) d imensions . B o t h o f these stabili ty condit ions are plotted s imultaneously i n this figure. T h i s d iagram (Figure 1.2) shows regions w h i c h have stable solut ions o n l y i n (z) or (r). Areas where the stabili ty regions overlap define condit ions where specif ic ions may be effectively trapped i n a l l three dimensions. A n enlarged v i e w o f the most important over lap reg ion (at l o w pos i t ive au and qu values) is shown i n Figure 1.3 [13]. F o r pract ical reasons, v i r tua l ly a l l i o n traps operate i n this stabili ty region. N o t e the iso-/3 lines o n Figure 1.3. These /? values are parameters specif ied by au and qu and define the frequency o f mot ion o f the trapped ions. W h e n qu < 0.4, the fundamental or "secular" frequency (a\) o f a trapped i o n osc i l la t ing w i t h i n the v o l u m e o f a quadrupolar f ie ld can be found f rom: 11 Figure 1.2 A graphical representation of the stable solutions to the Mathieu equation in both the radial (r) and axial (z) directions in terms of au and qu [13]. Figure 1.3 An enlarged view of the experimentally Important region Mathieu stability diagram [13]. - i 1 f — - i 1 — r i 0.2 0.4 0.6 0.8 1.0 1.2 1.4 13 " H u 2 2 Q Equ. 1.20 2 (Note: At qu values greater than 0.4, higher order terms become important and the ion motion becomes more complex. For this simplified analysis, only the case of when qu<0.4 will be considered). F o r a specif ied au and qu value, the i o n trajectory i n the radial plane resembles a i o n be ing contained i n a device without co l l i s ions w i t h neutral molecules or atoms. In practice, these instruments are almost a lways used w i t h hundreds o f thousands o f ions be ing trapped s imultaneously and w i t h background gas pressures o f 1 mTor r . U n d e r these condi t ions , the ions experience C o u l o m b i c repuls ion, w h i c h complicates any analysis . A d d i t i o n a l l y , co l l i s ions o f the neutral background gas w i t h the trapped ions also affect i o n mot ion . W h i l e a more detailed account o f these processes is beyond the scope o f this chapter, numer ica l methods are avai lable to quantitate these effects [5, 19-24]. The net results o f neutral co l l i s ions and C o u l o m b i c repuls ion p lay a very important role i n i o n trap operation and w i l l be discussed later. Nevertheless, the analysis described above is useful and is sufficient to predict most i o n trap properties. Lissa jous figure composed o f two frequency components: Equ. 1.21 Equ. 1.22 The analysis discussed above has been under the l i m i t i n g condi t ions o f a single 14 1.2.3 The 3D Quadrupole Ion Trap as a Mass Spectrometer The theory described i n the previous section shows on ly the condi t ions for t rapping ions effectively but makes no ment ion as to h o w this device m y be used as a pract ical mass spectrometer. Indeed, it is this very p rob lem: h o w to make a useful mass analyzer, w h i c h vexed researchers for almost 30 years. It was not un t i l 1983 w h e n George C . Stafford and his team at F inn igan M A T developed the "mass selective ins tab i l i ty" mode o f operation that the modern age o f i o n trap mass spectrometry real ly began [9, 10]. There have been a number o f addit ional methods suggested for operating the i o n trap [25], however , the majori ty o f commerc i a l and research devices ( i nc lud ing this one) s t i l l operate under this " instabi l i ty mode". The method o f mass selective instabil i ty can be clear ly understood by e x a m i n i n g the mod i f i ed stabil i ty d iagram i n F i g u r e 1.4. Th i s diagram is a plot o f the "pos i t ions" o f several ions o f different mass o n the az, qz graph for a specif ied R F voltage. In practice, for most i o n trap operation, there is no D C potential appl ied to the trapping field, so the az value is reduced to zero (this w i l l a lways be the case for the w o r k described i n this thesis). Thus the region o f stability w h i c h is most important is that w h i c h l ies o n the q -axis o f the stabil i ty d iagram between qz = 0 and qz = 0.908. These points mark the intersection o f the qz axis w i t h Bz =0 and fiz = 1 respectively. A s ment ioned earlier, an i o n w i l l be stable i f its az and qz values are such that they l ie w i t h i n the stabil i ty d iagram. Converse ly , i f the az and qz values are such that the pos i t ion o f the i o n lies outside the stabil i ty d iagram, then the i o n w i l l not be contained. Therefore, because o f the nature o f the az and qz equations, for a part icular R F voltage and ion trap radius there is a lower l imi t to the m / z ratio that m a y be trapped. 15 Figure 1.4 An enlarged view of the stability diagram showing the possible positions of several ions of different mass on the stability diagram under various RF voltage conditions. 1500 T h q = 0.908 1000 T h 500 T h <— - stable unstable • w w w fe i — f — _ w 3000 V -> S00OW 200 V •1 i i i i i i i i i I I I I 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 16 T h i s lower mass l i m i t represents the cut o f f point where ions have qz values greater than the qz s tabil i ty point. A l l ions o f mass higher than this w o u l d have 0< qz < 0.908 o n the stabil i ty d iagram and w o u l d therefore be contained. B y systematical ly increasing the R P "trapping vol tage", ions o f greater and greater mass become sequential ly unstable i n the z d i rec t ion as their qz values approach qz = 0.908. W h e n used as mass spectrometers, i o n traps are manufactured w i t h a number o f smal l holes d r i l l ed i n one o f the end caps. T h i s a l lows the ions that have gained unstable trajectories i n the z d i rec t ion to escape the i o n trap and strike a detector ( typ ica l ly an electron mul t ip l ie r ) . B y increasing the R F voltage and moni tor ing the detector output, one can, i n a straightforward way , obtain an entire mass spectrum as ions are ejected starting at the lowest mass to the highest mass. In the previous paragraph, a l o w mass l i m i t was described. There are also some fundamental l imi ta t ions to the heaviest i o n that m a y be effect ively trapped and detected by an i o n trap mass spectrometer. The p rob lem o f trapping heavy ions can be understood us ing the Dehmel t potential w e l l mode l . Th i s mode l suggests that the process o f an i o n be ing trapped is analogous to the process that occurs when a ba l l falls d o w n into a va l l ey . T h i s pseudo potential w e l l can be described as hav ing a depth: D 1 D Equ. 1.23 (Note: Again, this equation applies for qu <0.4) The pseudo potential w e l l can be thought o f as descr ibing the m a x i m u m kinet ic energy an i o n may have and s t i l l be contained i n the trap (i.e. an object w i l l be trapped i f it has less k ine t ic energy than the height o f the potential w e l l containing it). Therefore, i n an instrument w i t h fixed R F frequency, il, and size, z0, at a g iven m a x i m u m R F voltage, 17 V , as the mass o f an object increases, the size o f the potential w e l l conta in ing that i o n decreases. Therefore the ultimate h igh mass-trapping l imi t w i l l occur w h e n the thermal k ine t ic energy o f an i o n is greater than the pseudo potential w e l l depth. O f course, i f the instrument were designed specif ical ly for trapping very heavy ions, then one w o u l d s i m p l y choose a lower trapping frequency. Indeed, macroscopic objects, such as a l u m i n u m particles have been trapped w i t h devices operating at 148 H z [26]. T h i s example brings to the foreground one o f the classic problems o f the i o n trap: i f one were determined to trap very heavy ions (even on a molecular scale - i.e. proteins) then the trapping voltage appl ied w o u l d have to be large enough so that the potential w e l l depth was capable o f conta ining the ion . Unfortunately, at these large trapping voltages, the potential is raised so h igh that many l o w mass ions have qz values higher than qz = 0.908 and are not trapped. T h i s is one o f the fundamental downfal ls o f the i o n trap; the mass range o f ions that m a y be trapped s imultaneously w i l l be l imi t ed . W h i l e h igh mass trapping is a p rob lem w i t h i o n traps, it turns out that i o n ejection and detection is more often the l i m i t i n g factor i n observing h igh mass ions. T h i s occurs because there exists a fundamental l i m i t for the R F voltage w h i c h may by appl ied between t w o metal objects under m e d i u m vacuum condit ions before an electr ical discharge results. In an i o n trap w i t h the dimensions and pressures typ ica l ly used this value is about 7000 Vo-p eak- F o r instruments using the mass selective instabi l i ty mode o f operation, this impl ies that the heaviest i o n w h i c h may be ejected (i.e. hav ing its qz va lue raised to 0.908) is typ ica l ly about 650 T h . Note: this thesis will use the new convention for expressing the mass to charge ratio, m/z, as a Thomson, Th. 18 T h i s l i m i t was or ig ina l ly not a concern to instrument manufactures as i o n traps were in i t i a l ly created as detectors for gas chromatographs ( G C ) , w h i c h can typ i ca l ly on ly elute molecules w i t h masses be low this l imi t . Howeve r , w i t h the recent development and popular iza t ion o f methods for creating h i g h mass ions, electrospray ion iza t ion (ESI ) and matr ix assisted laser desorption ioniza t ion ( M A L D I ) for example, a need has arisen to adapt the i o n trap for the analysis o f h igh mass ions. The most popular method that is commerc ia l ly used today (and was used i n this thesis) to enhance the mass range o f the ion trap is k n o w n as resonant ejection [27]. T h i s method invo lves app ly ing a supplemental alternating potential to one or both o f the end caps. The frequency o f this secondary potential is chosen to match the frequency an i o n w o u l d have as it oscil lates i n the i o n trap w i t h a qz lower than 0.908. T h i s sma l l A C voltage ( typ ica l ly a few volts) is enough to excite and eject an i o n because i f the appl ied frequency is i n resonance w i t h the secular frequency, energy w i l l be deposited into the i o n m o t i o n and its orbit w i l l become larger and larger un t i l the i o n is eventual ly ejected. I f this supplementary A C voltage is appl ied as the R F voltage is ramped, and i f the frequency is chosen to be i n resonance w i t h ions o f a lower qz value, say qz = 0.303, then as the R F voltage is increased, ions w i l l come into resonance w i t h this appl ied voltage, as their qz values approach qz = 0.303, and w i l l be ejected at this point rather than at qz = 0.908. S ince the ions are be ing ejected at a q roughly a th i rd lower , we w o u l d observe a mass range increase o f approximately 3 times. T h i s method combined w i t h reducing the R F dr ive frequency has been used to extend the mass range substantially, f rom 650 T h to n o w over 70,000 T h [28, 29]. 19 The abi l i ty to add an auxi l ia ry frequency or wave fo rm between the end caps turns out to have many addi t ional uses besides just mass range extension. One o f the most important o f these is the abi l i ty to select ively excite specific ions i n the trap. T h i s ab i l i ty is leveraged to m a x i m u m effect dur ing the operation o f performing tandem mass spectrometry [30]. T a n d e m mass spectrometry ( M S / M S ) is the process whereby a specif ic i o n is fragmented and the resulting "daughter" ions detected. S ince no rma l mass spectrometry typ i ca l ly o n l y provides information on the total mass o f a molecu le and li t t le about its structure, M S / M S can provide the analyst w i t h a weal th o f in format ion about the sample. The most efficient means o f performing tandem mass spectrometry i n the i o n trap is by app ly ing a N o t c h e d B r o a d B a n d W a v e f o r m ( N B B W ) between the end caps. The method operates o n the pr inc ip le that at one particular trapping voltage every mass w i l l have a specif ic qz value, and that every qz value w i l l produce ions osc i l la t ing at unique secular frequencies. S ince a l l ions o f one part icular mass oscil late at one frequency, i f that frequency is appl ied between the end caps, then that specific i o n w o u l d be ejected f rom the device . T o perform M S / M S , we want the reverse o f this; we want to eject a l l other species f rom the trap and leave on ly the "parent" i o n o f interest. T o achieve this, a c o m p l e x wave fo rm is appl ied to the end caps that contain a l l frequencies except that o f the parent i on . T h i s effectively cleans the i o n trap o f a l l material but the i o n o f interest. Prac t ica l ly , this complex wavefo rm can be obtained by either taking the inverse Four ie r transform o f a notched frequency range (as i n the Stored W a v e f o r m Inverse Four i e r Transform ( S W I F T ) method [31, 32]), or more typ ica l ly , by s imp ly adding a number o f 20 discrete s ingle frequency components i n the t ime domain . The method o f t ime doma in wave fo rm summat ion was used to create the N B B W waveform i n this w o r k . Once this parent i o n is mass selected, an addi t ional s ingle frequency A C voltage is appl ied. T h i s t ime, however , the voltage used is m u c h lower than before, and instead o f the i o n be ing ejected f rom the trap, the resonant energy appl ied is o n l y enough to cause c o l l i s i o n induced dissocia t ion ( C I D ) [33]. Th i s dissociat ion is caused by the molecu la r ions c o l l i d i n g w i t h background gas neutral atoms or molecules and produces diagnost ic daughter ions. It is this abi l i ty to routinely and s imply perform M S / M S that t ruly sets the i o n trap apart f rom other mass spectrometers. W h i l e other instruments, namely l inear tr iple quadrupolar mass spectrometers (triple quads) and F T - I C R devices are also capable o f per forming M S / M S , the i o n trap, by its nature, is superior for this task. In fact, the N B B W / C I D methods described previous ly cou ld just as easi ly be appl ied to one o f the daughter ions result ing f rom a previous M S / M S cyc le . T h i s mul t ip le tandem mass spectrometry ab i l i ty ( M S N ) has been prev ious ly demonstrated up to M S 1 1 [34]. One final considerat ion i n the practical operation o f an i o n trap is that o f background "buffer" gas. It has been found empi r i ca l ly that the presence o f h e l i u m at a pressure o f ~1 m T o r r substantially improves the performance o f the i o n trap [35]. T h i s improvement o f performance results when the relat ively l ight H e atoms co l l ide w i t h the heavier molecu la r ions. These co l l i s ions have the effect o f transferring a po r t ion o f the k inet ic energy o f the or ig ina l i on to the H e atoms. The effects o f this " c o l l i s i o n a l c o o l i n g " on i o n trap performance are three fo ld . Fundamenta l ly , the existence o f this background gas is essential to the t rapping o f externally injected ions. Str ic t ly speaking, an externally created i o n c o u l d never be 21 trapped wi thout co l l i s ions w i t h buffer gas because as the i o n enters the trap, it gains a certain amount o f k ine t ic energy. I f this energy were perfectly conserved, then the i o n w o u l d s i m p l y continue t ravel ing right through the trap i n a manner analogous to a b a l l r o l l i n g into and out o f a symmetr ic va l ley w i t h no fr ict ional forces. The H e buffer gas, through co l l i s ions , can remove enough o f this energy to a l l o w the i o n to be trapped. O n c e trapped, co l l i s i ona l c o o l i n g continues to remove energy from the i o n causing the ampli tude o f the i o n mot ion to be d imin ished . Essent ia l ly , this m i n i m i z e s the i o n " c l o u d v o l u m e " to a smaller area at the center o f the trap. Th i s bunching o f ions produces increased sensi t ivi ty because the ions have a smaller radial dis t r ibut ion and thus have a higher l i k e l i h o o d o f passing through the smal l apertures at the center o f the end cap electrode w h e n ejected. F i n a l l y , co l l i s i ona l c o o l i n g also improves resolution. Ideal ly, ions are ejected f rom the i o n trap when they come into resonance w i t h the ejection point (either q z = 0.908 for normal operation or some lower q z value for resonance ejection). H o w e v e r , wi thout co l l i s i ona l coo l i ng , as ions approach this l imi t , they w i l l start to gain energy, as their secular frequency approaches the exci tat ion frequency. T h i s can cause the i o n c l o u d to expand and, wi thout co l l i s i ona l coo l ing (wh ich helps m i n i m i z e this energy), some percentage o f ions m a y be ejected earlier than the strict resonance point . T h i s produces a broader peak shape and thus lowers resolution. 1.3 Sampling and Ionization for Mass Spectrometry A s ment ioned previous ly , the i o n trap was o r ig ina l ly designed as a gas-sampl ing device . W h e n used i n this way , analyte molecules were s imp ly leaked into the i o n trap and ion ized by an electron gun (electron impact ion iza t ion - E I ) situated beh ind one o f 22, ... the end caps (through a smal l hole i n the end cap). Th i s configurat ion is ideal for the i o n trap because the ions are created i n the center o f the trap f rom l o w ve loc i ty gas species. T h i s s i tuat ion a l lows for the highest probabi l i ty o f trapping; and very good sensi t ivi ty results. There are, however , a wide range o f molecules and samples for w h i c h in t roduct ion into the i o n trap i n this way is not applicable. Fo r example , h i g h molecula r weight ( M W ) , non-volat i le , polar and/or thermally labi le molecules may undergo decompos i t ion dur ing the vapor iza t ion step. A s a result, direct sampl ing o f gas phase molecules by E I may not a lways be pract ical or even possible. A s a consequence o f this l imi ta t ion , a wide range o f vo la t i l i za t ion and ion iza t ion schemes have been developed or adapted to extend the capabil i t ies o f the i o n trap. Indeed, over the last 20 years, the ion trap has been coupled to v i r tua l ly every vo la t i l i za t ion and ion iza t ion source available. F o r l i q u i d samples this includes electrospray ion iza t ion (ESI) [36 , 37], chemica l ion iza t ion ( G l ) [38], and photo ion iza t ion (PI)[39]. " • v . -F o r s o l i d samples, a number o f protocols have been attempted inc lud ing ; secondary i o n mass spectrometry ( S I M S ) [40], laser desorption/laser ablat ion [41], fast a tom bombardment ( F A B ) [42], and matrix-assisted laser desorption ion iza t ion ( M A L D I ) [43, 44] . A l l o f these direct so l id sampl ing techniques, however , either produce c o m p l e x mass spectra result ing f rom extensive molecular fragmentation or require extensive sample preparation. One alternative for direct so l id sampl ing that has yet to be thoroughly explored i n an i o n trap is the method o f laser desorption/resonant two-photon ion iza t ion ( T w o laser so l id sampl ing - L 2 M S ) . Previous to this thesis there has been one demonstrat ion o f the two-laser method i n an i o n trap, however , this differed f rom the 23 w o r k presented here i n both a i m and scope - it was designed and bu i l t by surface scientists, w i t h the goal o f detecting adsorbates o n chemica l ly mod i f i ed metal surfaces [45 ,46 ] . • 1.3.1 Two-Laser Solid Sampling (L2MS) The two-laser method for so l id sampl ing has been k n o w n for some t ime. It was first developed for mass spectrometry o f invola t i le and/or thermally labi le organics i n the mid -1980 ' s [47-49]. Since that t ime it has been further developed and used for a var iety o f s o l i d samples inc lud ing : po lymers [50], soot and coal tar [51], dyes [52], b i o l o g i c a l molecules [53], and molecular adsorbates [46]. Two- laser mass spectrometry, as it has been p rev ious ly appl ied, is a three-step process ( F i g u r e 1.5). The first step invo lves us ing a pulsed I R laser ( typica l ly a CO2 laser) for thermal desorption o f the sample. I f the wavelength and power density are chosen correctly ( typica l ly around 1 0 6 W / c m 2 ) . thermal desorption w i l l produce a p lume o f intact neutrals w i t h very little ion iza t ion or sample decomposi t ion . The second step i n this process, invo lves the use o f a pulsed U V laser ( typ ica l ly a quadrupled N d : Y A G laser), to photoionize the analyte present i n the neutral desorbate p lume. A g a i n , i f the wavelength and power are chosen correct ly, the ion iza t ion process can be extremely gentle and selective and produce m a i n l y intact molecula r ions. F i n a l l y , i n v i r tua l ly every reported example o f the use o f this technique, the ions are detected us ing a t ime o f fl ight mass spectrometer. It is i n the temporal and spatial separation o f the desorption and ion iza t ion processes where the advantages o f the technique are gained. The separation a l lows an independent op t imiza t ion o f both the laser desorption and laser photo ion iza t ion steps, thus m a k i n g the process both extremely gentle and sensitive. The next two sections w i l l 24 Figure 1.5 Schematic cartoon of two-laser mass spectrometry as it has been previously applied. T O F M S © Pulsed IR © o o o t o @ w Pulsed U V •8^880 *8 880 25 describe these two processes independently and a th i rd w i l l describe the advantages o f per forming two-laser mass spectrometry i n an i o n trap. 1.3.2 Laser - Sol id Interactions The first step o f the two-laser method involves a short I R laser pulse interacting w i t h a s o l i d sample. T h i s process k n o w n as laser desorption has been examined for some t ime. Indeed, the observation o f the interaction o f laser l ight w i t h so l id samples was one o f the first experiments to be performed w i t h the newly invented laser. T h i s interest is evident b y the fact that on ly a ten year span elapsed between the t ime the ruby laser was d iscovered and the def ining text o n laser so l id interactions was publ i shed by Ready ( in 1971) [54]. W h i l e the study o f laser so l id interactions has cont inued over the laser 30 years there s t i l l remains a great deal w h i c h is not thoroughly understood. T h i s is due to the fact that when a h igh-powered short durat ion laser pulse interacts w i t h a so l id sample a number o f possible scenarios may result. These results depend on a mult i tude o f interconnected factors inc lud ing : the thermal conduct iv i ty o f the so l id , the heat capacity o f the so l id , the absorption coefficient o f the so l id at the laser wavelength, the laser pulse duration, the photon density, and wavelength o f the laser pulse [54]. A l l o f these factors can combine i n specif ic and compl ica ted ways to produce a variety o f consequences ranging f rom explos ive events and p lasma format ion a l l the way d o w n to gentle heating o f the surface. T o s impl i fy the analysis i n this section, we may d iv ide the possible outcomes o f laser s o l i d interactions into two classes. The first, ca l led laser ablat ion, results f rom the absorption o f a large amount o f energy by the surface. Th i s results, t yp i ca l ly , i n a very 26 energetic event. In contrast, when a re la t ively l o w flux o f energy is used, a less energetic process results - this is termed laser desorption. Lase r ablat ion results w h e n a relat ively h igh photon flux (10 W / c m ) is absorbed by a s o l i d surface i n a short t ime per iod [55]. T h i s energy in i t i a l ly induces a very r ap id rise i n temperature at the interface. Th i s intense, loca l i zed heat, often leads to me l t ing , vapor iza t ion , p la sma formation, and poss ibly absorption o f laser energy by the p lasma. T h i s high-energy process typ ica l ly produces a range o f very energetic particles that are ejected f r o m the surface. These particles may include everything f rom a tomic species, to molecula r fragments; i n neutral, radical , or ion ic form. Unde r these condi t ions , the su rv iva l o f any po lya tomic molecule into the gas phase is h igh ly un l i ke ly . A s a result, laser ablat ion has on ly found c o m m o n use for elemental analysis o f so l id samples. F r o m an instrumental point o f v i e w , laser ablat ion is a poor technique to couple to a mass spectrometer. T h i s is due to the fact that laser ablat ion produces particles w i t h a broad dis t r ibut ion o f k inet ic energies centered at very h igh values (1000 k e V kinet ic energies are not unusual) [56]. Since most mass spectrometers rely on either conta ining ions or d i rect ing them w i t h electric f ields, ablat ion products are often di f f icul t to analyze. F o r example , ions w i t h very h igh energy are v i r tua l ly imposs ib le to trap i n an i o n trap or F T - I C R and samples w i t h very broad spreads o f energies wreak havoc o n T O F analyzers. O n the other end o f the energy scale is laser desorption. Laser desorpt ion results w h e n the energy flux o f the laser is m u c h lower than the m i n i m u m for ablat ion/plasma format ion. Instead, what results can usual ly be described s imp ly as rapid heating. T h i s heating process can occur very q u i c k l y , w i t h rates o f 1 0 1 1 K / s typ ica l . U n d e r these ultra-fast heating condi t ions thermally labi le molecules can actually be desorbed before they 27 get a chance to decompose [57, 58]. Th i s counter intui t ive result occurs because under these condi t ions the temperature is q u i c k l y raised h igh enough so that decompos i t ion and desorpt ion are both energetically possible s imultaneously (/. e. they both have more energy than is needed to cross the act ivat ion barrier for the respective process). A t this point , frequency factors dominate the relative rates, and often desorption is favored over decompos i t ion . The laser desorption process produces predominant ly intact molecula r ions w h i c h are essentially "bo i l ed o f f the surface w i t h relat ively litt le k inet ic energy (< 10 e V ) . O b v i o u s l y , i f analysis o f molecula r adsorbates o n surfaces were the goa l , then finding the correct ly laser condit ions for the laser desorption regime to prosper rather than laser ablat ion w o u l d be cr i t ica l . A further analysis o f the laser desorption process as it spec i f ica l ly applies to the experimental w o r k done i n this thesis is contained i n C H A P T E R 4. Laser desorption, under the def ini t ion described above w i l l produce m a i n l y intact molecu la r species w i t h l o w kinet ic energies and w i t h very li t t le decompos i t ion or ion iza t ion . T h i s non-selective desorption process is ideal ly suited for combina t ion w i t h selective laser photoionizat ion. The photoionizat ion process w i l l be discussed i n S e c t i o n 1.3.3. A t this point , however , it w o u l d be instructive to compare "straight laser desorpt ion" as described i n this section w i t h a technique that has gained tremendous popular i ty o f late - M A L D I . M a t r i x Ass i s t ed Laser Desorpt ion Ionizat ion ( M A L D I ) was d iscovered i n the late 1980's by Tanaka and coworkers at Sh imadzu [59-61] and developed concurrent ly by H i l l e n k a m p and Karas i n Germany [62]. (Tanaka received the N o b e l P r i ze for this w o r k 28 i n 2002) . T h i s technique is based upon m i x i n g a smal l amount o f analyte w i t h a large excess o f so l id or l i q u i d matr ix that absorbs strongly at the laser wavelength. T h i s method produces intact molecula r ions direct ly f rom the desorption event. E x t r e m e l y h i g h molecu la r weight ions have been observed w i t h this method ( M W > 100 000 D a ) [63]. A m a z i n g l y , however , the mechanism o f ion iza t ion is s t i l l not w e l l understood. Researchers are s t i l l debating whether ions are "pre-formed" i n the matr ix or somewhere i n the desorbate p lume . Regardless, this abi l i ty to examine very h i g h M W species has greatly accelerated a number o f other fields o f research. N o w h e r e is this clearer than i n the w o r l d o f b iochemis t ry because n o w large proteins and D N A molecules may be examined d i rec t ly by mass spectrometry [63]. M A L D I , however , is not the ideal tool for a l l forms o f so l id analysis . M A L D I on ly works o n specific classes o f molecules , and d iscover ing the correct mat r ix for efficient i on iza t ion is bas ica l ly a matter o f t r ia l and error. A d d i t i o n a l l y , M A L D I is generally restricted to on ly l o o k i n g at relat ively h igh mass ions because the "ma t r ix" mater ial i t se l f is often easi ly ion ized and produces intense l o w mass peaks w h i c h are anathema to l o w mass analysis. F i n a l l y , M A L D I falls short f rom the perspective o f per forming direct analysis o f native samples. In order for M A L D I to be effective the sample must be co-deposited w i t h the analyte. O b v i o u s l y , i f performing direct analysis o f untreated samples o f moderate M W were, your goal then laser desorpt ion rather than M A L D I w o u l d be preferred. 