UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development of a laser ablation quadrupole ion trap mass spectrometer for direct spectrometry of solid… Gill, Christopher George 1994

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

Item Metadata

Download

Media
831-ubc_1994-894040.pdf [ 4.07MB ]
Metadata
JSON: 831-1.0059528.json
JSON-LD: 831-1.0059528-ld.json
RDF/XML (Pretty): 831-1.0059528-rdf.xml
RDF/JSON: 831-1.0059528-rdf.json
Turtle: 831-1.0059528-turtle.txt
N-Triples: 831-1.0059528-rdf-ntriples.txt
Original Record: 831-1.0059528-source.json
Full Text
831-1.0059528-fulltext.txt
Citation
831-1.0059528.ris

Full Text

DEVELOPMENT OF A LASER ABLATION QUADRUPOLE ION TRAPMASS SPECTROMETER FOR DIRECT SPECTROMETRY OF SOLIDSAMPLESByCHRISTOPHER GEORGE GILLB. Sc., Acadia University, 1988B. Sc. H., Acadia University, 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of ChemistryWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAJanuary, 1994© Christopher George Gill, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. it is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of CI1tn (S77 9The University of British ColumbiaVancouver, CanadaDate Y C A / S(Signature)DE-6 (2/88)AbstractElemental analysis of solid samples often involves a dissolution stepprior to analysis, which can introduce contamination and result in further dilutionof trace analytes. Methods of direct solid analysis obviate this step, and canprovide compositional information for solid materials. To achieve very lowdetection limits for analytes in solid samples requires special methodology.One way to achieve lower limits of detection is to increase the observation timefor an analyte. Quadrupole ion traps have demonstrated single ion detectionusing laser based ion detection methods, making the combination of laserablation for direct solid sampling and quadrupole ion trap mass spectrometry alogical choice for ultra-trace (sub-femtogram) analysis. This thesis examinesthe use of laser ablation coupled with quadrupole ion trap mass spectrometryas a potential analytical method for ultra-trace direct solid multielemental massspectrometry.This thesis develops an analytical method for direct solid massspectrometry, called Laser Ablation Ion Trap Mass Spectrometry (LAITMS).Through a series of investigations, a hybrid mass spectrometer has evolvedwhich embodies many characteristics that are desirable for LAITMS analysis.The function of this new spectrometer was subsequently investigated by aseries of parametric studies to determine optimum conditions for it’s operationas a simple mass spectrometer, capable of direct solid analysis.Parameters considered included laser irradiance & wavelength, ionstorage field characteristics and sample surface quality. The use of modifiedhyperbolic electrodes as well as cylindrical electrode configurations has beenexamined. Selective ionization schemes for analytes in solid samples hasbeen investigated by a two-color LAITMS experiment. The effect ofsynchronizing laser ablation with the phase of the ion storage field has alsobeen examined. The present LAITMS spectrometer has demonstratedpicogram detection limits for analytes present in samples of stainless steel.Future laser induced fluorescence ion detection schemes should realize muchlower detection limits.IIITable of ContentspageAbstract iiTable of contents ivList of Tables xList of Figures xiiList of Abbreviations xixAcknowledgments xxivCHAPTER 1INTRODUCTION 11.1 Mass Spectrometry 11.1.1 Historical Development of Mass Spectrometry 11.1.2 The Quadrupole Ion Trap 81 .1.2.1 Theory of Operation 91.1.2.2 Experimental Applications 151.1.2.2.1 Sample Introduction 151.1.2.2.2 Mass Spectrometric Capabilities 171.1.2.2.3 Ion Detection Methods 181.1.2.2.4 Alternate Ion Trap Geometry 191.2 Laser Ablation 201.2.1 Historical Perspective 201.2.2 Theoretical Considerations 211.2.3 Direct Solid Sampling Method for Solid Matrices 241.3 Scope of Thesis 25ivCHAPTER 2PRELIMINARY INVESTIGATIONS OF THE USE OF A RUBY LASERFOR DIRECT SOLID SAMPLING I IONIZATION INSIDE AQUADRU POLE ION TRAP MASS SPECTROMETER 272.1 Introduction 272.2 Experimental 282.2.1 Equipment 282.2.2 Samples 332.3 Results and Discussion 332.4 Summary 36CHAPTER 3INVESTIGATION OF THE USE OF Nd:YAG LASER ABLATION FORDIRECT SOLID SAMPLING / IONIZATION WITH QUADRUPOLEION TRAP MASS SPECTROMETRY 383.1 Introduction 383.2 Experimental 413.2.1 Ion Trap Mass Spectrometer 413.2.2 Laser and Optical Configuration 473.2.3 Samples 493.3 Results and Discussion 493.3.1 Calibration 493.3.2 Spectra Obtained for Solid Samples 523.4 Summary 57VCHAPTER 4INVESTIGATIONS OF AN IMPROVED ION TRAP MASSSPECTROMETER USING Nd:YAG LASER ABLATION FORDIRECT SOLID SAMPLING I IONIZATION 594.1 Improved Design Considerations 594.2 Experimental 614.2.1 Ion Trap Mass Spectrometer 614.2.2 Laser and Optical Configuration 654.2.3 Samples 674.2.4 Scanning Electron Microscopy (SEM) 674.3 Results and Discussion 684.3.1 Calibration 684.3.2 Spectra for Conducting Solid Samples 744.3.3 Spectra for Non-Conducting Solid Samples 784.3.4 SEM Analysis of Ablated Surfaces 854.4 Summary 93CHAPTER 5SAMPLE SURFACE PREPARATION, LASER IRRADIANCE,WAVELENGTH AND NUMBER OF SHOTS 965.1 Introduction 965.2 Experimental 985.2.1 Equipment 985.2.2 Samples 1005.3 Results and Discussion 1 015.3.1 Laser Irradiance 1015.3.2 Sample Surlace Preparation 1065.3.3 Laser Focus for Irradiance Control 1115.3.4 Detection Limits for NIST Stainless Steel Samples 114vi5.3.5 Laser Ablation by 532 nm and 266 nm Radiation 1225.4 Summary 133CHAPTER 6COMPARISON OF NORMAL AND STRETCHED HYPERBOLICELECTRODE GEOMETRY 1366.1 Introduction 1366.2 Experimental 1376.2.1 Equipment 1376.2.2 Samples 1396.3 Results and Discussion 1416.3.1 Storage Time 1416.3.2 Storage Potential 1496.3.3 Peak Shape and Resolution 1536.4 Summary 157CHAPTER 7EVALUATION OF TWO COLOR LASER ABLATION ION TRAPMASS SPECTROMETRY 1587.1 Introduction 1587.2 Experimental 1637.2.1 Equipment 1637.2.2 Samples 1687.3 Results and Discussion 1687.4 Summary 172viiCHAPTER 8INVESTIGATION OF THE USE OF AN ION TRAP WITHCYLINDRICAL ELECTRODE GEOMETRY 1748.1 Introduction 1748.2 Experimental 1798.2.1 Equipment 1798.2.2 Samples 1798.3 Results and Discussion 1 798.3.1 Comparison of Cylindrical and Hyperbolic ElectrodePerformance 1798.3.2 Electron Ionization and Buffer Gas Pressure 1828.3.3 Storage Time and Potential 1858.4 Summary 188CHAPTER 9LASER ABLATION SYNCHRONIZATION WITH RESPECT TO THERADIO FREQUENCY STORAGE FIELD 1909.1 Introduction 1909.2 Experimental 1919.2.1 Equipment 1919.2.2 Samples 1949.3 Results and Discussion 1 949.4 Summary 198CHAPTER 10CONCLUSIONS 19910.1 Generalities 19910.2 Understanding of LAITMS Developed by This Thesis 20010.3 Future Research Directions 204viiiREFERENCES .208APPENDIX I 218APPENDIX II 222ixList of TablesTable Description PageExperimental parameters for ruby laser ablation in thestorage volume of a quadrupole ion trap 31II Instrumental parameters for electron ionization calibrationexperiments 68III Instrumental parameters for LAITMS experiments 75IV Experimental parameters for LAITMS studies 99V Composition for NIST stainless steel samples 100VI Calculated isotopic concentrations and their correspondingLAITMS signals 120VII Material removed by laser ablation, calculated by the coneapproximation 121VIII Calculated limits of detection for atomic species in NIST 1155stainless steel 122IX Instrumental parameters for electron ionization experimentsused in the parametric investigations of the “normal” and“stretched” electrode geometry 138X Instrumental parameters used for LAITMS parametricinvestigations of the “normal” and “stretched” electrode geometry...1 39XI Elemental composition for metal alloy samples analyzed bythe LAITMS investigations 140XII Parameters used for the MacSimionTM 1.0 simulation studies 146XIII Resolution calculations for “normal” and “stretched” electrodeconfigurations 156XIV Transition probabilities for manganese(I) 161XV Instrumental and laser parameters for the two color LAITMSexperiment 167XVI Dimensions of the cylindrical ion traps used byMather et. a!. [133] 176XVII Operational parameters used for the cylindrical ion trapexperiments 178xXVIII Instrumental parameters for synchronization of the ablation laserwith the storage potential for LAITMS experiments 193XIX Elemental corn position of inconel 747 194xiList of FiguresFigure Description Page1.1 J. J. Thompson’s positive-ray parabola apparatus 31.2 Schematic diagram of Bainbridge’s first mass spectrograph 51.3 Cross sectional diagram of a typical quadru pole ion trapelectrode arrangement 101.4 Mathieu stability diagram for the three dimensional quadrupoleion trap 131 .5 Mathieu stability region near the origin for the three dimensionalquadrupole ion trap showing the iso-f3-lines 142.1 Schematic diagram of the optical arrangement for ruby laserablation of solid samples inside the storage volume of a quadrupole ion trap 302.2 Schematic diagram of the ruby laser trigger circuitry showingdetail of the micro relay chip 322.3 Mass spectra obtained by laser ablation of brass inside the storagevolume of a quadrupole ion trap for (a) one laser pulse on a freshsurface and (b) one laser pulse after 9 shots on the same samplelocation 343.1 Diagrams which show (a) the schematic ion trap configuration withmodifications allowing laser ablation inside the storage volume,and (b) a cross sectional view and three dimensional representation of the ring electrode 423.2 Block diagram of the ion trap mass spectrometer system 443.3 Timing diagram for the ion trap experiments 463.4 Optical configuration for laser ablation ion trap massspectrometry 483.5 Electron ionization (El) mass spectra used for calibration of(a) carbon tetrachioride and (b) xenon 503.6 LAITMS spectra of soft solder at mass/charge range (a) 50-140amu, and (b) 170-240 amu 533.7 LAITMS spectrum of silver solder 55xii4.1 Cross sectional diagram of the ion trap electrodes showing theoptical path through the ion storage volume 624.2 Diagrams of the ion trap vacuum manifold (i. front view, ii. rearview) used for LAITMS experiments. A = radio frequency (RE)power supply input feedthrough; B = channel electron multipliersignal out feedth rough; C = variable leak valve for gaseoussample introduction; D = variable leak valve for helium buffer gasintroduction; E = pirani cold cathode vacuum gauge; F = opticalports to enable probing studies of the stored ions; G = differentially pumped sample probe assembly; H = mechanical vacuumpump with in line liquid nitrogen trap 644.3 Illustration of the optical configuration used for the LAITMSexperiments 664.4 Electron ionization (El) mass spectrum used for calibration ofcarbon tetrachloride 694.5 Electron ionization (El) mass spectrum used for calibration of amixture of perfluoro-tri-N-butylamine (FC 43) and carbontetrachloride 704.6 Plot of average resolution (error = ±2a) for electron impact ionizedcarbon tetrachloride ions (m/z = 117, 119 and 121) for threereplicate experiments versus helium buffer gas pressure (1 bar =1X1O5Pa) 724.7 Plot of laser pulse energy (error = ±2a) versus oscillator flashlampvoltage averaged for 200 laser shots 734.8 LAITMS spectra obtained for (a) type 308 stainless steel usinglaser pulse irradiances of 5.2 X 101 Wcm2and (b) silver solder(irradiance = 3.2 X 1010 W•cm2) 764.9 LAITMS spectrum obtained for supposed thin Rh foil (actually Mo)using laser pulse irradiances of 1.44 X 1010 W•cm2 794.10 LAITMS spectrum obtained for a pure Mo metal rod sample usinglaser pulse irradiances of 1.28 X 1010 Wcm2 804.11 LAITMS spectrum obtained for a sintered ceramic sample(MACORTM) using laser pulse irradiances of 8.5 X 1010Wcm2 824.12 LAITMS spectrum obtained for a polyimide plastic sample(VESPELTM) using laser pulse irradiances of 1 X 1010 Wcm2 84xlii4.13 Scanning electron micrograph (SEM) of the surface of a type 308stainless steel sample which has been ablated by (a, b) 1 laserpulse and (c, d) 100 laser pulses 864.14 Scanning electron micrograph (SEM) of the surface of a silversolder sample which has been ablated by (a, b) 1 laser pulseand (c, d) 100 laser pulses 884.15 Scanning electron micrograph (SEM) of the surface of aMACORTM sample which has been ablated by (a, b) 1 (left) and100 (right) laser pulses, (c) ilaser pulse and (d) 100 laser pulses...904.16 Scanning electron micrographs (SEM) of the surface of aVESPELTM sample which has been ablated by 100 laser pulses,showing stereoscopic (a) right and (b) left views 925.1 LAITMS spectra obtained for samples of (a) silver solder usinglaser pulse irradiances of 3.2 X 1010 Wcm2, (b) silver solderusing laser pulse irradiances of 4.5 X 1010 W•cm2, (c) type 308stainless steel using laser pulse irradiances of 5.2 X 1010 Wcm2,and (d) type 308 stainless steel using laser pulse irradiances of8.1 X 1 10Wcm2 1025.2 Plots of relative signal intensities (error = ± 2G) for 52Cr, 55Mn,and 56Fe ions resulting from laser ablation of a stainless steel(SAM C1151) sample versus number of laser shots for (a) a fineabrasive paper polished surface, and (b) a diamond paste andalumina polished surface 1075.3 Scanning electron micrograph (SEM) of a type 308 stainless steelsample which was polished with fine abrasive paper; evident inthe picture are laser ablation craters for pulse irradiances of5.2 X 1010 Wcm2resulting from (a) one laser pulse and (b)100 laser pulses 1105.4 Scanning electron micrograph of stainless steel (SAM 1155)sample polished with diamond paste and alumina which showsthe laser ablation craters formed by 100 laser shots but varyingthe irradiance by increasing the laser beam waist at the samplesurface. This was accomplished by translating the focusing lenstowards the sample surface. Crater 1 was for ablation at the focalpoint of the lens; moving in a clockwise manner shows theablation craters formed as the lens was translated towards thesample. Spectrum for crater A was not stored 112xiv5.5 Plot of relative signal intensity for 52Cr, 55Mn, and 56Fe ionsresulting from laser ablation of a stainless steel (SRM 1155)which was polished using diamond paste and alumina versusdistance of lens translation towards the sample surface. Theseintensities result from summing the spectra obtained for25 laser pulses 1135.6 LAITMS spectra obtained for stainless steel (SRM 1155) from thelens translation study for (a) lens position 4, showing spacecharged peaks, (b) lens position 8, exhibiting good resolution,and (c) lens position 9 where the signal intensities are reduced.Refer to Figure 5.4 for a picture of the ablation craters for thedifferent lens positions 11 55.7 Plot of relative signal intensity and mass resolution versus lenstranslation distance (lens position) towards the sample surfacefor (a) 52Cr ions and (b) 56Fe ions resulting from laser ablationof a stainless steel (SRM 1155) which was polished using diamond paste and alumina 1185.8 LAITMS spectra obtained for samples of silver solder using(a) 532 nm laser ablation and (b) 266 nm laser ablation 1245.9 LAITMS spectra obtained for samples of MACORTM using(a) 532 nm laser ablation and (b) 266 nm laser ablation 1265.10 LAITMS spectra obtained for samples of VESPELTM using(a) 532 nm laser ablation and (b) 266 nm laser ablation 1295.11 LAITMS spectra obtained for samples of Cl 151 stainless steelusing (a) 532 nm laser ablation and (b) 266 nm laser ablation 1316.1 Plot of integrated ion signal intensity for electron ionization ofcarbon tetrachloride (CCl3 ions, error = 2a) versus storage timeobtained with an ion trap with “normal” electrode spacing,obtained for 6 replicate experiments of 50 scans each 1426.2 Plot of integrated ion signal intensity for electron ionization ofcarbon tetrachloride (CCI3+ ions, error = 2cr) versus storage timeobtained with an ion trap with “stretched” electrode spacing,obtained for 6 replicate experiments of 50 scans each 1436.3 Plot of integrated ion signal intensity for LAITMS (532 nm) ofC1151 atomic ions from stainless steel versus storage time obtainedwith an ion trap with “normal” electrode spacing, obtained for 6replicate experiments of 50 shots each, using a storage potentialof25OV(0-p) 144xv6.4 Plot of integrated ion signal intensity for LAITMS (532 nm) ofCl 151 atomic ions from stainless steel versus storage time obtainedwith an ion trap with “stretched” electrode