1.3.3 Laser Photoionization Af te r molecu la r desorption, the next step i n two laser mass spectrometry i s photo ioniza t ion . Photo ioniza t ion , and mul t iphoton ion iza t ion ( M P I ) i n part icular, is a 29 versatile ion iza t ion technique w i t h unique properties when compared to electron gun ion iza t ion for mass spectrometry. Since M P I depends in t r ins ica l ly on the electronic properties o f the molecu le , it a l lows for an addi t ional l eve l o f select ivi ty i n the analysis . A d d i t i o n a l l y , M P I is versatile i n that it is capable o f both soft (no fragmentation) and hard (extensive fragmentation) ion iza t ion depending o n the laser power used. The ion iza t ion potential o f most Organic molecules is o n the order o f 7-13 e V . Therefore, s ingle photon ion iza t ion w o u l d require a laser beam i n the v a c u u m U V energy region. In contrast, because many organic molecules have electronic transitions i n the U V / V i s i b l e reg ion o f the spectrum, the poss ib i l i ty exists o f performing mul t ipho ton ion iza t ion . A s impl i f i ed Jablonski diagram depict ing the various possible modes o f ion iza t ion is p rov ided i n Figure 1.6. M u l t i p h o t o n ion iza t ion is a technique that depends o n a molecu le absorbing two or more photons f rom one or more intense v i s i b l e / U V lasers. T h e most efficient f o r m o f this ion iza t ion occurs w h e n one o f the lasers is chosen to have a frequency (energy) that matches a real electronic transit ion i n the molecule . W h e n this occurs, the molecu le can potent ia l ly remain i n this exci ted state l ong enough to absorb a second photon and become ion ized . T h i s method is ca l led resonance enhanced mul t iphoton ion iza t ion ( R E M P I ) - Figure 1.6.c [64]. O f course, there are some specific requirements i n order for this process to be effective. A molecule w i l l on ly become ion ized w h e n the s u m o f the absorbing photon energies is greater than the ioniza t ion potential . A d d i t i o n a l l y , i f the laser wavelength does not match a real electronic transition, then the p robab i l i ty o f ion iza t ion decreases dramat ical ly- Figure 1.6.b. Therefore, by s imp ly choos ing the laser wavelength prudently, certain species can be select ively ion ized over others. T h i s t ruly 30 Figure 1.6 Simplified Jablonski diagram for a number of possible photon/molecule interactions, (a) Single photon ionization (b) No ionization (c) Resonant two photon ionization (d) Non Resonant multiphoton ionization (e) Two photon resonant ionization. (a) Single Photon Ionization Ionization Continuum (b) No Ionization UJIIglJIIIl mm wmrn (c) Resonant two (e)Non (f) Two photon photon Ionization Resonant MPI Resonant ionization 31 unique abi l i ty o f R E M P I to provide very selective ion iza t ion can be used to preselect w h i c h ions are to be analyzed by the mass spectrometer. Bes ides straight R E M P I , there are other means o f achieving mul t ipho ton ion iza t ion . These inc lude , two photon resonant ion iza t ion ( F i g u r e 1.6.e) and non-resonant mul t ipho ton ion iza t ion ( F i g u r e 1.6.d). B o t h o f these scenarios, however , re ly on the creat ion o f a very short- l ived "v i r t ua l " exci ted state. In order for ion iza t ion to be achieved a second photon must interact w i t h this v i r tual state before it relaxes ( l ifet imes t y p i c a l l y <10" 1 5 seconds). Th i s makes the eff ic iency o f the process far less favorable than R E M P I . Regardless, total ly non-resonant mul t iphoton ion iza t ion is possible w i t h very h i g h laser powers . T h i s is useful i n situations where a broad analysis is required. In addi t ion to being a very selective i o n source for mass spectrometry, R E M P I offers other advantages. One advantage is that control o f molecula r fragmentation is relat ively straightforward. R E M P I is inherently a very gentle technique because almost no addi t ional energy is deposited into the molecule during ioniza t ion . In fact, over a w ide range o f molecula r species R E M P I has been shown to provide very soft i on iza t ion at l o w laser powers (<10 W / c m ) [65]. A s a result, the observed mass spectra are often greatly s impl i f i ed because they contain peaks corresponding on ly to molecular ions. In contrast, the more energetic E I process u t i l i z ing 70 e V electrons produces molecula r ions and often many fragment ions as w e l l . A s a result, E I mass spectra o f mixtures o f more than a few compounds are often imposs ib le to decipher because o f the mul t i tude o f peaks. A d d i t i o n a l l y , w i t h an electron gun, very li t t le latitude is avai lable i n terms o f selecting the electron's energy because the ion iza t ion cross section drops dramat ica l ly at lower electron energies. W i t h M P I on the other hand, by s imp ly adjusting the laser 32 power the analyst has the opt ion o f observing just the molecular parent or a number o f fragment ions. The R E M P I process is also a very efficient means o f producing ions. Ion iza t ion efficiencies o f molecules i n the laser beam path are very h igh . It has been estimated that 10-100 % o f naphthalene molecules i n the laser path can become ion ized [66]. B y compar i son E I sources typ ica l ly ionize less than 1/10,000 o f those molecules that enter the ion iza t ion region[65]. The ultimate l i m i t to i o n product ion i n R E M P I is a matter o f fundamental constants. These include the absorption cross section o f the molecu le at the laser wavelength and the relative rate o f radiationless decay. A number o f mode ls and theories that describe the R E M P I process and the effects o f substituent atoms o n efficiencies are avai lable i n the literature [67, 68]. 1.4 Coupling Two-Laser Solid Sampling with an Ion Trap A s ment ioned previous ly , the i o n trap is a device capable o f a w i d e range o f experimental and operational parameters. In addi t ion, two-laser so l id sampl ing is a very effective means o f performing direct so l id sampl ing o f m e d i u m M W molecules . T h i s sect ion w i l l n o w describe the potential advantages to be gained by c o m b i n i n g the t w o techniques. Two- lase r so l id sampl ing ( L 2 M S ) h a s almost exc lus ive ly been performed w i t h a t ime-of-f l ight ( T O F ) apparatus [64, 69, 70]. Prev ious workers have shown that L 2 M S w i t h a T O F is capable o f very l o w detection l imi t s ; for example, attomolar detection levels have been publ i shed for the analysis o f ani l ine absorbed on s i l i c o n [71]. Furthermore, the opt ical selectivi ty afforded by the ion iza t ion stage a l lows preferential ion iza t ion o f analyte molecules over matr ix species that do not have an electronic 33 t ransi t ion at the laser wavelength. T h i s selectivity has a l l owed L 2 M S to be appl ied to a w i d e variety o f analytes and samples ranging f rom b io log i ca l l y important molecules [69, 70] to environmental contaminants on so i l and soot [51, 72]. These workers have also shown that L 2 M S is capable o f direct analysis o f samples that have not been pretreated. S ince , a l l o f the advantages associated w i t h this technique arise p r imar i l y f rom the sampl ing method rather than the mass spectrometer, one w o u l d expect that this method w o u l d be advantageous regardless o f the type o f mass spectrometer ( T O F vs. i o n trap). . Therefore, the quest ion o f w h y perform this experiment i n an i o n trap rather than a T O F real ly becomes one o f what are the pros and cons o f i o n traps vs. T O F ' s T ime-of - f l igh t mass spectrometers and i o n traps are both amenable to operat ion w i t h transient ion iza t ion sources. That is to say, for each instrument, a complete mass spectrum can be obtained f rom a single laser cyc le . T h i s abi l i ty is very important for laser sampl ing , and argues against, for pract ical reasons, per forming L 2 M S w i t h a scanning device l ike a l inear quadrupole since thousands o f laser shots w o u l d be required to obtain one mass spectrum. T i m e o f fl ight devices have greater v a c u u m system requirements and demand faster electronics than i o n traps. O n the other hand, iOn traps require h i g h voltage R F power supplies. H o w e v e r , the net cost o f a h igh performance T O F is t yp i ca l ly at least twice as expensive as an i o n trap. The m a i n advantage o f the T O F instrument over the i o n trap is its ab i l i ty to acquire a huge mass range w i t h very h igh mass accuracy (precise mass values). One o f these advantages, however , is negated w i t h this sampl ing system, by the fact that L 2 M S 34 is o n l y useful i n sampl ing species up to mass 9000 a m u [57, 73]. Th i s mass range is w i t h i n reach o f the i o n trap. The i o n trap, on the other hand, has many, advantages over the T O F w i t h respect to L 2 M S o f real samples. The most important o f these is the i o n traps abi l i ty to per form M S " , w h i c h is extremely useful i n elucidat ing the components o f mixtures. T h i s advantage is accented for L 2 M S sampl ing by the nature o f the i o n source. Because L 2 M S is such a soft ion iza t ion technique and produces m a i n l y molecula r ions , the ab i l i ty to per form M S / M S o n these ions to conf i rm their identity i n complex mixtures is very valuable . T h i s is especial ly true because the samples are often analyzed " w h o l e " , w i t h no previous chromatography or separation steps. Therefore, samples w i l l a lmost a lways conta in complex mixtures. The second key advantage o f the i o n trap over the T O F w i t h two-laser sampl ing is the fact that the ions do not have to be detected immedia te ly after the laser event. T O F ' s do not have any i o n storage abi l i ty , so whatever ions are formed f rom the two-laser event must be detected immedia te ly . In contrast, the i o n trap is capable o f select ively storing and manipula t ing the produced ions. F o r example, it w o u l d be possible to col lec t the ions f rom several laser shots to increase the total i o n count, and thus sensi t ivi ty. Furthermore, by careful addi t ion o f a N B B W between laser cycles , l o w concentration species can be select ively preconcentrated i n the gas phase before detection. T h i s o f course w o u l d be imposs ib le w i t h a t ime-of-f l ight M S . F i n a l l y , it should be mentioned, that two-laser sampl ing even addresses and m i n i m i z e s the two potential downfal ls that are inherent to the i o n trap. The first o f these concerns typ i ca l ly associated w i t h i o n traps is the p rob lem o f space charge. I f too many 35 ions are contained s imultaneously w i t h i n the trap then C o u l o m b i c repuls ion between ions w i l l cause a deleterious effect. Th i s "space charge" p rob lem causes broadening and shift ing o f mass peaks i n the observed spectra. Two- lase r sampl ing addresses this p rob lem, because the ion iza t ion process is very selective. A s a result, one should , i n theory, be able to ionize on ly molecules o f interest wh i l e leaving unwanted matr ix mater ial i n the neutral state, thus m i n i m i z i n g the total number o f ions i n the trap and as a consequence, the space charge. The other concern typ ica l ly associated w i t h i o n traps is that o f inject ion eff ic iency. I f an externally created i o n is to be trapped, it must first penetrate a h i g h voltage alternating electric field. T h i s field w o u l d alternately either reject the i o n or accelerate it through the trap depending o n the relative phase angle. A s a result, there is on ly a very sma l l acceptance w i n d o w w i t h the correct phase angle for t rapping. Consequent ly , it has been estimated that less than 5 % o f external ly created ions are eff iciently trapped [74, 75]. W h i l e some researchers have used "sudden onset" or fast swi tch ing power supplies to increase this eff iciency [76], these specia l ized devices are the except ion. B y compar ison , w i t h two-laser sampl ing , it is possible to pos i t ion the U V laser so that ion iza t ion occurs p r imar i ly i n the center o f the trap. T h i s should provide for the m a x i m u m possible trapping eff iciency. 1.4.1 This Work W i t h the above just if icat ions, motivat ions, and theory i n place, w e may n o w reexamine the goals o f this work . Th i s thesis w i l l describe the creation o f a nove l two-laser so l id sampl ing i o n trap mass spectrometer. T h i s instrument should , i n theory, be 36 capable o f a l l the experiments prev ious ly done on a L 2 M S - T O F device w i t h the added ab i l i ty o f be ing able to perform M S N and other post-laser sampl ing i o n manipula t ions . T o demonstrate the effectiveness o f this device, three m a i n experimental paths were fo l l owed . The first o f these i n v o l v e d the examinat ion o f envi ronmenta l ly important contaminants; p o l y c y c l i c aromatic hydrocarbons ( P A H s ) and polychlor ina ted b iphenyls ( P C B s ) o n s o l i d samples. The second course o f research, concerned invest igat ing the poss ib i l i ty o f examin ing b io log i ca l l y important compounds; this w o r k demonstrated the abi l i ty to detect a drug molecu le di rect ly o n b io log ica l tissues w i t h no sample pretreatment. F i n a l l y , the th i rd course o f research i n v o l v e d the appl ica t ion o f a th i rd pulsed laser to probe ions that had been formed by L 2 M S . T h i s process a l l owed for the semi-quantitative ident i f icat ion o f two P A H isomers direct ly i n less than f ive minutes wi thout any chromatography. 37 CHAPTER 2 Experimental 2.1 Introduction The pr imary goal o f this thesis was to develop and characterize a new scient if ic instrument. T h i s new device , the two-laser i o n trap mass spectrometer, was constructed over a span o f two years, and improvements were made as the experiments evo lved over the next three years. T h i s chapter w i l l describe the general components and operating parameters o f this n e w l y bui l t system. T h e overa l l instrument was constructed us ing a variety o f electronic s ignal generators and controllers, mechanical devices such as vacuum pumps, opt ica l components such as lenses, pr isms, and lasers, a l l operating under the control o f specia l ly designed software. The discuss ion i n this section w i l l be l imi t ed to on ly those components w h i c h were created or modi f i ed spec i f ica l ly by , or for, the author; so c o m m e r c i a l l y avai lable components w i l l be identif ied but not described i n detai l . A l s o , the d iscuss ion i n this chapter w i l l focus on ly o n details regarding the overa l l funct ioning o f the instrument. Operat ing parameters and condit ions for particular experiments w i l l be found i n the relevant chapters. A n o v e r v i e w o f some o f the more important components o f the system is s h o w n i n F i g u r e 2 .1 . 38 Figure 2.1 A conceptual representation of the two-laser ion trap system at UBC. T i m i n g Elec t ron ics 39 The successful operation o f this device requires many components to per form i n synchroniza t ion w i t h precise t iming . These include: • The h i g h voltage radio frequency electronics that were used for i o n trapping and ejection. • The laser systems, w h i c h include both electronic and opt ica l components . • The function generators that produce the auxi l ia ry waveforms for i o n manipula t ion . • The t i m i n g electronics that includes a number o f d ig i ta l delay generators that provide precise t im ing for the entire system. • The software to operate the entire system and col lect and process the r a w data. W i t h each o f these systems, the author was required, w i t h the help o f the U B C mechanica l and electr ical shops, to either mod i fy or synthesize n e w components . In order to c lar i fy the discuss ion, a descript ion o f the various components w i l l be b roken d o w n into sub headings. 2.2 The lon Trap Electrodes and Vacuum Manifold T h e phys ica l v a c u u m mani fo ld and the i o n trap electrodes are i n their th i rd generation o f use, and were machined by the members o f the M e c h a n i c a l Services Shop (Department o f Chemis t ry , U B C ) several years ago [77, 78]. A schematic d iagram o f the v a c u u m man i fo ld i nc lud ing the i o n trap electrodes is shown i n F i g u r e 2.2. The v a c u u m chamber and i o n trap electrodes were both machined f rom stainless steel and designed w i t h opt ical experiments i n m i n d . The i o n trap electrodes used i n these experiments were o f the " i d e a l " quadrupolar geometry. T h i s geometry is described as " i d e a l " when compared to the "stretched" 40 Figure 2 . 2 Vacuum manifold for the two-laser ion trap system at UBC. 41 geometry, w h i c h is typ ica l o f commerc ia l devices. In the ideal case, the end caps are separated b y the distance specif ied for quadrupolar fields (see S e c t i o n 1.2.1 i n the Introduction) whereas the "stretched" geometry has end caps that have been separated b y a further 10%. It has been empi r i ca l ly determined by some groups that this stretching improves the performance (by m i n i m i z i n g higher order fields) o f the instrument under some circumstances, however , it had been found i n earlier work , that stretching these specif ic electrodes d i d not improve the performance o f this part icular dev ice [77]. The i o n trap electrodes are he ld i n pos i t ion by M a c o r ® machinable glass ceramic spacers, w h i c h provide both r ig id i ty and electrical i so la t ion for the electrodes. The central r ing electrode has a radius o f 10.00 m m and extended hyperbol ic surfaces to provide a more homogeneous quadrupolar f ie ld . T h i s center r ing electrode also had four 2.5 m m diameter holes d r i l l e d through it. These holes, two o f w h i c h are perpendicular and two o f w h i c h are para l le l to the hor izonta l (i.e. a l l are 90° f rom each other) are d r i l l e d through the center o f the electrode a l l poin t ing toward the geometric center o f the t rapping vo lume . The use o f these holes w i l l be described shortly. The v a c u u m mani fo ld was designed to a l l o w a sample probe to be inserted into the i o n trap f lush w i t h the inside o f the r i ng electrode without the need to break the vacuum. T o achieve this, a differentially pumped vacuum inter lock was cus tom bui l t to a l l o w the probe to pass through one o f the 2 . 5 m m holes and into the interior o f the trap. The probe was cus tom bui l t w i t h a M a c o r ® spacer between the probe tip and the base o f the probe ( F i g u r e 2.3). T h i s a l lowed the probe t ip to float at the same potential as the central r ing electrode w h i l e p rov id ing electrical i so la t ion to the v a c u u m chamber as a whole . 42 Figure 2.3 Solid sample probe for the two-laser ion trap system at UBC. 30 cm 1 1; 1 i 1.5 mm 1 T 1 mm 4 Replaceable Probe T i p T J • Macor Spacer Spr ing Guide 43 The sample probe was inserted into the i o n trap through a differential ly p u m p e d v a c u u m chamber. A V a r i a n mode l S D - 9 0 rotary vane pump ( V a r i a n V a c u u m ; T o r i n o , Italy) was used to rough out the sma l l v o l u m e o f air before it was opened to the h i g h v a c u u m chamber. T h i s a l l owed for the direct and easy changing o f samples wi thout the need to break the h i g h vacuum. Since these experiments in t r ins ical ly demanded that the l ight output f rom several lasers be directed to the interior o f the i o n trapV the vacuum housing was designed w i t h several opt ica l ports. The hous ing has four opt ical ports located in l ine w i t h the asymptotes o f the i o n trap. T h i s a l l owed two direct l ines o f sight d iagonal ly through the i o n trap, between the end caps and the r ing electrodes. A fifth opt ical port was p laced i n l ine w i t h the center o f the r ing electrode opposite the sample probe. T h i s a l l o w e d opt ica l access through the two 2.5 m m holes d r i l l ed i n the r ing electrode to the sample probe. The s ix th opt ical port was located 90° f rom the fifth and was i n l ine w i t h the two holes d r i l l e d through the r ing perpendicular to the hor izonta l . F i g u r e 2.4 shows the path o f the desorbing and i o n i z i n g lasers through the trap. The detector for the i o n trap, a channel electron mul t ip l ie r , was p laced into a recess beh ind one o f the end caps and located in l ine w i t h the center o f the trap. A n electron impact (EI) ion iza t ion source was mounted i n the other end cap. T h i s a l l o w e d electron impact to be used as an alternative ion iza t ion method, w h i c h p roved useful for cal ibrat ion. The v a c u u m man i fo ld was evacuated us ing a V a r i a n "Nav iga to r " turbomolecular p u m p ( M o d e l T V 3 0 1 , V a r i a n V a c u u m ) w h i c h was backed by a mode l S D - 4 0 rotary vane pump ( V a r i a n V a c u u m ) . The turbo molecular pump was control led by software located 44 Figure 2 . 4 Close up view of the IR and UV lasers as they pass through the ion trap. 45 on Compute r 2 through a serial connect ion {there were two personal computers used in this work, which will be designated as Computer 1, a 350 MHz Pentium II, and Computer 2, a 133 MHz Pentium I). A " v a c u u m safety uni t" consis t ing o f a relay that prevented any h i g h voltage discharges i n the event that the v a c u u m was lost was added by the electr ical services shop (Department o f Chemis t ry , U B C ) . Pressure i n the man i fo ld was measured us ing a Ba lze r s (Pfeiffer-Balzers; Nashua , N H , U S A ) mode l I K R 020 P i r a n i -c o l d cathode gauge head that was read us ing a Balzers mode l P K G - 0 2 0 meter. W i t h this vacuum system, the lowest uncorrected pressure that c o u l d be achieved was approximate ly 10"7 torr. Str ict ly speaking, a P i r an i - co ld cathode gauge is sensit ive to gas compos i t ion , however , since i n a l l o f this work , the background gas consisted p r imar i l y o f he l i um, an appropriate convers ion factor cou ld be used to find the exact pressure. In a l l o f the experiments performed i n this thesis, the background h e l i u m pressure was he ld constant at approximately 1 mTor r . F i n a l l y , it should be noted, that this v a c u u m mani fo ld was also equipped w i t h two needle leak valves. The first va lve a l l owed for a specified pressure o f h e l i u m (Praxair , h igh puri ty H e (99.995 % ) , Edmon ton , Canada) to fill the mani fo ld vo lume . The second va lve p rov ided for a means o f adding calibrant gas into the chamber w h e n required. T h i s gas was t yp i ca l ly CCI4 (carbon tetrachloride, O m n i s Solvent Grade, B D H C h e m i c a l s , Toronto , Ont . Canada) , w h i c h has k n o w n peak posi t ions when ion ized by the electron gun. Thus a l l o w i n g the spectrum to be mass calibrated. 46 2.3 Electronics and Timing W i t h respect to the t im ing and electronics, we m a y d iv ide the instrument into two parts. The first o f these was the electronics for the i o n trap operation, the second was for the laser operation. In practice, these two systems were complete ly integrated, but for the d iscuss ion here, it is instructive to describe them separately. 2.3.1 Ion Trap Electronics A s ment ioned i n the introduction, mass spectral data are col lected f rom the i o n trap by ramping the R P voltage (either w i t h or without a secondary wavefo rm, depending o n the mode) and s imultaneously detecting the ejected ions us ing the electron mul t ip l ie r . T o produce rel iable mass spectral data, it is essential that this ramping process be hardware t imed w i t h the data col lec t ion . T o achieve this, in-house wri t ten programs operating o n a 350 M H z Pen t ium II computer (Computer 1) were used. Instrument control and data acquis i t ion software were wri t ten i n the L a b V i e w (Nat iona l Instruments, A u s t i n , T X , U S A ) p rogramming environment by the author. ( A more detailed descr ip t ion o f the software appears i n Sec t ion 2.4) The digital- to-analog ( D A C ) and analog-to-d ig i ta l ( A D C ) conversions for the m a i n instrument control took place o n a s ingle P C I board, w h i c h was also capable o f accepting and produc ing a variety o f d ig i ta l pulses ( P C I - M I O - 1 6 X E , N a t i o n a l Instruments). The D A C board control led the R F voltage us ing a 0-5 V control s ignal . T h e R F power supply was an Extranuclear Laboratories Inc. (Pit tsburgh, P A , U S A ) quadrupole power supply ( M o d e l 001-1) that was modi f i ed for use w i t h an i o n trap. Capaci tance matching was achieved us ing a H i g h - Q - H e a d (model 012-16), also f rom Extranuclear 47 Laborator ies . The m a x i m u m R F output achievable w i t h this system was approximate ly - 3 0 0 0 Vo-p (zero-to-peak) and the frequency used was f ixed at 0.967 M H z . In the simplest mode o f operation, the R F power supply was set to an appropriate " t rapping vol tage". Af ter the R F source reached this level (less than 1ms was need to stabi l ize at this voltage, but often 10ms was used) the ion iza t ion event occurred. T h i s was either ion iza t ion by an electron gun (EI) or by the two-laser process. The n e w l y created ions-were then a l lowed a " c o o l i n g pe r iod" o f between 50-1000 ms (user set delay) where the col lected ions co l l ide w i t h the h e l i u m buffer gas at a constant R F voltage. Af ter the coo l ing per iod, the R F potential was ramped to higher values to generate a mass spectrum (mass selective instabil i ty mode). T h i s ramp was often accompanied by a single frequency secondary wavefo rm that a l l o wed for mass range extension (resonance ejection mode). Resonance ejection was achieved us ing a Stanford Research Systems (Palo A l t o , C A , U S A ) mode l D S 3 4 5 funct ion generator. The ejected ions were detected us ing an E T P (Sydney, Aust ra l ia ) mode l A F 1 3 8 electron mul t ip l i e r he ld at - 1 . 7 k V . The detector was he ld at this voltage by a m o d e l 205a-03r (Bertan H i g h Vo l t age , V a l h a l l a , N Y , U S A ) D C power supply. The s ignal f rom the detector was ampl i f i ed w i t h a gain o f 10 7 V / A us ing a K e i t h l e y (C leve land , O H , U S A ) M o d e l 428 current amplif ier . The A D C on the P C I board then sampled this s ignal and reported it to the software. W h e n required, a secondary, complex waveform, cou ld also be appl ied to an end cap, after ion iza t ion , but before detection. Th i s wavefo rm was used to isolate the mass to charge ratio o f interest i n the trap. T h i s technique used the a N o t c h e d B r o a d B a n d W a v e f o r m ( N B B W ) method o f i o n isolat ion. In practice, this wave fo rm was formed by 48 s u m m i n g hundreds o f discrete single frequency components i n the t ime d o m a i n into a single composi te waveform. Th i s wavefo rm was then loaded onto a PCI -321 arbitrary funct ion generator ( P C I Instruments Inc, A k r o n , O H , U S A ) P C I card. The software used to create this wave fo rm was wri t ten by Sc iex ( P C I 3 W A V E , Sc iex , C o n c o r d , O n t , Canada) i n the B a s i c p rogramming language. The wavefo rm calculat ions were performed o n a secondary, computer (Computer .2) because the processing overhead required was so large as to s l o w the function o f a s ingle computer. Occas iona l ly after appl icat ion o f the N B B W waveform, a th i rd funct ion generator was used. T h i s function generator (a second Stanford D S 3 4 5 ) was used to apply a s ingle frequency wave fo rm burst at the secular frequency o f the iori o f interest. T h i s burst was typ i ca l ly lower i n voltage than the N B B W voltages, and was used to induce C I D (co l l i s i on induced dissociat ion) o f the isolated ions. 2.3.2 The Laser Timing In i t i a l ly , two N d : Y A G lasers were used dur ing the course o f this work . The first o f these was used as the desorption laser and was operated at the N d : Y A G fundamental wavelength o f 1064 n m w i t h a 10 ns pulse wid th . T h i s laser, a D C R - 2 A (Spectra Phys i c s Lasers , M o u n t a i n V i e w , C A , U S A ) was directed, unfocussed, into the i o n trap us ing two quartz pr isms ( M e l l e s Gr io t , Rochester, N Y , U S A ) . A series o f quartz sl ides were p laced i n the l ine o f the laser path i n order to s ignif icantly reduce the laser power. T h i s was necessary because the m i n i m u m energy for rel iable las ing was far greater than that needed for desorption. The average laser power measured i n the trap at the probe surface was typ i ca l ly 10 5 W / c m 2 . S ince the beam diameter was m u c h larger than" the 2 . 5 m m apertures i n the electrodes, the aperture size was used to calculate the power density. 49 The second laser used i n this work , the U V photo- ioniza t ion laser, was a N d : Y A G laser f rom L u m o n i c s Corpora t ion ( H Y 400, L u m o n i c s , 10 ns pulse). The laser energy was frequency quadrupled (internal to the laser housing) to provide output at 266 n m . T h i s l ight was directed by two s i lvered mirrors (Me l l e s Gr io t ) through a 1000 m m b i -convex lens ( M e l l e s Gr io t ) . Th i s loosely focused l ight was directed a long the d iagonal between the end caps and the r ing electrode (See F i g u r e 2.4). T y p i c a l energies for the U V laser were on the order o f 40 pJ/pulse. It is important to have a basic understanding o f the solid-state lasers used i n this w o r k i n order to appreciate the electrical requirements o f the system. N d : Y A G lasers are large s o l i d state devices. Th i s means, i n this case, that the lasing m e d i u m - n e o d y m i u m atoms (Nd) , are contained i n a so l id matr ix- a Y t t r i u m A l u m i n u m Garnet ( Y A G ) . In the case o f the lasers used here, the N d atoms are exci ted by a pulsed flash lamp. Once exci ted, the N d atoms w i l l begin to radiat ively relax i n a l l directions but no las ing w i l l occur. T h i s is because the laser cavi ty is in i t i a l ly spoi led by an opt ical shutter. T h i s shutter, a Q- swi t ch , is on ly opened to a l l o w lasing once the N d atomic energy popula t ion invers ion is at its m a x i m u m . Once opened, the Q- swi t ch a l lows a very short, h i g h intensity - 1 0 ns laser pulse to exit the laser. " There are a couple o f key factors that must a lways be taken into account w h e n cons ider ing these types o f lasers. The flashlamps, w h i c h are used to excite the atoms, produce a tremendous amount o f heat. Th i s heat is constantly removed f rom the laser r o d by water c o o l i n g . T h i s be ing said, dur ing normal operation, the N d : Y A G r o d heats up appreciably. T h i s heating causes the laser rod to change its opt ical properties. Laser designers take this fact into account, and as a result the opt ical configurat ion and 50 al ignment are based o n the behavior o f the rod at h igh temperatures. Therefore, these lasers o n l y run eff icient ly at a specif ic rod temperature. T h i s was an important considerat ion when designing the system, because the lasers were designed to operate at 20 H z (based o n the laser rods rece iv ing a certain amount o f energy f rom the flashlamps and thus operating at a certain temperature) whereas the i o n trap c o u l d on ly operate at a m a x i m u m frequency o f 1 H z . Therefore, the lasers c o u l d not s imp ly just be fired when required by the i o n trap because the laser energy and beam shape w o u l d be unstable (for thermal reasons). T o account for this frequency mismatch , a user bui l t pro tocol was required. In the s implest case, the f lashlamps for both N d : Y A G lasers were l i nked and a l l owed to fire at the D C R - 2 A internal c l o c k rate o f 20 H z . O f course, because the Q-switches were not be ing fired, no laser l ight is emitted. The f lashlamp f i r ing was used s i m p l y to keep the laser rods at an appropriate temperature. The pulse train f rom the D C R - 2 A internal c l o c k was moni tored w i t h a home bui l t "trigger mon i to r " (Electronics shop, Chemis t ry Department, U B C ) . W h e n laser pulses were required, an "enable" pulse was sent to. the "trigger moni tor" , w h i c h w h e n enabled, a l l owed , o n receipt o f the next flash l amp pulse, a pulse to be fed into a d ig i ta l delay generator ( M o d e l 400, B e r k l e y N u c l e o n i c s Corpora t ion , San Rafael , C A , U S A ) w h i c h tr iggered the Q-swi tches for the two lasers at two separate programmable delays. B y t r igger ing the Q-Swi tches independently i n this way , a very stable and rel iable delay between the two laser shots was achieved. A t i m i n g d iagram for the entire sys tem inc lud ing the lasers is shown i n F i g u r e 2 .5 . 51 ure 2.5 Timing diagram for the two-laser system used at UBC. D G D 535 A RF Software controlled Waveform Jr D C R 2A Flashlamp linked to HY-400 Flashlamp j Trigger Monitor Input 1 Enable Pulse From D G D 535 B Output pulse From Trigger Monitor B N C Output 1 To D C R - 2 A Q-Switch B N C Output 2 To HY-400 Q-Switch D G D 535 D DS345 Function Generator 142 us -i I L n •30 us i 1 1 1 1 1 1 i i i i 0 50 100 150 200 250 300 350 400 450 500 Time (ms) riming Cartoon IR l ' \ i t 52 The typ ica l delay between the I R and U V laser pulses was user set to 30 ps (jitter < l p s ) . H o w e v e r , w i t h some smal l modif icat ions , the laser delay for the U V cou ld be easi ly var ied between - 1 0 to 400 ps w i t h respect to the I R laser pulse. The o p t i m u m delay t ime o f 30 ps was determined experimental ly; h o w this t ime was chosen is described i n C H A P T E R 4. W h i l e the typ ica l power densities and t ime delays for the lasers are l is ted above, i t should be ment ioned that the exact power outputs and the exact t ime delays were accurately measured for each mass scan. T h i s was accompl i shed by ins ta l l ing two fast photodiodes ( M e l l e s Gr io t ) each o f w h i c h was used to detect the ar r ival o f one o f the laser pulses. The photodiodes were connected to a 400 M H z digi ta l osc i l loscope ( M o d e l T D S 380, Tek t ron ix , W i l s o n v i l l e , O R , U S A ) that recorded both the intensity o f the lasers and the t ime delay between shots. The osci l loscope then reported these results through a G P I B interface to the computer. In this way , informat ion about the laser intensity and the laser delays c o u l d be bundled w i t h the mass spectral data for each mass scan for subsequent s ignal processing. The G P I B interface board was a N a t i o n a l Instruments A T -G P I B / T N T , and was contro l led by user wri t ten software o n Compute r 1. 2.4 Software for lon Trap Operation T h i s unique instrument required the creation o f custom or ig ina l software. The software was wri t ten by the author i n the graphical based language " L a b V i e w " . T h i s type o f p rog ramming lends i t se l f very w e l l to data acquis i t ion applications and is extremely f lex ib le . H o w e v e r , because the result ing "code" is i n a graphical format it cannot eas i ly be transposed and printed i n an appendix for a thesis. Instead, this section w i l l s i m p l y attempt to show the m a i n operating w i n d o w ( F i g u r e 2.6) and describe some o f the 53 Figure 2.6 Primary operating window for Ion Trap control software. Written in the Lab View environment. 54 specif ic requirements and routines o f the program. A s impl i f i ed informat ional f l o w chart depic t ing h o w some o f the m a i n sub-routines are related is shown i n F i g u r e 2.7. The software was required to meet several operating condi t ions. It was required to be extremely f lexible and easily adaptable, w h i l e at the same t ime be stable enough to provide continuous error free operation. T h i s software was designed w i t h an " ins ide-out" approach, where the core function o f the program - data acquis i t ion was wri t ten first, f o l l o w e d by several higher order layers such as the v i sua l interface, f i le input/output, and inter-instrument control . The core requirement o f any software designed to run an i o n trap is the ab i l i ty to control the voltage o f the R T and synchronously col lect the data from the current ampl i f ier . In L a b V i e w , this was done by log ica l ly l i n k i n g a number o f sub-routines, w h i c h w o u l d sequentially load a wavefo rm onto a hardware buffer, wai t for a trigger, and then per form a simultaneous buffered write w i t h a buffered input read; both d r iven b y a hardware c lock . The ab i l i ty to rapid ly and synchronously write and read this data at a rate o f 50,000 samples/second was essential to recording a stable mass spectrum. I f the read and wri te functions weren ' t perfectly synched then the mass peaks w o u l d arrive at different t imes (positions) o n the mass spectra. The second key considerat ion to this "core funct ion" was the abi l i ty to easi ly change the R F wave fo rm that was being applied. F o r instance, the trapping voltage, the durat ion o f the coo l i ng , and the slope o f the R F ramp were a l l required to be user adjustable. T o make this possible, a rapid "wave fo rm ca lcu la t ing" routine was added before the scan function. T o a first approximat ion, the rate at w h i c h the R F voltage was to be ramped was required to meet certain criteria. 55 gure 2.7 Flow chart for the lon Trap software. User Controlled Trigger on Interface Wait for Trigger on Software Initialize Oscilloscope o Initialize Function Generator ^ Output to Oscilloscope through GPIB Output to Function Generator through GPIB Calculate R F waveform Load Waveform to Buffer . O Trigger from DGD 535 Digital Delay Generator Input from Current Amplifier through ADC Wait for Hardware Trigger S y n c h r o n o u s B u f f e r e d O u t p u t / I n p u t Convert Input Data to String and Display Poll Oscilloscope to see if Laser Data Received Output to RF generator through DAC Output to Oscilloscope through GPIB Data from Oscilloscope through GPIB • Download Laser Data From Oscilloscope Bundle Mass Spectral String with Laser Data and Header Info Repeat from A Save Data 56 F o r instance, it has been empi r ica l ly found that a suitable scan rate for the R F voltage is achieved for an ion trap when the ions are ejected at a rate o f 5000 amu/second. S ince the R F power supply used i n this w o r k was capable o f approximate ly a 3000 Vo-p m a x i m u m , this l imi t s the m a x i m u m mass w h i c h may be ejected to approximate ly 300 a m u (without resonance ejection). Therefore, the length o f the fu l l mass range scan should take approximately: 3 0 0 a m u * 1 S £ C ° n d = 0.060seconds Equ. 2.1 5000 a m u Since this corresponds to scanning the fu l l scale i n a t ime o f approximate ly 0.060 seconds, and the R F was dr iven by a control voltage o f 0-5 V , the control vol tage shou ld be ramped at approximately: 5 V / 0 . 0 6 0 seconds = 83.33 V / second . A l s o , because w e want to sample data at a rate sufficient to produce 10 data points/peak, i f w e are scanning 5000 amu/second, we w o u l d want to acquire 50,000 points/second. T h i s impl ies that the buffered wave fo rm should be loaded such that each step w o u l d increase the control voltage by: 83.33 V t 1 second : — = 0.00166 V/s tep « 1 . 6 6 mv/step Equ. 2.2 1 second 50,000 points In addi t ion to def ining the R F ramp, the control voltage also had to be constructed such that the base " t rapping" leve l cou ld be easi ly set. A l s o , it is occas ional ly useful , w h e n ejecting heavy ions, to first eject a l l o f the l o w mass components. T h i s " c l ean ing" pulse was also set to be user configurable. In a l l , the entire wave fo rm that is calculated and loaded onto the output buffer is between 5000-50,000 points. The user controls for this funct ion are o n the y e l l o w box o n the left hand side o f the operating w i n d o w . The 57 col lec ted data result ing f rom the R F ramp, the mass spectrum is d i sp layed i n the large b lack area i n the top left o f the w i n d o w ( F i g u r e 2.6). Bes ides the "core" operations, the i o n trap software was also required to p rov ide many secondary functions. These can be grouped into two categories: c o m m u n i c a t i o n w i t h the hard dr ive (file I /O) , and communica t ion w i t h other instruments through the G P I B interface. The fi le input/output protocol used i n this software was required to be user configurable. Therefore, routines were wri t ten into the code to a l l o w the user to either operate under the "s ingle scan and name" mode (a pop-up w i n d o w asks for a f i le name), or under the "autoname" mode (where a naming prefix may be added, and then the computer sequential ly numbers and names a l l the rest o f the fi les). The first o f these functions is useful w h e n discrete, "one t ry" experiments were preferred, whereas the autoname funct ion was useful for co l lec t ing a large amount o f data automat ical ly over t ime. The f i le I /O functions are d isp layed i n the b r o w n region at the bo t tom o f the operating w i n d o w . The software was also wri t ten to automatical ly create a new directory for every day o f experiments (labeled w i t h the current date) and then store a l l data f rom that days w o r k into this f i le . The software was also required to communicate w i t h two external devices. T h i s communica t i on was achieved through a G P I B , I E E E 488.2 interface. T h i s mode o f communica t ion was hardwired into both the osci l loscope (Tektronix , T K D S 380) and a funct ion generator (Stanford Research, D S 345). The read, wri te , and operat ional functions o f these devices were hard set by the manufacturers. Therefore, the software was s i m p l y required to c a l l the functions through the G P I B interface. The I E E E 488.2 58 standard requires that each data transfer packet be g iven a specific " loca tor" header. Therefore, it is possible to string both the osci l loscope and funct ion generator to the same G P I B card because the data is on ly read by the appropriate device. The osc i l loscope for example, was in i t i a l ly set w i t h a reset string f o l l o w e d by a configurat ion string-; w h i c h identif ied the operational voltage range, t ime domain , and tr igger parameters. Af t e r the osc i l loscope received the photodiode data, a second c a l l funct ion ini t iated a transfer o f the data to Computer 1 for processing. The control options for the osc i l loscope are shown i n the blue box on the right hand side o f the software w i n d o w (Figure 2.6). Rout ines were wri t ten into the software to calculate the power o f the laser shots and the de lay t imes between them. S i m i l a r l y , frequency and voltage values were sent to the S D S 345 function generator v i a the G P I B . F i n a l l y , the col lected data from a l l sources, i nc lud ing the mass spectral data, the laser powers and delays, as w e l l as header informat ion ( w h i c h l isted a l l the user set exper imental condit ions) are bundled and saved i n a spreadsheet format w h i c h is easi ly read by a c o m m e r c i a l spreadsheet program, E x c e l (Microsof t C o r p . R e d m a n , W A , U S A ) . 2.5 Overall System and Timing N o w that the subcomponents have been described, the overa l l data f l o w and operation can be appreciated. Figure 2.5 showed the required t i m i n g for the system w h i l e Figure 2.8 shows the b lock diagram o f the flow o f informat ion and energy. The t i m i n g system is p r imar i l y dr iven by the internal c lock o f a D G D 535 (Stanford Research Systems) d ig i ta l delay generator. T h i s 4-channel delay generator sends the in i t ia t ion signals to the software, the laser control system, and the synthesized funct ion generators. 59 Figure 2.8 Logical block diagram showing the flow of energy and information for the two-laser ion trap system at UBC. I R U V L A S E R L A S E R Photodiode 1 Photodiode 2 ION TRAP Cur ren t Ampl i f i e r Isolation A m p l i f i e r R F Power Supply t T D A C Oscilloscope A D C D A C Funct ion Generator Computer 1 Computer 2 Timing Electronics Legend: IR Laser UV Laser RF Voltage Digital Signal Analogue Low Voltage GPIB Interface 60 The t i m i n g d iagram i n Figure 2.5 shows the order o f operations for the "s imples t" case, i.e. there are no secondary waveforms and on ly two laser shots are used. T h i s case is d i sp layed as a " T i m i n g Car toon" at the bot tom o f Figure 2.5. Cartoons, o f this type w i l l be used throughout the rest o f the thesis, so the reader m y q u i c k l y observe the general order o f operations without the need for a page-long t i m i n g diagram. A s it stands, the majori ty o f the experiments discussed i n this thesis are based on this core t i m i n g function, w i t h addi t ional features and components being added where needed. S i m i l a r l y , Figure 2.8 shows a s impl i f i ed b lock diagram d i sp lay ing the general f l o w o f informat ion and energy through the system. F o r many o f the experiments described i n the f o l l o w i n g chapters, extra components were added as demanded by var ious experimental protocols. However , wh i l e some extra devices were added as required, for the most part, the core system described above was used for a l l the experiments described herein. 61 Chapter 3 Two Laser lon Trap Mass Spectrometry for the Analysis of Environmental Samples 3.1 Introduction T h i s two-laser i o n trap system was designed o r ig ina l ly w i t h the goal o f e x a m i n i n g envi ronmenta l ly important contaminants direct ly i n so l id matrices. Spec i f i ca l ly , one important f ami ly o f pollutants was considered: the p o l y c y c l i c aromatic hydrocarbons ( P A H s ) . P A H s represent a large fami ly o f molecules that are produced via both natural and anthropogenic processes and are found ubiqui tously throughout the environment [79, 80]. Figure 3.1 shows the structures o f the 6 P A H s p r inc ipa l ly used i n this work . P A H s are p r i m a r i l y formed via the incomplete combust ion o f hydrocarbons inc lud ing w o o d , coa l , o i l , or gas used i n the generation o f heat or electrici ty, or for motor ized vehic les , and dur ing pet ro leum c rack ing [81]. O n l y a smal l fraction o f the total P A H load comes f rom natural sources; p r inc ipa l ly from forest fires or the decomposi t ion o f biomass i n swaps and bogs [82]. P A H s are capable o f t ravel ing either attached to aerosols (for example soot particles) or deposited i n sludge and wastewater. The aerosol materials are capable o f t ravel ing long distances and have been found throughout the atmosphere i n remote oceanic [83] and polar atmospheres [84]. P A H s have long been k n o w n to have potential mutagenic and carc inogenic properties. A s early as the 1930's , some P A H s were named as potential carcinogens. In fact, the first k n o w n chemica l carcinogen was the P A H dibenz(a,h)anthracene w h i c h was . 6 2 Figure 3.1 The six polycyclic aromatic hydrocarbons (PAHs) primarily used in this work. Acenaphthene 154 amu Chrysene 228 amu Phenanthrene Pyrene 178 amu 202 amu Benzo(a)pyrene Coronene 252 amu 300 amu 63 isolated f rom a synthetic tar compound [85]. Subsequent w o r k has ident i f ied many other P A H s to be potential an imal carcinogens [80, 86]. These compounds can be harmful ly administered either top ica l ly (via direct contact) or as fine particulates deposited i n the lungs [87]. A d d i t i o n a l l y , P A H s are also capable o f entering the body i n d r i n k i n g water or i n s o l i d matrices such as food or so i l [82]. Once ingested, P A H s typ i ca l ly migrate to the fatty tissues and have been shown to affect l iver and k idney function. A s a result o f the demonstrated carcinogenic , mutagenic, and i m m u n o t o x i c effects o n animals and humans, the U S Envi ronmenta l Protect ion A g e n c y ( U S - E P A ) has labeled a number o f P A H s as " P r i o r t y - 1 " pollutants [88]. T h i s designation requires that the concentrat ion o f these compounds be c lose ly moni tored and l imi t s are set for human exposure. Unfortunately , however , the task o f determining P A H contaminant levels , par t icular ly i n s o l i d matrices, is not a t r iv i a l one. Curren t ly there exist several methods avai lable for the analysis o f P A H s i n sol ids . The usual method for P A H analysis involves some form o f extraction f rom the matr ix , t yp i ca l ly us ing a Soxhlet , f o l l owed by separation via chromatography, usual ly gas chromatography ( G C ) , w i t h detection us ing mass spectrometry ( M S ) [89, 90] . T h i s method, however , is l imi t ed ; due to the l o w vola t i l i ty o f P A H s , the detectable mass range us ing most G C / M S instruments is l imi t ed to around 300 amu. Furthermore, the l o w concentrat ion o f P A H s often requires that a pre-concentration step be employed i n addi t ion to sample clean up pr ior to analysis. A l o n g w i t h so lub i l i ty problems, the combina t ion o f a l l o f these pre-treatment procedures means that a fu l l analysis can take days to complete. W h i l e other separation and detection methods have been suggested, none are used rout inely for analysis [91-93]. 64 Anothe r important considerat ion i n the f ie ld o f P A H analysis o f environmental samples is the issue o f sample preparation. T y p i c a l l y , most forms o f P A H analysis require that the samples are cleaned and the P A H s removed from the s o l i d matr ix . T h i s process is neither inexpensive nor fast, and certainly not t r iv i a l because one sample often contains P A H s i n a variety o f phys ico-chemica l states [94]. Once the P A H s have been extracted f rom their native matrix, the analyst must also be concerned w i t h issues o f so lub i l i ty , storage, biotransformation, and photo-degradation. A s a so lu t ion to these persistent problems several groups have attempted direct analysis . S o m e o f the more successful applications include secondary i o n mass spectrometry ( S I M S ) [40], fast a tom bombardment ( F A B ) [42], and laser desorpt ion ( L D ) [41]. These techniques however , are often non-selective and produce complex mass spectra w i t h a var iety o f peaks result ing from molecula r fragmentation o f both the analyte and the matr ix . Two- lase r mass spectrometry has been appl ied to P A H analysis successfully by several groups [72, 95, 96]. These groups have performed P A H analysis i n both ar t i f ic ia l [96] and natural matrices [95]. Perhaps the bes t -known example o f this technique was the examina t ion o f M a r t i a n meteorites b y Zare ' s group [97, 98]. It should be noted that i n a l l o f these examples , the mode o f M S used was a lways a T O F . W h i l e a T O F is very useful w h e n used i n a two-laser experiment, it is not capable o f M S " - a feature that m a y prove useful for obta ining structural and poss ib ly i someric information. F i n a l l y , f rom a method development point o f v i e w , the fact the P A H s had been previous ly examined w i t h the two-laser technique w i t h a T O F provides an excellent base for the development and compar i son to this instrument. 65 T h i s chapter w i l l describe the early stages o f the development o f the instrument for environmental analysis. Th i s includes a smal l section devoted to demonstrat ing the effectiveness Of the current i on trap software by operating w i t h the electron gun mode o f ion iza t ion - this process was important for cal ibrat ion because the i o n trap software used here was newly created. The remainder o f the chapter w i l l be dedicated to descr ib ing the effectiveness o f the two-laser i o n trap system for the analysis o f P A H s d i rec t ly o n s o l i d materials. The bu lk o f the material reported i n this chapter was prev ious ly publ i shed by Specht and Blades i n the Journal of the American Society for Mass Spectrometry [99]. 3.2 Experimental The overa l l system design was described i n C H A P T E R 2. Therefore this sect ion w i l l be l im i t ed to a specific descr ipt ion o f the experimental techniques and materials used dur ing the w o r k described i n this.chapter. In this chapter, two distinct set-ups were employed . The i o n trap was in i t i a l l y operated w i t h the " t radi t ional" electron gun mode o f ioniza t ion , whereby analyte molecules were leaked into the vacuum chamber and ion ized by electron impact ion iza t ion by an electron gun located behind one o f the end caps. T h i s method is useful for p r o v i d i n g mass cal ibra t ion and demonstration o f the in i t i a l effectiveness o f the n e w ion trap software.. The i o n trap was also operated i n the newly developed two-laser set-up for the direct analysis o f so l id samples. In the electron gun mode o f operation, the p r imary sample used was carbon tetrachloride - CC1 4 (Omniso lve grade, BDH Chemica l s , Toronto , Ont.) . F o r the so l id sampl ing L 2 M S method the so l id PAH samples were created by first preparing standard solutions o f the P A H s i n H P L C grade hexane. The hexane and a l l o f the P A H s were 66 acquired f r o m S i g m a - A l d r i c h ( M i l w a u k e e , W I , U S A ) and were used as received. The s ix P A H s used i n this chapter were acenaphthene (mass 154 amu), phenanthrene (178 amu), pyrene (202 amu), chrysene (228 amu), benzo(a)pyrene (252 amu), and coronene (300 amu). The product ion and creation o f the so l id samples was part o f the w o r k descr ibed i n this chapter and therefore informat ion concerning this is located i n the 'Resul t s and D i s c u s s i o n ' Section 3.3.2. 3.3 Results and Discussion 3.3.1 Test Of Instrument Effectiveness The first stage i n the development o f this system was to test the effectiveness o f the n e w l y bui l t i o n trap system. Th i s was achieved through the use o f a standard electron gun arrangement. The electron gun was used to ion ize a test gas ( C C L 4 ) that had been leaked into the v a c u u m chamber at a pressure o f 0.1 mTor r . Figure 3.2 shows a mass spectrum (average o f one hundred spectra) that results f rom the electron impact ion iza t ion o f C C 1 4 us ing a 70 e V electron gun. T h i s spectrum was col lec ted under the t i m i n g regime shown at the bot tom o f the figure (this convention will be used throughout the thesis). The spectrum contains 4 p r imary peaks result ing f rom the 4 poss ible combinat ions o f the chlor ine isomers o f mass 35 (75.77 % abundance) and mass 37 (24.23 % abundance). The average resolut ion o f these peaks is (m /AmFWHivi) is 246. The reader should also note that under this s imple mode o f operation the mass range was l im i t ed to 280 T h as the h igh mass l imi t . Once the normal mode o f operation was established, the more sophisticated resonance ejection mode o f operation was examined. T h i s method o f operation, w h i c h is 67 Figure 3.2 Mass spectrum of CCI4 ionized by electron impact ionization, resulting in the formation of CCI3+ ions. 119 . 117 RF Voltage 68 n o w standard o n most commerc ia l devices, ut i l izes, a s ingle frequency wave fo rm, i n conjunct ion w i t h the R F voltage ramp i n order to eject species at a q value o f less than 0.908. T h i s should, i n theory, increase the mass range o f the device because ions are ejected to the detector at a lower q-value, and thus at the m a x i m u m value o f the R F potential appl ied to the trap, a higher mass i o n may be ejected. A l s o , this mode o f operat ion should produce higher resolut ion peaks because the ions are ejected dur ing a nar row range o f R F voltages (i.e. they are more coherently ejected). Figure 3.3 is the mass spectrum (average, o f one hundred spectra) that results f rom the electron impact ion iza t ion o f CCI4 under ident ical condit ions to those used for Figure 3.2 except that a single frequency wavefo rm is appl ied between the end caps dur ing the R F ramp. Th i s wavefo rm has a frequency o f 247 k H z and exci ta t ion voltage o f 6.25 Vpeak-to-peak- The addi t ion o f this wavefo rm causes the ions to be ejected at a reduced q va lue o f 0.62, and consequently an increase i n the mass range to a value o f ~ 405 T h . A l s o , note that the average resolut ion o f the C C l 3 + peaks has increased f rom 246 to 301 m / A n i F W H M -Anothe r important feature found i n most i o n trap mass spectrometers is the ab i l i ty to perform i o n i so la t ion w i t h i n the trap by the addi t ion o f supplemental waveforms. T h i s abi l i ty was demonstrated on the CCI4 sample by app ly ing a notched broad band wave fo rm (NBBW) dur ing the ion-coo l ing per iod before the R F ramp. Figure 3.4 shows a mass spectrum (average o f one hundred spectra) that resulted f rom the appl ica t ion o f a N B B W wave fo rm after a C C L 4 sample was ion ized by electron impact ion iza t ion . T h e N B B W w i n d o w was chosen to have a notch w i t h no frequency components between 84.5 k H z and 86.5 k H z and the overa l l wavefo rm had a voltage 69 Figure 3.3 Mass spectrum of CCI 4 ionized by electron impact ionization and ejected from the ion trap by the resonance ejection mode of operation. • 130 150 Mass/Charge E - ° U N Resonance 1 ' ' Ejection fl : 4 J RF Voltage 70 Figure 3.4 Mass spectrum of CCI4 ionized by electron impact ionization with a single NBBW cycle after ionization. The NBBW cycle caused the removal of all mass components below 119 Th and above 120 Th. 50 70 90 110 130 Mass/Charge 150 170 190 E-GUN I | 1 RF Voltage I 71 range no larger than 1.4 V p . p . T h i s notch wavefo rm effectively removed a l l other ions i n the i o n trap except that at 119 T h . F i n a l l y , one addi t ional abi l i ty was required to be demonstrated for this i o n trap system. Af te r the N B B W ion isola t ion wavefo rm is appl ied, typ ica l ly , one w o u l d apply a secondary s ingle frequency component at the secular frequency o f the i o n o f interest i n order to induce co l l i s iona l d issociat ion and thus gain daughter i o n informat ion. The l og i ca l extreme o f this appl ica t ion occurs w h e n the single frequency wave fo rm voltage appl ied is large enough so that no daughter ions remain, but rather the parent i o n is comple te ly removed. F o r demonstration purposes, a single frequency wave fo rm o f this nature was appl ied to an electron impact ion ized CCI4 sample, but w i t h no N B B W wave fo rm appl ied. Th i s a l lows one to examine the selectiveness o f the appl ied wavefo rm. F i g u r e 3.5 shows a mass spectrum (average o f one hundred spectra) that results f rom the electron impact ion iza t ion o f CCI4 f o l l owed by the appl icat ion a single frequency wave fo rm at 85.20 k H z , w i t h a voltage o f 1 .5V P . P . N o t e that this wave fo rm effect ively removes the peak at 119 T h wh i l e leaving its neighbors untouched. T h i s single frequency was useful not on ly i n that it showed that a single frequency/voltage combina t ion c o u l d be chosen to complete ly remove a single i o n w i t h + / - 2 T h unit resolut ion, but it also a l l o w e d the user to determine the precise q value under w h i c h the ions were trapped. D u e to the nature o f the R F circui t ry it is v i r tua l ly imposs ib le to di rect ly measure the exact R F voltage appl ied to an ion trap. A s a result, determining the t rapping R F leve l , and thus the trapping q value is very diff icul t . A s a solut ion to this p rob lem most workers instead deduce a q value and thus the trap R F level by determining the frequency o f a k n o w n i o n by selective ion ejection. F o r example, us ing the data 72 Figure 3.5 Mass spectrum of CCI4 ionized by electron impact ionization and with removal of 119 Th by a single frequency ejection after ionization. 50 70 90 110 130 Mass/Charge 150 170 190 RF Voltage. 73 described i n F i g u r e 3.5 one can calculate the trapping q value for the 119 T h i o n was 0.23 and by k n o w i n g the i o n trap dimensions,that the R F trapping voltage was approximate ly 283 volts . 3.3.2 Sample Preparation Once the in i t i a l problems o f software design and i o n trap functionali ty were w o r k e d out, the next major challenge concerned that o f sample preparation. A s ment ioned previous ly , the two-laser method has been k n o w n for some t ime, and natural ly, a number o f sample preparation schemes have developed over the years [78, 100-102]. There are several requirements that must be met w i t h respect to sample preparation i n this system. Ideally, a system should be designed so that a sample may be inserted direct ly into the mass spectrometer without m u c h pre-treatment or preparation. M e c h a n i c a l l y , this presents a challenge as the samples to be investigated (soi l or sediments for example) are often i n a loose powder form. T h i s p rob lem is compounded by the fact that, i n the case o f this instrument (and many others), that the sample must be inserted perpendicular to the hor izontal . Th i s impl ies that some means must be developed i n order to contain a powder sample on a ver t ical surface. In addi t ion to the issue o f preparing " rea l " powder samples, such as soi ls and sediments, the question o f what an appropriate "test" sample should consist o f was also addressed. Ideal ly, the test sample w o u l d provide a means o f p roduc ing a s imple and reproducible s ignal so that the mult i tude o f instrumental parameters c o u l d be op t im ized w i t h ease. M a n y groups have confronted the "test" sample p rob lem by s i m p l y d i s so lv ing the analyte o f interest i n an appropriate solvent, then deposi t ing a s m a l l drop o f the 74 so lu t ion onto a sample probe and a l l o w i n g the solvent to evaporate [103]. A l t e rna t ive ly , other groups have gone the route o f producing th in po lymer f i lms w h i c h were impregnated w i t h the sample o f interest [100, 101]. W h i l e both o f these techniques produce very stable and reproducible sources o f analytes, neither addresses the p r o b l e m o f prepar ing a real environmental sample, i.e. a so l id powder such as a so i l or sediment. It was therefore decided early i n this w o r k that the method o f sample in t roduct ion and test sample selection should be engineered simultaneously so that the f ina l procedure w o u l d be amenable to " r ea l " powder samples and that the test samples w o u l d be an accurate representation o f the same. N u m e r o u s protocols were examined before a suitable method was decided upon. In terms o f addressing the p rob lem o f sampl ing a powder material o n a ver t ica l surface, a number o f procedures were attempted. These inc luded the use o f g lyce ro l as a b i n d i n g agent w h i c h was used to adhere the powder sample to a ver t ical surface. It has also been suggested by L u b m a n that the use o f g lycero l enhances the desorption process because o f a favorable absorption by the a lcohol groups o f the g lycero l at the desorption laser wavelength [102]. W h i l e this m a y be true.in the case o f desorption w i t h a CO2 laser (10.6 um) it was found to be unsuitable when used w i t h the N d : Y A G desorpt ion laser (1064nm). M o r e important ly, the g lycero l sample preparation method was found to be imprac t ica l and introduced numerous technical problems such as contaminat ion, unpredictable sample modi f ica t ion , and most importantly samples s l ipped o f f the probe. The second mechan i sm for sample preparation w h i c h was attempted concerned the use o f double s ided tape. It has been suggested that a powered sample c o u l d be s i m p l y adhered to one side o f the tape and then attached to the probe surface. W h i l e w i t h 75 this procedure it was possible to obtain a 2-laser mass spectrum, the quest ion o f the poss ible interferences by the adhesive matr ix l imi t ed the effectiveness o f this approach. The method that was the most successful i n v o l v e d mechanica l ly pressing the powder into a smal l sample cup machined into the end o f the sampl ing probe. T h i s method, w h i c h is routine i n F T - I R measurements (i.e. w i t h the creation o f K B r pellets), was able to re l iab ly produce stable samples that maintained their phys i ca l structure w h e n he ld ver t ica l ly . The phys ica l design and manufacturing o f the sample probe cup and the press were done i n cooperat ion w i t h the mechanica l services group at U B C . The final des ign is shown i n F i g u r e 3.6. The powder samples were first p laced into the sample cup w h i c h was then inserted into the press. The press was designed w i t h mechanica l stops i n place to insure that every sample was compressed to the same extent. Once the sample preparation question had been resolved the next concern i n v o l v e d sample development. Ideally, when testing and des igning a new instrument, there are several requirements w h i c h must be met for selecting a "test" sample. F o r example , the ideal sample w o u l d produce a long lasting stable s ignal , so that instrument parameters m a y be var ied and the result ing effect on the s ignal can be investigated. Secondly , a test sample should also m i m i c , as c losely as possible , potential analytes and matrices. S ince the in i t i a l goal o f this instrument was to test for P A H s o n environmenta l matrices, this obv ious ly i m p l i e d the creation o f powder sample that had been sp iked w i t h P A H s . Several so l id matrices were considered and tested, but the one that was f ina l ly used was activated charcoal . Ac t iva t ed charcoal is an inexpensive, carbon based mater ial 76 Figure 3.6 Mechanical press and sample cup used in the creation of solid samples for the two-laser system at UBC. s imi la r i n structure to graphite, w h i c h is easily attainable and k n o w n to have excel lent chemica l adsorpt ion properties [104]. Test samples were produced by first creat ing a hexane so lu t ion o f the P A H s o f interest and then sp ik ing a k n o w n v o l u m e o f the so lu t ion onto a we ighed amount o f activated charcoal . The samples were then sonicated i n a c losed container for 30 minutes to insure good m i x i n g and the solvent then a l l o w e d to s l o w l y evaporate o f f at r o o m temperature leaving the P A H s deposited on the charcoal matr ix . T h i s s imple method produced very reproducible samples, w h i c h were long l i v e d (for thousands o f laser shots), easi ly managed, and had roughly the same mechanica l properties as so i l or sediment. 