spacing, obtained for 6replicate experiments of 50 shots each, using a storage potentialof400V(0-p) 1456.5 Cross sectional diagram through an ion trap with (a) “normal”electrode geometry and (b) “stretched” electrode geometry thatshow equipotential contour lines for the storage fields created bysimulation using MacSimionTM 1.0. The contour lines represent 10 volt increments for the phase when the ring electrodeis 100 V and the endcaps at ground 1476.6 MacSimionTM 1.0 ion trajectory simulation for a single ion ofmlz = 109 amu positioned initially at the center of the storage fieldfor (a) “normal” and (b) “stretched” electrode geometry. Parameters used for this simulation are given in Table 6.4. TheMacSimionTM program does not compensate for the effects ofhelium buffer gas upon the ion trajectories developed bythe simulation 1486.7 Plot of integrated ion signal intensity obtained by LAITMS(532 nm) of atomic ions from an inconel 747 sample versusstorage potential obtained using an ion trap with “stretched”electrode spacing, obtained for 5 replicate experiments of 50shots each, using a constant storage time of 5 ms 1506.8 Plot of integrated ion signal intensity obtained by LAITMS(532 nm) of atomic ions from an inconel 747 sample versusstorage potential obtained using an ion trap with “normal”electrode spacing, obtained for 5 replicate experiments of 50shots each, using a constant storage time of 5 ms 1516.9 Mass spectral peaks for CCl3 (119, 121, 123 amu) ionsobtained by electron ionization of carbon tetrachloride for both“normal” and “stretched” electrode geometry. The abscissa isgiven in data points to facilitate resolution calculations, withm/z directly proportional to data number 1546.10 Mass spectral peaks for 56Fe, 58Ni and 60Ni ions obtainedby LAITMS (532 nm) for both “normal” and “stretched” electrodegeometry. The abscissa is given in data points to facilitate resolution calculations, with m/z directly proportional to data number 1557.1 Typical stability diagrams for ion storage in an ion trap for ions ofmass/charge 18 and 28 amu 1597.2 Grotrian diagram for Mn(I) 1627.3 Schematic diagram of the optical configuration used for the2 color LAITMS experiment 164xvi7.4 Box diagram of the electronics and instrumentation used for the2 color LAITMS experiment 1667.5 Plot of 55Mn / 56Fe ion signal intensity ratio (± 2a) versus dyelaser delay relative to laser ablation, averaged for 5 separateexperiments 1697.6 Diagram of the dye laser pulse delay generator output whichshows the sporadic pulse shift “bug” produced by the electronics....1707.7 Diagram of the dye laser pulse delay generator which shows theshorter delays used for the second experiment 1707.8 Plot of 55Mn / 56Fe ion signal intensity ratio (± 2a) versus dyelaser delay relative to laser ablation for the second investigation,averaged for 5 separate experiments 1718.1 Detailed schematic diagram of the cylindrical ion trap electrodes ....1778.2 Typical electron ionization mass spectrum obtained for xenonisotopes using an ion trap with cylindrical electrodes.Experimental parameters used are given in Table 8.2 1808.3 Typical electron ionization mass spectrum obtained for xenonisotopes using an ion trap with hyperbolic electrode geometry 1818.4 Plot of Xe ion intensity versus electron gun filament current.Each point represents the average of 100 scans 1838.5 Plot of Xe ion intensity versus helium buffer gas pressure. Eachpoint represents the average of 100 scans 1848.6 Plot of Xe ion intensity versus ion storage time. Each pointrepresents the average of 100 scans 1868.7 Plot of Xe ion intensity versus ion storage potential. Each pointrepresents the average of 100 scans 1879.1 Schematic diagram which summarizes the function of theelectronics used for laser ablation synchronization with respectto the phase of the ion storage field 1929.2 Plot of 56Fe, 58Ni and 60Ni ion signal intensities resultingfrom LAITMS of inconel 747 versus phase delay values forlaser ablation relative to the phase of the ion storage field 1959.3 Plot of 56Fe, 58Ni and 60Ni ion signal intensity RelativeStandard Deviation (RSD) values for ions resulting from LAITMSof inconel 747 versus phase delay values for laser ablationrelative to the phase of the ion storage field 196xvii10.1 Experimental arrangement for laser induced fluorescencedetection of La and Gd by LAITMS of solid samples 206xviiiList of AbbreviationsA ampere (Cs1)a thermal diffusivity (m2s1)AAS Atomic Absorption SpectroscopyAES Atomic Emission SpectroscopyAFS Atomic Fluorescence Spectroscopyamu atomic mass unitAPP Abrasive Paper Polisheda.u. arbitrary unitsau Mathieu equation parameter (dimensionless)B uniform magnetic field (kgC1.sl)Bar pressure unit (1 X iO Pa)ion trajectory stability value (dimensionless)C Coulombc analyte concentrationCCD Charge Coupled DeviceCCP Capacitively Coupled PlasmaCEM Channel Electron MultiplierCID Collision Induced DissociationCW continuous waveDC direct currentDI de-ionizedDPP Diamond and alumina Paste PolishedE uniform electric field (Vm2)e elementary charge (1 .6021 X 1019 C)El Electron IonizationxixEmin minimum absorbed irradiance for solid evaporation (W-cm2)eV electron-Volt (1 .602189 X i019 J)applied storage potential (U, V)Fr-ICR Fourier Transform-Ion Cyclotron ResonanceFT-ICRMS FT-ICR Mass SpectrometryFWHM Full Width at Half Maximumg gramGDMS Glow Discharge Mass SpectrometryHz hertz (s1)ICP Inductively Coupled PlasmaICP-AES ICP-Atomic Emission SpectroscopyICPMS ICP Mass SpectrometryICR Ion Cyclotron ResonanceITDTM Ion Trap Detector (Finnigan Corp.)ITMSTM Ion Trap Mass Spectrometer (Finnigan Corp.)J joule (kgm2s)K degrees kelvink constantL literLA-ICPMS Laser Ablation-ICPMSLAITMS Laser Ablation Ion Trap Mass SpectrometryLAMMA Laser Ablation Micro-Mass AnalyzerLMSA Laser Micro Spectral AnalysisLOD Limit of Detectionlaser plasma wavelength (nm)Lpm liters per minuteLPP Laser Produced PlasmaxxL latent heat of vaporization (Jkg1)M particle massm mass of an ionm meterm/z mass-to-charge ratioMACORTM sintered ceramic of proprietary composition (Corning Ltd.)MALDI Matrix Assisted Laser Desorption Ionizationmm. minuteMS Mass SpectrometryMS-MS tandem mass spectrometryMS tandem mass spectrometryn critical electron number density for LPP formation (m3)Nd:YAG Neodymium:Yttrium Aluminum Garnetne electron number density (m3)NIST National Institute of Standards and Technology (U. S. A.)Pa pascal (kg.m1-s2)PBBO 2-[1 , 1 ‘-biphenyl]-4-yl-6-phenyl-benzoxazolePRA Photochemical Research Associates Ltd.q elementary unit chargeQIT Quadrupole Ion TrapQS Q-SwitchQUISTOR QUadrupole Ion STORep mass density of a solid target (kg.m3)r radial electrode coordinateRE Radio Frequency (Hz)RIMS Resonance Ionization Mass Spectrometryinternal radius of ring electrodexxiRSD Relative Standard Deviation (%)RSDB Relative Standard Deviation of the Background (%)S seconda standard deviationSBR Signal to Background Ratio (%)SIMS Secondary Ion Mass SpectrometrySRM Standard Reference MaterialSSMS Spark Source Mass Spectrometryt timeTbp boiling point temperatureti duration of laser pulse used for ablationTmp melting point temperatureTOE Time-of-FlightTorr pressure unit (133.32 Pa•Torrl)U DC component of (volts)U. B. C. University of British ColumbiaU. N. B. University of New BrunswickV RF component of I (volts)V voltv velocity (scalar)v:v composition by volumeVDC direct current voltage (volts)VESPELTM polyimide of proprietary composition (Dupont Ltd.)W Watt (J.s-1)(0u,n nth order frequency of a stable ion trajectoryx distanceMathieu equation parameter (2t12, radians)xxiiy distancez axial electrode coordinatez0 1/2 of the endcap electrode spacingz0’ 1/2 of the stretched endcap electrode spacingdegrees centigradeohm (V.A-i)radial frequency of applied potential (cI0, radians s)(O-p) 0 to peakxxiiiAcknowledgmentsThis thesis is dedicated to my Grandfather, Percy Bendell, whoencouraged my interest in science and supported me throughout my education.A thesis represents a turning point in one’s life, yet it takes the concerted effortsof many to actually complete the work required. There are many individualswhose expertise in diverse fields has helped to make my endeavors possible.I would like to thank my research supervisor, Dr. Michael Blades, whoallowed me freedom to pursue research in his laboratories and providedguidance when it was required. Thanks are also due to the members of hisresearch group for all the assistance while writing this manuscript, and for thefriendship that they have given. Special thanks are extended to Alison C. Tang,my friend and companion, who gave both loving support and helpful proofreading during the preparation of this thesis.There are many members of the chemistry department that have helpedme to finish this work. Thanks are due to Dr. C. Brion, Dr. M. Comisarow, Dr. A.Bree, Dr. A. Merer, Dr. A. Adam, Dr. G. Eigendorf, and B. Clifford for their advice,assistance and helpful equipment loans over the course of my studies. Gratefulacknowledgment is given to M. Vagg and D. Slovarty for their precisionmachining of the quadrupole ion traps and quick assistance when needed.Credit is also due to S. Rak for his glass work, M. Carlisle in the ElectronicsShop, and S. Rollinson for her illustrations of the new spectrometer. Others atU. B. C. who have contributed time and effort include Dr. P. Lawrence forcomputer software/hardware advice, Dr. A. Mitchell for helpful discussions onmetallurgy and a sample of inconel 747, and M. Mager and M. Weis fortechnical assistance regarding scanning electron microscopy. Thanks areextended to Dr. R. March at Trent University for inspiration and long discussionsxxivregarding ion trapping, and Dr. N. Nogar at Los Alamos National Laboratoriesfor helpful discussions about both ion traps and laser ablation.Acknowledgment is made to the University of British Columbia, theNatural Sciences and Engineering Research Council , the American ChemicalSociety, the Federation of Analytical Chemistry and Spectroscopy Societiesand the Bruker-Spectrospin Corporation for financial support over the course ofthese studies.xxv1CHAPTER 1INTRODUCTION1.1 Mass SpectrometryIn its simplest form, a mass spectrometer performs three essentialfunctions. Firstly, a sample is converted to gaseous ions (in a vacuumenvironment), then these ions are separated or analyzed according to theirmass-to-charge ratios, and finally the analyzed ions are detected by a devicecapable of measuring the number of ions which strike it. The detector output isconverted into a plot of intensity versus mass-to-charge ratio (a mass spectrum),which provides information about the sample based upon the ions created fromit. Closer inspection of any one of these functions shows that it is much moreintricate than the simplistic description given above. Mass spectrometers aregenerally complex analytical instruments, requiring complex electronics,vacuum technology and precisely made components. Despite all thecomplexity however, mass spectrometry is a powerful analytical tool which canbe applied to determinations of both atomic and/or molecular information foralmost any sample whether it is a solid, liquid or gas. The next section of thisthesis presents an historical perspective of mass spectrometry leading up to thedevelopment of quadrupole ion trap mass spectrometry.1.1.1 Historical Development of Mass SpectrometryMass spectrometry predates the majority of instrumental methodsavailable to analytical chemists today. In 1898, Wien [1, 2] demonstrated thedeflection of cathode rays (‘Kanalstrahlen’) by magnetic and electric fields. This2observation was used to deduce that these rays were composed of positivelycharged particles with specific charge values much smaller than that ofelectrons. J. J. Thompson [3] had demonstrated the existence of electrons, andupon continuing his investigation of the positive cathode rays, he began todiscover their remarkably complex nature. His investigations were conductedusing the positive-ray parabola apparatus [4, 5] which is schematically depictedin Figure 1.1. This experimental analyzer arrangement was first employed byKaufman [6] for his studies of cathode rays. Cathode rays were formed in thedischarge bulb A, and then were passed as a collimated beam to the analyzerregion through a long narrow tube. The beam was acted upon by magnetic andelectric fields, which analyzed the beam into its various charge-to-masscomponents.In a uniform electric field E of length x, a particle of mass M and charge qexperiences an acceleration in the z direction of Eq/M. If the velocity of theparticle is given by v, and its time spent in the field given by x/v, the particlewill emerge from the field with a vertical displacement z given by:z =(Eq/M)(x/v)2 1.1Simultaneously, in a uniform magnetic field B, the particle experiences anacceleration in the y direction of Bqv/M and emerges with a lateraldisplacement (for small deflections) given by:y = -Bqx2/Mv 1.2From Equations 1.1 and 1.2, the following relationship can be obtained:y2kz 1.3c)Figure1.1:J.J.Thompson1spositive-rayparabola (FromAston, 1942)apparatus.Thecomponentsareasfollows:A,dischargetube;B,cathode;C,water jacketcoolingcathode;D,anode;E,gasinlet;F, pumplead;G,photographicplate;I,magneticshield;M,M’,magneticpoles;N, Nmica (for insulatIon);P,Psoft iron(servesbothascondenserplatesandtodefinethemagneticfield).FDE4where the constant k depends upon the specific charge of the ion, theapparatus geometry, and the magnitude of the applied deflecting fields. Theresult is that particles of the same q/M strike the photographic plate G (Fig. 1.1)in a parabola whose vertex lies at the undeflected position of the particles.There is a different parabola for each value of q/M represented in the positive-ion beam. Thompson’s research studied various gases introduced to thedischarge, whereby he proved that the positive rays were the massivefragments remaining when one or more electrons were removed from a neutralatom or molecule. This work provided strong evidence that neon exists as twoisotopic forms, one of atomic weight 20 and a rarer variety with an atomic weightof 22. This was the first indication that isotopes exist among the stableelements.The increase in mass spectrometric research which evolved from theseexperiments was enormous. F. W. Aston, who was encouraged by Thompsonwhile a research student at Cambridge, developed a mass spectrograph in1919 [7] that achieved much better resolution than the Thompson Parabolaapparatus. Aston’s first experiments provided unequivocal proof [8] that neonhad two isotopes. Further investigation of other atoms such as chlorine,mercury, nitrogen and the noble gases lead him to deduce that all atomspossessed integrally related mass values. Although further improvements weremade to the Aston mass spectrograph, other researchers were active in thefield. The Canadian researcher A. J. Dempster (at the University of Chicago,U.S.A.) [9-12] also constructed a mass spectrograph which he used todetermine isotopic ratios for Mg, Li, K, Ca and Zn. In the early 1930’s, K. T.Bainbridge [13-18] constructed a mass spectrograph based upon the Dempstertype semicircular instrument geometry. A schematic diagram of the Bainbridgemass spectrograph is given in Figure 1.2. The semicircular geometry utilized a5Figure 1.2: Schematic diagram of Bainbridge’s first mass spectrograph. (FromAston, 1942) The components are as follows: Si, S2, and S3, slits;Pi, P2, Wien velocity filter condenser plates, Plate, photographicemulsion used for ion detection. As indicated, the radius ofcurvature in the magnetic analyser is linear with ion mass for aconstant degree of ionization.