3.3.3 Two Laser Ion Trap Mass Spectra Several hundred different c h a r c o a l / P A H standards were examined us ing the sample preparation procedure described above. A typ ica l example o f the mass spectrum obtained f rom one o f these samples containing five P A H s (acenaphthene, phenanthrene, pyrene, chrysene, and benzo(a)pyrene) a l l at a concentration o f approximate ly 25 p ino le P A H / g r a m o f charcoal is shown i n F i g u r e 3.7. The operating condi t ions under w h i c h this data was col lected is shown at the bottom o f the figure. N o t e that the spectrum is re la t ively " c l ean" i n that is shows very few peaks other than for the 5 P A H s . It shou ld also be noted that very li t t le fragmentation was observed i n any o f the trials performed. A n expanded v i e w o f this spectrum w h i c h shows the detail around the P A H central mass peaks is also shown i n F i g u r e 3 . 8 . There are smal l peaks at M - l and M - 2 around each parent peak that represents the loss o f one or two hydrogen atoms f rom each o f the P A H s . T h i s type o f fragmentation is typ ica l i n two laser mass spectrometry o f P A H s and is dependant o n the laser power and wavelength[51, 72, 95 , 105]. 78 Figure 3.7 Typical mass spectrum collected with the two-laser mode of ionization of a sample of charcoal spiked with five PAHs. 50' 100 150 200 250 Mass/Charge 300 350 400 IR UV RF Voltage 79 Figure 3.8 Expanded view of Figure 3.7 between mass 130-265 Th. 130 140 150 160 170 . 180 190 200 210 220 230 240 250 260 Mass/Charge IR UV 1 RF Voltage 80 In order to test the longevi ty , stabili ty, and reproducibi l i ty o f the standard "test" sample several trials were performed where the one standard was exposed to several thousand laser shots. A n example o f this type o f experiment is shown i n Figure 3.9, w h i c h is a p lo t o f the magnitude (integrated pyrene peak areas) o f the pyrene s ignal as a funct ion o f the number o f laser shots for a sample containing 10 u m o l pyrene/gram o f charcoal . T h i s plot , typ ica l o f these experiments, shows a relat ively dramatic in i t i a l drop i n pyrene s ignal w h i c h then levels o f f to fo rm a stable constant s ignal . T h i s data suggests, that there are l i k e l y two types o f pyrene environments i n the charcoal sample. F o r one adsorpt ion site, the pyrene is weak ly bound and it is desorption f rom this site that is p r imar i l y responsible for the pyrene signal dur ing the first 500 laser shots. F o r the other type o f adsorption site pyrene is bound more strongly and produces the re la t ively constant s ignal observed i n the later tens o f thousands o f mass spectra. W h i l e there is no further evidence p rov ided here to conf i rm this hypothesis, the result is nonetheless very useful. The flat region o f the curve i n Figure 3.9 impl ies that this sample preparation can prov ide a stable source o f s ignal so that instrumental parameters and operating condi t ions m a y be op t imized without fear o f other sample effects. 3.3.4 Effect of IR Power on Observed Signal It is w e l l k n o w n that the magnitude o f the I R power plays an important role i n the desorption process, therefore the effect o f I R intensity on the P A H signals was investigated. A n example o f this experiment is shown i n Figure 3.10 for the case o f phenanthrene at a concentration o f 10 u m o l phenanthrene/gram o f charcoal . Figure 3.10 shows an experiment where, l ike above, a sample was exposed to several thousand laser shots. F o r the first 2500 shots ( IR laser power approximately 1.2* 10 5 W / c m 2 ) the s ignal 81 Figure 3.9 Magnitude of the pyrene peak area as a function of the number of laser cycles for a charcoal sample spiked with pyrene. 0 10 20 30 40 100's of laser shots 82 Figure 3.10 Effect of IR power on observed signal for phenanthrene as a function of the number of laser shots for three IR laser powers. • Measured IR Power (Left Hand Units). • Integrated Peak Areas (Right Hand Units). 20 40 60 80 100 100's of laser shots 83 was observed to decrease and leve l o f f as before. A t this point , however , the I R laser intensity was increased by 3 3 % ( IR laser power approximately 1.6*10 W / c m ). A s a results o f this increase i n I R power "the phenanthrene signal increased sharply, and then began the d o w n w a r d trend as before. The same phenomenon was observed a th i rd t ime w h e n the laser power was increased yet again by 8 0 % w i t h respect to the or ig ina l ( IR laser power approximately 2.2* 10 5 W / c m 2 ) . These observations are consistent w i t h bu lk sampl ing theory [106-108]. B r i e f l y , i n order for a P A H to be desorbed f rom the surface o f a bu lk so l id , the temperature must increase enough to make the process statistically probable. D u r i n g the course o f a 10 ns laser pulse, the laser induces a very rapid heating i n the so l id . The rate o f heating as w e l l as the temporal and spatial properties o f the temperature i n the so l id are affected by a number o f factors i nc lud ing the absorption coefficient o f the material at the laser wavelength , the thermal conduct iv i ty o f the sample, and the heat capacity o f the mater ial [109, 110]. Several excellent treatments o f this heating phenomenon show that, t yp i ca l ly , the bu lk sample w i l l rise to a temperature that makes the desorption probable for o n l y a few tens o f nanoseconds dur ing and after the laser pulse [108, 111]. A l s o , w h e n a greater I R power is appl ied to the sample, a higher temperature w i l l be achieved at a greater b u l k depth. Therefore, we may conclude that i n the case o f a stepped I R profi le , we are observing sampl ing f rom greater and greater depths w i t h each increase i n laser power . The signals a l l q u i c k l y decay, as seen previous ly , because as the number o f laser shots increases, the sampl ing area becomes depleted i n weak ly bound P A H s . F i n a l l y , it is noted that i n the th i rd I R power increase step, the increase i n phenanthrene s ignal is not 84 as pronounced as the second step. Th i s is most l i ke ly due to the fact that the P A H must desorb f rom a greater depth i n order to be sampled. 3.3.5 Effect of UV power on Observed Signal The U V power is also w e l l k n o w n to affect the type and magnitude o f observed signals i n two laser mass spectrometry. A s a result, a systematic study was undertaken to examine the effect o f the U V power o n the signal f rom several P A H s . V i r t u a l l y ident ica l results were observed for a l l o f the f ive P A H s examined. F i g u r e 3.11 is a p lo t o f the U V power vs . integrated peak areas for the f ive P A H s . The graphs exhibi t a l inear relat ionship between s ignal ampli tude and laser energy at energies b e l o w about 80 pJ , however , at higher energies, a l l o f the signals begin to leve l o f f (note: the line drawn through the data has no physical meaning and is only meant to guide the readers eye). A t pulse energies higher than about 100 uJ , fragmentation products f rom the P A H s beg in to appear i n the mass spectra. Th i s w o u l d indicate that at h i g h photon fluxes, the neutrals are either absorbing three photons causing fragmentation, or ions formed prev ious ly are absorbing a photon to induce fragmentation. F i n a l l y , it should also be noted, that at these h i g h fluxes, the space charge l i m i t was easily reached. So any addi t ional ions w h i c h were formed were faced w i t h a large C b u l o m b i c repuls ion. In general, the relat ionship between the signal intensity and U V energy is s imi l a r to that observed by Z e n o b i ' s group [96] and is consistent w i t h the theoretical treatment by Johnson and Ot i s [112]. B r i e f l y , the results can be expla ined i f one considers the process i n v o l v e d i n resonance enhanced two-photon ioniza t ion . I f the absorpt ion cross section at the exci ta t ion wavelength used is identical for both the ground state and the 85 Figure 3.11 UV power (uJ) vs. integrated peak areas for five PAHs. The x-axis is in the units of uJ of UV energy in the trap volume and the y-axis is in terms of integrated peak areas. 86 Figure 3.11 continued. Chrysene 900 Benzo(a)Pyrene 5 -. 4 _ 0 100 200 300 400 500 600 700 800 900 87 exci ted state o f the molecule , then one w o u l d expect to see a quadratic dependence o f the s ignal o n the U V energy. If, however, the absorption cross section at the exci ta t ion wavelength was m u c h greater for one state than the other, a "rate determining step" w o u l d be observed, and the signal w o u l d become l inear w i t h U V intensity. The latter scenario is what was observed i n the case o f a l l f ive P A H s examined. It is imposs ib le to k n o w , w i t h this experimental set up, w h i c h absorption process was rate determining. The analy t ica l ly important result f rom this experiment is that it a l lows us to select the reg ion around ~ 40 pJ/pulse as the o p t i m u m U V laser energy for analyt ical determinations o f P A H s because the signal/energy ratio i n this region is l inear and thus easi ly accounted for. 3.3.6 Semi-Quantitative Analys is It has l ong been k n o w n that laser sampl ing is a less than ideal means o f per forming quantitative analysis. The shot-to-shot fluctuations i n the lasers and noise induced by uneven sample preparation typ ica l ly renders fu l l quantitative analysis an imposs ib i l i ty . Regardless, it was s t i l l an important parameter to investigate w i t h this newly created system since semi-quantitative analysis is possible. T o evaluate this aspect, a series o f standards o f k n o w n concentration were prepared cover ing the concentrat ion range o f 1 to 25 u m o l pyrene/gram o f charcoal . The measured s ignal vs . k n o w n concentrat ion o f the sample is plotted i n F i g u r e 3.12. T h i s p lo t exhibi ts a good l inear relat ionship between concentration and observed s ignal w i t h and R 2 value o f 0.99. The error bars are based on measuring five replicates o f the same sample and are ± 5%. The measured signals were col lected by averaging the s ignal (integrated peak area) for 88 Figure 3.12 Concentration vs. measured signal for a series of pyrene/charcoal standards. 1000 mass spectra after the sample had been exposed to 500 laser shots (i.e. once the s ignal was i n the flat part o f the l i fet ime curve). It should be noted, that this ca l ibra t ion curve demonstrates on ly the l inear response o f the system and not the absolute detection l imi t s . T h i s can be appreciated w h e n one considers that the magnitude o f the peak area is a funct ion o f both I R and U V laser energy as w e l l as the concentration o f the analyte and the nature o f the matr ix . A s a result, the magnitude o f the analyte signal can be s ignif icant ly changed for a g iven concentration by s imply increasing the I R and U V power. Ca l ib ra t i on curves have been observed prev ious ly for the two-laser technique. H o w e v e r , most have used an internal standard to compensate for shot to shot fluctuations i n the laser energy [53]. Regardless, a useable cal ibra t ion curve was obtained i n this experiment, w h i c h demonstrates that the technique is useful for, at least, semi-quantitative analysis. O f course, i n cases where a matr ix match was hard to find, the method o f standard additions w o u l d be preferred. 3.3.7 Selective Ion Accumulat ion T o demonstrate the unique capabil i t ies o f the ion trap, a series o f experiments were performed to evaluate the poss ib i l i ty o f select ively accumulat ing the results o f several laser shots. Ex te rna l waveforms were used to isolate specific l o w concentrat ion species i n the i o n trap. These species were pre-concentrated i n the gas phase by per forming mul t ip le laser s h o t / N B B W sequences w h i l e mainta in ing the R F voltage at a steady t rapping leve l . T h i s procedure begins w i t h desorption and ion iza t ion laser shots f o l l o w e d by a short N B B W pulse that is used to clear the trap o f a l l species but the l o w concentrat ion analyte. Successive laser s h o t / N B B W pulse combinat ions are then used to 90 b u i l d up a large analyte popula t ion i n the trap. Wi thout the N B B W pulses between the laser shot combinat ions , a l l species i n the trap w o u l d b u i l d up equal ly, and as a result the l o w intensity s ignal w o u l d be swamped and space charge w o u l d render the resul t ing spectrum meaningless. A s a demonstrat ion o f this method, a charcoal sample was prepared w h i c h contained five P A H s : acenaphthene, phenanthrene, pyrene, benzo(a)pyrene, and coronene a l l at a concentrat ion o f approximately 25 p m o l P A H / g r a m charcoal . A s ixth P A H , chrysene, was also sp iked on this sample, at a concentration o f approximately 1 u m o l P A H / g r a m charcoal . A n example o f a mass spectrum o f this sample col lected i n the no rma l manner is shown as the so l id l ine i n Figure 3.13. The chrysene s ignal is seen to be o n l y s l igh t ly above the base l ine i n this spectrum. The gray l ine i n Figure 3.13 shows a spectrum where seven laser s h o t / N B B W cycles were used to accumulate a large chrysene i o n populat ion. In this spectrum, the chrysene signal is n o w far above base l ine and easi ly quantifiable. The condit ions under w h i c h these two spectra were acquired are s h o w n at the bot tom o f the figure. Figure 3.14 shows the average chrysene peak height (normal ized relative to a single laser cycle) as a function o f the number o f cycles . T h i s l ine has a slope o f 0.9991 and a R o f 0.996 indica t ing that the bui ld-up is very linear and that successive laser shots do not substantially interfere w i t h the trapped ions. Th i s procedure cou ld become analy t ica l ly useful i n situations where l o w concentration species are to be examined by M S / M S . O b v i o u s l y , a s ignal bui ld-up l ike this w o u l d be imposs ib le i n a T O F M S . 91 Figure 3.13 Mass spectrum from a sample containing 6 PAHs (chrysene depleated) [black line]. After chrysene gas phase pre concentration [gray line]. 100 150 200 250 Mass/Charge 300 350 N o pre concentration IR U V R F V o l t a g e W i t h pre concentration IR UV . ,|R UV 92 Figure 3.14 Average chrysene peak height normalized relative to a single laser cycle as a function of the number of laser cycles. 0 1 2 3 4 5 6 7 8 9 10 11 12 Number of Cycles 93 3.3.8 SRM 1994 Analysis The f inal sets o f experiments were a imed at determining whether or not real s o l i d samples c o u l d be analyzed w i t h the proposed protocol . T o this end, a standard reference mater ia l , S R M 1944 N e w Y o r k - N e w Jersey R i v e r Water Sediment, was examined i n a manner ident ical to the determinations o f the charcoal standard. The resul t ing spectrum produced w i t h no s a m p l e p r e t r e a t m e n t is shown i n F i g u r e 3.15. T h i s figure shows the uncorrected mass spectral data that results when a real sediment sample conta in ing a series o f P A H s a long w i t h several other classes o f contaminants is analyzed. T a b l e 3.1 shows a l ist o f some o f the P A H compounds k n o w n to exist i n the sample w i t h their cert if ied concentrations. It was beyond the scope o f this in i t i a l study to quantify and conf i rm the concentrations o f a l l the P A H s i n this sample. Instead this should s i m p l y be accepted as a p r o o f o f concept to stimulate further invest igation. In the case o f this experiment, no real secondary layer o f select ivi ty, beyond opt ica l select ivi ty, is required i n the analysis o f the spectra; however , it w o u l d be a re la t ively t r i v i a l matter to perform N B B W iso la t ion o n selected mass peaks for further daughter analysis and structural conf i rmat ion. Th i s conf i rmat ion o f mass selected peaks is imposs ib le w i t h most convent ional two-laser systems because they a l l a lmost exc lus ive ly use T O F mass spectrometers for analysis. 94 Figure 3.15 Two laser mass spectrum observed from a sample of standard reference material 1944a. I ii i i iLlLilU j 50 100. 150 200 250 300 350 400 450 500 Mass/Charge IR UV RF Voltage 95 Table 3.1 PAH compounds certified to be contained in SRM 1944a. P A H s certified i n S R M 1944a - N e w Y o r k / N e w Jersy R i v e r water sediment M a s s (amu) Component M a s s F rac t ion ( m g / k g ) 128 Naphthalene 1.65 154 Acenaphthene 0.57 178 Phenanthrene 5.27 178 Anthracene 1.77 192 Methylphenantherene 7.88 202 Fluorantherene 8.92 202 Pyrene 9.7 206 Dimethylphenantherene 11.94 228 • Chrysene 4.86 228 Benz(a)anthracene 4.72 252 Benz(x)fluoranthene where x = bj,k,a 9.04 252 . Benzo(x)pyrene where x=a,e 7.58 252 Perylene 1.17 276 Benzo(ghi)perylene 2.84 276 Indeno(l ,2,3-cd)pyrene 2.78 96 3.4 Conclusions R E M P I - I T M S can be effectively used to direct ly obtain mass spectra for P A H s i n soi ls and s imi la r materials. The R E M P I method provides excellent select ivi ty for P A H s coupled w i t h very h igh ion iza t ion eff iciency. The i o n trap was demonstrated to have many advantages over tradit ional T O F instruments w i t h respect to two-laser analysis ; however , a few problems w i t h i o n traps remain. Ion traps typ ica l ly have a lower resolut ion, mass accuracy, and mass range than reflectron T O F ' s . W h i l e the samples examined here were relat ively h igh i n concentration there is no reason to bel ieve that by s imp ly increasing the laser i r radiat ion area or ion iza t ion t ime that sensi t ivi ty o n par w i t h other environmental mass spectrometric methods cou ldn ' t be achieved. F i n a l l y , the reason w h y no M S / M S data was col lected o n any o f the P A H s s h o w n i n this chapter should be noted. P A H s are one o f a very smal l class o f compounds for w h i c h M S / M S is poor at d is t inguishing between isomers because most isomers have re la t ively s imi la r fragmentation patterns. Therefore, future chapters w i l l address these two issues. A demonstrat ion o f the M S / M S capabili t ies o f this two-laser i o n trap system w i l l be shown w i t h an analyt ica l ly relevant sample i n C H A P T E R 5. A l s o , the issue o f d is t inguish ing between P A H isomers w i t h this system w i l l be addressed i n C H A P T E R S 6 and 7. 97 Chapter 4 Desorption Profiles of PAHs from Activated Charcoal 4.1 Introduction D u r i n g the course o f w o r k described i n C H A P T E R 2 and C H A P T E R 3 the quest ion o f the o p t i m u m time delay between the f i r ing o f the desorption laser and the ion iza t ion laser arose. K n o w l e d g e o f the effect o f this delay is important for two reasons. P r i m a r i l y , the determination o f the ideal delay is important for instrument development i n that it helps op t imize the result ing signal . In addi t ion, this data can also be used to provide insight into the desorption process. The study o f desorption o f molecules and atoms f rom surfaces is a re la t ively mature field [106, 113]. H o w e v e r , when this desorption is induced by a laser pulse there are further factors w h i c h must be considered [54]. O w i n g to the great variety o f possible combinat ions o f laser wavelength, temporal widths , and pulse energies, it is diff icul t to describe a l l desorption events w i t h a single comprehensive mode l . A d d i t i o n a l l y , the compos i t i on o f the desorbate matr ix further complicates analysis. F o r example , the electr ical conduct iv i ty , thermal properties, and opt ical cross section a l l p lay a role i n determining the mode o f desorption. In terms o f laser wavelengths, two descript ive regimes are c o m m o n ; either resonant laser desorption by direct v ibra t ional or electronic exci ta t ion o f the adsorbates [114, 115], or exci ta t ion by laser-induced thermal desorption ( L I T D ) (i.e. indirect heating o f the analytes). In general, w h e n the desorption is performed w i t h a pu lsed I R laser (as 98 i n the case o f this w o r k ) a l l adsorbed species, regardless o f structure, are affected due to the intermolecular coup l ing and surface heating effects by adsorbate-surface coupl ings [116]. A s a result o f this non-select ivi ty, L I T D is preferred for routine analysis o f u n k n o w n analytes and matrices [64]. In the literature, numerous mechanisms have been presented to exp la in the observat ion that thermal ly labi le , polar, and non-volat i le compounds can be desorbed wi thout fragmentation by L I T D . Zare and L e v i n e suggested a bottleneck i n the f l o w o f energy f rom a rap id ly heated surface through the surface-adsorbate bond to the internal bonds o f the desorbing molecules [117, 118]. Th i s effect is proposed based o n a mi sma tch i n frequency between internal v ibra t ional modes o f the molecu le and the l o w frequency surface-adsorbate mode. T h i s effect w o u l d most l i k e l y on ly become important for situations w i t h h igh heating rates ( 1 0 1 1 K / s ) and very w e a k l y bound molecules . The bu lk o f the current literature, however , suggests that most L I T D experiments can be expla ined us ing c lass ica l transit ion state theory [57, 73, 119-121]. It is proposed, that, under these circumstances, thermal equ i l ib r ium is maintained, and thermal desorption is i n compet i t ion w i t h thermal decomposi t ion where the relative rates are determined by frequency factors ( A ) and act ivat ion energies ( E A ) . F o r most systems, at moderate temperatures, decomposi t ion dominates over desorption. H o w e v e r , at h i g h temperatures, desorption is expected to be the predominant mode. Therefore, i n situations such as L I T D , where the temperature rises very q u i c k l y , it is postulated that i t is possible to desorb thermally labi le molecules w i t h litt le decomposi t ion . The w o r k i n this chapter was or ig ina l ly performed for pract ical reasons; to determine the ideal t ime between the I R desorption laser and the U V ion iza t ion laser. 99 T h i s was achieved by incrementing the delay t ime between laser shots over many thousands o f col lected mass spectra. T h i s delay data, or t ime-of-f l ight data ( T O F ) as it is often ca l led , provides some insight into the desorption process. It also suggests possible fundamental l imi t s to future work . 4.2 Experimental The experiments described i n this chapter were performed o n the two-laser system discussed i n C H A P T E R 2 and C H A P T E R 3. In this experiment, a s ingle sample was analyzed over several thousand-laser cycles . T h i s sample was prepared as described i n C H A P T E R 3, and contained f ive P A H s ; Acenaphthene (154 amu), Phenanthrene (178 amu), Pyrene (202 amu), Chrysene (228 amu), and Benzo(a)pyrene (252). The delay between the desorption laser ( N d : Y A G @ 1064 nm) and the ion iza t ion laser ( N d : Y A G @ 266 nm) was precisely regimented by control o f each laser 's Q -S W I T C H . T h i s y ie lded excellent t im ing resolut ion w i t h jitters o f less than l p s . T o ensure accurate t i m i n g , each laser shot was recorded us ing the detector/oscil loscope system described i n C H A P T E R 2. A s a result the actual delay between every laser shot was recorded; it was this value w h i c h was used i n further calculat ions. 4.3 Results and Discussion T h e p r imary goal o f this chapter was to determine the ideal t ime delay between the f i r ing o f the desorption laser and ion iza t ion laser. T h i s experiment was performed by first preparing a charcoal standard containing five P A H s . The sample was then exposed to t w o thousand laser shots i n order to reach the "steady s igna l " seen p rev ious ly 100 ( C H A P T E R 3). The U V laser delay was then var ied between - 5 to 400 ps w i t h respect to the I R f i r ing t ime. The delay settings for acquir ing the data were set non-sequential ly i n order to a v o i d systematic effects. A "reference" delay t ime o f 30 ps was also examined several t imes throughout the course o f the experiments i n order to track any s ignal decrease caused by sample degradation. M a s s spectral data were acquired for at least 100 laser shots at each delay setting. The integrated peak areas for each o f the f ive P A H s as a funct ion o f delay t ime between desorption and ion iza t ion is plotted as data points i n F i g u r e 4.1. Several interesting features may be not iced i n the r aw data. The p r imary use o f this data is that it a l lows us to determine the o p t i m u m delay t imes between laser shots for each P A H ( T a b l e 4.1). The average op t imum delay t ime was approximate ly 27 ps between laser shots; however , a smal l mass dependant var ia t ion was observed. F o r example the lightest P A H used, acenaphthylene, had ah o p t i m u m delay t ime o f 25.9 ps w h i l e the heaviest P A H , benzo(a)pyrene had an o p t i m u m delay t ime o f 28.4 ps. Ano the r interesting feature, w h i c h was observed i n a l l five figures, occurred at approximate ly 60 ps. A t this t ime delay, a secondary peak was observed as a shoulder. T h i s structure was most l i k e l y due to the reflect ion o f neutral species f rom the end cap electrodes and back into the center o f the trap. Neut ra l reflections o f this type have been prev ious ly reported i n s imi la r experiments [122]. In addi t ion to the pract ical aspect o f this experiment, this data may be used to ga in some insight into the desorption process. Exper iments o f the type described i n this chapter have recently been used to provide a means o f observing the energy d is t r ibut ion o f molecules desorbing f rom surfaces [56, 58, 71 , 73 , 107, 122-126]. H o w e v e r , these 101 Figure 4.1 Integrated peak areas vs. IR-UV delay time for Acenaphthene, Phenanthrene, Pyrene, Chrysene, and Benzo[a]pyrene. 100 200 Time Delay (usee) Acenaphthene 300 400 50 100 150 200 250 300 350 Time Delay (usee) Phenanthrene 100 200 300 400 Time Delay (usee) Pyrene 102 F i g u r e 4.1 C o n t i n u e d . 400 100 200 300 400 Time Delay (usee) Benzo(a)pyrene Table 4.1 Experimentally determined ideal delay times between laser events. P A H M a s s I d e a l T i m e D e l a y (us) Acenaphthylene 154 25.5 Phenanthrene. ' 178 : ' \ 25.9 Pyrene 202 27.5 Chrysene 228 27.5 Benzo(a)pyrene 254 28.4 103 workers col lec ted this data on instruments designed w i t h this experiment i n m i n d (i.e. desorption events were detected by laser ion iza t ion i n the extraction reg ion o f a t ime o f f l ight M S ) . A s imi la r analysis can be performed here, however , there are several caveats because the data was col lected i n a less than ideal environment. In this experiment, the analyte species are desorbed f rom a charcoal matr ix o n the r ing electrode o f the i o n trap. Once desorption occurs, the neutral molecules are free to traverse the trap vo lume . T y p i c a l angular spreads for desorption o f this type are approximate ly 4 0 ° (so l id angle representing 90 % o f material) [107]. Depend ing o n the sample (concentration o f analyte, matr ix type, etc.) the desorbing species m a y co l l ide w i t h each other. I f this occurs, the co l l i s ions w i l l t yp i ca l ly have a net forward m o m e n t u m and thus increase the ve loc i ty perpendicular to the surface (this effect is termed "stream veloci ty") [127] . In addi t ion to co l l i s ions w i t h ejected neutrals, the analytes m a y also experience co l l i s ions w i t h the H e buffer gas w i t h i n the trap. H o w e v e r , at the pressures used i n this work , the mean free path is o n the order o f 30 c m , so co l l i s ions dur ing the or ig ina l expansion are un l ike ly . Af t e r a set delay t ime the U V i o n i z i n g laser is passed through the i o n trap a long the asymptote between the end caps and the r ing electrode. O b v i o u s l y , on ly neutrals that are i n the laser path dur ing the 10 ns w i n d o w i n w h i c h the laser fires can be ion ized . Once the analytes are ion ized , the question o f i o n trapping becomes important. The t rapping eff ic iency depends on several factors i nc lud ing the i on ' s k ine t ic energy,.angular ve loc i ty , and the R F phase angle when ion iza t ion occurs [128-130]. S ince 100 averages were acquired at each t ime delay, we can assume that the gains and losses associated w i t h phase angle w i l l balance each other out at a l l locat ions except near 104 the center o f the trap. Therefore, we w i l l make the s imp l i fy ing assumption that most o f the observed s ignal results from ions created at the center o f the i o n trap; 10 m m f rom the probe surface. The second issue related to i o n trapping concerns possible d i sc r imina t ion based o n k inet ic energy. Trapp ing efficiency is greater at the center o f the trap for ions w i t h lower k ine t ic energy [74]. Therefore, one might expect an observed s ignal d i sc r imina t ion as a funct ion o f k inet ic energies (i.e. a reduced s ignal at short delay t imes). O v e r the typ ica l range o f k inet ic energies observed here, between 0.01 to 5 e V , this is most l i k e l y not important. Howeve r , i n the worst possible case (i.e. fast, heavy ions) , such as w h e n benzo(a)pyrene arrives i n the center o f the trap i n 1 ps or less, the i o n w o u l d have a k ine t ic energy o f approximately 130 e V and w o u l d most l i k e l y not be trapped. W i t h the above caveats relating to the l imitat ions o f the col lected data i n place w e m a y beg in to analyze the desorption profiles. W h i l e a variety o f experimental data and theories relat ing to laser induced desorption have been prev ious ly observed, w i t h i n the last five years many groups have settled o n a c o m m o n mechan i sm for L I T D . These researchers have proposed that laser desorption o f analytes f rom a surface is s i m p l y a thermal process [57, 122]. In fact, Zare et a l . have recently shown that i n the case o f ani l ine desorbing f rom a sapphire surface, the analytes were i n almost complete translat ional , v ibra t ional , and electronic thermal equ i l i b r ium w i t h the surface [122]. In order to proceed w i t h a thermal analysis o f the desorption event a few addi t ional s i m p l i f y i n g assumptions must be made: (1) the laser induces a rapid heating o f the sample o n the order o f 10 7 - 10 9 K / s [111]. Th i s heating is fast enough that the desorpt ion pathway can become favorable i n preference to the decompos i t ion process 105 [131]. The sample then cools almost as rapidly as it is heated. S ince desorption is on ly statist ically favorable w h i l e the temperature is elevated, i n essence, this impl ies that a l l desorpt ion events are almost concurrent w i t h the laser shot [132, 133]. (2) The rma l e q u i l i b r i u m is mainta ined by the molecules o n the so l id dur ing the process o f laser heating [125]. (3) F i n a l l y , analytes desorb w i t h mic roscop ic revers ib i l i ty , meaning that the translational energy w i t h w h i c h the molecules desorb is an accurate representation o f the sample surface [122]. W i t h these assumptions i n place, w e may beg in to examine the T O F profi les for the P A H analytes. I f the analytes desorb i n equ i l i b r ium w i t h the surface, and i f w e assume that ion iza t ion and efficient capture o f ions on ly occurs at the center o f the i o n trap, then we w o u l d expect to see a M a x w e l l - B o l t z m a n n ( M - B ) l ike d is t r ibut ion o f the desorbing analytes w i t h a translational temperature representative o f the surface temperature w h e n desorption occurred. S ince we k n o w the f l ight t ime and the distance each analyte travels, we cou ld , i n pr inc ip le , turn each T O F dis t r ibut ion into a k ine t ic energy prof i le (as is seen i n most M - B distributions). Instead, we w i l l re-write the M - B equation to vary signal. intensity (gas phase density) as a funct ion o f t ime instead o f energy. T h i s manipu la t ion has been performed prev ious ly by Assche r et a l . [134], and used by Zare et a l . [122]. The manipulated equation has the fo rm be low: Where C is a constant, t is the t ime delay between laser pulses, m is the mass o f the desorbate molecu le , k is the B o l t z m a n n constant, T is the temperature, v s is the stream ve loc i ty term, d is the distance between desorption and ion iza t ion events, and g(t) the V m(d - tvs) 2kTt2 J Equ. 4.1 106 observed s ignal . The curve-fi t t ing program M i c r o c a l O r i g i n ( M i c r o c a l Software Inc., Nor thhampton , M A , U S A ) was used to best fit the M - B distributions. These fits are s h o w n as the s o l i d l ines on F i g u r e 4.1. Th i s data was w e l l fit wi thout the need for the "stream v e l o c i t y " term indica t ing that few co l l i s ions between analytes occurred after desorpt ion and before ion iza t ion . The M - B temperature for each o f these P A H s as found by the curve-f i t t ing program is shown i n T a b l e 4.2. The M - B fits w o r k very w e l l for a l l analytes but one. The lightest P A H used, acenaphthene, tended to have an unusual ly h igh signal for the ta i l ing edge o f the t ime dis t r ibut ion. T h i s is most l i k e l y due to the fact that acenaphthene, the smallest P A H used, is the P A H w h i c h is easiest to desorb. One may assume, therefore, that dur ing the c o o l i n g per iod after the laser pulse, acenaphthene continued to be desorbed at some s m a l l rate. In fact, samples w h i c h contained the s l ight ly smaller P A H , naphthalene, c o u l d not be used i n pract ical applications because the naphthalene desorbed f rom the activated charcoal under the reduced pressure o f the i o n trap, and large signals were observed by laser ion iza t ion wi thout any need for desorption. F o l l o w i n g this l og ica l argument, that smaller P A H s are easier to desorb, then one w o u l d expect that larger P A H s , hav ing greater mass and more intermolecular forces, w o u l d be harder to desorb. In this context, we w o u l d not expect the heavier P A H s to desorb appreciably unt i l the surface temperature was h igh enough to make this process statist ically favorable. B u t since the analytes desorb i n thermal equ i l ib r ium, one m a y expect this higher temperature weight ing to be reflected i n the translational energy profi les (i.e. the T O F distr ibutions). In fact, as shown i n F i g u r e 4.2, a plot o f the M - B translational temperature vs. P A H mass is linear. Th i s graph can prov ide insight into 107 Table 4.2 Table of PAH mass and experimentally determined Maxwell-Boltzmann temperature for laser desorption from activated charcoal. P A H M a s s T e m p e r a t u r e ( K ) Acenaphthylene 154 686 7 - 2 4 Phenanthrene 178 765 7 - 23 Pyrene 202 817 7 - 25 Chrysene 228 923 7 - 3 0 Benzo(a)pyrene 254 963 7 - 39 Figure 4.2 Chart of experimentally determined Maxwell-Boltzmann temperature vs. PAH mass. 100 150 200 Molar Mass 250 300 108 in termolecular interactions and may provide a clue as to the ult imate mass l i m i t for desorption. It should be noted that a plot o f mass vs. temperature is l inear i n this case because the analytes are a l l part o f a homologous series. I f the analytes were w i l d l y different, then the graph w o u l d most l i k e l y not be linear. In this case, however , the graph points the w a y towards what cou ld be a mass l i m i t for desorption w i t h respect to the temperature elevat ion induced by the I R laser. Presumably , there are some P A H s for w h i c h the intermolecular bonding is greater than the energy p rov ided by the laser heating. In these situations, two-laser mass spectrometry may not be applicable. 4.4 Conclusions The goal o f the w o r k i n this chapter was to determine the o p t i m u m t ime delay between desorpt ion and ion iza t ion events i n the two laser system. The experimental data presented here suggests that the o p t i m u m time delay is approximately 27 ps. In addi t ion to the pract ical aspects o f these experiments, this data can p rov ide some insight i n to the desorption process o f P A H s f rom an activated charcoal surface. T h i s data, w h i l e severely l imi t ed by experimental constraints suggests that the desorpt ion process induced by a 10 ns pulse N d : Y A G laser is purely thermal i n nature. T h i s is consistent w i t h results observed by Zare [122] and more recently by Z e n o b i [57, 58, 73 , 111, 131]. Furthermore, there appears to be a l inear relat ionship between the mola r mass o f the P A H and the M - B based translational temperature. T h i s type o f l inear relat ionship has been p rev ious ly observed for a series o f polyethylene g l y c o l units, and is most l i k e l y due to the var ia t ion i n desorption rates as a function o f temperature for the homologous series [73]. 109 F i n a l l y , this type o f experiment may prove useful i n the f ie ld o f atmospheric science, where P A H s are o f great interest. Spec i f ica l ly , the knowledge o f P A H adsorpt ion and desorption rates f rom graphite and soot partials are a key to many p o l l u t i o n models [135-137]. The laser desorption data p rov ided by experiments s imi l a r to that described here may provide a means o f accessing these important constants. 110 Chapter 5 Detection of the Drug Spiperone on Biological Matrices 5.1 Introduction D u r i n g the process o f pharmaceutical research and development there are several key chemica l attributes that must be considered when selecting lead compounds for further study. These include receptor/drug specif ici ty, potency, potential tox ic i ty at therapeutic concentrations, and f inal ly , the issue o f transport and ava i lab i l i ty o f the agent at the site o f action. T y p i c a l l y , i n vi t ro cel l-based experiments are used to examine the b i n d i n g efficiencies o f a drug w i t h a receptor site. W h i l e these studies provide a good ind ica t ion o f drug act ivi ty, little" knowledge is gained concerning transport and accumula t ion . S i m i l a r l y , metabol ic p rof i l ing o f body fluids performed us ing standard analyt ica l techniques ( N M R , M S , etc.) produces informat ion o n b ioava i l ab i l i ty o f specif ic drugs, but y ie lds litt le informat ion regarding the transport o f the drug into tissues i n the intracellular/extracellular spaces. T o address the important issue o f whether a drug reaches a site o f act ion a variety o f sensitive microprobe techniques have been developed [138-141]. Indeed, the mapping o f targeted compounds i n tissues has many appl icat ions beyond just those i n drug discovery. F o r example, altered chemica l distr ibutions are diagnostic for diseases such as stroke [142], cancer, and A l z h e i m e r ' s disease [143]. Furthermore, microprobe techniques have also found use i n the detection o f trace metals i n in f lamed tissues surrounding replacement jo ints [144]. I l l The term microprobe s imp ly defines a type o f instrument that provides spatial analyt ica l in format ion concerning so l id samples o n the mic roscop ic scale. T y p i c a l l y these devices combine an x - y stage for rastering a two-dimens iona l sample and a detection element used to record the analyt ical information. These methods have been performed w i t h a variety o f analyt ical schemes inc lud ing opt ical methods and electron scattering techniques [145]. B y far, the most c o m m o n technique used to date invo lves secondary i o n mass spectrometry [146, 147]. One o f the most powerfu l combinat ions o f these analysis techniques has been the advent o f the laser/mass spectrometer based microprobe [138, 140, 148-150]. The laser permits excel lent control o f the sample spot size and input energy w h i l e the mass spectrometer provides very sensitive detection w i t h informat ion o n molecula r compos i t ion . V i r t u a l l y , a l l types o f mass spectrometer have been descr ibed as detectors for laser microprobes . B y far , they most c o m m o n l y ut i l ize t ime-of-f l ight systems [151, 152] however , successful devices have also been reported w i t h F T - I C R [149, 153, 154], and tr iple quadrupole systems [155]. The usual pros and cons apply to the applicat ions o f the var ious mass spectrometric methods to this problem. F o r example typ ica l T O F devices provide excellent mass range and sensit ivity, but have no abi l i ty to perform M S / M S - something that can be very useful i n analysis o f complex samples l ike tissues. S i m i l a r l y , F T - I C R s are superb at M S / M S and have excellent resolut ion, but are expensive. T o date, the most undeveloped method i n this f ie ld invo lves the combina t ion o f the i o n trap mass spectrometer w i t h laser ioniza t ion . In terms o f the laser-sampling step, these methods almost a l l exc lus ive ly use M A L D I as the means o f ion iza t ion [62, 63]. The M A L D I technique is k n o w n to be a 112 very soft ion iza t ion source and excellent at vo l a t i l i z i ng / i on i z ing large molecula r weight compounds . H o w e v e r , the M A L D I method is l imi ted by the fact that the tissue surface must first be chemica l ly modi f i ed by the addi t ion o f the matr ix material . A d d i t i o n a l l y , once the matr ix is appl ied, the sample surface often becomes opaque so correla t ion w i t h v i s i b l e surface features becomes diff icul t . A s an alternative to the M A L D I method, straight laser desorption o f analytes f rom b io log ica l surfaces has been attempted. W h i l e this technique is useful i n the analysis o f transit ion metal ions [148], the ion iza t ion eff ic iency for most organics is very l o w [139]. T o address this issue, the Y o s t group at the Un ive r s i t y o f F l o r i d a , has constructed a microprobe system based o n laser desorption fo l lowed by chemica l ion iza t ion for use i n an i o n trap [156]. T h i s method has proved useful i n the detection o f pharmaceut ical agents f rom b i o l o g i c a l materials. A s an alternative to chemica l ion iza t ion , this chapter w i l l focus o n the use o f the two-laser mode o f ion iza t ion for the analysis o f the drug Spiperone on b i o l o g i c a l samples direct ly i n the vo lume o f the i o n trap. The pharmaceut ical compound Spiperone was in i t i a l ly developed as an ant ipsychotic drug. It is part o f a class o f compounds k n o w n as azipirones, w h i c h are s imi la r i n structure to serotonin, and so b i n d to 5 - H T receptors i n the central nervous system [157]. It has been shown through i n v i v o studies, that Spiperone not on ly binds to the 5 - H T I A receptors, but also 5-HT2, and dopamine2 receptors [158]. S ince these receptors are at var ious densities throughout the brain, there is considerable interest i n determining the concentration o f this drug i n different cerebral regions. T h i s chapter describes the use o f the two-laser i o n trap system for the detection o f the drug Spiperone di rect ly on a b io log ica l matr ix . Th i s study had two goals. P r i m a r i l y , 113 the detection o f pharmaceutical compounds on complex matrices is an important p r o b l e m i n analyt ical chemistry, and one that is w e l l addressed us ing the two-laser method. The gentle desorption process combined w i t h specific ion iza t ion and the M S / M S capabil i t ies o f the i o n trap are a perfect solut ion to the p rob lem o f detection o f analytes f rom a compl ica ted s o l i d matr ix . The second role o f this chapter was to demonstrate the potential scope o f the instrument and methodology described i n this thesis. C H A P T E R 3 described the use o f the two-laser system for environmental samples. Howeve r , because the analytes used were not compat ib le w i t h M S / M S , a fu l l demonstration o f the potential was not possible . In the case o f the Spiperone detection o n a b io log i ca l matr ix , however , the use o f M S / M S is essential for complete characterization. 5.2 Experimental The experiments described i n this chapter were performed o n the home bui l t two-laser i o n trap mass spectrometer system described i n C H A P T E R 2 and C H A P T E R 3. The basic electronic and opt ical systems were conserved. H o w e v e r the var ious waveforms used for i o n manipula t ion were changed as required. A s has already been ment ioned the pharmaceutical compound used for the study descr ibed i n this chapter was Spiperone ( S i g m a C h e m i c a l , S t .Lou i s , M o , U S A ) . S l i gh t l y ac id ic ( 1 % acetic ac id by vo lume) aqueous solutions o f Spiperone were prepared as spikes for the so l id samples. Standard charcoal standards o f Spiperone were prepared as described i n C H A P T E R 3. 114 T w o types o f b io log i ca l tissues were examined i n this study. The samples used i n this w o r k were f rom the bra in and l iver o f male Sprague D a w l e y rats (225-250 grams). The animals were p rov ided by G r e g B o w d e n o f the U B C department o f Neurosc ience through the U B C A n i m a l Care Center. The animals were sacrif iced by a s l o w increase i n CO2 levels i n their enclosures. The l iver and bra in o f four rats were r emoved and immedia te ly flash frozen i n - 3 0 ° C 2-methyl-butane. The frozen tissue was then stored i n a - 7 0 ° C freezer. F i g u r e 5.1 and F i g u r e 5.2 show representative samples o f the b ra in and l ive r tissues as col lected by the author. Samples were prepared by first s l i c ing a 2 m m thick piece o f f rozen tissue. The tissue was then sonicated i n the Spiperone solut ion (~0.7g Spiperone/kg frozen tissue) for 15 minutes. The solutions were then poured off, and a tissue p l u g was inserted into the probe t ip. The sample was then a l lowed to dry i n a fume hood for 1 hour pr ior to inser t ion into the i o n trap. 5.3 Results and Discussion The pr imary goal o f this chapter was to investigate whether the two-laser method combined w i t h an i o n trap mass spectrometer possess enough select ivi ty and specif ic i ty to observe a pharmaceut ical compound i n the presence o f a real b i o l o g i c a l matr ix . H o w e v e r , before this c o u l d be achieved, the first task was s i m p l y to observe a two-laser mass spectrum o f Spiperone on the standard charcoal matr ix . 115 Figure 5.1 Photo of brain tissue extracted from a male Sprague Dawley rat used in this work. 116 A two laser mass spectrum o f a charcoal based Spiperone sample was col lec ted and is s h o w n i n Figure 5.3. The spectrum shows the molecular peak o f Spiperone at 395 T h , i n addi t ion to a smal l fragment peak at 340 T h . The peaks observed b e l o w 330 T h were caused by l inger ing contaminat ion i n the i o n trap f rom the mult i tude o f previous experiments us ing P A H s . Unfortunately, the P A H s and other organic contaminants released f rom the pump o i l were a persistent p rob lem throughout this project. T h i s contaminat ion speaks to two w e l l - k n o w n facts about this type o f w o r k ; (1) P A H s are k n o w n to be " s t i c k y " and are excellent at contaminat ing mass spectrometer . systems, and (2) the two-laser method is extraordinari ly efficient at i o n i z i n g P A H s . O f course routine c leaning was performed to help m i n i m i z e this p rob lem. Unfor tunate ly , the ideal so lu t ion to this contaminat ion problem, w h i c h involves heating the v a c u u m system, was unavai lable due to problems i n the v a c u u m chamber design. Regardless o f the contaminat ion problem,-Figure 5.3 does indeed demonstrate that Spiperone can be observed by the two-laser method. H o w e v e r , i n cases where the sample concentrat ion is smal l , it w o u l d be preferable to increase the observed s ignal by pre-concentrat ing the i o n popula t ion i n the gas phase. Figure 5.4 demonstrates the effect o f 1, 5, and 10 laser cycles fo l lowed by a N B B W pulse on the observed Spiperone s ignal . In this case the N B B W pulse was appl ied at the end o f the laser cycles rather than between cycles . The experiments were performed i n this w a y because it was emp i r i ca l l y determined that the net improvement was greater w i t h a single N B B W cyc le rather than mul t ip le N B B W cycles . T h i s is most l i k e l y because Spiperone is a re la t ively fragile molecu le , and even a w i d e N B B W w i n d o w caused a degree o f s ignal loss. 117 Figure 5.3 Two laser mass spectrum of Spiperone on a charcoal matrix. RF Voltage 118 Figure 5.4 Two laser mass spectrum of Spiperone on a charcoal matrix with 1, 5, and 10 laser cycles followed by a single NBBW pulse. 10 laser cycles 5 laser cycles 1 laser cycle 375 385 395 405 415 Mass/Charge IR UV IR UV IR UV IR UV IR UV mm # RF Voltage 119 Once a large Spiperone i o n popula t ion was acquired and isolated, the next stage was to col lec t a M S / M S spectrum. Figure 5.5 demonstrates the result o f a c o l l i s i o n induced d issocia t ion ( C I D ) cyc le after N B B W isolat ion. There are f ive p r imary daughter ions that result f rom C I D o f the Spiperone sample. These occur at 280 T h , 261 T h , 165 T h , 146 T h , and 109 T h . One possible fragmentation is shown i n Figure 5.5. It should also be noted that two pairs o f peaks are observed (280 T h , 261 T h and 165 Th ,146 T h ) w h i c h are both separated by 19 mass units. These most l i k e l y result f rom the loss o f f luorine dur ing fragmentation. Th i s suggests that the charge center is most l i k e l y o n the aromatic r ing connected to the fluorine. The d i f f icul ty i n fu l ly assigning the daughter spectrum is typ ica l o f c o m p l e x molecules o f this type. D u r i n g the C I D process a number o f unusual fragmentation pathways are possible [159]. A d d i t i o n a l l y , rearrangement reactions f o l l o w e d by e l iminat ions are often observed; this further complicates fragment assignments [159]. Regardless o f any missed assignments, this M S / M S spectrum can be used as a useful diagnost ic too l for conf i rming the presence o f Spiperone i n a compl ica ted matr ix . The next step i n the method development was the analysis o f b lank b i o l o g i c a l tissues. Figure 5.6 and Figure 5.7 show typ ica l 2-laser mass spectra o f rat bra in and l iver tissues. These samples represent very complex matrices w i t h an abundance o f molecula r divers i ty . It is important to note that when a l iver sample was examined by the method o f laser desorpt ion/chemical ion iza t ion a peak was observed at v i r tua l ly every mass [156]. In contrast, the selectivi ty prov ided by the U V ion iza t ion scheme used here produces at least some re l i e f f rom the over abundance o f matr ix ions. ...120 Figure 5.5 Two laser mass spectrum of Spiperone on a charcoal matrix with five laser cycles followed by a single NBBW pulse and a collision induced dissociation (CID) waveform. 50 100 150 200 250 . 300 350 400 450 Mass/Charge RF Voltage 121 Figure 5.6 Two laser mass spectrum of a slice of brain tissue from a male Sprague-Dawley rat. RF Voltage 122 Figure 5.7 Two laser mass spectrum of a slice of liver tissue from a male Sprague-Dawley rat. F i n a l l y , samples o f bra in and l iver w h i c h had been sonicated i n a Spiperone so lu t ion o f ~0 .7mg/kg tissue were examined. Figure 5.8 represents the observed spectra o f a rat b ra in sample sp iked w i t h Spiperone col lected by two-laser mass spectrometry. Figure 5.9 shows the same sample after i o n accumula t ion and i so la t ion w i t h a N B B W pulse. F i n a l l y , Figure 5.10 demonstrates the result ing M S / M S spectra f rom the peak at 395 T h . S i m i l a r l y , data col lected for a sp iked l iver sample are s h o w n i n Figure 5.11, Figure 5.12, and Figure 5.13. Figure 5.8 and Figure 5.11 both show a smal l peak i n the mass spectrum resul t ing f rom Spiperone at 395 T h , however , is w o u l d be dif f icul t to con f i rm the analytes presence us ing this data alone. B y the appl ica t ion o f mul t ip le laser cycles f o l l o w e d by a single N B B W pulse a large ion popula t ion at 395 T h c o u l d be acquired. F i n a l l y , C I D o f the 395 T h peak i n Figure 5.10 and Figure 5.13 conf i rms the presence o f Spiperone i n the tissue. 5.4 Conclusions The goal o f this chapter was to demonstrate the unique capabil i t ies o f the t w o -laser system w i t h an ion-trap mass spectrometer. T h i s was achieved by co l l ec t ing spectra and ident i fy ing the molecu la r ly fragile pharmaceutical compound Spiperone on one o f the most complex matrices possible, b io log ica l tissue. B y way o f compar i son , this analysis w o u l d have been imposs ib le w i t h the tradit ional two-laser T O F set-up, as a T O F device c o u l d neither precoricentrate ions nor col lect M S / M S data. 124 Figure 5.8 Two laser mass spectrum of a slice of brain tissue from a male Sprague-Dawley rat which had been spiked with a Spiperone solution. 395 Mass/Charge R UV RF Voltage 125 Figure 5.9 Two laser mass spectrum with 5 laser cycles followed by a single NBBW pulse of a slice of brain tissue from a male Sprague-Dawley rat that had been spiked with a Spiperone solution. 50 100 150 200 250 300 Mass/Charge 350 400 450 IR UV IR UV IR UV IR UV IR UV RF Voltage* 126 Figure 5.10 Two laser mass spectrum with 5 laser cycles followed by a single NBBW pulse and a single CID waveform of a slice of brain tissue from a male Sprague-Dawley rat which had been spiked with a Spiperone solution. 50 100 150 200 250 300 Mass/Charge 350 400 450 127 Figure 5.11 Two laser mass spectrum of a slice of liver tissue from a male Sprague-Dawley rat which had been spiked with a Spiperone solution. Figure 5.12 Two laser mass spectrum with 5 laser cycles followed by a single NBBW pulse of a slice of liver tissue from a male Sprague-Dawley rat that had been spiked with a Spiperone solution. 200 250 300 Mass/Charge 450 IR UV IR UV IR UV IR UV IR UV RF Voltage 129 Figure 5.13 Two laser mass spectrum with 5 laser cycles followed by a single NBBW pulse and a single CID waveform of a slice of liver tissue from a male Sprague-Dawley rat which had been spiked with a Spiperone solution. 