+Plate6uniform magnetic field to produce directional focusing of ions from a mono-energetic ion source. He modified Dempster’s design by adding a secondelement to the device, a Wien velocity filter. This was an additional section ofcrossed and coterminous electric and magnetic fields through which the ionsmust pass. The velocity filter obviated the need for a mono-energetic ionsource. For an ion to pass through the velocity filter and enter the magneticanalyzer, it was required that the forces exerted by the Wein velocity filterbalance, orEq=Bqv 1.4This required the transmitted ions to possess a unique velocity v = EIB whichresulted in the radius of curvature for magnetic deflection by the analyzer beinginversely proportional to the q/M value for the ion. The mass scale wastherefore linear, which was convenient for the photographic detection methodsused.Bainbridge used his instrument for isotopic mass determinations,including the newly discovered heavy hydrogen atom and the atoms involved inthe nuclear reaction1H+7Li—>24e 1.5This research provided [181 the first experimental proof of Einstein’s massenergy relationship. In addition, isotopic determinations were made using thisinstrument for Zn, Ge and Te.The pioneering research of Thompson, Aston, Dempster and Bainbridgewas done in spite of grave limitations in experimental technique and with only7primitive knowledge of ion optics. The 1930’s heralded dramatic changes inthese conditions. Studies of the focusing properties of magnetic and electricfields were conducted by Herzog and Mattauch, whose definitive works [19, 20]provided general focusing equations for homogeneous electric and/or magneticfields. Further technical improvements were made for vacuum technology, ionsources and electrical detector systems. In 1935, Nier [21] combined all ofthese advances to his benefit and used his instrumentation to demonstrate thesensitivity and accuracy that could be achieved in the determination of isotopicabundances. This seminal work set a standard which subsequent researchershave steadily improved.The availability of commercial instruments in the early 1950’s allowednon-physicists to use mass spectrometry to study phenomena of interest tothem, yet with little concern for the basic instrumentation. Mass spectrometryquickly became a standard method for analytical measurement, and has seen avariety of instrument types develop and become commonplace in manylaboratories. As the technology and design of mass spectrometers improved,the originally expensive and experimentally limited equipment has becomeaffordable and easy to use to the extent that undergraduate analyticallaboratories are often equipped with simple mass spectrometers for laboratoryexperiments. Many new types of mass spectrometers have been developed tosuit a variety of experimental situations. They resolve charged atoms ormolecules using magnetic and/or electric fields, or on the basis of theirvelocities in time of flight mass spectrometry [22, 23]. Other researchersdeveloped mass spectrometers based upon the storage of ions using magneticand electric fields. These included the quadrupole ion trap [24], the Penningtrap [25], and the ion-cyclotron resonance [26] mass spectrometers. A veryimportant variation of the ion cyclotron resonance technique was developed at8The University of British Columbia in 1976 by Comisarow and Marshall [27]called Fourier Transform -Ion Cyclotron Resonance (FT-ICR) Spectroscopy,which studied the ions confined in an ion cyclotron mass spectrometer on thebasis of their cyclotron frequencies. FT-ICR has developed into a potent massspectrometric method, capable of providing mass resolutions in excess of 108.The advent of the quadrupole ion trap mass spectrometry in the early 1950’s[24] eventually lead to the development of versatile, compact and inexpensivemass spectrometers for use in a wide variety of mass spectrometricexperiments. Clearly, a discussion of all of the methods of mass spectrometrywhich evolved from Thompson’s positive ray parabola is beyond the scope ofthis thesis, which investigates quadrupole ion trap mass spectrometry. The nextsection of this thesis presents the quadrupole ion trap mass spectrometer alongwith the theory of its operation, some experimental applications, and severalanalytical uses.1.1.2 The Quadrupole Ion TrapIn 1989, the Royal Swedish Academy of Sciences awarded the NobelPrize in Physics to Wolfgang Paul of the University of Bonn and Hans Dehmeltof the University of Washington. The awards were made “for the developmentof the ion trap technique. . . which has made it possible to study a singleelectron or single ion with extreme precision.” [28]. This award has broughtmuch attention to ion trap mass spectrometry, which has seen intense researchactivity since the advent of commercial instrumentation in 1983 [29, 30]. Toillustrate this activity, between 1989 and 1992 some 300 publications onquadrupole ion trap mass spectrometry appeared in the literature. An excellentreview of quadrupole ion trap mass spectrometry was recently published by R.E. March [31] as well as a comprehensive treatment of the field in a text by9March and Hughes [32]. The following sections of this thesis present a concisereview of quadrupole ion trap mass spectrometry to provide readers with anunderstanding of the method.1.1.2.1 Theory of OperationA detailed account of the theory of quadrupole ion trap operation is givenby March and Hughes [32], with a more concise summary presented in therecent review article by R. E. March [31]. This section provides a concise reviewof the theory presented therein to develop an understanding for this thesis.The quadrupole ion trap, in essence, is composed of a three electrodearrangement: Two end-cap electrodes and a central ring electrode. Aschematic diagram of a cross section through an ion trap is given in Figure 1.3.The central ring electrode has an inner surface described by a hyperboloid ofone sheet and is located in space between the endcaps, which are defined bytwo hyperboloids of two sheets. The geometry of the electrodes is described bythe following equations [33]End Caps:r21.6r0 z0Ring:r2 z21.7r0 z0For the quadrupole ion traps presented by this thesis, r0 and z0 are related by:2 2r0z 1.80EndCapRingEndCap—Figure1.3:Crosssectionofatypical quadrupoleiontrapelectrodearrangement. (FromMarchandHughes,1989)11where r0 is the internal radius of the ring electrode, z0 is half the distancebetween the end caps, with r and z denoting electrode surface coordinates. Todefine a quadrupole ion trap, only r0 needs to be selected; most ion trapexperiments have used r0 values from 1-25 mm. For clarity, the devicedescribed in this theoretical treatment will be called an ion trap. Physicists referto the device as the “Paul trap”, whereas chemists have adopted Dawson’s [34]description of it as the “quadrupole ion trap” or Todd’s acronym “QUISTOR” [35].The mathematics of ion trajectory stability within the ion trap follows theMathieu second-order differential equation [36] which was developed over 120years ago to describe the vibrations of stretched membranes. The Mathieuequation used to describe ion motion is given by:d2u+ (a - 2qcos2)u = 0 1 .9where u represents the radial and axial coordinate axes, r and z respectively,and is a dimensionless quantity defined by ptJ2, p being the radialfrequency of the applied potential For an ion trap operated by groundingthe endcaps and applying Io to the ring electrode (as is the case for thisthesis), solutions to Equation 1.9 give the dimensionless quantities a and qu:-8eUaz = -2ar mr02p2 1.10-4eVq=-2qr=2 1.11where V is the zero-to-peak amplitude of the potential to , U is the d.c.component of the potential c10 , e is the proton charge and m the mass of an ionconfined by the ion trap. For an ion to be stored within an ion trap, it must12describe a trajectory which is stable both axially and radially as defined by thesolutions of the Mathieu equation given in Equations 1.10 and 1.11.If the dimensionless parameters au and q are plotted for an ion withmass m and charge e, regions of stability and instability exist for both axial andradial motion. As previously stated, for ion storage within an ion trap, both axialand radial stability are required. Thus, it is the overlap of the stability regions forboth radial and axial motion which define the requirements for ion storage.Figure 1 .4 is an au versus qu plot of the stability regions for both axial andradial ion motion, showing the regions of z-stable and r-stable overlap. Anenlarged representation for the overlap region near the origin is given by Figure1.5. An understanding of the ion motion stability diagram given by this Figure isessential to understanding ion behavior within an ion trap.The stability diagram given in Figure 1.5 applies to ion traps with any r0value and for the total range of mass/charge values. For an ion which has aand q values within the stability region a pseudopotential well exists; if thekinetic energy of the ion does not exceed the limits of this well, the ion is stored.According to March [31], the stability diagram represents “a fishing net with anodd shaped rim and where the depth of the net varies from about 1eV near theorigin to more than 10 eV near the f3 = 1 stability boundary.” Figure 1.5 showstypical Iso-13zand I5O3r lines. The values of f3 and 13r determine the frequencyspectrum of the stable ion trajectories. The parameter 13L is a function of thestability parameters a and q, with the frequency spectrum of the stable iontrajectories defined by:= (n +13u/2)2 n -oo, ,-1, 0, 1,. . . oo 1.1213Figure 1.4: Mathieu stability diagram for the three dimensional quadrupole iontrap. (From March and Hughes, 1989)151050—5S—10stable—15140.20.10—0.1—0.2—0.3—0.4—0.5—0.6—0.7azFigure 1.5: Mathieu stability region near the origin for the three dimensionalquadrupole ion trap showing the iso-p-lines. (From March andHughes, 1989)0.6 0.0.51.0 00.10.20.30.50.60.70.80.91.00.2 0.4 0.6 0.8 1.0 1.2 1.4 1.615where o is the nth order frequency, u = x, y or z, n has integer values and 13uis a function of defined a and qu stability parameters. For n = 0, thefundamental frequency w in either radial or axial direction is described by:w,0= 13up/2 1.13The ion trajectory in r, z space resembles a Lissajous figure composed of twofrequency components and w, describing the secular motion with asuperimposed micromotion frequency of p127t (Hz), with the Wro and Oz,ocomponents given by:(Or,o = = f3z(p12 1.14The theoretical treatment given by this section applies only to an ion trapwhich confines a single ion. Although this theory is applicable to researchinvolving single-ion systems, the theory does not account for analyticalsituations of multiple ion storage, which require a consideration of space chargeeffects arising from the confinement of many charged particles. Although thecomplex interactions arising from the storage of multiple ions within the ion trapare interesting, their discussion is beyond the scope of the theoretical treatmentgiven by this thesis. Further discussion of the ion trajectories for multiple ionstorage is given by Benilan and Audoin [37].1.1.2.2 Experimental Applications1.1.2.2.1 Sample IntroductionIon traps have a great diversity of experimental applications. One of thedifficulties with ion trap mass spectrometry is introducing the ions to the storage16field inside the ion trap electrode arrangement. Many schemes for overcomingthis difficulty have been explored. The first commercial ion trap, the FinniganTMIon Trap Detector (ITD), was employed [38] as a detector for gaschromatography. In this trap, ions were generated from gaseous analyte byelectron ionization directly in the storage volume of the ion trap. This eliminatedthe complexities of having an external ion source, then injecting the ionsthrough the electrodes for storage. Ion injection has been investigated for iontraps but has not yet seen wide application. This is because the storage ofinjected ions is phase dependent [39, 40] and has been shown to haveexperimental efficiencies of typically less than 10 %. A problem with electronionization of atoms or molecules with low volatility arises from theirgeneration/transport to the trap as neutrals. Thermal methods for volatilizationdegrade large molecules such as proteins and, although ElectrothermalAtomization can be employed for external generation of atomic species, poortransport efficiencies are observed for the neutral atoms [41] in their delivery tothe storage volume. Large biomolecules can be successfully introduced to theion trap for mass spectrometry by methods such as electrospray [42, 43],membrane interfaces [44, 45] or by Matrix Assisted Laser Desorption Ionization(MALDI) methods [46-49]. Atomic ions can be generated externally and theninjected into the trap for analysis [50, 51]. These ions can also be createddirectly from low volatility samples by laser desorption/ablation directly withinthe storage volume of the ion trap [52, 53]. This method will be presented andexplored by this thesis.Other methods of sample introduction/ionization possible with ion trapsinclude chemical ionization methods. A pulsed reagent gas can be ionized byelectron ionization methods. Allowing for a suitable reaction period, the reagentions can then ionize neutral analyte species by gas phase ion chemistry [54].17This facile method can be used to ionize fragile molecules which would befragmented by normal electron ionization methods, and also allows ionchemistry studies [32] to be conducted with ion trap mass spectrometry. Laserlight can be employed for photolonization of neutral species introduced to theion trap and, in recent experiments at Los Alamos, ion tomography studies [55,56] have been accomplished by laser photodissociation methods within thestorage volume of an ion trap.1.1.2.2.2 Mass Spectrometric CapabilitiesQuadrupole ion trap mass spectrometry allows a wide variety of massspectrometric techniques to be performed with relatively inexpensiveequipment. Much of the fundamental ion trap research in the last few years hasbeen directed towards the exploitation of the many mass spectrometric methodspossible with ion traps. In the recent review by March [31], these new methodsare explained. Most of the scan methods investigated for analysis of stored ionsinvolve the manipulation of the electric fields used for ion confinement. Thisrequires complex electronics and sophisticated computer control over theexperiment, but relatively few new instrumental components. Some of the manypossible mass spectrometric methods include tandem mass spectrometry (MSfl)experiments, selective ion storage, selective ion ejection, mass range extensionby both ion trap electrode modification and by electric potential “RE-Tickle”methods, mass range extension by scan rate control, and ion generation feedback systems to control and quantify the number of ions generated and storedwithin the ion trap. These new experimental methods reflect the great interest inion trap mass spectrometry. Ion trap mass spectrometers are simple, compact,robust, extremely capable and versatile, providing many mass spectrometric18methods with a single instrument; ion traps are becoming an important tool forthe analytical chemist.1.1.2.2.3 Ion Detection MethodsQuadrupole ion traps store ions within a discrete volume, making theminteresting to chemists for both mass spectrometry and gas phase ion chemistryinvestigations, as well as to atomic physicists for probing the structure of singlecharged species. Many of the experimental applications listed above rely uponion detection by ejecting the stored species through the end caps of the ion trap,and then detecting them using an ion detector (such as a channel electronmultiplier) which amplifies the current created by the impinging ions by anelectron cascade amplification. For detection, ions must successfully passthrough apertures in the endcap and, because the ions may be ejected througheither endcap, only a fraction of the ejected ions is actually converted to ananalytical signal. Because chemists are usually interested in the analysis ofmany ions within a sample, this method of ion detection has proven to be auseful means of providing signals corresponding to stored ions. For chemistsinterested in pursuing very low detection limits, and for atomic physics research,this method of ion detection does not provide sufficient sensitivity for a few to asingle ion stored in a trap. Physicists have demonstrated the ultimatespectroscopic limit of detection [57], single ion detection, with ion traps. Theyhave achieved this by utilizing optical detection methods, such as laser inducedfluorescence. Optical detection methods have the advantage in that the storedion(s) may be observed for a longer period (or many times); integrating thesignal(s) obtained produces the enhanced sensitivity for this method. Inaddition, all the stored ions may be