50 100 150 200 250 300 Mass/Charge 350 400 450 130; Furthermore, the data col lected here, wh i l e useful as an instrument characterizat ion too l , may also have some potential use i n real pharmaceut ical appl icat ions. The drug examined here, Spiperone, for example is o f current interest i n the f ie ld o f neurobio logy [158]. Interestingly, the Spiperone sample concentrations used i n this p re l imina ry w o r k were not totally out o f l ine w i t h those o f pharmacolog ica l studies. F o r example i n this chapter samples were created w i t h 0 .7mg Spiperone/kg o f tissue, w h i l e a recent report observed effects on discharge rates o f seven medul la ry 5 - H T cel ls i n a cat w i t h intra venous doses o f Spiperone at levels o f l m g drug/kg o f an imal [157]. O f course, to be a true microprobe technique the method should be able to scan over a range o f tissue locations. W i t h the current set-up this cou ld on ly be achieved by co l lec t ing a series o f sample plugs for sequential analysis. H o w e v e r , a future device c o u l d perform true m i c r o scanning by p lac ing the sample outside the i o n trap and r e l y i n g on i o n optics to transport the analyte ions. Th i s w o u l d , o f course, come at the expense o f reduced sensi t ivi ty, so a trade o f f w o u l d have to be made. In addi t ion to l o o k i n g at tissue samples, these techniques m a y also prove invaluable for the direct analysis o f bacteria. There has been great interest i n recent years i n the f i e ld o f rapid bacterial detection, speci f ica l ly as it relates to potential cases o f tox ic i ty (i.e. bioterrorism). The two-laser method combined w i t h an i o n trap may be useful as a rap id probe because o f its abi l i ty to provide relat ively selective ion iza t ion combined w i t h the abi l i ty to col lect M S / M S data. 131 Chapter 6 Optical Spectroscopy in an Ion Trap 6.1 Introduction T o this point i n the thesis, the goal o f the w o r k was to investigate var ious facets o f two-laser i o n trap mass spectrometry. Th i s chapter describes the addi t ion o f a th i rd laser to the system i n an attempt to broaden the already wide capabil i t ies o f the device . Pure mass spectrometry alone w i l l a lways be l imi ted i n its abi l i t ies by the fact that it provides li t t le structural informat ion directly. M u l t i p l e stages o f M S ( M S 1 1 ) m a y be used to address this p rob lem. Howeve r , direct gas phase molecula r structures are s t i l l d i f f icul t to determine. R E M P I ion iza t ion described i n this thesis and as appl ied b y m a n y others helps address this p roblem; by us ing a tunable laser source and moni to r ing the appearance o f the resul t ing i o n s ignal a large amount o f spectroscopic (and structural) informat ion about the parent molecule can be gained [69, 70]. It turns out, however , that w h e n ion iza t ion occurs direct ly after the laser desorption process, v i r tua l ly a l l o f the spectroscopic informat ion is lost due to the large internal temperature o f the analytes [96]. A s a solut ion to this persistent p rob lem, most workers n o w use a supersonic beam to c o o l analyte molecules after desorption. U s i n g this c o o l i n g technique, comprehensive spectral l ibraries have been acquired over the last thirty years [65, 160]. There remain , however , some analytes that have broad spectral features even under jet condi t ions either because their spectra are diffuse and un-resolvable, or because molecu la r species become increasingly diff icul t to c o o l as they increase i n s ize [68, 161, 162]. A s an alternative to re ly ing o n the spectroscopic properties o f the R E M P I process 132 to p rov ide structural informat ion and selectivity, the poss ib i l i ty o f p rob ing the molecula r ions w i t h a th i rd pulsed laser has been investigated. The analytes that were selected for this study, the P A H s , were chosen la rge ly because they are o f great interest i n two fields o f science. A s discussed i n C H A P T E R 1, P A H s are o f interest to environmental scientists because they are k n o w n carcinogens, and are ub iqui tous ly found throughout the environment [79, 80]. P A H s , and spec i f ica l ly P A H cations, have also experienced m u c h interest by the astrophysical commun i ty . T h i s is due to the current speculat ion that P A H s are expected to represent a large por t ion o f the carbon present i n interstellar space [163, 164]. Spect roscopica l ly , wh i l e neutral P A H s absorb l ight i n the U V , their cations exhibi t absorption bands i n the v i s ib le and near I R range. Ast rophys ic is t s are interested i n these part icular features as they make P A H s good candidates as a poss ible source o f the diffuse interstellar bands ( D I B s ) [165]. The D I B s consist o f a large number o f absorption l ines superimposed on the interstellar ext inct ion curve [166]. S ince their d i scovery i n the 1920 's [167], the identity o f these D I B carriers has remained an important and diff icul t p rob lem i n astronomy [168]. These ~ 300 absorption features are generally correlated w i t h dust extinction, and are presumed to relate to interstellar dust c louds . T h e detection o f substructure i n some D I B s has lead to the conc lus ion that molecules or ions m a y be the source o f some o f these bands [169]. The D I B p rob lem is made more diff icul t by the fact that as t ronomical surveys over large wavelength ranges and directions i n space suggest that a l l measured D I B s originate f rom different carriers (i.e. there are no two locations i n space where the ratios o f the peak heights are identical) . Howeve r , some D I B s show s imi la r behavior i n 133 different as t ronomical environments and may arise f rom structurally related species [170]. D u e to the h igh U V f lux and long mean free path i n interstellar space, it has been suggested that P A H cations may be the source o f many o f these D I B s . A s a result, many groups have sought to ga in an understanding o f the phys ica l chemistry o f gas phase P A H cations. T h e reactivi ty and photo stability o f these molecules and cations has been studied extensively by photo fragmentation studies experimental ly f rom the V U V [171], to the U V w i t h lamp irradiat ion [172], synchrotron radiat ion [173], a l l the w a y into the I R w i t h mul t ipho ton dissociat ion [174]. The photo stability o f these cations has also been investigated theoretically [173, 175, 176]. V i r t u a l l y a l l cat ion spectroscopic techniques have been appl ied to the P A H s . The most c o m m o n o f a l l o f these (by far) involves co l lec t ion o f the ca t ion spectra i n a rare-gas so l id matr ix [177]. T h i s method involves co-deposi t ing analyte ions w i t h a noble gas onto an ul t ra-cold w i n d o w . A n opt ical absorption spectrum is then acquired i n the no rma l w a y using a lamp source and a monochromator . T h i s technique, however , has many problems; it is diff icul t ( i f not impossible) to identify exact ly the nature o f the ions deposited o n the surface. F o r example fragments f rom the ion iza t ion process can also be deposited. Furthermore, the noble gas matr ix tends to induce peak loca t ion shifts depending o n the method o f matr ix formation and the phys ica l loca t ion o f the analyte i n the matr ix . At tempts are made to account for this by c y c l i n g through a series o f different matrices ( A r , N e , H e ) and t ry ing to predict the loca t ion o f a matr ix free peak. W h i l e this method is not totally satisfactory (the peak locations i n free space result on ly f rom 134 calculat ions) , due to its ease o f use, it has become very routine and a c o m m o n source o f astrophysical data. A s an alternative to this method, a number o f gas phase methods have been attempted. Recent advances include us ing cavi ty r ing -down spectroscopy o n molecu la r beams [178] and a rare-gas Complex photodissociat ion complex technique [179, 180]. Recen t ly it has even been suggested that is may be possible to observe florescence direct ly f rom large gas phase cations [181].. One possible method o f obtaining gas phase spectra, w h i c h has largely been ignored for P A H cat ion analysis, is the method o f photodissocia t ion ( P D ) spectroscopy. V i s i b l e photodissociat ion studies have been performed i n the gas phase since the late 1970's us ing a variety o f techniques [182]. In fact, opt ical p robing o f trapped gas phase ions was one o f the or ig ina l uses o f the i o n trap [11]. M o s t o f the recent w o r k i n this f ie ld has focused o n observing weak ly -bound complexes where the absorption o f a single photon induces dissociat ion. T h i s w o r k has p r imar i ly focused o n clusters o f sma l l gaseous molecules or molecular-rare gas and metal - l igand complexes where non-vola t i le species can be vapor ized by laser ablat ion and the gaseous l igand introduced through a supersonic beam[183]. Recent ly , even electrospray has been used as a source for the study o f F e + 3 complexes by P D spectroscopy [184]. The technique, as used by the most p ro l i f i c worke r i n this f ie ld , R . Dunbar , invo lves performing P D spectroscopy o f var ious molecula r gaseous ions introduced into an I C R trap w i t h ion iza t ion by electron impact , and then us ing a tunable dye laser for dissociat ion [182, 185]. The advantage o f these traps is the low-pressure condit ions a l l o w for v i r tua l ly col l i s ionless studies - for example t ime dependant energy relaxat ion experiments. I C R was also used by the group o f 135' B o i s s e l to trap laser desorbed and ion ized P A H cations, us ing the broadband emis s ion o f a X e lamp for d issocia t ion [105, 186-189]. B y compar ing a statistical mode l w i t h the results f rom mul t ipho ton absorption induced fragmentation o f the isolated ions they were able to obtain osci l la tor strengths o f the v i s ib le absorption processes. A d d i t i o n a l l y , P D spectroscopic studies have also been performed on molecular ions i n an i o n trap both w i t h l amp sources and tunable lasers [12]. T h i s chapter demonstrates the use o f an R P i o n trap for the co l l ec t ion o f v i s ib l e absorpt ion spectra o f large gas phase cations o f non-volat i le organics through Resonance Enhanced M u l t i p h o t o n Dis soc ia t ion ( R E M P D ) . The addi t ion o f a th i rd tunable laser to the i o n trap system a l lows for the poss ib i l i ty o f performing spectroscopy o n the trapped cations. The spectra were observed by scanning the laser wavelength and observing the fragmentation products that result f rom the trapped ions. A s w i t h R E M P I , the R E M P D process is selective based on resonant absorption o f a single photon by the trapped i o n , f o l l o w e d by several successive non-resonant photon absorptions that lead to fragmentation. The bu lk o f the material presented i n this chapter has been p rev ious ly publ i shed by R o l l a n d , Specht, Blades , and Hepburn , i n Chemical Physics Letters [190]. 6.2 Experimental The experiments described i n this chapter used mod i f i ed equipment based o n the two laser system described i n C H A P T E R 2. The pr imary addi t ion to this system was a th i rd laser; a Spectra Phys ics P D L - 3 D y e Laser (Spectra Phys ics , M o u n t a i n V i e w , C A , U S A ) pumped by a Quante l Y G 6 6 0 (Quantel Lasers , L e s U l i s , France) N d : Y A G . In addi t ion to this dye laser, the electronics required to define the t i m i n g regimes were also added. 136 These lasers were added to the opt ical table set up as shown i n Figure 6.1 and Figure 6.2. The output f rom the doubled N d : Y A G ( Y G 6 6 0 ) beam was first spli t us ing a die lectr ic coated mi r ro r that reflected o n l y the 532 n m l ine w h i l e a l l o w i n g the fundamental ( @ 1064 nm) to pass through into a beam dump. T h i s 532 n m beam was then directed into the dye laser by a 90° quartz p r i sm. The output f rom the dye laser was passed into the v a c u u m man i fo ld by three quartz pr isms ( M e l l e s Gr io t ) after be ing shaped by a telescope external to the dye laser (two lenses suppl ied by M e l l e s Gr io t ) . T h e dye laser beam entered the v a c u u m chamber through a quartz w i n d o w at the top o f the mani fo ld . It then interacted w i t h the i o n c loud v o l u m e through a 2.0 m m diameter hole d r i l l ed i n the r ing electrodes 90° f rom the sample probe. T h i s is schemat ical ly represented i n Figure 6.3. T y p i c a l energies used were o n the order o f 1 mJ/pulse i n the v i s ib l e and near I R ranges. A s i n C H A P T E R 2, the laser power was measured for every cyc le by p l ac ing a pyroelectr ic power meter in- l ine w i t h a reflection o f the dye laser. The s ignal f rom this meter was recorded and stored o n d isk w i t h each mass spectrum us ing the osc i l loscope system detailed i n C H A P T E R 2. A power meter was used i n this w o r k rather than the photodiodes used earlier because the photodiode response is related to wavelength ; w h i c h o f course becomes a p rob lem when using a tunable dye laser. S ince the i o n trap operates on a 1 H z cyc le , and the N d : Y A G / D y e Laser can operate at 20 H z , it was possible to accumulate the products o f a number o f photo fragmentation laser shots, thus increasing the fragmentation y i e l d . Exper iments , therefore, were performed w i t h f ive or more photo fragmentation laser shots, obtained by 137 Figure 6.1 Diagram of the three-laser set-up at UBC. 138 Figure 6.2 Enhanced photo of the three-laser set-up at UBC. 139 Figure 6.3 Diagram of the IR desorption, UV photoionization, and visible photo fragmentation lasers interacting with the interior of the ion trap. 140 act ive ly t r iggering the Q - S w i t c h o f the N d : Y A G pumped laser w i t h a B N C m o d e l 555 d ig i ta l delay generator (Berk ley N u c l e o n i c s Corp . ,San Rafael , C A . U S A ) . F o r these experiments, the dye laser was scanned i n 2 n m steps over a 450 n m range. A t each wavelength , 50 mass spectra were recorded. The wavelength scan range o f 575 n m to 980 n m was achieved by the use o f 8 laser dyes (al l laser dyes were suppl ied by E x c i t o n Inc., Day ton , O H , U S A ) . T h e y were: K i t o n R e d (578-606 nm) , D C M (605-670 nm) , L D S 678 (660-720nm), L D S 750 (705-750 nm), L D S 751 (715-792nm), L D S 821 (785-85 l n m ) , L D S 867 (830-916nm), L D S 925 (890-980nm). 6.3 Results and Discussion 6.3.1 Photodissociation of Trapped Cations T h e goa l o f this chapter was to demonstrate the poss ib i l i ty o f ga in ing spectroscopic informat ion o f P A H cations i n an i o n trap. T o demonstrate the usefulness o f this method, two P A H isomers were chosen for analysis. They were anthracene ( D 2 h symmetry ; 7 .45eV IP ; 178 amu) and its i someric partner phenanthrene ( C 2 V symmetry ; 7.86 e V IP ; 178 amu) shown i n Figure 6.4 Figure 6.5 is a mass spectrum o f a sample o f pure phenanthrene f o l l o w i n g normal 2-laser mass spectrometry under the condit ions shown at the bot tom o f the figure. T h i s spectrum shows the parent i on at 178 T h i n addi t ion to some fragmentation products that result f rom the 2-laser process o n top o f the regular background s ignal . C l ea r ly , w h e n deve lop ing this technique (where the observation o f any smal l s ignal change is useful) it was essential to first isolate the parent i o n and remove any and a l l other ions f rom the i o n trap i n order to observe any change w h i c h results f rom the appl ica t ion o f the th i rd laser. 141 Figure 6.4 The PAH isomers at 178 Th phenanthrene and anthracene. Phenanthrene Anthracene i 7 8 a m u 178 amu" 142 Figure 6.5 Two laser mass spectrum of a sample of phenanthrene on activated charcoal. R F Vo l tage 143 Figure 6.6 shows the spectrum that is observed after the appl ica t ion o f a N B B W pulse before the ions were ramped out o f the i o n trap. W i t h the addi t ion o f this wave fo rm, the i o n trap was ready to begin to col lect fragments f rom the in t roduct ion o f the th i rd laser. It turns out, that w i t h the current dye laser powers avai lable, that a s ingle th i rd laser shot produced o n l y a m i n o r change i n the total i o n populat ions. Therefore, i t was decided early o n i n this work , that mul t ip le th i rd laser shots cou ld be used i n order to b u i l d up an appreciable daughter i o n signal . Figure 6.7 is the spectrum that is observed after 5 laser shots at 892 n m interact w i t h the isolated phenanthrene ions. F o u r m a i n fragmentation channels were observed for phenanthrene (and anthracene); they were due to loss o f - 1 H , - 2 H , -C2H2, and poss ib ly - 1 0 H . The fragmentation w i l l be discussed i n the Section 6.3.2. S i m i l a r l y , a sample o f anthracene was exposed to 5 laser shots at 648 n m after an ident ical i so la t ion routine and the result is shown i n Figure 6.8. In order to just i fy the use o f mul t ip le laser shots and to insure that there were no other effects o f the third laser on the daughter i o n populat ion, a study was performed where the number o f laser shots was var ied between 0-4. The effect o f the number o f shots at 648 n m o n a sample o f isolated anthracene cations is shown i n Figure 6.9 and Figure 6.10. Figure 6.9 focuses o n the daughter i o n popula t ion around 152 T h whereas Figure 6.10 focuses o n the daughters w h i c h result f rom the loss o f one or t w o hydrogens. C l e a r l y , it can be seen that addi t ion o f mul t ip le laser shots o n l y improves the s ignal and has v i r tua l ly no effect o n any newly created daughter ions. T o demonstrate this point , the area o f the daughter s ignal at 152 T h is plotted vs. the number o f laser shots i n Figure 6.11. T h i s figure suggests that there is no cumulat ive effect o f mul t ip le laser shots (other than the obvious one) and that it is not l i k e l y that the parent ions have any m e m o r y o f 144 Figure 6.6 Two laser mass spectrum of a sample of phenanthrene on activated charcoal with the addition of a NBBW isolation pulse. 100 110 120 130 140 150 160 Mass/Charge 170 180 190 200 RF Voltage 145 Figure 6.7 Two laser mass spectrum of phenanthrene on activated charcoal with the addition of a NBBW isolation pulse followed by 5 photodissociation laser shots (892 nm). 146 Figure 6.8 Two laser mass spectrum of a sample of anthracene on activated charcoal with the addition of a NBBW isolation pulse followed by 5 photodissociation laser shots (682 nm). 120 140 160 Mass/Charge 180 200 RF Voltage 147 Figure 6.9 Daughter ion population observed over 0-4 photodissociation laser shots (focusing around 152 Th) on a sample of anthracene. i ^ 4 Pulses I 3 Pulses I 1/ 2 Pulses ' .11 1 Pulse AJ / VA 0 Pulses 145 150 155 160 M a s s / C h a r g e Figure 6.10 Daughter ion population observed over 0-4 photodissociation laser shots (focusing around 178 Th) on a sample of anthracene. 4 Pulses I 3 Pulses 2 Pulses 1 Pulse ] 0 Pulses ni XJ i 160 165 170 175 180 185 M a s s / C h a r g e 148 Figure 6.11 Normalized areas for the ratio of 152/178 vs. the number of photodissociation laser shots on a sample of anthracene. 7 -, 0 1 2 3 4 5 6 Number of Photodisociation Laser Shots 149 previous laser shots; i.e. the parent ions l i k e l y do not retain any energy between laser shots, and that the 50 ms between laser shots is enough t ime for the ions to radiat ively or c o l l i s i o n a l l y relax. 6.3.2 Fragmentation Pathways in Anthracene and Phenanthrene Exper iments concerning the photofragmentation o f gas phase cations have been performed for some t ime. Spec i f ica l ly , the photo stabil i ty o f P A H monocat ions have received a great deal o f interest due to their potential astrophysical role [165]. It has even been proposed that certain fragments o f P A H cat ion may i n fact be more stable than the parent i o n (and thus more abundant i n deep space)[ l73] . In this work , the fragmentation o f an isolated cat ion begins w i t h the absorpt ion o f a photon. Af te r the absorption o f this first photon, a compet i t ion takes place, between the reemiss ion o f the energy (fluorescence) and the redistr ibut ion o f this energy among the v ibra t ional degrees o f freedom o f the electronic ground state (internal convers ion) . W h e n internal convers ion occurs, the photon energy is stored i n the i on , and can be r emoved one o f two ways i n an i o n trap, either radiat ively v i a infrared vibrat ional fluorescence or c o l l i s i o n a l l y w i t h background buffer gas. B o t h o f these loss processes are considered s low, especial ly w i t h respect to the w i d t h o f a N d : Y A G laser pulse (10 ns). The absorption o f a second photon is then possible before the energy o f the first has dissipated. T h i s process can continue unt i l the net energy w i t h i n the i o n is large enough for the dissocia t ion rate to become comparable w i t h the c o o l i n g rate. W e m a y state that the magnitude o f photodissociat ion for a P A H ca t ion depends on: (a) the exci ted state energies being attainable by the incident radiat ion, (b) the exci ted 150 '• .' state l ifet imes are sufficiently long to ensure mul t iphoton absorption, and (c) a sufficient density o f exci ted states levels to enable mul t iphoton absorption to occur [172]. The cond i t ion described i n (a) is used to gain spectroscopic informat ion, because the fragmentation eff ic iency is related to the abi l i ty o f the i o n to gain energy (i.e. the laser wavelength is i n resonance w i t h an electronic transit ion i n the cation). D u r i n g the w o r k performed i n this chapter, v i r tua l ly ident ical fragmentation patterns were observed for both the anthracene and phenanthrene cations. The two p r imary patterns m a y be described as be l low: c u H ' o - > c i 4 # 8 + + H i - Equ. 6.1 c u H ' o -> c u H 9 + + H \ Equ. 6.2 C u H w -+CX2Hl+ +C2H2 EqU_ 6.3 These react ion patterns are w e l l established and have prev ious ly been exper imenta l ly observed [171, 191, 192] and theoretically described [173]. M u c h w o r k has been done examin ing the hydrogen loss process, but to date, it is s t i l l unclear as to whether both hydrogens depart s imultaneously (Equ. 6.1) or sequentially (Equ. 6.2)[171]. Unfortunately , w i t h the experiments discussed here, it was imposs ib le to shed any l ight o n this debate. H o w e v e r , future w o r k m a y be performed to isolate the s ingle H loss ca t ion i n order to examine its photo stabili ty independently. P rev ious w o r k has shown that the Appearance Energies ( A E ) for the processes i n Equ.6.1 and Equ.6.3 are v i r tua l ly ident ical to each other and s imi la r for both isomers 151 [171]. T h i s A E value o f - 1 5 . 0 e V was der ived f rom a series o f experiments where the arr ival o f fragments was moni tored as a function o f photon energy (for a s ingle photon experimental 71]. W e may conclude, therefore, that i n this work , fragmentation requires a m i n i m u m o f between 6 to 11 (visible/near IR ) photons, depending o n the laser wavelength used. A second issue o f m u c h debate concerning these fragmentation processes is that o f i somer iza t ion . It is possible that, f o l l o w i n g the loss o f acetylene (C2H2), the resul t ing fragment f r o m both phenanthrene and anthracene isomerize to an ident ica l fo rm. A g a i n , to date, there has been no agreement on the identity o f this species, or even i f i somer iza t ion occurs. H o w e v e r density functional theory suggests that the most l i k e l y c o m m o n product is either b ipheny lene # + or acenaphthylene* + [193]. T h i s quest ion is o f part icular astrophysical importance, because it is this type o f stable fragment w h i c h some suggest m a y represent a large por t ion o f astronomical P A H s [194]. A m o d i f i e d vers ion o f this experiment w o u l d be able to unambiguous ly solve this p rob lem and many l i ke it because the daughter ions that result f rom photo fragmentation can be further mass selected and subjected to addi t ional laser spectroscopy. The mass spectra i n Figures 6.6 and 6.7 show one addi t ional interesting feature. In both cases, the photo-fragmentation results i n the appearance o f a mino r peak at 168 T h . T h i s peak may correspond to the loss o f a l l 10 hydrogens f rom the parent cat ion. W h i l e this unusual result has never been reported for phenanthrene or anthracene it has been observed i n the case o f coronene and naphtha[2,3-a]pyrene [195]. It is d i f f icul t to exp la in this observation, but it may be attributed to the co l l i s iona l c o o l i n g process i n the i o n trap; most comparable experiments (where H - loss was not observed) were 152 performed i n I C R mass spectrometers where the pressure was three orders o f magni tude lower than i n the i o n trap. There may be a mechanism whereby co l l i s ions w i t h the re la t ively smal l H e buffer gas atoms favors energy transfer into C - H bond stretching modes, however , at this point, this is on ly speculation. Further experiments w i t h a var iety o f different buffer gases (Neon , A r g o n , etc.) at a variety o f pressures may shed l ight o n this observation. 6.3.3 Visible Spectra of Phenanthrene and Anthracene cations The use o f the R E M P D process as a spectroscopic too l has a 30 year his tory [182]. In order for photodissocia t ion to take place, the cat ion must absorb enough energy to make the rate o f d issocia t ion comparable to the relaxation process. In theory, this rate o f energy absorption is a function o f the exci ta t ion wavelength. W h e n the exc i t i ng laser wavelength becomes s imi la r i n energy to an electronic transit ion o f a ca t ion o f interest a dramatic improvement i n the energy transfer w i l l occur. Therefore, a photodissocia t ion spectrum (i.e. a plot o f the ratio o f daughter ions/parent ions vs. wavelength) should y i e l d an accurate representation o f the cat ionic absorption spectra; assuming a few caveats are met. The most important condi t ional requirement is that the photon f lux must by large enough so that sufficient energy is deposited into the cat ion to induce fragmentation. T h e second cond i t ion is that we must also assume that fragmentation is the pr imary result o f energy deposi t ion (i.e. instead o f fluorescence). A s a first demonstration o f the wavelength selectivi ty o f the photodissocia t ion process two wavelengths were chosen to examine the fragmentation eff ic iency o f the anthracene cat ion. A t each o f these wavelengths (720 n m and 740 nm) the laser power 153 was var ied and the fragmentation eff iciency recorded. The resul t ing data, F i g u r e 6.12, demonstrates two key points. Firs t , that the fragmentation process is linear w i t h respect to laser power . T h i s result conf i rms that the fragmentation rate depends p r imar i ly o n the absorption o f a s ingle photon. T h i s is important for two reasons; it validates the w o r k i n g theory that s ingle photon absorpt ion is the ra te- l imit ing step, but it is also useful exper imenta l ly because it means that sma l l changes i n laser power can be easi ly accounted for. The second point that F igure 6.12 demonstrates is that the fragmentation eff ic iency is a funct ion o f wavelength. Presumably , more fragmentation occurs at 720 n m rather than at 740 n m because it is closer to being i n resonance w i t h a real electronic transi t ion o f the cation. Therefore, energy is more efficiently transferred into the cat ion; the result be ing more fragmentation. The log ica l extension o f this experiment was to scan the dye laser over a large range o f wavelengths at a constant power and record the fragmentation eff ic iency o f the anthracene cat ion. The result o f these experiments is shown i n F i g u r e 6.13. E a c h point a long the curve represents the magnitude o f the peak area for a fragment at 152 T h d i v i d e d by the peak area at 178 T h averaged over 50 mass spectra. In this spectrum, data points were taken every 2 n m due the large t ime required to perform these experiments manua l ly . The l ines connect ing the points i n this figure are five point m o v i n g averages f rom each dye laser scan. A n interesting aspect o f this spectrum is its s imi lar i ty to that recorded i n a rare gas matr ix at l o w temperature. F o r compar ison, the v i s ib le spectra o f anthracene recorded i n 154 Figure 6.12 Fragmentation efficiency (ratio of 152/178 Th) vs. energy at two different photodissociation laser wavelengths (720 nm and 740 nm) for a sample of anthracene. 1.2 -, -i 1 : i i i i i i 0 2 4 6 8 10 12 14 16 Energy Measured External to Trap (mj) 155 , Figure 6.13 Top spectrum- Photodissociation spectra of the anthracene cation (ratio of 152/178 Th vs. wavelength). Bottom spectrum - Anthracene cation spectrum acquired in a frozen argon matrix at 12 K [196] 712 nm | I I , I I I M | i i I I I I I i | i I I I M I M I I i I I I i I I I | l ' I I M I I ! I I I i M I M I | I I I I . I I I I I I M I i M I I | I M I i M I I I M I I I I 500 600 700 800 900 wavelength (nm) 156 an A r matr ix at 12 K has been inc luded on F i g u r e 6.13 [196]. The s imilar i t ies between the t w o sets o f experiments are encouraging. Bes ides a smal l shift i n peak locat ions and relative intensities, the two sets o f data are i n good agreement consider ing the vast ly different methods o f determination. Shifts o f this magnitude are t yp ica l i n frozen matrices. M a n y other workers us ing a variety o f methods and solvent med ia have recorded s imi la r data. A peak at 722 n m was first observed by A n d r e w s and coworkers [197] and ascribed by them to the a l l owed D 2 ( A u)<— D o ( B 2 g ) t ransit ion i n the anthracene cat ion. Photoelect ron spectral bands recorded i n the gas phase suggest a t ransi t ion at 704.5 n m [196]. Har t ree-Fock calculat ions performed by V a l a and co-workers agree w i t h this assignment and predict a peak between 717 nm-724 n m [196]. O w i n g to the s imi la r i ty i n spectral structure, the peak at 712 n m w i l l be assigned i n this w o r k to the transi t ion 2 A U <— B 2 g . T e n addi t ional component bands o f decreasing intensity, bui l t o n the 7 2 2 n m band (712nm i n this case) have also been prev ious ly observed [196]. Consequent ly , the peaks at 6 4 8 n m and 6 0 0 n m w i l l , for now, be attributed to v ib ron ic bands. In a s imi la r manner, the gas phase photo fragmentation spectrum for the ca t ion o f phenanthrene was also recorded. The result ing spectrum is shown i n F i g u r e 6.14. A g a i n , the bot tom trace represents the v i s ib le spectra o f the phenanthrene cat ion, this t ime recorded i n a neon matr ix at 4.2 K by the A l l a m a n d o l a group [165]. Here , as above, the two sets o f data bare remarkable resemblances. In the frozen matr ix , the longest wavelength peak is found at 898.3 n m where as this w o r k has found a peak m a x i m u m at 892 n m . O w i n g to the s imilar i t ies between the two sets o f data, the peak at 892 n m is assigned to the same transi t ion as that at.898 n m by the A l l a m a n d o l a group: 157 Figure 6.14 Top spectrum- Photodissociation spectrum of the phenanthrene cation (ratio of 152/178 Th vs. wavelength). Bottom spectra-Phenanthrene cation spectrum acquired in a frozen neon matrix at 4.2 K [165]. 892 nm 700 750 800 850 900 950 1000 wavelength (nm) 158 D2(2A2)<—Do(2Bi). The remain ing peaks at higher energies are assigned to a v ibra t iona l progress ion bui l t o n the 892nm peak as above. One devia t ion between the two spectra is the relative increase i n magnitude o f the peaks at 874 n m (880 n m i n the matr ix) . Interestingly, the Ar-phenanthrene + photodissocia t ion spectra recorded by Brech ignac and P i n o show very s imi la r results to those reported here (their pr imary transit ion occurs at 891 nm) however their recorded data shows no peak at 876 n m [ l 80]. A t this point it is diff icul t to account for these differences. 6.4 Conclusions The goal o f (his chapter was to demonstrate this system's abi l i ty to acquire direct gas phase spectroscopic informat ion relating to large P A H cations. F i g u r e 6.13 and F i g u r e 6.14 demonstrate that the P A H cations phenanthrene and anthracene can be spectroscopical ly evaluated i n this manner. The observed opt ical spectra exhibi t spectral features very s imi la r to those obtained by matr ix i so la t ion methods. Howeve r , w i t h matr ix techniques, the exact loca t ion o f the peaks i n the gas phase must be hypothesized because o f unpredictable matr ix shifts. Furthermore, w i t h the matr ix method the exact identity o f the absorbing species is never certain. In contrast, w i t h this technique the user has excellent control over the identi ty o f the species analyzed. The poss ib i l i ty even exists o f mass selecting daughter ions or other species that exist on ly as radical cations as candidates for further analysis . F i n a l l y , because the source o f ions is the two-laser sampl ing method a very w i d e variety o f potential analytes is available for analyses. 159 The R E M P D method does, unfortunately have a few l imita t ions . Prac t ica l ly , the technique, as it is currently performed (wi th a dye laser), is cumbersome and t ime consuming . H o w e v e r , the next generation o f this device cou ld be greatly improved w i t h the addi t ion o f a solid-state tunable laser - an opt ical parametric osci l la tor ( O P O ) - , w h i c h has the abi l i ty to scan over a large wavelength range under computer control . It w o u l d then be possible to acquire data for both spectroscopic purposes as w e l l as analyt ica l applicat ions. A scenario cou ld even be imagined where real envi ronmenta l samples were examined and certain isomers were mass selected and then ana lyzed by opt ica l means direct ly i n the trap; the i o n trap, i n this case, act ing as a cuvette. W h i l e the opt ical bands observed i n this chapter were o f s imi la r l ine widths to those i n the so l id matrices, they were certainly m u c h wider than those observed i n a free jet expansion. In the rare gas matr ix , the analyte was relat ively c o o l ( 1 0 K ) . H o w e v e r ; the spectral features were broadened by co l l i s ions w i t h the matr ix . O n the other hand, i n this work , the broadening was due to the large internal temperatures o f the analytes; assuming the analyte ions are fu l ly equil ibrated w i t h the background buffer gas, the internal temperature o f the ions is most l i k e l y ~ 3 0 0 K [198-202]. In contrast, w i t h free jet expans ion techniques (i.e. cavi ty r ing d o w n spectroscopy), the ions are not o n l y super coo led but are also i n col l is ion-free environments. Therefore the spectra are often very sharp and the peaks occur at the true gas phase locations. The one disadvantage o f beam methods, however , is that a relat ively smal l f lux o f ions is avai lable for analyses, so the sensi t ivi ty is often l imi ted . A s a result, experiments o f this type are far f rom routine. The R E M P D technique for P A H photodissociat ion experiments, w h i l e mature i n the f ie ld o f I C R spectrometry, is relat ively new w i t h the i o n trap. W i t h the current 160 development o f smal l inexpensive solid-state lasers, a w i d e variety o f future poss ibi l i t ies exist. F o r example , a series o f laser diodes cou ld be inc luded i n the v a c u u m man i fo ld at careful ly selected wavelengths w h i c h a l l o w the user to rap id ly and rout inely ga in op t ica l ly informat ion o n the trapped cations seamlessly w i t h every mass spectra; thus ach iev ing true mul t id imens iona l analyses. 161 Chapter 7 Semi-Quantitative Determination of PAH Isomers Directly from Solid Matrices 7.1 Introduction M a s s spectrometry provides a fast and effective means o f ana lyz ing a w i d e variety o f samples over a large range o f condit ions and concentrations. E v e n though mass analyzers, par t icular ly h igh resolut ion machines such as F T - I C R , are capable o f exquisi te select ivi ty they are not able to resolve mixtures o f structural isomers because they occur at prec ise ly the same mass to charge. A variety o f supplemental techniques have been developed hand i n hand w i t h the mass spectrometer to a l l o w structural characterizat ion ( M S / M S ) [12], however , there remain a few types o f samples and classes o f molecules where an acceptable strategy has yet to be developed. One important class o f these molecules , where structural determination is c r i t i ca l to the complete analysis , is the f a m i l y o f p o l y c y c l i c aromatic hydrocarbons ( P A H s ) . P A H s are an environmenta l ly important class o f compounds where i somer ic determinations are important [79, 80]. A s discussed i n C H A P T E R 3 these molecules are produced m a i n l y by the combus t ion o f hydrocarbons and are found ubiqui tous ly throughout the environment. The l eve l o f tox ic i ty o f the var ious P A H s can vary w i d e l y between isomers, and as a result, isomer d i sc r imina t ion is essential to whatever s o l i d analysis technique is used. F o r example, benzo[a]pyrene is k n o w n to be a strong carcinogen, w h i l e its isomer benzo[e]pyrene is n o n carcinogenic [203]. A d d i t i o n a l l y , the abi l i ty to easi ly discr iminate between isomers may prove useful w h e n t ry ing to determine 162 the source o f contaminants, as the i someric ratio is often indicat ive o f the p roduc t ion process [79]. Curren t ly there are a number o f methods avai lable for the analysis o f P A H s i n so l id samples. The technique that is most routinely used typ ica l ly invo lves a l i q u i d extract ion o f the P A H s f rom the so l id , f o l l owed by separation v i a gas chromatography ( G C ) w i t h detection by mass spectrometry [89]. The most c o m m o n ion iza t ion scheme i n these analyses is electron impact (EI) . Howeve r , the E I mass spectra o f most P A H isomers are v i r tua l ly ident ical . A s a result, isomer differentiation relies o n differ ing G C retention t imes and matches w i t h pure standards. E v e n i f the mass spectrometry were done i n a device capable o f M S / M S such as an i o n trap or tr iple quadurpole, P A H isomers often have ident ical fragmentation patterns [204]. Al te rna t ive ly , some success has been rea l ized i n P A H isomer differentiatiori by observing the results o f gas phase chemistry [205-207]. The phys ica l process o f extraction and G C analysis is often t ime consuming and expensive. A s a result a number o f direct so l id analysis methods have been attempted. Some o f the more successful direct so l id sampl ing techniques appl ied to this p r o b l e m inc lude secondary i o n mass spectrometry ( S I M S ) [40], fast a tom bombardment ( F A B ) [ 4 2 ] , and laser desorption (LD) [41 ] . These techniques, however , are non-select ive i o n sources and typ ica l ly produce complex mass spectra w i t h a variety o f molecu la r fragments and li t t le i someric information. One direct so l id sampl ing technique that has r isen to prominence over the last twenty years is that o f laser desorption, resonant two-photon ion iza t ion mass spectrometry ( L 2 M S ) [ 7 2 , 96]. Th i s technique, relies on the non-specif ic desorpt ion 163 ( typ ica l ly w i t h a pulsed I R laser) o f analyte f rom the so l id f o l l o w e d by selective pho to ion iza t ion o f the analyte by a second pulsed laser ( typ ica l ly a pulsed U V laser). The resul t ing ions are then, i n v i r tua l ly every reported case, examined by a t ime o f f l ight mass spectrometer ( T O F ) [208]. The ion iza t ion stage provides a h i g h degree o f select ivi ty because ion iza t ion w i l l on ly occur i f the molecule has a strong absorpt ion at the ion iza t ion laser wavelength. Thus , the poss ib i l i ty exists o f p rov id ing a second d imens ion to the M S because this ioniza t ion step depends in t r ins ica l ly o n the molecu la r spectroscopy o f the molecules . W h i l e this step may be used to discr iminate between isomers i n theory, most groups have not pursued this and instead have been content to on ly use one wavelength (266nm) for the ion iza t ion step[64]. T h i s is understandable f rom a pract ical standpoint (tunable laser sources i n the U V are cumbersome) and f rom an analyt ica l point o f v i e w because the goals o f the experiments are often to get a picture o f the overa l l P A H content (and most P A H s have strong absorptions at 266nm) . T o the extent o f us ing the select ivi ty o f the ion iza t ion step, Z e n o b i and coworkers have examined the absorption spectra o f P A H s direct ly after laser desorption, and found these w a r m molecules to produce spectra w h i c h were extremely broad due to the large internal temperatures o f the molecules f o l l o w i n g laser desorption. Isomer d i sc r imina t ion di rect ly thus m a y not be possible [96]. O n the other end o f the temperature regime, L u b m a n and coworkers have achieved h igh resolut ion, h igh ly specific ion iza t ion o f peptides and neurotransmitters b y insert ing a supersonic jet between ion iza t ion and desorption [209]. S i m i l a r l y , Sh lag ' s group observed i somer selective ion iza t ion o f polychlor ina ted b iphenyls ( P C B s ) b y s imi la r means [160]. The supersonic jet provides extremely efficient c o o l i n g o f the 164 analytes. H o w e v e r , this apparatus comes at the expense o f becoming somewhat more ana ly t ica l ly unpract ical . A s an alternative to creating the second d imens ion o f analysis i n the ion iza t ion step, the poss ib i l i ty o f p robing the molecular ions w i t h a third pulsed laser has been investigated. T o accommodate this, a system was buil t , w h i c h a l lows L 2 M S to occur direct ly i n the v o l u m e o f an i o n trap. The analyte ions o f interest can be mass selected i n the i o n trap and probed w i t h a th i rd laser. C H A P T E R 6 demonstrated that the P A H isomer cations phenanthrene and anthracene (178 Th) have strong dist inct absorpt ion i n the v i s ib l e and near I R i n the gas phase ( F i g u r e 7.1). These spectra were recorded by the method o f Resonance Enhanced M u l t i p h o t o n Di s soc i a t i on ( R E M P D ) , where the degree o f fragmentation was observed as a function o f wavelength. The absorption spectra alone can a l l o w d i sc r imina t ion between isomers; the i o n trap act ing as a cuvette for the samples. H o w e v e r , even w i t h a state o f the art solid-state laser, it w o u l d take approximate ly 20 m i n to acquire a useful spectrum o f each P A H . A s a result, this chapter describes a secondary method o f us ing the gas phase cat ion absorpt ion spectral informat ion to identify isomers. The method relies o n j ud i c ious ly choosing one wavelength and mon i to r ing the fragmentation eff ic iency o f the mass selected isomers. I f a sample contained two isomers then the i somer that is more i n resonance w i t h the laser wavelength w o u l d have a greater degree o f fragmentation. Thus , by examin ing the relative degree o f fragmentation, the contr ibut ion o f each isomer to a specific mass peak can be determined. In this chapter, an internal standard is also used, to provide semi quantitative concentration in format ion o n two P A H isomers direct ly f rom a so l id sample i n less than five minutes. 165 Figure 7.1 Anthracene and phenanthrene photofragmentation spectra acquired in an ion trap by the REMPD method. 166 7.2 Experimental T h e experiments described i n this chapter were carr ied out o n an i n house bu i l t i o n trap designed for the direct analysis o f so l id samples us ing the two-laser method o f s o l i d sampl ing . The use o f this instrument for the direct analysis o f P A H s o n s o l i d matrices was described i n CHAPTER 2 and CHAPTER 3. A th i rd laser was added to this system to a l l o w wavelength selective photofragmentation and was described i n CHAPTER 6. The methods for co l lec t ing and manipula t ing the ions i n this system were a l l p rev ious ly k n o w n , however , the combina t ion presented here is somewhat unique. Sample molecules are first desorbed f rom a s o l i d matr ix into the gas phase b y laser desorption us ing the unfocussed 1064 n m beam o f a N d : Y A G laser ( D C R - 2 A , Spectra Phys ic s , M o u n t a i n V i e w , C A ) . The I R power reaching the probe surface was typ i ca l ly 10 s W / c m 2 . T h e I R laser induces a rapid heating o f the sample surface and produces a p lume o f desorbed neutrals. A p p r o x i m a t e l y 30 ps after the desorption event; the majori ty o f the neutrals have reached the center o f the i o n trap. These neutrals are then select ively photo ionized i n the center o f the i o n trap by 1+1 Resonance Enhanced M u l t i p h o t o n ion iza t ion ( R E M P I ) . T h i s ion iza t ion is achieved by passing the 4 t h harmonic o f a second N d : Y A G laser (Lumon ic s , H Y 400, ~ 40 pj/pulse) through the neutral p lume. The ions were trapped us ing an R F quadrupole power supply operating at 0.967 M H z at a h e l i u m buffer gas pressure o f - 10" Torr . Once trapped, a 40 ms N B B W , pulse was appl ied to the end caps to isolate a specific mass range. In the case o f these experiments the pulse was designed to el iminate a l l species w i t h mass lower that 178 a m u and higher 167 than 202 amu. T h i s wavefo rm proved useful for pract ical reasons dur ing the experiments (it a l l o w e d easy v i e w i n g o f any fragment ions produced) but was not essential. F i n a l l y , the th i rd laser beam, a N d : Y A G (Quantel Y G 6 6 0 - 532 nm) p u m p e d dye laser (Spectra Phys i c s P D L - 3 ) was sent ver t ica l ly through the r i ng electrode to interrogate the i o n c loud . T y p i c a l energies were 1 mJ/pulse for the v i s ib le to near I R ranges. The energy f rom each o f these laser shots was recorded and bundled w i t h each mass spectrum, thus a l l o w i n g for power fluctuations to be accounted for dur ing data analysis . Exper iments were performed by a l l o w i n g the i o n c l o u d to interact w i t h 20 laser shots each cyc le . T h e result ing ions were then resonantly ejected f rom the i o n trap. T h e samples used i n this w o r k were a l l manufactured i n the lab and are o f the type p rev ious ly characterized i n C H A P T E R 3. B r i e f l y , samples were prepared by d i s so lv ing a P A H (Anthracene, Phenanthrene, Pyrene) a l l f rom S i g m a A l d r i c h ( M i l w a u k e e , W I ) i n hexane ( H P L C grade, S i g m a A l d r i c h ) . These solutions were then sp iked o n to a p rev ious ly we ighed amount o f activated charcoal . The samples were then sonicated for 30 m i n , and then the hexane a l l owed to evaporate off. The charcoal samples were then mechan ica l ly pressed into a smal l sample cup and inserted into the mass spectrometer. The samples used i n this study were a l l i n the concentration range o f between 5 - 250 p m o l / g r a m charcoal . 7.3 Results and Discussion Exper iments were performed to determine the potential u t i l i ty o f us ing a single wavelength laser to discr iminate between isomeric P A H cations stored i n an i o n trap. The f o l l o w i n g equations describe h o w a single laser wavelength c o u l d be used to d iscr iminate between two isomers; however , i n pr inc ip le (n) wavelengths c o u l d be used 168 to discr iminate between (n+1) isomers. The f o l l o w i n g analysis describes the specif ic case o f ident i fy ing the concentrations o f phenanthrene and anthracene (178 amu) direct ly i n a s o l i d that has been sp iked w i t h a k n o w n concentration o f an internal standard, pyrene. I f w e assume that the observed peak intensity result ing f rom two laser mass spectrometry o f pyrene (Ip y) is , o n average, the product o f the pyrene concentrat ion, [Py] , t imes the desorption/ionizat ion/trapping efficiencies (g iven the s y m b o l : ap y ) . T h e n w e m a y mathemat ica l ly wri te that: i„y=VPy}(*„y) E q u 7 1 S i m i l a r l y , for anthracene ( A ) and phenanthrene (P) we may wri te : IA=[A](aA) E q u 7 - 2 /,=m(«;0 Equ. 7.3 Afte r the i o n c loud has been irradiated by the thi rd laser, a certain percentage o f each o f the P A H s w i l l have been removed based o n the fragmentation eff ic iency o f that P A H at the specif ied wavelength. F o r example, the s ignal observed for pyrene after the f i r ing o f the th i rd laser w i l l be: Ipy = [Py](apy)(JragPy) ' E q u . 7.4 W e m a y therefore write that the ratio o f signals observed at mass 178 (for a mixture o f anthracene plus phenanthrene) vs. mass '202 (pyrene) w i t h two-laser mass spectrometry without the third laser to be: Ratio of observed Anthracene (I,) + Phenantherene (I,,) Equ. 7.5 signals at mass 178 /202 = K— Pyrene (I,,y) 169 Or in equation format: Ratio of observed signals at mass 178/202 = + Equ. 7.6 [Py](aPJ Similarly, after the third laser interacts with the ion cloud and a certain amount of fragmentation has occurred the ratio of signals would be as follows: Ratio of observed signals at [A](aA)(fragA) + [P](cc„)(fragP) Equ. mass 178/202 with third laser on = A/^I & / L J I I/U <s / ~i [PyYaiy)(fragPy) In principle, the only two unknowns that are being observed in Equ.7.6 and Equ.7.7 are the concentrations of the isomers phenanthrene and anthracene. Therefore, since we have two equations and two unknowns this problem should in theory be tractable. From a practical point of view, many of the constants listed above must be determined for the specific experimental system being used. For example, the term (apy) the desorption/ionization/trapping efficiency of the PAH is system dependant and must be determined empirically. This can be achieved in a straightforward manner, by simply performing a few routine calibration curves. For example, if a series of standards were made which contained only phenanthrene and pyrene and where the ratio of the concentrations of phenanthrene/pyrene were systematically varied, then a plot of the ratio of observed signals vs. the ratio of concentrations would produce a straight line, and Equ.7.6 would be reduced to: Ratio of signals 178/202 with the laser off = ( „ V r m A P \ a P y J JZ1 Equ. 7.8 170 The slope o f this graph then w o u l d s imply be (ap)/(apy). S i m i l a r l y , i f a series o f anthracene and pyrene standards were produced, then the slope o f the ratio o f observed signals vs . ratio o f concentrations w o u l d y i e l d (a^/fapy). I f both sets o f the cal ibra t ion curves were also acquired w i t h the th i rd fragmentation laser on, then, two addit ional constants c o u l d be acquired. F o r example , i f y o u plot ted the ratio o f observed signals i n the phenanthrene/pyrene mixture vs. the ratio o f concentrations w i t h the th i rd laser on then: Ratio of signals 178 /202 with laser on (aP)(fragP) y(aPy)(fragPy) Equ. 7.9 Therefore, the slope o f the phenanthrene/pyrene ratio o f observed signals vs . the phenanthrene/pyrene ratio o f concentrations w i t h the th i rd laser o n w o u l d y i e l d a straight l ine numer ica l ly equal to [(o.p)(fragP)]/[(apy)(fragPy)]. S i m i l a r l y for samples w h i c h contain o n l y anthracene and pyrene, w i t h the third laser on , the ratio o f observed signals at mass l78 /mass 202 vs. ratio o f concentrations o f anthracene/pyrene w o u l d produce a straight l ine graph w i t h the slope equal to [(aA)(fragA)]/[(apy)(jragPy)]. T o summarize , the slope o f the f o l l o w i n g cal ibrat ion curves y ie lds the constants described i n Table 7.1. F i n a l l y , i f we define two addi t ional terms, namely, the ratio o f observed peaks at mass 178/202 for an u n k n o w n mixture w i t h the laser o f f ( R N L ) and w i t h the laser o n ( R L O ) , then we m a y beg in to rewrite equations (Equ. 7.6) and (Equ. 7.7). Af t e r several pages o f algebra, we can rearrange the above equations to y i e l d the concentrat ion o f phenanthrene and anthracene di rect ly f rom a s ingle series o f measurements. T h e phenanthrene concentrat ion becomes: 171 Table 7.1 Summary of symbols and reference slopes used in this chapter. Observed Slopes P h y s i c a l M e a n i n g S y m b o l used Slope Phenanthrene/Pyrene N O L A S E R (a , ) / / ( « / > ) S P P y N Slope Phenanthrene/Pyrene L A S E R O N (a,,)(fragP) (a,>y)(fragPy) S P P y O Slope Anthracene/Pyrene N O L A S E R (<*A)/ S A P y N Slope Anthracene/Pyrene L A S E R O N (aA)(fragA) (aPy)(fragPy) S A P y O 172 Unkown Phenanthrene Concentration [P] (SAPyN)(RLO)[Py] - (SAPyO)(RNLJ[Py] (SPPyO)(SAPyO)-(SPPyN)(SAPyO) Equ. 7.10 The anthracene concentration also becomes: Unkown Anthracene Concentration [A] = ( R N L ) [ P y ] - [ P ] ( S P P y N ) SAPyN Equ. 7.11 Ideal ly, the cal ibrat ion curves w o u l d be at least five or s ix points i n scope. H o w e v e r , w i t h the system currently employed , the power drift o f the th i rd laser over the t ime required to take several hundred mass spectra l imi t s the number o f ca l ibra t ion points that can be useful ly employed . Therefore a l l ca l ibra t ion curves s h o w n are based o n o n l y three sample points . E a c h point is the average o f 50 mass spectra. Figure 7.2 shows a (50 cyc le averaged) mass spectrum demonstrating the effect o f the th i rd fragmentation laser o n the phenanthrene s ignal . The laser wavelength (892 nm) was chosen to be i n resonance w i t h a phenanthrene cat ion electronic absorption hence the relative s ignal o f phenanthrene is reduced by 5 0 % when the third laser interacts w i t h the i o n c loud . B y w a y o f compar ison , Figure 7.3 shows a sample o f anthracene and pyrene w i t h and wi thout the thi rd laser where the third laser reduces the signal o f anthracene by o n l y 2 0 % . Figures 7.4 shows the cal ibrat ion curves for anthracene and phenanthrene w i t h respect to the pyrene concentrations w i t h the thi rd laser o n and off. T h e slopes o f these curves p rov ide the values for the constants f rom Table 7.1. W i t h these constants determined, w e m a y n o w analyze a sample that contains a mixture o f the two isomers (anthracene and phenanthrene). Figure 7.5 shows the mass spectra o f the mix ture sample 173 Figure 7.2 Two laser mass spectrum of a sample of phenanthrene and pyrene on activated charcoal with the addition of a NBBW laser pulse. The gray line is with the addition of 20 photofragmentation laser shots at 892 nm. 170 180 190 2 0 0 2 1 0 Mass/Charge IR UV VISIBLE RF Voltage 174 Figure 7.3 Two laser spectrum of a sample of anthracene and pyrene on activated charcoal with the addition of a swift laser pulse. The gray line is with the addition of 20 photofragmentation laser shots at 892 nm. 175 Figure 7.4 Calibration curve for anthracene and phenanthrene measured relative to pyrene with and without the addition of the photofragmentation laser. 176 Figure 7.5 Two laser mass spectrum of an "unknown" sample of anthracene, phenanthrene, and pyrene on activated charcoal with the addition of a NBBW laser pulse. The gray line is with the addition of 20 photofragmentation laser shots at 892 nm. 170 180 190 Mass/Charge 200 210 RF Voltage 177 w i t h and wi thout the th i rd laser on. The th i rd laser effectively reduces the ratio o f peaks b y about 3 9 % - consistent w i t h a sample conta ining both anthracene and phenanthrene. F i n a l l y , by insert ing the relevant values into Equ .7 . 10 and Equ .7 . 11 we m a y calculate the concentrat ion o f the P A H s . Since this " u n k n o w n " mixture was produced i n the lab, we can compare the calculated values to the actual concentrations. The results are presented i n T a b l e 7.2. B y examin ing this chart, we can determine the relative effectiveness o f this procedure. The values reported here are relat ively close to those o f the actual so lu t ion concentrations. The absolute errors i n this measurement were o n the order o f 1-2 umole /g ram o f charcoal . Th i s translates into percent errors o f between 2 -23%. W h i l e these errors are re la t ively large i n compar ison to those obtained i n G C / M S experiments, the speed at w h i c h this data was acquired places the value i n a different area. Where G C / M S is useful for quantitative analysis, this technique w o u l d instead f ind use w h e n large numbers o f s o l i d samples were i n need o f analysis , and a s imple " Y E S / N O " answer relevant. F o r examples, i n cases where a hazardous s p i l l occurred, and a large number o f samples needed to be pre-screened for further analysis and possible site remediat ion. 7.4 Conclusion In this chapter, a third laser was added to the two-laser i o n trap system, w i t h the hopes that it c o u l d be used for semi-quantitative analysis o f P A H isomers. C H A P T E R 6 demonstrated that it was possible to obtain spectroscopic informat ion on two P A H isomers di rect ly i n the vo lume o f an i o n trap. W h i l e this may be spectroscopical ly useful, the method, as it was currently performed, was analyt ica l ly unpract ical because o f the t ime i n v o l v e d to scan the wavelength range o f interest. Instead, this chapter focused o n 178 Table 7.2 Summary of observed data for the "unknown" sample. P A H Calcula ted concentration (|j.mole/gram charcoal) A c t u a l concentration (u.mole/gram charcoal) Abso lu te Er ro r ((rmole/gram charcoal) Percent E r r o r Phenanthrene 7.433 9.773 2.34 23 .94% Anthracene 51.08 49.98 1.10 2 . 2 1 2 % 179 j u d i c i o u s l y choos ing a single wavelength, and then carefully measuring the relative degree o f fragmentation for the isomers o f interest. S ince the fragmentation eff ic iency at a selected wavelength depends o n the spectral properties o f the analyte it a l l o w e d i somer ic differentiation. T h e results f rom this chapter showed that it is poss ible to obtain at least s emi -quantitative informat ion concerning i somer ic compound concentrat ion direct ly f rom a s o l i d sample i n less than five minutes. Th i s abi l i ty to rapidly sample so l id materials m a y prove useful i n a number o f environmental applicat ions. The technique, as it was presented here, produced measurement errors m u c h larger than those typ ica l i n tradit ional G C / M S methodologies. These are most l i k e l y acceptable, because the goal o f this technique was to be a rapid complement to G C / M S for s o l i d samples rather than a replacement. Th i s being said, the technique c o u l d s t i l l be vast ly improved . These improvements w o u l d most ly stem f rom engineering issues rather than scientif ic ones. F o r example the stabili ty o f the a l l three lasers p lays a huge r o l l i n the s ignal measurement. The use o f an internal standard helps m i n i m i z e this p rob lem, but s t i l l , more stable lasers are desirable. Secondly , m u c h w o r k c o u l d be done i n the select ion o f the ideal wavelength and power o f the fragmentation laser. The wavelength used i n this w o r k was chosen because it represented the longest wavelength ( lowest energy) that matched a resonance transit ion, w i t h the idea that this w o u l d m i n i m i z e non-resonant fragmentation. H o w e v e r , no attempt was made to determine the ideal number o f laser shots or laser power that produced the best result. F o r example, a perfectly engineered system w o u l d comple te ly remove one isomer w h i l e leaving the other total ly untouched. F i n a l l y , i n a 180 dedicated commerc i a l system, it c o u l d be possible to insert a single diode laser for a m i n i m a l cost that cou ld be used to selectively photo- d is t inguish a select i somer pair. 181 Chapter 8 Conclusions 8.1 Generalities T h i s thesis presented the development o f an analyt ical device capable o f p r o v i d i n g mass spectrometric data direct ly f rom so l id samples. T h i s new instrument is based o n the coup l ing o f the ion trap mass spectrometer w i t h the method o f two-laser s o l i d sampl ing . The nove l combina t ion o f these methods increases the already w i d e range o f capabil i t ies o f the i o n trap. The in i t i a l goal o f this w o r k was to s imp ly b u i l d and characterize a nove l instrument. Howeve r , as it often happens i n doctoral research, dur ing the process o f instrumental characterization, many more questions arose f rom the results than c o u l d poss ib ly be answered dur ing the normal tenure o f a graduate student. Therefore, choices had to be made concerning w h i c h paths to f o l l o w and w h i c h to leave for the next group. There two schools o f thought regarding these choices. Some w o u l d argue, that one should f o l l o w a singular l ine o f questioning only , and pursue this l ine to the end. O n the other hand, one c o u l d instead, f o l l o w many paths, and therefore gain a m u c h broader understanding o f a diverse range o f phenomena. O v e r the last f ive years, I have general ly chosen the later opt ion. Because o f this choice, three separate facets o f research have evo lved f rom this work . These include: environmental analysis, b i o l o g i c a l analysis , and a nove l spectroscopic too l . In the f o l l o w i n g sections each o f these areas w i l l be discussed i n terms o f results achieved, conclus ions drawn, and potential for future work . 182 8.2 Environmental Analysis The most natural appl ica t ion o f the two-laser i o n trap instrument is for direct analysis o f s o l i d environmental samples. The method o f two laser so l id sampl ing is a useful too l for the examinat ion o f environmental ly relevant materials because it features a gentle and non-selective desorption process fo l l owed by a h igh ly selective and sensitive ion iza t ion scheme i n laser based R E M P I . Th i s pulsed i o n source is a perfect complement to the i o n trap mass spectrometer because its select ivi ty helps m i n i m i z e space charge concerns and ion iza t ion can occur i n the center o f the i o n trap where trapping eff ic iency is greatest. A d d i t i o n a l l y , the i o n trap is an ideal device to couple to this very gentle ion iza t ion method because it easi ly a l lows for M S / M S to be performed, thus p r o v i d i n g excellent structural informat ion o f the ion ized species. A s demonstrated i n C H A P T E R 3, the instrument developed here, was s h o w n to be capable o f detecting a w ide variety o f P A H molecules direct ly f rom a s o l i d sample. These test samples were useful i n characterizing the instrument, but also had some environmental applicat ions. T o demonstrate the appl icabi l i ty o f this device to real w o r l d analysis , a test sample o f N e w Y o r k R i v e r water sediment was analyzed and the presence o f a range o f contaminants was conf i rmed. In terms o f environmental moni tor ing , there are several advantages to this two-laser i o n trap system. F r o m an analysis point o f v i e w , perhaps the greatest advantage is speed and ease o f sample preparation. The two laser method, because o f it select ivi ty , is capable o f direct ly sampl ing materials i n their raw form w i t h m i n i m u m sample preparation. A d d i t i o n a l l y , the i o n trap is able to acquire meaningful data d i rec t ly f rom this ion iza t ion technique i n seconds. F o r example, the r iver water sample analysis was 183 accompl i shed i n less than five minutes (from sample preparation to the t ime data was col lected) . B y compar i son , tradit ional methods for the analysis o f P A H s i n s o l i d materials t yp i ca l ly i nvo lve several laborious steps, inc lud ing sample c lean up, solvent extraction, and G C / M S analysis w h i c h take over 24 hours to complete. O f course a f u l l G C / M S analysis w o u l d provide a greater weal th o f sample information, but at the cost o f increased t ime and expense. Therefore, these two techniques should be v i e w e d as be ing complementary. W i t h the two-laser system used to rapid ly pre-screen samples for further analysis b y G C / M S . The two-laser i o n trap system is capable o f p rov id ing a very rapid ident i f ica t ion o f a s o l i d sample, and as shown i n C H A P T E R 3, m a y be used to p rov ide at least s e m i -quantitative informat ion. A d d i t i o n a l l y , w i t h the M S / M S capabil i t ies o f the trap, structural conf i rmat ion m a y be accompl i shed direct ly . E v e n i n the worse cases, where M S / M S is not applicable, and the semi-quantitation not v a l i d , this instrument w o u l d be valuable as a rapid pre-screening device to determine i f a sample shou ld be subjected to further analysis . T h e developed system is, o f course, not without some disadvantages. In terms o f cost and size, the current vers ion o f the instrument is neither cheap nor sma l l . The device , as it stands n o w , contains components w i t h an o r ig ina l cost o f approximate ly t w o hundred thousand dollars . A d d i t i o n a l l y , this apparatus alone fills an entire r o o m and requires several 20-amp power l ines and two sources o f water -cool ing. H o w e v e r , m u c h o f this cost and size m a y be reduced dramatical ly by u t i l i z i n g current developments i n the field o f electronics and lasers. F o r example , desorption and ion iza t ion c o u l d be ach ieved w i t h a s ingle w e l l designed N d : Y A G laser. In fact current laser technology has reached a 184 point where the opt ical specifications required for this w o r k are avai lable i n an air coo led laser the size o f a text book w h i c h runs o f f l i n e voltage. Furthermore, wel l -engineered power supplies n o w a l l o w an entire i on trap system to be enclosed i n a structure rough ly the size o f two computer cases. Therefore, it is not diff icul t to imagine the poss ib i l i ty o f construct ing a vehicle-portable device capable o f rapid analysis i n the f ie ld . In terms o f future environmental applications, one important quest ion concerns the issue o f sensi t ivi ty. Mechan i sms for i m p r o v i n g sensit ivi ty can best be rea l ized by cons ider ing current areas o f s ignal loss. The observed s ignal can be descr ibed as: Observed signal = [Concentration of Sample'.on Surface] [Desorption Efficiency] [Ionization Efficiency] [Trapping Efficiency] [Detection Efficiency] M a n y o f these factors are due to the intr insic properties o f the materials (desorption eff ic iency is largely related to the identify o f the analyte and mat r ix for example) , however , there is one area w h i c h may be greatly improved ; the i on i za t ion eff ic iency. A s shown i n CHAPTER 4, after desorption, analytes continue to pass through the center o f the i o n trap for approximately 300 ps. H o w e v e r , the i o n i z i n g laser pulse lasts for on ly 10 ns. A d d i t i o n a l l y , the laser beam intersects on ly a very sma l l por t ion o f the total desorption plume. Therefore, an obvious and immediate w a y to increase the sensit ivi ty o f the device is to improve the spatial and temporal overlap between the analyte ions and the laser beam. There are many methods that may be e m p l o y e d to improve this overlap, for example, a set o f two mirrors c o u l d be arranged so that the laser beam is reflected mul t ip le t imes through the center o f the trap i n a w a y analogous to that o f cavi ty r ing d o w n spectroscopy. 