observed, which eliminates the signallosses incurred by ion ejection prior to detection mentioned previously.19There have been efforts to observe the stored ions within a quadrupoleion trap by in situ detection methods. This is the oldest method of ion detectionemployed by ion traps. It was first described by Paul and Steinwedel [58] andwas followed by experimental demonstration of the principles in 1959 byFischer [59] and then again in 1967 by Rettinghaus [60]. Fisher’s experimentsdetected ions within the trap by a complex resonance circuit which did notprovide a wide mass range or high degree of sensitivity. Rettinghausdeveloped a sensitive resonance detection circuit for ion detection within an iontrap. Although able to demonstrate detection limits of about 4 ions with hisexperiment, his methods had difficulty with mass discrimination of the storedions. In 1987, Syka and Fies [61] reported that they had demonstrated massselective detection of ions within an ion trap by methods similar to those usedwith Fourier Transform-Ion Cyclotron Resonance (FT-ICR) spectroscopy [62,63]. Their instrument measured the induced image current across the endcapsof the trap and, upon Fourier Transformation of this current, a mass spectrum forstored toluene ions was obtained. Although the mass resolution for this methodwas poor compared to conventional ion trap mass spectrometry, the resultswere attributed to space charge effects within the stored ion cloud, similar tothose observed with FT-ICR instruments at high ion concentrations; the resultsmay be improved by reducing the number of ions stored within the trap.1.1.2.2.4 Alternate Ion Trap GeometryQuadrupole ion traps are constructed with very precisely machinedhyperbolic electrodes, made to exacting tolerances, in an effort to achieveuniform quadrupole electric fields for ion storage and analysis. Chapter 8 of thisthesis discusses the use of cylindrical ion trap electrodes and presentsexperimental results obtained in this laboratory. Cylindrical ion trap electrodes20are much simpler to construct and are therefore less expensive than thehyperbolic electrodes used by conventional ion traps. Cylindrical ion trapshave been studied by Lagadec and colleagues at Orsay [64] where they foundthat cylindrical ion traps were able to store a greater number of ions thanhyperbolic ion traps of comparable dimensions. The mechanical simplicity ofcylindrical ion trap electrodes suggests that they might be very suitable foroptical probing experiments, where numerous apertures through the electrodesfor ion excitation and observation are required.Recently it was announced [65] that FinniganTM ion traps wereconstructed with a stretched endcap spacing. This trade secret provided someimproved performance characteristics, which are investigated and discussedfurther in Chapter 6 of this thesis. The stretched endcap configuration wasintroduced to eliminate mass defects observed for heavier ions stored within theion trap. This empirical geometry modification was found to improve thereproducibility for ion trap mass spectrometry at the expense of resolution.1.2 Laser Ablation1.2.1 Historical PerspectiveSince the discovery of laser action in ruby crystals [66], there has beenextensive use of lasers in science. For additional reading, there are manyexcellent books on lasers and their applications for analytical chemistry [67-72].The first use of laser produced plasmas (LPP) in chemical analysis appeared in1962 when Bretch et. a!. [73] proposed laser microspectral analysis (LMSA) asa useful analytical tool based upon atomic emission spectroscopy (AES) ofLPP. Advantages of using a laser produced plasma as a source/samplingmethod are those of speed and simplicity. Laser produced plasmas sample the21surface of a solid directly, allowing spot analysis with minute sample destruction(—50 pm crater) of both conducting and dielectric materials.Laser produced plasmas used as (atomic) sources, and in materialssampling, have undergone several stages of evolution. Emission studiesdominated the development of LPP as sources, but were quickly followed bytechniques based on absorption and fluorescence. It is only recently that LPPhave been used as a sampling technique for other analytical methods such asinductively and capacitively coupled plasmas (ICP, CCP) and massspectrometry (MS).There are many thorough reviews on the plasmas generated by theinteraction of laser light with solids: Radziemski & Cremers (1989)[74],Moenke-Blankenburg (1986)[75], Adrain & Watson (1984)[76], Fürstenau (1979,1981)[77, 78], Hughes (1975)[79], Krokhin (1972)[80], DeMichelis (1970)[81],and Ready (1965, 1971, 1978)[82-84]. The most recent articles emphasizeanalytical applications of LPP while the earlier reviews concern high densityplasmas of interest for laser induced fusion research. A recent paper by Balazs[85] models the processes which occur during the laser ablation of solids withstrong agreement to experimental findings. In addition, Blades et. al. publisheda paper [86] in which the role of weakly ionized plasmas, including laserplasmas, was examined for material sampling and analysis.1.2.2 Theoretical ConsiderationsThe interaction of a high power laser pulse impinging upon a solid targetis a complex process dependent upon the thermal properties of the target aswell as the wavelength, duration, and irradiance of the laser pulse. Initially,intense, localized heating causes a rapid surface temperature increase of about1010 K-s1 . This heat is then conducted to the target interior producing a thin22molten layer below the surface. Eventually, the energy absorbed by the targetexceeds the latent heat of vaporization. This results in evaporation of materialfrom the target surface. At this stage, evaporation processes govern surfacetemperature, and heat conduction acts only to preserve the thin molten layerbehind the evaporation front.Since vaporization of the target relies upon the energy deposited uponthe molten layer exceeding the latent heat of vaporization (Lv), Krokhin [80]gives the threshold condition:Emin= P’-”1ti1/2J 1.15where Emin is the minimum absorbed irradiance for evaporation, p is the massdensity of the target, a is the thermal diffusivity and ti the laser pulse duration.Below the threshold no evaporation of sample will occur. The mode ofoperation of the laser is an important consideration since t1 can vary greatly. Anormal mode laser pulse consists of a series of output pulses over 150- 300 ps,whereas a Q-switched (QS) mode laser pulse is a single, higher intensity pulse,typically less than 100 ns in duration.After the sample has been vaporized, plasma production must beconsidered. For irradiances just above the vaporization threshold (10- 100GW.m2), evaporation proceeds at the boiling temperature of the target, giving alow density plasma which is transparent to the laser beam. The expansionvelocity and temperature of this plasma depend on the thermal properties of thetarget, and not the laser irradiance. These low density and temperatureplasmas are of use in atomic absorption spectroscopy (AAS), but their low23temperatures may give increased matrix effects making them undesirable asanalytical sources.Alternatively, irradiances well over the threshold (>1 PWm2)generateextremely dense, high temperature plasmas, such as those used in laserinduced fusion studies. Plasmas produced by irradiances between theseextremes (100 GWm2-1PW-m2)are generally the LPP used in LMSA andother methods. The vapor pressure of these plasmas can be so high that laserabsorption is an important process, assumed to occur in a thin, dense, partiallyionized layer between the solid and vapor phases. As the electron density (ne)of this layer approaches the critical electron density (ne) the target becomesshielded from laser light. Furthermore, light is reflected by the plasma if 2 isgreater than the plasma wavelength 2 [79]Xp 10’5(ne) 1.16where n is in units of m3 and w- is in units of nm. Blackbody radiation of thehot gas is the dominant radiative process until the plasma volume increasessuch that 1e is less than n. After the electron density has fallen below n, thevapor becomes highly absorbing and enters a “self-regulating’ regime. Laserlight is absorbed by inverse bremsstrahlung, which causes heating andexpansion, further reducing the vapor density. This allows more light to reachthe target and repeats the process, with the critical density boundaryapproaching the laser at velocities as high as 10 kms1. This process continuesuntil the laser pulse is removed, after which decay and cooling of the plasmaensues.The high expansion velocity and irradiance of the plasma produces asignificant ablation pressure, which can affect the laser-target interaction. For a24laser operating in the normal (non-QS) mode, the continuous heating/cooling ofthe target results in a pool of molten material behind the evaporation front. Thismolten pool is subsequently ejected by the high ablation pressure to formcraters 50-1 00 im deep. In the QS mode, shorter lived plasmas are generatedwith significantly higher ablation pressures. Evaporation of the target occurs ata much higher temperature than the normal boiling point, with most materialremoved to the vapor phase. The craters formed in the QS mode are typicallymuch shallower (a few jim) than those produced by a laser operating in thenormal mode.1.2.3 Direct Solid Sampling Method for Solid MatricesThe use of laser produced plasmas as sources and in material samplinghas many advantages. Because small amounts of practically any solid samplecan be readily analyzed with minimal preparation, errors associated withsample preparation are minimized. The emission of LPP can be studied directlyfor a sample; often enhancements in plasma emission are achieved by usingadditional excitation such as spark, microwave or radio frequency (RE)discharges applied to the plasma. Similarly, absorbance and fluorescencebased techniques can also be applied to LPP but they must be specialized todeal with the relatively short lived and dynamic nature of the plasma. In AAS noadditional excitation of the LPP is required, but for atomic fluorescencespectroscopy (AFS), an additional source of radiation (le.: dye laser) is usuallyapplied to the LPP.Laser produced plasmas can also be used as a sampling method forInductively Coupled Plasma (ICP) and Capacitively Coupled Plasma (CCP)spectroscopy. An ablation cell is employed for this purpose in which the LPP isentrained in a stream of inert gas and then swept into the source as an aerosol25of the sample. A similar arrangement is used in MS except that the bulk of theinert gas is removed from the analyte gas stream prior to introduction into themass spectrometer with a skimmer device. Another interesting use of LPP as aMS sampling technique is that of generating the plasma inside a quadrupoleion trap, allowing several modes of ionization (including laser ablation). Thismethod of direct solid mass spectrometry has been called Laser Ablation IonTrap Mass Spectrometry (LAITMS) [52, 53] and is examined by this thesis.1.3 Scope of ThesisIn this laboratory, we are primarily concerned with the development,characterization and application of plasma based methods for atomicspectroscopy/spectrometry. Important figures of merit for atomic analyticalmethods include simultaneous multielemental capabilities, broad coverage ofthe elements of the periodic table, low detection limits, high selectivity forelements of interest, good accuracy and precision, large dynamic range,freedom from chemical and spectral interferences, ability to accommodate gas,liquid or solid samples easily, high throughput rate and ease of automation forrepetitive analysis. For this project, the specific aim was to develop an atomicmass spectrometric method capable of direct solid analysis of conductors andnon-conductors; one that could provide multielemental detection for traceanalyte species in real sample matrices. It was decided that coupling laserablation for direct solid sampling and ionization with the advantages ofquadrupole ion trap mass spectrometry could address this problem. This thesisexamines the union of laser ablation and ion trap mass spectrometry, andprovides an understanding of some of the associated chemical and physicalprocesses which occur for this method.26For this thesis, Chapter 2 presents preliminary investigations of pulsedruby laser ablation directly within the storage volume of a simple ion trap massspectrometer. Chapter 3 presents further investigation of the laser ablation iontrap combination, using a Q-switched Nd:YAG laser for direct solid samplingand ionization. This chapter also demonstrates potential for the method, butpresents experimental problems. Chapter 4 addresses the experimentalproblems encountered earlier by the development of a new spectrometercapable of specific investigation of Laser Ablation Ion Trap Mass Spectrometry(LAITMS). From Chapter 4 onward, the new spectrometer is used andimproved upon for the direct analysis of solid samples. Chapter 5 evaluates theeffects of sample surface preparation, laser irradiance, laser focus and numberof laser shots upon signals obtained by LAITMS As well, estimated detectionlimits for atomic analytes with the new spectrometer were determined. Chapter6 compares the performance of normal and stretched electrode geometry uponthe operation of the quadru pole ion trap, and how this affects massspectrometry of ions created by laser ablation. Chapter 7 presents anexperiment which demonstrates selective ionization of a single analyte by usinga tunable dye laser, providing enhanced ion signals for that analyte. Chapter 8surveys of the potential use of cylindrical ion trap electrodes for LAITMS,detailing experiments which demonstrate the potential advantages anddisadvantages of using cylindrical ion trap electrodes. Chapter 9 presentsinvestigations where the laser ablation event is synchronized with the phase ofthe applied storage potential. The final chapter summarizes the informationdeduced by each chapter of the thesis, and presents future research directionsfor fundamental LAITMS research.27CHAPTER 2PRELIMINARY INVESTIGATIONS OF THE USE OF A RUBY LASERFOR DIRECT SOLID SAMPLING I IONIZATIONINSIDE A QUADRUPOLE ION TRAP MASS SPECTROMETER2.1 IntroductionFor atomic mass spectrometry, analyte must be converted to gas phaseatomic ions for analysis. Unfortunately, many samples of analytical interestexist as liquids or solids. Analysis of liquids involves some type of vaporizationstep to convert the analyte to a gaseous form. In addition, analytical methodsfor solids typically involve a dissolution step, whereby the sample is convertedto a liquid solution form; this may introduce contamination and also furtherdilute trace analyte species. Methods for direct solid analysis which sample thesolid material directly obviate this pre-dissolution step and its associatedproblems.Many direct solid atomic mass spectrometric methods, such as SparkSource Mass Spectrometry (SSMS) and Glow Discharge Mass Spectrometry(GDMS) are available, but can only be used with solid samples which areconductors or can be combined with a suitable conducting matrix. The REGDMS [51] can analyze non-conducting samples, although the methodprovides only bulk solid analytical information. Direct solid analytical methodsbased upon laser ablation sampling, such as Laser Micro Mass Analyzer(LAMMA) or Laser Ablation-Inductively Coupled Plasma Mass Spectrometry(LA-ICPMS) methods, are capable of direct solid atomic mass spectrometry forboth conducting and non-conducting sample matrices, while also providing28spatial information regarding analyte distribution within a bulk sample.Koppenaal [87] gives a fundamental review of these methods. A limitation ofthese techniques arises during the analysis of trace atomic analytes: lasersampling (and ionization) is a pulsed sampling method, which limits theobservation time for analyte in the mass analyzers used for these methods.One solution to this problem is to increase the observation time for traceanalytes by employing a mass analyzer which stores the analyte. This massspectrometric method would allow increased observation of trace analytes,hopefully increasing the sensitivity of the method for these species.Ion