185 Future w o r k o n this device should first focus o n i m p r o v i n g the sensi t ivi ty and ease o f use for environmental applications. Once this is accompl ished there are a number o f possible col labora t ion and appl icat ion projects that w o u l d be o f interest to the scientif ic communi ty . F o r example, i n results not presented here, the two-laser i o n trap has been s h o w n to be effective i n detecting polychlor ina ted biphenyls ( P C B s ) d i rect ly o n so l id matrices. T h i s abi l i ty cou ld be o f use to workers s tudying bioreactor- induced degradation o f P C B s o n arctic soi ls . Current ly , reactor eff iciency cannot be measured i n real t ime; instead samples must be analyzed after a lengthy sample pretreatment and extract ion procedure. Here , the two-laser i o n trap, w i t h its ab i l i ty to ga in analyt ical in format ion direct ly f rom solids o n a t ime scale o f minutes c o u l d prove useful. 8.3 Biological Analysis The second appl icat ion, w h i c h naturally grew out o f the development o f the two-laser i o n trap, is that o f detection o f b io log i ca l l y relevant compounds . Here , the select ivi ty afforded by the two-laser process proved valuable i n s i m p l i f y i n g the observed spectra that resulted f rom direct analysis o f complex matrices. In this w a y c o m p l e x b i o l o g i c a l matrices such as tissues could.be assayed direct ly. A d d i t i o n a l l y , the implementa t ion o f the i o n trap as the mass spectfometric detector a l l ow ed for conf i rmat ion o f analyte compos i t ion by M S / M S . T h i s system has many advantages over tradit ional means o f spatial ly resolved pharmaceut ical analysis . Compared to other direct analysis techniques ( S I M S , M A L D I ) the two laser i o n source is very rapid, selective, and sensitive to organic compounds . A d d i t i o n a l l y , because o f the pulsed nature o f the device, it is possible to preconcentrate 186 ions di rect ly i n the gas phase to improve sensit ivi ty. F i n a l l y , because analysis is performed i n the i o n trap, the abi l i ty to acquire M S N data is a lways avai lable . The system as it is currently implemented does have some l imita t ions . The sample de l ivery mechan i sm l imi ts sample sizes to those less than 2 m m square. A l s o , i n order for efficient ion iza t ion to occur the analyte o f interest must have a strong absorpt ion at the laser wavelength . W i t h the fixed wavelength system currently avai lable , this l imi t s analysis to those species w h i c h are either h igh ly conjugated or conta in aromatic r ings (and absorb at 266 nm). Future w o r k i n this area should first focus on i m p r o v i n g the divers i ty o f appl icable samples and analytes. Th i s may be achieved by performing the laser desorption/laser ion iza t ion steps outside o f the trap vo lume. B y locat ing the sample probe outside the v o l u m e o f the i o n trap the sample size may be increased dramat ical ly . Implement ing this change w o u l d require the addi t ion o f a stage o f i o n optics to transport the ions into the trap. In addi t ion to m o v i n g the sampl ing locat ion, the addi t ion o f a tunable solid-state laser w o u l d greatly increase the versati l i ty o f the device . In this w a y , the i o n i z i n g laser cou ld be tuned to an absorption o f the analyte o f interest and thus increase the sensi t ivi ty and select ivi ty. W i t h the above changes i n place, a number o f sample types and exper imental protocols become possible . F o r example, a device o f this type may prove useful i n t o x i c o l o g i c a l or pharmaceutical studies. The.spatial resolut ion p rov ided by the laser sampl ing c o u l d also be useful i n probing b ind ing affinities for a variety o f b i o l o g i c a l l y active agents direct ly f rom relevant, matrices. F i n a l l y , this type o f set-up m a y even have 187 use i n the analysis o f who le cel ls i n cases o f food po ison ing or even bio- ter ror ism where a rap id analysis is c r i t ica l . 8.4 Spectroscopy of Trapped Ions The f inal course o f research that was pursued dur ing this doctoral w o r k i n v o l v e d the creat ion o f a unique too l for opt ica l spectroscopy. T h i s was achieved by the addi t ion o f a third, tunable, laser to the p rev ious ly established two-laser i o n trap system. T h e addi t ion o f this th i rd laser a l l owed the co l lec t ion o f spectroscopic data o f mass selected gas phase ions. C H A P T E R 6 demonstrated, for the first t ime, v i s ib le spectral features for the P A H isomer cations phenanthrene and anthracene direct ly i n the gas phase by the method o f resonance enhanced mul t iphoton dissocia t ion ( R E M P D ) . The col lec ted spectra were observed to be very s imi la r to those recorded i n frozen noble gas matrices. T h i s nove l combina t ion o f techniques may have some advantages as a spectroscopic too l for large gas phase ions. F o r example , because the analyte ions are created by two-laser so l id sampl ing , a w ide variety o f sample molecular weights are possible . A d d i t i o n a l l y , because the ions are created and stored i n an i o n trap mass spectrometer, it is possible to mass select ions o f interest for spectroscopic evaluat ion. Furthermore, the fragment i o n that result f rom photofragmentation may also be trapped for further analysis , thus increasing the magnitude o f potential uses i n the f i e ld o f phys ica l chemistry. The device as it currently stands is not without some l imitat ions. F o r example , the co l l ec t ion o f data i n this w o r k was very laborious. The experiments presented i n C H A P T E R 6 were performed w i t h a chemica l dye laser, where a different dye had to be 188 used rough ly every 20 n m . Further more, the large number o f data files had to be col lec ted and processed ind iv idua l ly by hand. Therefore, future w o r k on this device should focus around the moderniza t ion and automation o f the instrument. F o r example , the addi t ion o f a tunable solid-state laser, such as an opt ical parametric osci l la tor , w o u l d greatly improve the ease o f use and speed o f the device. W i t h this type o f laser it w o u l d be possible to fu l ly automate data co l lec t ion because wavelength select ion can be computer control led. F i n a l l y , future w o r k should focus o n the analysis o f as t rophysical ly meaningful samples that are di f f icul t to analyze i n standard ways . F o r example it has been suggested that one o f the possible carriers o f the D I B s are radical fragments o f larger P A H s . T h i s instrument is idea l ly suited for such analysis because analyte species can be examined first o n a mass basis and then opt ica l ly . In addi t ion to unique fragments, this too l w o u l d also be useful for the examinat ion o f larger ion ic species that are not easi ly vapor ized into the gas phase. F o r example , the large aromatic carbon structure Buckminster ful lerene , Cgo, was examined by two-laser mass spectrometry w i t h an eye for future examina t ion by the opt ica l R E M P D method. The result ing two-laser spectrum for C6o + is shown b e l o w i n Figure 8 .1. C l e a r l y , the parent cat ion or any o f its daughter ions are avai lable for future analysis by opt ica l means. T h i s large molecule is typ ica l o f those that are not easi ly analyzed by other gas phase methods, because o f its l o w vapor pressure. In addi t ion to the phys ica l chemistry applications o f this device, another l ine o f research c o u l d also focus on the analyt ical capabil i t ies o f the instrument. Resul ts f rom CHAPTER 7 suggest that it may be possible to use the gas phase opt ica l properties o f different isomers to dis t inguish samples where the isomers have s imi la r M S / M S C I D 189 Figure 8.1 Two laser mass spectrum of Buckminsterfullerene. ' c ro u ^0 -t—' c (D 1_ 3 O 400 500 600 mass/charge 700 800 spectra. The l imi t ed results from this chapter show that the three-laser i o n trap c o u l d potent ial ly be used to p rov ide semi quantitative information o f P A H s d i rec t ly f rom a so l id surface i n less than five minutes. 190 REFERENCES (1) T h o m s o n , J . J . Philos. Mag. 1897, 44, 293. (2) T h o m s o n , J . J . Rays of Positive Electricity; Longmans Green and C o : L o n d o n , 1913. (3) M c D o w e l l , C . A . Mass Spectrometry; M c G r a w - H i l l : N e w Y o r k , 1963. (4) D a w s o n , P . Quadrupole Mass Spectrometry and its Applications; E l sev ie r : A m s t e r d a m , 1976. (5) M a r c h , R . E . Int. J. Mass Spectrom. Ion Process. 1992,118-119, 71 . (6) M a r s h a l l , A . G . ; Hendr i ckson , C . L . ; Jackson, G . S. Mass Spec. Rev. 1998,17,1. (7) C o m i s a r o w , M . B . J.Chem.Phys. 1978, 69, 4097. (8) Cotter , R . J . Time of Flight Mass Spectrometry; A m e r i c a n C h e m i c a l Socie ty : Washing ton , D C , 1997. (9) Stafford, G . C ; K e l l e y , P . E . ; Bradford , D . C . Am. Lab. 1983, 51 . (10) Stafford, G . C ; K e l l y , P . E . ; Reyno lds , W . E . ; S y k a , J . E . P . 31st Conf. Am. Soc. Mass Spectrom. and Allied Topics 1983. (11) Dehmel t , H . G . Advances in Laser Spectroscopy; P l enum: N e w Y o r k , 1983. (12) M a r c h , R . E . ; T o d d , J . F . J . Practical Aspects of Ion Trap Mass Spectrometry; Volume II: Ion Trap Instrumentation; C R C Press: B o c a Raton , 1995. (13) M a r c h , R . E . ; Hughes , R . J . Quadrupole Storage Mass Spectrometry; W i l e y : N e w Y o r k , 1989. (14) M a r c h , R . E . Int. J. Mass Spectrom. 2000, 200, 285 . (15) P a u l , W . ; Ste inwedel , H . German Patent: Aparatus for separating charged particles of different specific charges. 1956, 944,900. 191 (16) B e r k l i n g aus L e i p z i g , K . , Phys ika l i sches Institut der Univers i ta t , B o n n , 1956. (17) M a t h i e u , E . J. Math. Pure Appl. (J. Liouville) 1868,13, 137. (18) M c L a c h l a n , N . W . Theory and Applications of Mathieu Functions; C la rendon: O x f o r d , 1947. (19) Ju l ian , R . ; N a p p i , M . ; W e i l , C ; C o o k s , R . G . J. Am. Soc. Mass Spectrom. 1995, 6, 51. (20) M a e n o , Y . ; Tach ikawa , M . ; M o r i w a k i , Y . ; S h i m i z u , T . Jap. J. Appl. Phys. , Part 2 1995, 34, L 174. (21) Goer inger , D . E . ; D u c k w o r t h , D . C ; M c L u c k e y , S. A . J. Phys. Chem. A 2001, 705, 1882. (22) L o n d r y , F . A . ; A l f r e d , R . L . ; M a r c h , R . E . J. Am. Soc. Mass Spectrom. 1993, 4, 687. (23) A l f r e d , R . L . ; L o n d r y , F . A . ; M a r c h , R . E . Int. J. Mass Spectrom. Ion Process. 1993,725,171. (24) M a r c h , R . E . ; T o d d , J . F . J . Practical Aspects of Ion Trap Mass Spectrometry. Volume 1. Fundamentals of Ion Trap Mass Spectrometry; C R C Press: B o c a Raton , 1995. (25) F ischer , E . Z. Phys. 1959,156, 1. (26) Wuerker , R . F . ; Shel ton, F L ; Langmui r , R . V . J. Appl. Phys. 1959, 30, 342. (27) Ka i se r , R . E . ; L o u r i s , J . N . ; A m y , J . W . ; C o o k s , R . G . Rapid Commun. Mass Spectrom. 1989,1989, 225. (28) Ka i se r , R . E . ; C o o k s , R . G . ; Stafford, G . C ; S y k a , J . E . P . ; Hemberger , P . H . Int. J. Mass Spectrom. Ion Process. 1991,106,19. 192 (29) Ka i se r , R . E . ; L o u r i s , J . N . ; A m y , J . W . ; C o o k s , R . G . Rapid Comm. Mass Spectrom, 1989, 3, 225. (30) Johnson, V . J . V . ; Y o s t , R . A . ; K e l l y , P . E . ; Bradford , D . C . Anal. Chem. 1990, 6 2 , 2 1 6 2 . (31) S o n i , M . ; C o o k s , R . G . Anal. Chem. 1994, 66, 2488. (32) Doroshenko , V . M . ; Cotter, R . J . Rapid Comm. Mass Spectrom. 1996,10, 65. (33) L o u r i s , J . N . ; C o o k s , R . G . ; S y k a , J . E . P . ; K e l l e y , P . E . ; Stafford, G . C ; T o d d , J . F . J.Anal.Chem. 1987, 59, 1677-1685. (34) L o u r i s , J . N . ; Brodbel t , J . S.; C o o k s , R . G . ; G l i s h , G . L . ; V a n B e r k e l , G . J . ; M c L u c k e y , S. A . Int. J. Mass Spectrom. Ion Process. 1990, 96, 117. (35) D a w s o n , P . Et.; Wheten, N . R . J. Vac. Sci. Technol. 1968, 5, 11. (36) V a n B e r k e l , G . J . ; M c L u c k e y , S. A . ; G l i s h , G . L . Anal. Chem 1991, 63, 1098. (37) V a n B e r k e l , G . J . ; G l i s h , G . L . ; M c L u c k e y , S. A . Anal. Chem 1990, 62, 1284. (38) Brodbel t , J . S.; L o u r i s , J . N . ; C o o k s , R . G . Anal. Chem. 1987, 59, 1278. (39) M i l l e r , C . M . ; Noga r , N . S.; Gancarz , A . J . ; Shie lds , W . R . Anal. Chem. 1982, 54, 2377. (40) D a y , R . J . ; Unger , S. E . ; C o o k s , R . G . Anal. Chem. 1980, 52, 5 5 7 A . (41) Ba lasanmugam, K . ; V i s w a n a d h a m , S. K . ; Hercules , D . M . Anal. Chem. 1986, 55 , 1102. (42) D u k e , G . Org. Mass Spectrom. 1984,19, 242. (43) Q i n , J . ; Cha i t , B . T . Anal. Chem. 1996, 68, 2108. (44) Doroshenko , V . M . ; Cotter, R . J . Anal.Chem. 1996, 68, 463 . (45) K o r n i e n k o , O . ; A d a , E . ; Han ley , L . Anal. Chem. 1997, 69, 1536. 193 (46) K o r n i e n k o , O . ; T i n k a , J . ; Wijesundara, B . J . ; Hartley, L . Anal. Chem. 1998, 70, 1208. (47) Weyssenhoff , H . V . ; Seltze, H . L . ; Schlag , E . W . Z . Naturforsch, Teil A 1985, 40A, 61 A. (48) Tembreu l l , R . ; L u b m a n , D . M . Anal. Chem 1986, 58, 1299. (49) Grotemeyer , J . ; Wa l t e rm , K . ; B p e s l , U . ; Sch lag , E . W . Org. Mass Spectrom. 1986, 2 7 , 6 4 5 . (50) A n e x , D . S.; de V r i e s , M . S.; K n e b e l k a m p , A . ; B a r g o n , J . ; Wendt , H . R . ; H u n z i n k e r , H . E . Int. J. Mass Spectrom. Ion Process. 1994,131, 319. (51) Da l e , M . J . ; Jones, A . C ; Po l l a rd , S.; Langr idge-Smi th , R . R . Environ. Sci. Technol. 1993 ,27 , 1693. (52) Da l e , M . ; Jones, A . ; Langr idge -Smi th , P . R . R . ; Cos te l lo , K . F . ; C u m m i n s , P . G . Anal. Chem. 1993, 65, 793. (53) Zare , R . N . ; H a h n , J . H . ; Zenob i , R . Bull. Chem. Soc. Jpn. 1988, 61, 87. (54) Ready , J . F . Effects of High Power Laser Radiation; A c a d e m i c Press: N e w Y o r k , 1971. (55) Ver tes , A.-; G i jbe l s , R . ; A d a m s , F . Laser Ionization Mass Analysis; W i l e y -Interscience Publ ica t ions : N e w Y o r k , 1993. (56) Hansen , S. G . J. Appl. Phys. 1989, 66, 3329. (57) Handschuh , M . ; Net tesheim, S.; Z e n o b i , R . J. Chem. Phys. 1997,107, 2603 . (58) Handschuh , M . ; Net tesheim, S.; Z e n o b i , R . J. Phys. Chem: B 1999,103, 1719. (59) T<maka,¥L. Shimadzu Hyoron 1997, 54,9. (60) Tmaka, K.Bunseki 1996, 253. 194 (61) Tanaka , K . Nippon Iyo Masu Supekutoru Gakkai Koenshu 1993,18, 49 . (62) Karas , M . ; Baclrmanri , D . ; Bahr , U . ; Hi l l enkarnp , F . Int. J. Mass Spectrom. Ion Process. 1987, 75, 53. (63) H i l l enka rnp , F . ; Karas , M . ; Beav i s , R . ; Chai t , B . T . Anal. Chem. 1991, 63, 1 1 9 3 A . (64) Z e n o b i , R . Fres. J. Anal. Chem. 1994, 346, 506. (65) L u b m a n , D . Lasers in Mass Spectrometry; O x f o r d Un ive r s i t y Press: N e w Y o r k , 1990. (66) L u b m a n , D . M , ; N a a m a n , R . ; Zare , R . N . J. Chem. Phys. 1980, 72, 3034. (67) D ie t z , T . G . ; Duncan , M . A . ; L i v e r m a n , M . G . ; Smal l ey , R . E . J. Chem. Phys. 1980, 73, 4816. . (68) Tembreu l l , R . ; S i n , C . H . ; Pang , H . M . ; L u b m a n , D . M . Anal. Chem. 1985, 84, 2911 . * •• V •' -(69) L u b m a n , D . Mass Spec. Rev. 1988, 7, 535. (70) L u b m a n , D . Mass Spec. Rev. 1988, 7, 559. (71) V o u m a r d , P . ; Zhan , Q . ; Z e n o b i , R . Chem. Phys. Lett. 1995, 239, 89. (72) Da l e , M . ; Jones, A . ; Po l l a rd , S.; Langr idge-Smi th , P . Analyst 1994,119, 571. (73) Handschuh , M . ; Net tesheim, S.; Z e n o b i , R . Chem. Phys. Let. 1997, 275, 93 . (74) M c L u c k e y , S. A . ; V a n B e r k e l , G . J . ; Goer inger , D . E . ; G l i s h , G . L . Anal. Chem. 1994, 66, 6 8 9 A . (75) W e i l , C . ; N a p p i , M . ; C l e v e n , C . D . ; W o l l n i k , H . ; C o o k s , R . G . Rapid Commun. Mass Spectrom. 1996,10, 742. (76) D i n g , L . ; K a w a t o h , E . ; Tanaka, K . ; Smi th , A . J . ; K u m a s h i r o , S. Proc. SPIE. 1999, 3777, 144. 195 (77) G i l l , C . G . Thesis , U B C , Un ive r s i t y o f B r i t i s h C o l u m b i a , Vancouve r , 1994. (78) R o b b , D . R . Thesis , U B C , Un ive r s i t y o f B r i t i s h C o l u m b i a , Vancouve r , 1997. (79) C l a r , E . Polycyclic Hydrocarbons; A c a d e m i c Press: L o n d o n , 1964. (80) C o o k e , M . ; Denn i s , A . J . Polynuclear Aromatic Hydrocarbons: a Decade of Progress; Battel le Press: C o l u m b u s , O H . , 1988. (81) F in layson-Pi t t s , B . J . ; Pitts, J . N . Science 1997, 276, 1045. (82) M e n z i c , C . A . ; P o t o c k i , B . B . ; Santodonato, J . Environ. Sci. Technol. 1992, 26, 1278. (83) Masc l e t , P . ; P i s t ikopoulos , P . ; Beyne , S.; M o u v i e w , G . Atmos. Environ. 1988, 4, 639. (84) Jaffrezo, J . L . ; Masc le t , P . ; C l a i n , M . P . ; Wor tham, H . ; Beyne , S.; Cachie r , H . Atmos. Environ. 1993, 27A, 2781 . (85) K e n n a w a y , E . L . , Hieger , I. Br. Med. J. 1930, ii, 1044. (86) Conney , A . H . Cancer Res. 1982, 42, 4875. (87) G r i m m e r , G . G . ; Naujack , K . W . ; Dettburn, G . Toxicol. Lett. 1987, 35, 117. (88) K i e t h , L . H . , Te l l i a rd , W . A . Environ. Sci. Technol. 1979,13, 416. (89) M c D o w , S. R . ; Gige r , W . ; Burtscher, H ; Schmidt-Ot t , A . ; S iegmann, H . C . Atmos. Environ. 1990, 24a, 2 9 1 1 . : r " (90) G ige r , W . ; Schaffner, C . Anal. Chem. 1978, 50, 243 . (91) Niessner , R . ; Robers , W.' ; K r u p p , A . Fres. J. Anal. Chem. 1991, 341, 207. (92) M e l l o n e , A . ; Smi th , B . M . ; Winefordner , J . D . Talanta 1990, 37, 111. (93) Robbat , A . ; L i u , T . Y . ; A b r a h a m , B . M . Anal. Chem. 1992, 64, 1477. 196 (94) P o l l a r d , S. J . T . ; Hrudey , S. E . ; Fuhr , B . J . ; A l e x , R . F . ; H o l l o w a y , L . R . ; Tosto , F . Environ. Sci. Technol. 1992, 26, 2528. (95) Gi l le t te , J . S.; L u t h y , R . G . ; Clement , S. J . ; Zare , R . N . Environ. Sci. Technol. 1999 ,55,1185. (96) Haef l iger , O . P . ; Z e n o b i , R . Anal. Chem. 1998, 70, 2660. (97) M c K a y , D . S.; G i b s o n , E . K . , Jr.; Thomas-Kepr ta , K . L . ; V a l i , H . ; Romanek , C . S.; Clemet t , S. J . ; C h i l l i e r , X . D . F . ; M a e c h l i n g , C . R . ; Zare , R . N . Science 1996, 275 , 924. (98) Z e n o b i , R . ; P h i l l i p p o z , J . M . ; Buseck , P . R . ; Zare , R . N . Science 1989, 246, 1026. (99) Specht, A . A . ; Blades , M . W . V . J. Am. Soc. Mass Spectrom. 2003,14, 562. (100) R o b b , D . ; B lades , M . W . J. Am. Soc. Mass Spectrom. 1997, 8, 1203. (101) Haef l iger , O . P . ; Z e n o b i , R . Rev. Sci. Instrum. 1998, 69, 1828. (102) L i , L . ; L u b m a n , D . Appl. Spectrosc. 1988, 42, 418 . (103) H a h n , J . H . ; Z e n o b i , R . ; Zare , R . N . J.Am. Chem. Soc. 1987,109, 2842. (104) Me i j e r , G . ; D e V r i e s , M . S.; Hunz ike r , H . E . ; Wendt , H . R . Appl. Phys. B 1990, 7557,395. (105) B o e l l , H . ; Pueffel , P . ; Weickhardt , C ; B o e s l , U . AIP Conf. Proc. 1995, 329, 68 . (106) Redhead, P . A . Vacuum 1962,12, 203. (107) C o w i n , J . P . ; Auerbach , D . J . ; Becker , C . A . ; Whar ton , L . Surf. Sci. 1978, 78, 545. (108) P h i l i p p o z , J . M . ; Z e n o b i , R . ; Zare, R . N . Chem. Phys. Lett. 1989,158, 12. (109) Cavanagh , R . R . ; B u n t i n , S. A . ; R ich t e r , ,L . J . ; K i n g , D . S. At. Mol. Phys. 1990, 2 4 , 3 6 5 . (110) Cotter , R . J . ; Tabet, J . C . Int. J. Mass Spec. Ion Phys. 1983, 55 , 151. 197 (111) Net tesheim, S.; Z e n o b i , R . Chem. Phys. Lett. 1996, 255, 39. (112) Johnson, P . M . ; Ot i s , C . E . Ann. Rev. Phys. Chem. 1981, 32, 139. (113) Schmidt , L . D . Cat. Rev. Sci. Eng. 1974, 9, 115. (114) C h u a n g , T . J . ; S e k i , H . ; Huss la , I. Surf. Sci. 1985,158, 525. (115) K a r a s , M . ; Bachmanh^ D . ; H i l l e n k a m p , F . AnahChem. 1985, 57, 2935. (116) Huss l a , I.; S e k i , H ; Chuang , T . J . ; Gor te l , Z . W . ; Kreutzer , H . J . ; R i e r c y , P . Phys. Rev. B. 1985, 32, 3489. (117) Lucchese , R . R . ; T u l l y , J . C . J. Chem. Phys. 1984, 81, 6313. (118) Zare , R . N . ; L e v i n e , R . D . Chem. Phys. Lett. 1987, 136, 593. (119) T u l l y , J . C . Surf. Sci. 1994, 299, 667. (120) H a l l , R . B . ; Santolo, A . M . D . Surf. Sci. 1984,137, 421 . (121) Decker t , A . A . ; George , S. M . Surf. Sci. 1987,182, 215. (122) M a e c h l i n g , C . R . ; Clemett , S. J . ; Enge lke , C . F . ; Zare, R . N . J. Chem. Phys. 1996, 104, 8768. (123) Struck, L . M . ; Richter , L . J . ; B u n t i n , S. A . ; Cavanagh , R . R . ; Stephenson, J . C . Phys. Rev. Lett. 1996, 77, 4576. (124) C a m p b e l l , E . E . B . ; Her te l , I. V . ; K u s c h , C ; M i t z n e r , R . ; Win te r , B . Synth. Met. 1996, 77, 173. (125) Burgess , D . , Jr.; V i swana than , R . ; Huss l a , I.; Stair, P . C ; W e i t z , E . J. Chem. Phys. 1983, 79, 5200. (126) Watanabe, K . ; L i n , M . C ; G r u z d k o v , Y . A . ; Mat sumoto , Y . J. Chem. Phys. 1996, 104,5974. (127) Hess , P . ; Springer: B e r l i n , 1989. 198 (128) R o b b , D . R . ; B lades , M . W . Int. J. Mass Spectrom. 1999,190, 69 . (129) M c L u c k e y , S. A . ; G l i s h , G . L . ; A s a n o , K . G . Anal. Chim. Acta 1989, 225, 25 . (130) M c i n t o s h , A . ; D o n o v a n , T . ; Brodbel t , J . Anal. Chem. 1992, 64, 2079. (131) V o u m a r d , P . ; Z e n o b i , R . J. Chem. Phys. 1995,103, 6795. (132) N a k a y a m a , T . Surf. Sci. 1983,133, 101. (133) D y e r , P . E . ; Sr in ivasan, R . Appl.Phys-.A. 1986, 48, 445 . (134) Asscher , M . ; Z i m m e r m a n n , F . M . ; Springstenn, L . L . ; Hous ton , P . L . ; H o , W . J. Chem. Phys. 1992, 96, 4808. (135) Goss , K . ; Schwarzenbach, R . P . Environ. Sci. Tech. 1998, 32, 2025. (136) Corne l i s sen , G . ; V a n Noor t , P . C . ; Parsons, J . R . ; Gbvers , H . A . J . Environ. Sci. Tech. 1997, 31, 454. (137) G h o s h , U . ; T a l l e y , J . W . ; L u t h y , R . Environ. Sci. Tech. 2001, 35, 3468. (138) V a n v a e c k , L . ; Struyf, H . ; V a n r o y , W . ; A d a m s , F . Mass Spectrom. Rev. 1994,13, 189. (139) V a n V a e c k , L . ; Gi jbe l s , R . Fresenius. J. Anal. Chem. 1990, 337, 755. (140) V a n v a e c k , L . ; Bennett, J . ; Lauwers , W . ; Vertes , A . ; Gi jbe l s , R . Mikrochim. Acta 1990, 3, 283. (141) T o d d , P . J . ; Schaaff, T . G . ; Chaurand, P . ; C a p r i o l i , R . M . J. Mass Spectrom. 2001, 36, 355 . (142) Porce l la t i , G . Phospholipids in the Nervous System; R a v e n Press: N e w Y o r k , 1985. (143) Soderberg, M . ; E d l u n d , C . ; Alafuzof f , I.; Kr is tensson, K . ; Da l lne r , G . J . Neurochem. 1992,59, 1646. 199 (144) D e n o l l i n , S.; Poels , K . ; Vanvaeck , L . ; Dec le rck , N . ; B a k k e r , A . ; D u w e l , V . ; Vandeve lde , D . ; V a n m a r c k , E . Path. Res. Prac. 1997,193, 313. (145) O d o m , R . W . Appl. Spectrom. Rev. 1994, 29, 67. (146) T o d d , P . J . ; M c M a h o n , J . M . ; Short, R . T . ; M c C a n d l i s h , C . A . Anal. Chem. 1997, 69, 5 2 9 A . (147) Chandra , S.; Smi th , D . R . ; M o r r i s o n , G . H . Anal. Chem. 2000, 72, 1 0 4 A . (148) V a n v a e c k , L . ; Gi jbe l s , R . Anal. Chem. 1990, 743. (149) V a n v a e c k , L . ; V a n r o y , W . ; Struyf, H . ; A d a m s , F~.; Caravat t i , P . Rapid Commun. Mass Spectrom. 1993, 7, 323. (150) Vanvaeck , L . ; Poels , K . ; D e n o l l i n , S.; H a c h i m i , A . ; Gi jbe l s , R . Cell Bio. Int. 1997 ,27,635. (151) V o g t , H . ; He inen , H . J . ; M e i e r , S.; Wechsung , R . Fres. Z. Anal. Chem. 1981, 308, 195. (152) He inen , H . J . ; M e i e r , S.; V o g t , H . ; Wechsung , R . Int.J. Mass Spectrom. Ion Proc. 1983,47,19. (153) Brenna , J . T . ; Creasy, W . R . ; M c B r a i n , W . ; Sor ia , C . Rev. Sci. Instrum. 1988, 59, 873. (154) Pel let ier , M . ; K r i e r , G . ; M u l l e r , J . F . ; W e i l , D . ; Johnston, M . Rapid. Comm. Mass Spectrom. 1988, 2, 146. (155) M c M a h o n , J . M . ; Short, R . T . ; M c C a n d l i s h , C . A . ; Brenna , J . T . ; T o d d , P . J . Rapid Commun. Mass Spectrom. 1996,10, 335. (156) Troendle , F . J . ; Redd ick , C . D . ; Y o s t , R . A . J. Am. Soc. Mass Spectrom. 1999,10, 1315. 200 (157) Escandon , N . ; Z i m m e r m a n , D . C ; M c C a l l , R . B . J.Pharmacol. Exp. Ther. 1994, 268,441. (158) R ive t , J . M . ; Lejeune, F . ; Gobert , A . ; Canton , H . ; L a v i e l l e , G . ; Peg l i on , J . L . ; M i l l a n , M . J . Soc. Neurosci. Abstr. 1992,18, 1521. (159) M c L a f f e r t y , F . W . ; Turecek, F . Interpretation of Mass Spectra, 4 th ed.; U n i v e r s i t y Science B o o k s : N e w Y o r k , 1996. (160) Z i m m e r m a n n , R . ; B o e s l , U . ; Weickhardt , C ; L e n o i r , D . ; S c h r a m m , K . W . ; Ket t rup , A . ; Sch lag , E . W . Chemosphere 1994, 29, 1877. (161) B r a d y , B . B . ; Peteanu, L . A . ; L e v y , D . H . Chem. Phys. Lett. 1988,147, 538. (162) L i , L . ; L u b m a n , D . M . Appl. Spectrosc. 1989, 43, 543. (163) Leger , A . ; Puget, J . L . Astron. Astrophys. 1984,137, L 5 . (164) Sel lgren , K . Spectrochim. Acta A 2001, 57, 627. (165) Sa lama, F . ; Job l i n , C ; A l l a m a n d o l a , L . J . J. Chem. Phys. 1994,101, 10252. (166) Jenniskens, P . ; Desert, F . X . Astron. Astrophys. Suppl. Ser. 1994,106, 39. (167) Heger , M . L . Lick. Obs, Bull. 1921,10, 146. (168) Wese lak , T . ; Schmidt , M . ; K r e l o w s k i , J.Astron. Astrophys. Suppl. Ser. 2000,142, 239. . . . (169) F o i n g , B . F L ; Ehrenfreund, P . Nature 1994, 369, 296. (170) K r e l o w s k i , J . ; Wester lund, B . E . Astron. Astrophys. 1988,190, 339. (171) L i n g , Y . ; L i f s h i t z , C . J. Phys. Chem. A 1998,102, 708. (172) B o i s s e l , P . ; de Parseval , P.-; M a r t y , G . ; Lefevere, G . J. Chem. Phys. 1997,106, 4973. (173) Joch ims , H . W . ; Baumgar te l , H . ; L e a c h , S. Astrophys. J. 1999, 512, 500. 201 (174) Oomens , J . ; van R o i j , A . J . A . ; Me i j e r , G . ; v o n He lden , G . Astrophys. J. 2000, 542, 404. (175) L i f s h i t z , C . Int. Rev. Phys. Chem. 1997,16, \ 13. (176) Kos te r , G . ; M a r t i n , J . M . L . ; L i f s h i t z , C . J. Chem. Soc.-Perkin Trans. 1 9 9 9 , 2, 2383. (177) M e y e r , B . Low Temperature Spectroscopy; E l sev ie r : N e w Y o r k , 1971. (178) R o m a n i n i , D . ; B ienn ie r , L . ; Salama, F . ; K a c h a n o v , A . ; A l l a m a n d o l a , L . J . ; S toeckel , F . Chem. Phys. Lett. 1 9 9 9 , 303, 165. (179) P i n o , T . ; B o u d i n , N . ; Brechignac , P . J. Chem. Phys. 1 9 9 9 , 111, 7337. (180) Brech ignac , P . ; P i n o , T . Astron. Astrophys. 1 9 9 9 , 343, L 4 9 . (181) W r i g h t , K . ; B lades , M . W . , 51 Confrence o f the A m e r i c a n Socie ty for M a s s Spectrometry, M o n t r e a l , Canada 2003. (182) Dunbar , R . C . Int. J. Mass Spectrom. 2000, 200, 571. (183) D u n c a n , M . Int. J. Mass Spectrom. 2000,200, 545. (184) Spence, T . G . ; Burns , L . A . ; Posey, J . J. Phys. Chem. A 1997 ,101 , 139. (185) Dunbar , R . C . Anal. Instrum. 1988,17, 113. (186) Heger , H . J . ; Z i m m e r m a n n , R . ; Dorfner, R . ; B e c k m a n n , M . ; G r i e b e l , F L ; Ket t rup , A . ; B o e s l , U . Anal. Chem. 1 9 9 9 , 71, 46. (187) Heger , H . J . ; Z i m m e r m a n n , R . ; Dorfner, R . ; Ket t rup , A . ; B o e s l , U . AIP Conf. . Proc. 1 9 9 8 , 454, 305. (188) Z i m m e r m a n n , R . ; Heger , H . J . ; Dorfner, R . ; Boesl,- U . ; B lumens tock , M . ; L e n o i r , D . ; Ket t rup , A . Comb. Sci.Tech. 1998,134, 87. 202 (189) Job l i n , C ; M a s s e l o n , C ; B o i s s e l , P . ; de Parseval , P . ; M a r t i n o v i c , S.; M u l l e r , J . Rapid Comm. Mass Spectrom. 1997,11, 1619. (190) R o l l a n d , D . ; Specht, A . A . ; Blades , M . W . ; Hepburn , J . Chem. Phys. Lett. 2003, 373, 292. (191) E k e r n , S. P . ; M a r s h a l l , A . G . ; Szczepansk i , J . ; V a l a , M . Astroph. J. 1997, 488, L 3 9 . (192) R o b s o n , L . ; Tasker , A . D . ; L e d i h g h a m , K . W . D . ; M c K e n n a , P . ; M c C a n n y , T . ; K o s m i d i s , C . ; Tza l las , P . ; Jaroszynski , D . A. ; . Jones , D . R . Int. J. Mass. Spectrom. 2002, 220, 69. (193) L i n g , Y . ; M a r t i n , J . M . L . ; L i f s h i t z , C . J. Phys. Chem. 1997, 219. (194) Hudg ins , D . M . ; Bauschl icher , C . W . , Jr.; A l l a m a n d o l a , L . J . Spectrochim. Acta. Part A 2001, 57, 907. (195) E k e r n , S. P . ; M a r s h a l l , A . G . ; Szczepanski , J . ; V a l a , M . J. Phys. Chem. A 1998, 102, 3498. (196) Szczepansk i , J . ; V a l a , M . ; T a l b i , D . ; Par ise l , O . ; E l l i nge r , Y . J. Chem. Phys. 1993, 98, 4494. (197) A n d r e w s , L . ; F r i edman , R . S.; K e l s a l l , B . J . J. Phys. Chem. 1985, 89, 4016. (198) Goer inger , D . E . ; A s a n o , K . G . ; M c L u c k e y , S. A . Int. J. Mass Spectrom. 1999, 182, 275 . (199) Gronert , S. J. Am. Soc. Mass Spectrom. 1998, 9, 845. (200) H u i - F e n , W . ; L i - W e i , C . ; Y a - P i n g , L . J. Chin. Chem. Soc. (Taipei) 1999, 46, 923 . (201) L u n n e y , M . D . N . ; Buch inger , F . ; M o o r e , R . B . J. Mod. Opt. 1992, 39, 349. (202) Nourse , B . D . ; Kent tamaa, H . I. J. Phys. Chem. 1990, 94, 5809. 203 (203) Ha rvey , R . G . In ACS Symposium Series 283; A m e r i c a n C h e m i c a l Socie ty : Wash ing ton , D C , 1985. (204) Shushan, B . ; B o y d , R . K . Org. Mass Spectrom. 1980,15, 445 . (205) M o s i , A . A.; C u l l e n , W . R . ; Eigendorf , G . K . J. Mass Spectrom. 1997, 32, 864. (206) S i m , P . G . ; E l s o n , C . M . Rapid Commun. Mass Spectrom. 1988, 2, 137. (207) S temmier , E . A . ; Buchanan , M . V . Org. Mass Spectrom. 1989, 24, 94. (208) B o e s l , U . J. Mass Spectrom. 2000, 35, 289. (209) L i , L . ; L u b m a n , D . M. Rev. Sci. Instrum. 1988, 59, 557. 204 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0061200/manifest

Comment

Related Items