storage atomic mass spectrometry has been demonstrated by severalauthors for both Fourier Transform-Ion Cyclotron Resonance (FT-ICR) MassSpectrometry [88, 89] and Quadrupole Ion Trap (QIT) Mass Spectrometry [50].Both authors were interested in using ion storage mass spectrometry forinvestigating the gas phase ion chemistry of atomic species, but theirexperiments demonstrated that storage mass spectrometry could be used foranalysis of atomic species resulting from laser sampling methods. Thefollowing chapter presents the initial experiments conducted in this laboratorywhereby atomic mass spectrometry was attempted by using laser ablation tosample and ionize solid materials directly inside the storage volume of an iontrap. Although ruby laser ablation coupled with quadrupole ion trap massspectrometry was previously published in an earlier thesis [41], theexperimental results were not very satisfactory. The results presented in thischapter are for spectra not examined in the previous thesis; this chapterestablishes the interest for pursuing an understanding of the processesinvolved in Laser Ablation Ion Trap Mass Spectrometry (LAITMS) for directmass spectrometry of solid samples.292.2 Experimental2.2.1 EquipmentThe ion trap used for these experiments was initially developed [41] foruse as a mass spectrometer for neutral atomic species created in a graphitefurnace electrothermal atomizer. It was modified by machining 1.5 mm diameterapertures through the ring electrode to facilitate laser ablation of solid metalpins. The ion trap mass spectrometer and its operation have been describedelsewhere [41, 53] and in Chapter 3. For these experiments, a TRG DataSystems model 104A ruby laser was used. This laser operated in free runningmode, providing pulses of up to 1 Joule lasting about one ms, with a duty cycleof one pulse per minute. Because a low efficiency forced air cooling systemwas used by the laser, a low duty cycle was required to prevent heat damage tothe ruby crystal.The low shot repetition rate of the ruby laser made alignment for thisexperiment difficult. By removing an access panel on the back of the ruby laser,a Melles Griot (Irvine, CA) model 05-LHR-1 11 helium-neon (He-Ne) laser wasaligned so that the beam was directly down the bore of the ruby rod and opticsof the laser. Both lasers were fixed to a single 15 mm thick aluminum platewhich was mounted upon a 3 point kinematic mount. This mounting systemwas constructed in house by the author so that lasers could be made coaxial,then aligned with respect to the ion trap using the He-Ne laser alone. Aftercoaxial alignment of the lasers, the entire mounting plate was oriented to passthe focused He-Ne beam through the modified ion trap ring electrode (Figure2.1 shows the schematic optical arrangement). To facilitate this process, asimple photodiode and amplifier circuit was constructed (Electronics Shop, U.B. C.). By replacing the sample with the photodiode mounted in a TeflonTM pin,30Ion Trap in VacuumManifold30.0 cm Bi-Convex Lens1 J Ruby LaserFigure 2.1: Optical configuration for ruby laserablation-ion trap mass spectrometry.He-Ne Alignment Laser31the lasers could be oriented (using the He-Ne laser) until a maximum signalwas obtained with the circuit, establishing alignment for the laser ablationexperiments.To synchronize the laser ablation with the ion trap scan sequence, theruby laser was triggered using the computer. The signal normally used toactivate the electron beam source was used to trigger a micro relay chip(#PRMA 2A05) which then triggered the laser for ablation. This complex triggerarrangement was required because a +175 V output existed on the externaltrigger line from the ruby laser. Grounding this output caused the laser to fire. Asimple schematic for the laser trigger circuitry is given in Figure 2.2. Table I listsexperimental parameters.Table I: Experimental parameters for ruby laser ablation in the storagevolume of a quadrupole ion trap.Ruby Laser ablation---Wavelength 694.3 nmPulse width -1 msBeam waist at sample surface 100 jimIrradiance-1.3 X 10 Wcm2Storage and ejection---Storage time 5 msRF. 1.05MHzRE. storage potential * 250 VInitial RF. scan potential * 250 VFinal RE. scan potential * 3600 VRE. scan rate 2.54 X iO Vs1Ion detection--CEM voltage bias -1800 VCEM current amplification 1 X 1 O V A-1Number of scans per laser pulse 1Number of scans per spectrum 1*Measured (0 - peak)32LaserPowerSupplyNB: Not drawn to scaleRelay ChipData Acquisition ComputerFigure 2.2: Schematic diagram of ruby laser trigger circuitryshowing detail of the micro relay chip1’ ‘&—‘L, PRMA2A051iiLaser Head332.2.2 SamplesFor this study, brass metal was machined into 4.0 cm long metal pins witha 0.5 cm dia X 0.5 cm long knob on one end and a 1.45 mm X 1.0 cm pin on theother. The “thins end of the sample pin was inserted into the ring electrodewhile the knob prevented the pin from sliding too far into the electrodeassembly. The sample surface to be ablated was simply cleaned with fineabrasive paper prior to analysis.2.3 Results and DiscussionThe initial results for this experiment were not promising. The control ofthe ruby laser pulse energy was very crude, implemented by adjusting thevoltage supplied to the laser flashlamps by a dial on the laser control box.Pulse to pulse reproducibility for the ablation laser was not calibrated, but wassuspected to be very poor (at best ±20%) because the laser and associatedelectronics were very old. In addition, the low repetition rate of the laser meantthat the experiment was conducted for one laser shot at a time, not averaged forseveral laser ablation events. These undesirable experimental drawbacksundoubtedly limited the accuracy and precision of any possible results from theexperiment. The free running laser produced up to 1J output pulses, but overapproximately one ms. Assuming that the laser was focused to a 0.01 cmdiameter spot (based upon visual inspection of the sample surfaces), the laserproduced typical irradiances of 1.3 X107 W•cm2 . These low laser irradiancesremoved significant quantities of melted sample material (verified by deepcraters on the sample surfaces), and yet produced some ionized species whichwere subsequently stored in the ion trap. Although the ablation laser was firedduring the storage portion of the ion trap scan sequence (see Figure 3.3), thejitter of the laser relay trigger circuitry employed was estimated to be ±5 ms.4.’(0C04.’C0>4.’Cu0C)Cu&ZnIsotopes?300025002000150010005000120Figure2.3a:Massspectrumobtainedforlaserablationof brassinsidethestoragevolumeofaquadrupoleiontrapforonelaserpulseonafreshsurface.2030405060708090100110masslchargeIamuFigure2.3b:Massspectrumobtainedforlaserablationof brassinsidethestoragevolumeofaquadrupoleiontrapforonelaserpulseafter9shotsonthesamesamplelocation.tOC)Cu+64Zfl+66Zfl++63Cu+>(0,CCC)>(Ua)a:300025002000150010005000CHCICH2=C(OH)CH5H2=C(OH)OCH3+/CCI22030405060708090100110mass!chargeIamu12036The ions created by the laser solid interaction were stored for 25 ms; this jitterwould dramatically affect ion storage times (up to 20%). The stored speciesresulting from the laser ablation of a brass sample were then analyzed by iontrap mass spectrometry. The spectra obtained for single laser shots are given inFigure 2.3. These typical spectra illustrate that ions can be generated andstored by laser ablation of a solid sample directly inside the storage volume ofan ion trap. Figure 2.3a was obtained from a single ruby laser pulse focusedupon a fresh sample spot. A variety of unidentified mass peaks are observed inthis spectrum, likely the result of unknown impurities upon the sample surface.After 9 laser shots upon the same sample location, the mass spectrum shown inFigure 2.3b was obtained. Many of the supposed surface impurity ion signalsobserved in Figure 2.3a are much less intense, and ion signals for copper andzinc atomic ions are also observed. Analyte peak assignments are based uponanalytes known to be present in the sample. All matrix species identified inFigure 2.3 are tentative, based on possible fragmentation of organic species.Although promising, the results observed in Figure 2.3b were not easilyreproduced, likely the result of the non-optimized experimental arrangement.The spectra obtained for the laser ablation of brass (Fig. 2.3) exhibitedsignificant space charge [40] signal degradation, resulting from storing toomany ions from the laser ablation process. Ion signals for the atomic speciesare square topped, the result of ion signals which were too large to be digitizedby the data acquisition hardware.2.4 SummaryThis chapter introduced the concept of utilizing laser ablation for solidsampling and ionization directly within the storage volume of a quadrupole iontrap mass spectrometer. Although the experimental apparatus used an inferiorlaser and optical configuration, and had poor control over synchronization of thelaser ablation event, atomic mass spectra were obtained for a brass sample.The spectra obtained were space charge limited and obscured by possiblesurface contaminant fragment ions, yet they suggested that laser sampling Iionization directly within the storage volume of a quadrupole ion trap could be aviable method for mass spectrometry of solid samples. The experimentsconducted for this rather short chapter are not extensive. This thesis is devotedto the development and investigation of Laser Ablation Ion Trap MassSpectrometry (LAITMS) as a mass spectrometric tool for direct solid analysis.3738CHAPTER 3INVESTIGATION OF THE USE OF Nd:YAG LASER ABLATION FORDIRECT SOLID SAMPLING I IONIZATION WITHQUADRUPOLE ION TRAP MASS SPECTROMETRY3i IntroductionThere have been many significant advances in the area of atomicspectrochemical analysis during the past two decades which have led toanalytical techniques with lower detection limits, enhanced multielementcapabilities, improved precision & accuracy and more versatile samplingcharacteristics. Atomic mass spectrometry, in particular, has undergoneexplosive growth with the development and application of techniques such asInductively Coupled Plasma Mass Spectrometry (ICPMS), Laser Micro MassAnalyzer (LAMMA) methods, Glow Discharge Mass Spectrometry (GDMS), andthrough the continued use of Spark Source Mass Spectrometry (SSMS) andSecondary Ion Mass Spectrometry (SIMS). A fundamental review of recentactivity for these techniques has been published by Koppenaal [87].ICPMS is a very powerful technique for atomic mass analysis of solutionsamples. However, it utilizes relatively large amounts of sample in solution form(-1 mL) and requires the dissolution of solid samples prior to analysis. Toovercome this limitation, laser ablation and graphite furnace atomizationcombined with subsequent transport of the aerosol material to the ICPMS havebeen developed. SSMS and GDMS allow direct analysis of solids, but requireconducting samples or samples that can be incorporated in a suitableconducting matrix prior to analysis. Any sample preparation introduces not only39the risk of contamination but inherent dilution of trace analyte species, both ofwhich can be eliminated using a direct solid sampling method such as laserablation, which is suitable for both conductors and non-conductors. LAMMAallows direct analysis of both conducting and non-conducting samples usinglaser radiation for sampling and ionization, utilizing a time-of-flight massspectrometer for mass analysis [90]. One difficulty encountered with theLAMMA technique is that ion yields are strongly dependent on the samplematrix and are sensitive to the presence of elements with low ionization energy.In recent years, quadrupole ion traps have been utilized in a number ofapplications ranging from tissue extract analysis and isotopic dilution [91] tolaser desorption mass spectrometry of biochemical compounds [92]. A recentreview of advances in ion trap technology by Nourse and Cooks [93] discussesthe theory of ion trapping, describes the various operational modes (includingthe use of external ion sources) and outlines some ion chemistry which hasbeen studied with ion traps. Additional theoretical reviews of ion trap operationhave been given by other authors [33, 94, 95].One method of using an ion trap for atomic mass spectrometry is togenerate ions externally and, using ion optics, inject these ions into the trappingvolume. External ion injection methods suffer from several problems; firstly,most of the externally generated ions are not injected; secondly, only a fractionof the injected ions are actually trapped [39]. Computer simulation has shownthat injected ions only remain in the trapping volume after several collisions withbuffer gas molecules and are only stable during a short time window when theRE voltage is near zero [40]. Although quadrupole ion trap atomic massspectrometry of ions created by an external glow discharge has been recentlydemonstrated [51], a potentially more successful method for trace analysis is thegeneration of atomic ions directly inside the trap. Laser ablation of solids inside40an ion trap allows atomization and ionization directly within the trappingvolume.Freiser and associates [39, 88, 89] have studied both atomic and clusterion chemistry of metals using Fourier Transform Ion Cyclotron Resonance MassSpectrometry (FT-ICRMS) coupled with laser sampling. Their research relieson ablation and laser ionization of metal samples inside the trapping cavity of aspectrometer. This is related to the research focus for this thesis, which is thedevelopment of an instrument for atomic mass spectrometric analysis of solidsamples using laser ablation directly inside an ion trap mass spectrometer. Iontrap mass spectrometry provides an alternative to FT-ICRMS, while also havingpotential advantages. Ion traps allow higher operating pressures, are lower inprice, and are far more compact than FT-ICRMS instruments. However, currention trap mass spectrometry does not provide resolution as high as that availablefrom FT-ICRMS.Chapter 2 demonstrated the potential for direct atomic mass spectrometryof solid materials by laser sampling and ionization directly inside the storagevolume of a quadrupole ion trap. Although the experimental arrangement usedfor the previous chapter was not very successful because meaningful spectrawere not easily reproduced, the idea of direct solid sampling inside the storagevolume of an ion trap warrants further investigation. To this end, anexperimental system was constructed using simple components alreadyavailable in the laboratory. A Nd:YAG laser was borrowed for a short period(Dr. A. Merer, Chemistry Department, U. B. C.), and after an introduction to safeuse of high pulse energy lasers (Dr. A. Adam, now at the University of NewBrunswick), the new spectrometer was able to generate meaningful atomicmass spectra for metallic solid samples. For these experiments, samples weredirectly ablated and ionized in the ion trapping volume, providing some unique41characteristics and capabilities. This research has been published [53] and theexperimental details and results for the method, called Laser Ablation Ion TrapMass Spectrometry (LAITMS), are presented in this chapter.3.2 Experimental3.2.1 Ion Trap Mass SpectrometerFor all experiments described in this chapter, an ion trap of our owndesign was used, constructed by The Mechanical Services Shop (ChemistryDepartment, U.B.C.). The end cap and ring electrodes were made of type 304stainless steel, having complimentary surfaces defined by equations 1.6 to 1.8,where r0 is 10.00 mm. A schematic diagram of the electrode configuration isgiven in Figure 3.1. Figure 3.la is a schematic diagram of the arrangement ofthe three electrodes comprising the ion trap. In order to admit the sample, a 1.5mm diameter aperture was machined through the ring electrode, allowingsuitably sized sample pins to be mounted flush with the electrode inner surface.The laser beam aperture was drilled 180° to the sample position, with a 1.5 mmdiameter opening which flared in a conical shape allowing the laser beam to befocused at the sample pin tip (Figure 3.lb).An experimental block diagram for the ion trap system is provided inFigure 3.2. During preliminary studies and for calibration purposes, neutralanalyte molecules and/or atoms were ionized with electron ionization (El). Agated electron beam source was fashioned using a rhenium filamentsurrounded by a stainless steel ring. The ring was maintained at -70 VDC; thefilament remained at 0 VDC relative to the ring except when pulsed to -70 VDCusing a pulse amplifier built by the Electrical Services Shop (ChemistryDepartment, U.B.C.). For El purposes, the electron beam was pulsed for a 1 msinterval. Ionized analyte species were then stored in the ion trap by applying a42Sample PinApertureVespel TMSpacersElectron BeamSource Aperture—Endcap Electrode4— Ring ElectrodeLaser Aperture-a—— Endcap ElectrodeFigure 3.la: Schematic ion trap configuration with modificationsallowing laser ablation inside storage volume. Inpractice, the electron beam source and the channelelectron multiplier are recessed into the respective endcap apertures shown in dotted lines.Channel ElectronMultiplier Aperture43(i)(ii)SampleFigure 3.lb: A three dimensional representation (I) of the ringelectrode showing the laser aperture and across-sectional view (ii) of the ring electrodeshowing sample pin placement.Laser Input44Figure 3.2: Block diagram of the ion trap mass spectrometer system.45constant storage radio frequency (RE) potential (250 V, 5 ms) to the ringelectrode and grounding the end cap electrodes. The RE quadrupole supplyused was an Extranuclear Laboratories Inc. (Pittsburgh, PA) model number011-1, modified for operation from 0.6-3.0 MHz. Capacitance matching wasachieved using an Extranuclear Laboratories Inc. High-Q Head model number012-16. The operational frequency used for these experiments was 1.05 MHzwith a maximum RE output (0-p) of 3500 V.All experimental parameters were controlled using a Zenith DataSystems (St. Joseph, Ml) IBM-AT compatible computer equipped with aQuaTech PBX-721 (Akron, OH) data acquisition parallel expansion board, aQuaTech model DM12-b digital-to-analog converter and an RC Electronics(Santa Barbara, CA) model ISC-16 analog-to-digital converter. A timingscheme for the experiments is given in Figure 3.3. The electron beam wasturned off for all the laser ablation experiments. All software used for control ofexperimental parameters and data acquisition was written in Borland TurboPascal Version 3.0. A copy of the code used for data acquisition with the iontrap is given in APPENDIX I.Upon ramping the RE voltage above the storage potential (250-3500 V),the stored ions were sequentially ejected along the z-axis of the ion trapaccording to increasing mass to charge ratio. This ion trap operational modewas termed “mass selective instability” by Stafford [96] and co-workers. Theseions were detected using a Galileo Electro-Optics Corp. (Sturbridge, MA) model4870 channel electron multiplier. Prior to computer data acquisition, the analogsignal from the channeltron was amplified using a Keithley (Cleveland, OH)model 427 current amplifier. During the RE voltage ramp, the amplified signalis sampled 4096 times, the entire process requiring approximately 120 ms. Atmaximum ramp voltage the RE voltage was reduced to zero to allow any46Electron/LaserPulseRF VoltageData AcquisitionTriggerData CollectionB B.Iii. 1,11 . 1i. .IiI —Mass SpectrumFigure 3.3: Timing diagram for the ion trap experiments47remaining ions in the trapping volume to escape.To achieve the low pressure required for normal ion trap operation, theion trap was encased in a simple manifold constructed of stainless steel.Electrical feed through connections were made on the flange to which the iontrap was secured. The manifold was evacuated using an Edwards (Oakville,Ontario) model 100/300 diffusion pump filled with Varian (Lexington, MA)Santovac 5 diffusion pump oil. Oil back streaming from the diffusion pump wasminimized using an Edwards “Peltier Effect” thermo-electric cooling baffle,model DCB100. A Welch Scientific (Skokie, IL) model 1397 rotary vacuumpump was used to rough pump the system and back the diffusion pump. Theentire pump and baffle portion of the vacuum manifold was isolated from the iontrap chamber by an Edwards model QSB100 butterfly valve, allowing samplechanges without shutting down the entire vacuum system. Manifold pressurewas monitored with a Varian model 880 vacuum ionization gauge, using amodel 531 thermocouple and a model 564 broad range ionization gauge tube.Helium buffer gas was introduced with a Granville-Phillips (Boulder, CC)variable leak valve to maintain a constant experimental pressure of 2.5 mTorr(corrected for He [97]). Gaseous analyte was introduced by a second variableleak valve connected to a 1.5 mm OD stainless steel capillary passed throughthe ring electrode. Normally, gaseous analyte was introduced by head spacesampling above a liquid sample through which a small stream of helium wasaspirated at atmospheric pressure.3.2.2 Laser and Optical ConfigurationA Lumonics (Warwickshire, England) model HY400 Nd:YAG laserequipped with a frequency doubler (532 nm) was used for laser ablation.Pulses of 5 mJ lasting 10 ns were used in the experiments. Ablation laser pulse48lontrapLens (ft = 300.0 mm)Cube B.S. (12.5 mm)Mirror 2He-Ne LaserNd:YAG LaserMirror 1Figure 3.4: Optical configuration for laser ablation ion trap mass spectrometry.49energies were determined using a Scientech (Boulder, CC) model 365 LaserPower I Energy Meter with a model 380101 Optical Pyrometer Detector. Figure3.4 shows a schematic diagram of the optics used for laser ablation. Opticalalignment of the system was achieved using a Melles Griot (Irvine, CA) He-Nelaser (model 05-LHR-11 1). The Nd:YAG pulses were directed into the ion trapusing CVI Laser Corp. (Albuquerque, NM) model M318736 high energy lasermirrors. The focused beam diameter was 0.2 mm with an irradiance ofapproximately 1.6 XiO Wcm2.3.2.3 SamplesFor calibration purposes, xenon gas (10% in He, Spectra Gases Inc.,Newark, NJ) and carbon tetrachloride (Omnisolve grade, BDH Chemicals,Toronto, Ont.) were analyzed by electron ionization in the ion trap. Solidsamples used in this study included a silver solder and a Pb/Sn soft solderobtained from the Mechanical Services Shop (Chemistry Department, U.B.C.).All samples were freshly cleaned prior to analysis using very fine abrasivepaper, rinsed successively in concentrated nitric acid, de-ionized water thenacetone (Omnisolve grade, BDH Chemicals, Toronto) and finally, dried in air.3.3 Results and Discussion3.3.1 CalibrationExperimental characterization of this ion trap was accomplished usingelectron ionization (El). Gaseous Xenon and carbon tetrachloride (see Section3.2.1) were introduced to the ion trap via the leak valve and were subsequentlyanalyzed, allowing not only mass scale calibration but evaluation of theresolution [98] achieved by the ion trap mass spectrometer. Mass spectraobtained for these analytes are given in Figure 3.5. Figure 3.5a is0IC)z>14-’(I)CC>Q25x120151050117119cc13+cc12+828486121mass/chargeIamu123Figure3.5a:Electronionization(El)massspectrumusedforcalibrationofcarbontetrachioride.U,6000-129V1305000-131132Xeci4000-3000-1342000-136IIIIIIIIIIIIIII•IIIIIIIII••IIIIJIIII100105110115120125130135140145150masslchargeIamuFigure3.5b:Electronionization(El)massspectrausedforcalibrationofxenon.52an El spectrum of carbon tetrachioride obtained with a resolution of 221 for mlz82, 84. The mass spectrum of Xe is given in Figure 3.5b, showing signals forthe various isotopes of xenon. Because the mass scale in our experiments waslinear, calibration was possible using any two mass peaks. The ion trap usedfor these experiments had an upper mass limit of approximately 270 amu,adequate for the studies presented by this thesis.3.3.2 Spectra Obtained for Solid SamplesIon trap operation using LAITMS showed similarities with the El mode ofoperation. Both techniques require He buffer gas to increase the resolution ofhigher mass ions by restricting the kinetic energy distribution between ions [96].Also, when too many ions are stored in the trapping volume, resolutiondegrades in both modes as a result of space charge effects [40]. Signalsaturation during the El mode results from a combination of buffer gas pressure,ionization current from the filament, and the duration of the ionization; LAITMSsignal saturation resulted from excessive laser power and buffer gas pressure.An uncharacterized “surface” effect was observed for the first few laser pulsesresulting in noisy spectra and poor reproducibility, possibly due to impurities stillpresent on the sample surface. After 100 laser pulses, the signal obtained fromthe solid metal samples showed peaks for the various isotopes known to bepresent in the sample matrices. Because the samples used for theseexperiments were not previously characterized standards, the experimentalspectra were used to demonstrate the potential of the technique rather than togive quantitative information. As in the El mode, spectra were summed for anumber of scan sequences in LAITMS to increase the signal-to-noise ratio.Figures 3.6 and 3.7 depict spectra acquired for several metal samples, obtainedby summing the spectra from 100 consecutive laser pulse/scan cycles. FigureC)Lx,4000-120Sn3000-118Sn—.;;.2000-0116Sn—1Sn>01000-,)24Sn\•____4A0-11111111111ii.iiiiiiiiiiiiiii.IIIIIIIIIIIIII1111111IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII5060708090100110120130140mass/chargeIamuFigure3.6a:LAITMSspectraofsoftsolderatmass/chargerange50-140amu.U,6000-5000-z4000-4-.C.$3000-C—2O7rn+a)>0—IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII170175180185190195200205210215220225230235240mass/charge/amuFigure3.6b:LAITMSspectraofsoftsolderatmass/chargerange170-240amu.U,U,12x103-112CdllOCd+\+\10-109Agai8->1CC)65Cu4.’-66Zn/117C) U)L116Cd63CuSI./67Z&106Cd—1192JnjCI3+0-IIIIIIlIIIIIIItIIIIIIIIIIIII,IIIIIIIIIgIIIIIIlIIIIIIIIIIIIIIIIIIlIIIIlIIIIIIIIgII,IIIIIIII5060708090100110120130140mass/chargeIamuFigure3.7:LAITMSspectrumofsilversolder.563.6 shows the ion signals obtained for the Sn and Pb isotopic clusters for asoft solder sample (approximately 50/50 by weight). Figure 3.7 is the spectrumobtained from a silver solder sample, showing isotopic mass peaks for Cu-’-, Zn,Cdt, and Ag. Variation from literature abundances can be partially accountedfor by isobaric interferences from the sample (le. 114Cd and ll4Sn+) as well asionization of contaminants or residual analytes present as neutral species in thebuffer gas (ie. 119Sn and m/z=119 for residual CCI3+).One difficulty encountered with ion trap mass spectrometry is that thedynamic range available can be restricted as a result of space charge effects.High charge density can distort the quadrupole field, changing the stability limitsof the trap. Careful selection of the energy and power density of the incidentlaser pulses allows for control over the number of charged atomic speciesgenerated in the cavity of the trap, thus helping to reduce the effects of spacecharging. The laser used for these experiments was designed for pulses of 200mJ; the 5 mJ pulses used were made by simply reducing the discharge voltagesupplied to the flashlamps. This method of pulse energy regulation suffers fromsome degree of pulse to pulse energy variation, which contributes to thedegraded signal-to-noise ratio observed for the LAITMS spectra (Figures 3.6and 3.7) when compared with that of a El spectra (Figure 3.5). Otheroperational modes for ion traps involving DC as well as RF operation [91] allowmass selection of specific analyte ions (storing species of interest, reducingspace charge effects) as well as increasing the resolution achieved with an iontrap. Precise control of the laser irradiance should allow control over thequantity of material sampled and ionized, reducing space charge effects. Thistopic is further addressed in Chapter 4.573.4 SummaryA new method for direct solid atomic mass spectrometry, called LaserAblation Ion Trap Mass Spectrometry (LAITMS), has been presented in thischapter and published elsewhere [53]. The experimental spectrometerdeveloped for this chapter provided atomic mass spectra for solid metalsamples. At this stage of development it is not yet clear whether LAITMS couldbe a viable analytical methodology, but it could be a useful alternative to theuse of LAMMA and FT-ICRMS for investigating the products from laser ablationof solids [88, 89] and dried solution samples at pressures below io Torr. Also,laser sampling offers a practical method of creating analyte ions (from anexternally introduced solid sample matrix) directly in the trapping volume of anion trap. The use of an ion trap mass spectrometer allows the collection andstorage of ions generated by several laser pulses, thus offering the potential forenhancing sensitivity for trace elemental analysis. For example, ions generatedfrom successive firings of the laser may be stored and integrated inside thetrapping volume to enhance the signal-to-noise ratio during readout. Also,LAITMS may be used as a method for ablated neutrals mass spectrometrybecause the electron gun can be used to ionize neutral atoms and moleculescreated using laser ablation. Clearly, post ionization is an intrinsic feature ofLAITMS.Ions formed in the manner described in this chapter could be probedoptically, for example, through the use of fluorescence spectroscopy. It hasbeen shown that the combination of ion traps and atomic fluorescencespectrometry can be used for the detection of a single ion [57].For this chapter, the sensitivity of the LAITMS method has not been fullyinvestigated, nor have the operational characteristics (laser power, pressure,sample type, etc.) been studied in detail. Such studies are the basis for theremaining chapters of this thesis. In the next chapter, a new LAITMSspectrometer is developed which facilitates further investigation of this newanalytical method.5859CHAPTER 4INVESTIGATIONS OF AN IMPROVED ION TRAP MASS SPECTROMETERUSING Nd:YAG LASER ABLATION FOR DIRECT SOLID SAMPLING /IONIZATION4.1 Improved Design ConsiderationsChapter 3 presented published [53] experiments with a quadrupole iontrap mass spectrometer in which solid samples were directly sampled andionized by laser ablation inside the storage volume of the trap. Theseexperiments allowed atomic mass spectrometry of metal alloys anddemonstrated the potential for this new method of direct solid massspectrometry. This method was called Laser Ablation Ion Trap MassSpectrometry (LAITMS), and its development as a new analytical tool isinvestigated in the remaining chapters of this thesis. Several potentialadvantages of this approach to direct solid atomic mass spectrometry wereidentified. The use of an ion trap allows the collection and storage of ionsgenerated from several laser pulses, offering the potential for enhancingsensitivity. Also, ions created in this manner are available for alternate meansof detection, such as fluorescence spectrometry [99].The experiments of Chapter 3 and the first LAITMS publication [53] usedthe existing spectrometer which was modified to demonstrate the method. Theion trap itself was originally used for electron ionization of neutral analytes,although it had been modified to allow ruby laser ablation experiments. Theseexperiments are discussed in Chapter 2 to allow a chronological developmentof LAITMS in this thesis. Direct solid analysis using the spectrometer featured60in Chapter 3 was inhibited by many experimental obstacles. The massspectrometer did not facilitate sample changes, because solid sample pinswere inserted radially through the ring electrode prior to installation in thevacuum manifold. To change a sample, or expose a fresh sample surface to theablation laser, the entire ion trap was removed from the system, a new sampleinstalled, the trap replaced, the vacuum re-established and then the experimentrepeated. This fact alone made direct solid analysis with the Ch. 3 spectrometera curiosity rather than a method of analysis, typically because only one samplecould be analyzed in a working day. Optical alignment of the ablation laser alsopresented problems, because no line of sight existed through the ion trapmanifold assembly when the ion trap was in position. The optical arrangementused for the earlier investigations [53] with a Nd:YAG laser was highlyunsuitable, partially due to the temporary nature of these experiments (a shortloan period for the Nd:YAG laser) and the lack of proper optics and tables forthe experiment in our laboratory at this time. The ion trap vacuum manifold wasevacuated by using an oil diffusion pump with a Peltier effect cooled baffle. Thisvacuum pump took several hours to achieve a steady vacuum and introducedneutral contaminants to the manifold via back streaming through the bafflesystem, interfering with the performance of the ion trap mass spectrometer. Theion trap electrodes themselves were modified to pass a sample and ablationlaser beam, but no other optical path existed to allow optical detection schemessuch as laser induced fluorescence.The new spectrometer presented in this chapter was designed toaddress the problems listed above. Moreover, several other changes weremade to improve the spectrometer for future LAITMS experiments. Theimprovements included a differentially pumped sample probe, ion trapelectrodes with extended surfaces and optical paths for fluorescence or other61optical ion interrogation methods, an optical table for the entire experiment, animproved optical configuration for ablation, modified ion detection apertures inthe end cap electrodes and a turbomolecular vacuum pump. These and othermodifications are detailed in Section 4.2. This chapter investigates some of thedifficulties encountered in the early LAITMS experiments and exploresimprovements in the method through the design and development of the newspectrometer.4.2 Experimental4.2.1 Ion Trap Mass SpectrometerA new spectrometer was designed and constructed to improve upon theoriginal [53] in order to conduct the experiments (subsequently published [52])presented in this chapter. All machining for the new spectrometer was carriedout by the Mechanical Services Shop (Chemistry Department, U.B.C.); theelectronics used are described in Chapter 3 and have been previouslypublished [53], with all modifications done in house by the Electrical ServicesShop (Chemistry Department, U.B.C.).The hyperbolic end cap and ring electrodes used for the experiments in thischapter were machined from type 304 stainless steel, having complimentarysurfaces defined by equations 1.6 to 1.8, where r0 = 10.00 mm. Althoughdescribed by the same equations, the electrodes used for these experimentsdiffer from the electrodes used in Ch. 2 and Ch. 3 in that they have extendedhyperbolic surfaces carried out to a radius of 30.00 mm to provide morehomogeneous quadrupole electric fields for ion storage and analysis. Aschematic diagram of a cross section through these new electrodes is given inFigure 4.1. Experimentally, the new electrodes gave 15-20% better massresolution than the electrodes used for earlier investigations [53]. In addition,62Sample PinApertureElectron BeamSourceLaserApertureFigure 4.1: Cross sectional diagram of the ion trap electrodes showing theoptical path through the ion storage volume.To Ion detector63an optical path passing through the asymptote of the radially symmetrichyperbolic electrode surfaces has been established to facilitate future opticalprobing studies of stored ions. Electron ionization was performed using anelectron source identical to that described previously [53] in Chapter 3. Thesample and laser apertures machined through the ring electrode were similar tothose in the previous ion trap used for LAITMS, with inner diameters of 1.5 mmat the inner surface of the electrodes to minimize electric field distortion. Thechannel electron multiplier (CEM) mounting flange was recessed into the lowerend cap to position the detector as close to the center of the trap as possible.The ion trap described in this chapter was housed in a new vacuummanifold, constructed in house from stainless steel; schematic diagrams of thistrap and manifold are given in Figure 4.2. High vacuum was maintained in theion trap manifold by using a Pfeiffer-Balzers (Mississagua, Ont.) turbomolecularpumping system, consisting of a model TPU 170 turbomolecular pump, a TCP300 pump control system and a DUO 010 B rotary vane pump. The sampleintroduction port was differentially pumped using a Varian (Lexington, MA)model SD 90 rotary vane pump. To eliminate oil back streaming into thesample introduction probe, a liquid nitrogen trap was used. Manifold pressureswere monitored with a Balzers model PKG 020 pirani-cold cathode gaugemeter, equipped with a model IKR 020 cold cathode gauge head for the mainvacuum chamber and a model TPR 010 pirani gauge head for the sampleintroduction probe. All experiments were performed with the cold cathodegauge off because it’s operation saturated the OEM signal. Helium buffer gaswas introduced to the main vacuum manifold by using a Granville-Phillips(Boulder, 00) variable leak valve. The sample probe assembly was purgedwith helium via a needle valve during sample changes to prevent residual airintroduction. All helium used in these studies was pre-purified grade (99.996 %64(I)E(ii)C,JJ, Turbomolecular pumpFigure 4.2: Diagrams of the ion trap vacuum manifold Ci. front and ii. rear views)used for LAITMS experiments. A = radio frequency (RF) powersupply input feedthrough; B = GEM signal out feedthrough; C =variable leak valve for gaseous sample introduction; D = variableleak valve for He buffer gas introduction; E = pirani cold cathodevacuum gauge; F = optical ports to enable probing studies of thestored ions; G = differentially pumped sample probe assembly;H = mechanical vacuum pump with in line liquid nitrogen trap.FBD CFG65pure). Gaseous analyte for electron ionization experiments was introduced tothe ion trap via a second variable leak valve connected to a 1.5 mm ODstainless steel capillary which was passed through the ring electrode (aVESPELTM sleeve was used to provide electrical insulation) of the ion trap.Gaseous analyte was introduced by head space sampling above a liquidsample through which a small stream of helium was aspirated at atmosphericpressure.The electronics used for the operation of the new ion trap, describedherein, were the same as used previously [53], with modifications to adapt themfor use with this system. Included in the modifications were simple coaxialshielding for the Channel Electron Multiplier (CEM) signal out line and RadioFrequency (RE) input lines inside the manifold (to reduce possible noise in thesignal from electrical interferences). Ions were detected using a Detech(Brookfield, MA) model 401 CEM; the ion trap was operated using the “massselective instability” mode described by Stafford and co-workers [96], and alsoin Chapter 3.4.2.2 Laser and Optical ConfigurationA Lumonics (Warwickshire, England) model HY400 Nd:YAG laserequipped with a frequency doubler (532 nm) used for laser ablation with thenew spectrometer. Figure 4.3 shows a schematic diagram of the optics used forthese investigations. Optical alignment of the system was achieved using aMelles Griot (Irvine, CA) Helium-Neon laser (model 05-LHR-111). Because a10 % beam splitter was used to sample the ablation laser beam, lower energypulses (10 % of laser output) could be used for these experiments. Ablationlaser pulse energy calibration was accomplished using a Scientech (Boulder,CO) model 365 Laser Power I Energy Meter with a model 380101 Optical66High Energy LaserMirrorDifferentiallyPumped SampleProbeIon Trap &ManifoldQ-SwitchedNd:YAGLaserA = Spatial Filter30cm LensABeamDumpA10%B.S.AHe-Ne Alignment LaserFigure 4.3: Optical configuration for new LAITMS spectrometer.67Pyrometer Detector. In addition, the silica piano convex lens (f = 30.00 cm)used to focus the laser upon the sample surface was mounted upon atranslation stage to allow precise adjustment of the laser focus upon the samplesurface.4.2.3 SamplesInitial characterization of ion trap performance and mass scale calibrationwas accomplished by electron ionization of carbon tetrachloride (Omnisolvegrade, BDH Chemicals, Toronto, Ont.) and the mass scale calibrant FC 43(Perfluoro-tri-N-butylamine, Pierce Chemicals, Miami, FL).A variety of samples were analyzed by LAITMS for this chapter,including type 308 stainless steel, silver solder, rhodium foil, molybdenum,VESPELTM and MACORTM. All samples used for these studies were machinedinto 1 cm long pins which were 1 .48 mm diameter for at least half their lengthand 1.8 mm for the remainder to facilitate sample probe mounting forintroduction to the ring electrode. The rhodium foil sample was prepared bycementing a small disc of foil to the end of a type 308 stainless steel samplewith epoxy resin (The Borden Company, Toronto, Ont.). Prior to analysis, thesample surface was polished using methanol (Omnisolve grade, BDHChemicals, Toronto) wetted 400 and then 600 grade abrasive papers. The pinswere sequentially washed with de-ionized (Dl) water, concentrated nitric acid,Dl water, methanol, and then air dried. Samples were aligned flush with thering electrode inner surface using a helium-neon alignment laser.4.2.4 Scanning Electron Microscopy (SEM)The electron micrographs of the sample surfaces were obtained with theuse of an Hitachi (Tokyo, Japan) model S-2300 electron microscope68(Metallurgical Engineering, U. B. C.). VESPELTM and MACQRTM samples wereprepared for SEM by sputtering a thin film of Au/Pd (40% Au / 60% Pd) overthem with a Hummer-V Sputter Coater (Tecnics Inc., Alexandria, VA).4.3 Results and Discussion4.3.1 CalibrationInitial characterization of the new ion trap was achieved through electronionization (El) experiments in which carbon tetrachloride and FC 43 were usedas samples. Experimental parameters for the electron ionization experimentsare given in Table II.Table II: Instrumental parameters for electron ionization calibrationexperiments.Ionization---Electron energy 70 eVDuration of electron beam 1 msStorage and ejection---Storage time 25 msRE. 1.05MHzRE. storage potential * 250 VInitial RE. scan potential * 250 VElnal RE. scan potential * 3600 VRE. scan rate 2.54 X io V•s-1Ion detection--CEM voltage bias-1800 VCEM current amplification 1 x io V•A-1N umber of scans per spectrum 100* measured (0 - peak)cci33000-(U4-’Co2000-C0>4-’(U(U1000-cci2—w0—111111111111IIIIPIIIIIIIIIIIIIIIIIIIIIIIP1111111111liii1111111..I..••I707580859095100105110115120125130135140mass/chargeIamuFigure4.4:Electronionization(El)massspectrumusedforcalibrationofcarbontetrachioride.0CuCowCQ>CuC)Figure4.5:Electronionization(El)massspectrumusedforcalibrationofamixtureof perfluoro-tri-N-butylamifle(FC43)andcarbontetrachioride.CF380006000400020000cc13++g%C+ø315++CF+36++cc12÷C4F6I&+6080100120140160180200mass/chargeIamu71Figure 4.4 shows the El mass spectrum obtained for a carbontetrachioride sample, showing excellent resolution for the various fragment ions.A second sample consisting of 0.5 mL of FC 43 and 0.5 mL of carbontetrachloride was prepared and analyzed. The El mass spectrum for thissample is given in Figure 4.5. This spectrum shows the various ion signalsexpected for FC 43 and those for carbon tetrachloride observed in Figure 4.4.The positive identity of these various fragment ions allowed mass scalecalibration for the new quadrupole ion trap.The analytical performance of the new ion trap electrodes was alsoevaluated prior to LAITMS experiments by electron ionization (El) of carbontetrachloride samples. Experimental parameters for these studies are given inTable II. Figure 4.6 plots the mass resolution as a function of total pressure(corrected for He [97]) for El carbon tetrachloride mass spectra (m/z=1 17, 119,121), the average of three separate experiments (± 2cr). The resolving power ofthe ion trap was determined by using the 5% peak-height definition used byBradshaw and co-workers [100]. These experiments suggested that anoptimum manifold pressure of 0.5-1.5 pEar He (1 Bar 1 X iO Pa) provided aresolution of 240. This resolution and pressure are consistent with literaturevalues [96] for optimum ion trap operation.To calibrate the ablation laser output, the pulse energy for ablationpurposes was determined using a laser power meter located outside thevacuum manifold. The result of this calibration is given in Figure 4.7, which is aplot of laser pulse energy versus flashlamp voltage for the laser. Extrapolationfrom this curve allows direct calibration of the laser pulse energy used inLAITMS experiments. This energy calibration was repeated periodically toensure accurate readings because flashlamp output (and, thus, laser output forzCuC04-.0Cow240220200180160Figure4.6:Plotofaverageresolution(error=±2G) forelectronimpactionizedcarbontetrachlorideions (mlz=117,119and121)forthreereplicateexperimentsversusheliumbuffer gaspressure(1bar=1XioPa).0.51.01.52.02.5PressureImBarHeC)•PulseEnergy=168.5050.647095*X+O.OOO621337*X12I1.5--,E>0)U.a,Cwa,U,z00.5-111111111111111111111111111111111111111111111111111111I111111520530540550560570580OscillatorEnergyI VFigure4.7:Plotof laserpulseenergy(error=±2a)versusoscillatorflashlampvoltageaveragedfor 200lasershots.74a given setting) slowly decreased upon normal operation over an extendedperiod.4.3.2 Spectra for Conducting Solid SamplesPrevious studies reported [53] that the LAITMS spectra obtainedexhibited significant peak broadening and a degraded signal to noise ratioresulting from space charge effects in the trapping volume. This problem wasaddressed by adjusting the laser energy to reduce the amount of sample beingablated and ionized per laser pulse. For these investigations, the laser pulseenergy was adjusted by reducing the flash lamp voltage. The pulse-to-pulseenergy reproducibility problems previously encountered [53] were greatlyreduced by sampling 10 % of the laser output beam for ablation purposes,allowing flash lamp operation at higher voltages. Further investigation of theeffects of laser irradiance is given elsewhere [52] and in Chapter 5.LAITMS spectra were obtained for a variety of conducting samples.Experimental parameters for these experiments are given in Table Ill. Figure4.8 shows mass spectra for type 308 stainless steel and silver solder samples.These spectra show much better resolution and signal to noise ratios than thespectra presented in Chapter 3, due to the improved laser pulse-to-pulsereproducibility and cleaner vacuum environment provided by the newspectrometer used for these experiments. In addition, the analysis of stainlesssteel addressed a reviewer’s criticism of the initial publication [53] in which itwas pointed out that only samples with low melting points were analyzed.Clearly, the LAITMS spectrometer is capable of analysis of higher melting pointmetal materials.75Table Ill: Instrumental parameters for LAITMS experiments.Laser ablation---Wavelength 532 nmPulse width 10 nsBeam waist at sample surface 50 jimIrradiance refer to spectra (-P1 X 1010 W-cm2)Storage and ejection---Storage time 5 msRE. 1.05MHzRE. storage potential * 250 VInitial RF. scan potential * 250 VFinal RE. scan potential * 3600 VRE. scan rate 2.54 X io V•s-1Ion detection--CEM voltage bias -1800 VCEM current amplification 1 x io V-A-1Number of scans per laser pulse 1Number of scans per spectrum 100*Measured (0 - peak)Co54Cr+&5000-56Fe+4000-ci52Cr+3000-N002000-4e+053Cr+1000-\,58N1+/,...__-6ONi+88Sr4I___)Ni-*Jw1ii.-arkWJIWLb4JJ0—1111111I11111111111111IIllillIllIIIIIIIIIIIILIIIIIIIIIIIIIIIIIIJI111111111IIIJIIIItII•I2030405060708090100110masslchargeIamuFigure4.8a:LAITMSspectraobtainedfortype308stainlesssteelusinglaserpulseirradiancesof 5.2X1010•W•cm2N2500-113cd+114cd+112cd+2000-‘NlllCd+1500-llOCd+\U)C109A1o7Ag+116cd+1000-Cu106cd+/500-ZnZnN0-liii111IIII1111111IIIIIIIII1111111111111IIIIIIIIIIIIItIIIIIlIIIIIIIIIlIIIIIIIIIIIIII5060708090100110120130140mass!chargeIamuFigure4.8b:LAITMSspectraobtainedforsilversolderusinglaserpulseirradiancesof3.2X1010W•cm2.78A potential test of the usefulness of LAITMS arose in which a sampleused for surface chemistry experimental apparatus (Dr. K. Mitchell, ChemistryDepartment, U. B. C.) of unknown origin was analyzed by LAITMS. The sampleconsisted of a small shard of metal foil, 0.5 mm thick. A portion of the shard wasmounted upon a sample pin as described in Section 4.2.3. The sample wassuspected to be pure molybdenum foil; the mass spectrum for the foil is given inFigure 4.9. Upon mass scale calibration of the spectrum, the low mass peakswere found to be sodium and potassium (possible surface contaminants) andthe single higher mass peak corresponded to m/z = 103. Because rhodium ismonoisotopic with a mass of 103 amu, it was assumed that the unknown foilsample was, in fact, pure rhodium and not molybdenum. This could beconclusively verified by a second spectroscopic method, such as ICP-AES;however the isotopic singularity of the analyte signal at this mass providesstrong evidence of identity. As a verification experiment, a pure molybdenumsample was obtained (S. Rak, Glass Shop, U. B. C.) and analyzed by LAITMS.The spectrum for this sample is shown in Figure 4.10, and it is clear that thisspectrum is not the same as Figure 4.9. The ion signals observed were not forpure Mo, but instead oxides of the metal were observed. This can be attributedto the sample preparation method, which exposed the sample to water,methanol, nitric acid and air (see Section 4.2.3), allowing oxides to form on thesample surface prior to LAITMS analysis.4.3.3 Spectra for Non-Conducting SamplesCharacterization of non-conducting sample matrices was demonstratedby obtaining LAITMS spectra for a ceramic material (MACORTM) and athermally stable polymer (VESPELTM). The experimental parameters used forLAITMS of these samples are listed in Table Ill. Figures 4.11 and 4.12 are the0)5000-lO3Rh+4000-C>3000-CS2000-0Na1000--0-IlIlIlIJIlIlIJIIlpIlIlIllIlpIIlIllIllIllIllIlIIIJII,IJII,IIIIIIIIIIIIIIJIIIIIIIII2030405060708090100110120mass/charge/amuFigure4.9:LAITMSspectrumobtainedforsupposedthinMofoil(actuallyRh)usinglaserpulseirradiancesof1.44X1010W-cm2.0GD7000-MoO26000-5000-96Mo602+9BMo1602+4000-03000-92Mo16O2>02000-(98M0160HO)+1000-Moj/‘—b--fl-4fr-rJrfLLbJIJIp3-19bhr1r—t-JsLr1—tr0-IIiIIIII1111111111111111111III1111111111111111111IIIIIIII11111111111Ij•8090100110120130140150160mass/chargeIamuFigure4.10:LAITMSspectrumobtainedforapureMometalrodsampleusinglaser pulseirradiancesof1.28X1010W•cm2.81spectra obtained for these non-conducting samples; tentative peakassignments were based upon chemical species known to be present in thesamples. These peak assignments could easily be verified by using an ion trapcapable of MS/MS type experiments [101]. The direct solid analysesdemonstrated by the spectra in Figures 4.11 and 4.12 show the potential ofLAITMS for direct analysis of both conducting and non-conducting solid samplematrices.Different irradiances were required to generate LAITMS spectra for thevarious materials analyzed in this investigation. This may be attributed to thedifferent characteristics of the samples, which affects absorption of the incidentlaser beam at the sample surface (i.e. initial electron generation to facilitateplasma development) [102]. There are several ways to create these initialelectrons: (i) multiphoton ionization; (ii) thermal ionization of hot target vapor;(iii) shock heated ionization of the surrounding gas; and (iv) ionization of thesurrounding gas or target vapor by collisions with the thermionic electronsemitted from the hot sample surface.Since metals are surface absorbers for 532 nm radiation and cangenerate a vapor that becomes rapidly ionized [102], thermal ionization maydominate the initial electron production. Figure 4.11 is a LAITMS spectrumobtained for a MACORTM sample. MACQRTM is a white alum mo-siliconcarbide ceramic with a proprietary composition. Although ceramic materialssuch as alumina, magnesia and zirconia are relatively transparent to 532 nmradiation [103], sintered ceramics such as MACORTM are not because of thepresence of impurities and scattering resulting from material inhomogeneity.This means that sintered ceramic samples must be regarded as dielectrics and,hence, thermal ionization may occur for ceramics as well as metals. Ceramicsare unlike metals in that laser radiation penetrates below the surface andc.’J__AIC&39K15x103-á10-2+AISICAISiN2CD5-AIN&—----0-IIIIIIIIIIIIIII,IIIIIIIIIIIIIIIIIlI,IIIIIIIlIIII,IIIIIIIIIIIII,IlIII,IIItIIIIIII2030405060708090100110mass/chargeIamuFigure4.11:LAITMSspectrumobtainedforasinteredceramicsample(MACORTM)usinglaserpulseirradiancesof8.5X1010Wcm2.83absorbs in a volume, not only at the surface like a metal. As a result, for aconstant laser beam irradiance value, metal samples have higher powerdensity at the target surface than ceramic samples and thus breakdown willoccur more readily for metals than ceramic samples. Additionally, ceramicsamples usually have higher melting and boiling temperatures than metals (eg.for aluminum TMP = 933 K and Tp = 2740 K whereas for alumina TMP = 2345 Kand TBP = 3353 K) [104] . Therefore, ceramic samples require higherirradiances for laser plasma generation, consistent with the results of this study.The analysis of polymers by LAITMS constitutes an interestinginvestigation of laser pyrolysis, providing direct mass spectrometric informationfor solid polymeric samples. Figure 4.12 depicts a LAITMS spectrum obtainedfor a VESPELTM sample. VESPELTM is a heterocyclic polymer with a polyimidebackbone [105] which is thermally stable at temperatures as high as 500 °C, butforms a char above 800 °C. Folmer and Azarraga [106] used gaschromatography to compare the results of polymer pyrolysis by using filamentpyrolysis, tube furnace pyrolysis and laser pyrolysis. Their results indicated thatlaser pyrolysis resulted in product patterns which were simpler and morematerial specific than those obtained by other methods. These results wereattributed to the fact that fewer secondary reactions take place because of theextremely rapid heating and cooling rates obtained with laser pyrolysis. Theuse of an ion trap for laser pyrolysis mass spectrometry facilitates ion chemistrystudies of the pyrolysis products as well as allowing the possibility of relativelyinexpensive MS/MS type experiments [101] for pyrolysate fragment structuralidentification.&C2HN8000-6000->C4000-41K&C2H3N0/C2H5NC2HNO2000-23Na+/C2HNO/C2H3NOC2H5NO3______0-IIIIIIIIIIIIIIIIIIIIIIIIIII,IIIIIIIIIIIIIIIIIIIIIIlIIIIIIItIIIIIIIIIIIIIlIIIIIIIIIIIIIIIIIII2030405060708090100110mass/charge/amuFigure4.12:LAITMSspectrumobtainedforapolyimideplasticsample(VESPELTM)usinglaserpulseirradiancesof1X1010Wcm2.854.3.4 SEM Analysis of Ablated SurfacesThis section presents a brief survey of the various laser solid interactionsobserved for the samples analyzed during the course of these experiments.The various samples analyzed for this chapter were ablated by laser pulsesfocused upon their surface in the ion trap mass spectrometer. The resultingcraters were then analyzed by electron microscopy. Figures 4.13 to 4.16 arethe electron micrographs obtained for type 308 stainless steel, silver solder,VESPELTM and MACORTM. These electron micrographs illustrate that differentlaser solid interactions occur (also discussed in Section 4.3.3) in concurrencewith the nature of the solid substrate.Electron micrographs of type 308 stainless steel are given in Figure 4.13.For single laser pulses focused upon the sample surface the resultinginteraction is shown in Figure 4.13a,b. Slight melting and smoothing of thesolid substrate is observed, but very little material appears to have beenremoved by the ablation process. Upon repeated exposure to 100 laser pulses,a crater-like depression forms (Figure 4.13c,d) on the sample surface. Thiscrater formation shows that an appreciable amount of material is depositedaround the edge of the crater, forming a rim. This crater formation is suspectedto interfere with the production of ions from the sample surface and is discussedfurther in Chapter 5. Similar phenomena are observed for laser ablation of asilver solder sample, shown in Figure 4.14. For a single laser pulse impingingupon the sample surface, localized melting and smoothing is observed (Fig.4.14a,b) and for 100 laser pulses, a rimmed crater forms (Fig. 4.14c,d).Laser ablation of the sintered ceramic MACORTM is illustrated in Figure4.15. For laser ablation of this sample, a different type of laser solid interactionis observed. Craters from both 1 and 100 laser shots upon the sample surfaceare evident in the figure, but they are not smooth, unlike those formed for metal86(a)(b) — ; j-: —-%-...... .., .... *. .• •.. •‘ ... . .. ... .......I__L’La —— dI:*4ct*?b *; - S‘-4c:7;______-Figure 4.13a,b: Scanning electron micrograph (SEM) of the surface of a type308 stainless steel sample which has been ablated by 1 laserpulse at (a) X60 and (b) X800 magnification.87(c)(d)Figure 4.13c,d: Scanning electron micrograph (SEM) of the surface of a type308 stainless steel sample which has been ablated by 100 laserpulses at (c) X60 and (d) X500 magnification.88(a)(b)Figure 4.14a,b: Scanning electron micrograph (SEM) of the surface of a silversolder sample which has been ablated by 1 laser pulse at(a) X60 and (b) X800 magnification.0?I(c)(d)Figure 4.14c,d: Scanning electron micrograph (SEM) of the surface of a silversolder sample which has been ablated by 100 laser pulses at(c) X60 and (d) X600 magnification.89c-ae4/4.* a(a)(b)Figure 4.15a,b: Scanning electron micrograph (SEM) of the surface of aMACORTM sample which has been ablated by 1 (left) and 100(right) laser pulses at (a) X60 and (b) X400 magnification.90WA91(c)(d)Figure 4.15c,d: Scanning electron micrograph (SEM) of the surface of aMACORTM sample which has been ablated by (c) 1 and (d) 100laser pulses, both at X800 magnification.92(a)(b)Figure 4.16a,b: Scanning electron micrographs (SEM) of the surface of aVESPELTM sample which has been ablated by 100 laser pulses,both at X800 magnification, showing stereoscopic (a) right and(b) left views.93ablation. Sintered ceramics are constructed by fusing ceramic powdercompositions into a solid substrate in a furnace. The craters have jaggededged fragments over their surfaces, implying that the localized heating causedby the formation of a laser produced plasma induces shattering or cracking ofthe ceramic particles (evident in the micrograph). No crater rim is observed,suggesting also that some fragmented ceramic particles may simply fall awayinto the spectrometer. Further discussion of the laser-solid interaction forsintered ceramics is given in Section 4.3.3.The final laser-solid interaction presented in this study is for laserpyrolysis of the polyimide VESPELTM; the SEM of a VESPEL surface and laserablation crater from 100 laser pulses is given in Figure 4.16. Significantlydifferent from laser ablation of metal samples, no crater rim formation isobserved. This is because for laser interaction with the solid polymer, laserpyrolysis occurs; the laser sampled material is converted to gaseous material(ionized and neutral), and does not melt as observed for metal samples. Laserpyrolysis using LAITMS was discussed further in Section 4.3.3, but it is clearthat the analysis of polymers by LAITMS provides information which can beused to characterize solid polymer samples.4.4 SummaryThis chapter presented a new spectrometer for Laser Ablation Ion TrapMass Spectrometry (LAITMS). The spectrometer featured many improvementsover the instrument used for earlier LAITMS experiments [53] presented inChapter 3. Included in the improvements was a new ion trap electrodeassembly with extended hyperbolic surfaces (to provide more uniformquadrupole electric fields, resulting in 15-20 % better resolution, and allowingan optical path through the storage volume for optical probe experiments), a94differentially pumped sample probe (to facilitate sample positioning andchanges), an improved optical arrangement (providing better laser irradiancereproducibility and control), and a new vacuum manifold with optical ports and aturbomolecular pump (producing a clean and stable vacuum environment formass spectrometry). The new spectrometer produced LAITMS spectra withreduced space charge interferences, likely the result of improved laser pulsereproducibility. In addition, the spectra obtained showed better signal-to-noiseratios than obtained previously [53], attributed to laser pulse stability and theclean vacuum environment of the new spectrometer. A metallic shard ofunknown composition was correctly identified by LAITMS based upon its massspectrum (and that of a known sample), demonstrating one of the potential usesof a direct solid sampling method such as LAITMS. Both conductors (metalsand alloys) and non-conductors (ceramic and polymer) were analyzed with thenew spectrometer. The nature of the laser I solid interaction was found to besample dependent. Examination of the ablated surfaces of metals by electronmicroscopy showed a melting with material deposition around the ablationcrater, a fractured crater with no crater rim deposition for a sintered ceramic anda deep crater cut in a polyimide sample, also without rim deposition. Thephysical and chemical processes occurring during the laser / solid interactiondetermine the nature of the interaction and, thus, laser irradiance is sampledependent.The new LAITMS spectrometer has demonstrated that the method isviable as a means of direct solid mass spectrometry. The method has thedisadvantages of a restricted sample (pin) geometry, a limited dynamic rangeresulting from space charge effects and possible sample matrix interferencessuch as the presence of easily ionized elements. Additionally, sampling andionization are not independent because they result from the same laser pulse.95Advantages of LAITMS for direct solid analysis include collision induceddissociation (CID) and MS/MS experiments for identification of sample matrixspecies, ablated neutrals mass spectrometry by using electron ionization,integration of sample ions from several laser pulses before detection toenhance sensitivity and optical detection schemes such as laser inducedfluorescence, to achieve very low detection limits. The next chapter providesfurther understanding for LAITMS through its discussion of the use of the newspectrometer as a tool to study fundamental physical parameters such as theablation laser irradiance and wavelength, the sample surface quality and thedetection capabilities for this new method.96CHAPTER 5SAMPLE SURFACE PREPARATION, LASER IRRADIANCE,WAVELENGTH, AND NUMBER OF SHOTS5.1 IntroductionPreceding chapters of this thesis developed laser ablation ion trap massspectrometry (LAITMS) as a method for direct solid analysis. Chapter 4presented a new spectrometer specifically for LAITMS and provided spectra fora wide array of solid sample matrices. These spectra illustrated the utility ofLAITMS for direct solid analysis of both conducting and non-conductingmaterials. The spectrometer also exhibited better resolution than achieved byearlier experiments, and the new optical configuration provided better control ofthe laser irradiance so that the space charge problems encountered previously[53] could be eliminated by careful adjustment of the laser power.Initial interest was devoted to experimental design improvement so thatthe spectrometer was capable of obtaining satisfactory, reproducible massspectra for solid materials. Some of the difficulties encountered with direct solidanalysis by LAITMS include an “uncharacterized surface effect” described inprevious investigations [53], as well as laser irradiance control to limit spacecharge effects. In order to investigate how these parameters affectedspectrometer performance, experiments were designed to test the phenomena.Laser irradiance is an important parameter that must be fully understoodto exploit the analytical capabilities of LAITMS. The laser irradiance upon asolid sample determines the nature of the resulting interaction. For low laserirradiances, laser-solid interaction may cause only a localized warming of the97material without disrupting the solid phase. Higher irradiances (1X105 to lxi 08Wcm2)may cause sufficient localized heating that material may be evaporatedor desorbed from the sample surface. Special sample matrices have beenrecently developed [46-49] to assist in the desorption and ionization of largebio-molecules. These matrices typically contain a substrate which eitherdonates or accepts protons or cations to/from the analyte molecules of interestto facilitate their conversion into gas phase ions. This method has been calledMatrix Assisted Laser Desorption (MALDI), and has been developed to theextent such that commercial MALDI instrumentation is now available. Whenlaser irradiance reaches a certain threshold value (.-1X109Wcm2), a localizedplasma is generated above the sample surface. This intense, hot plasma istransient, sustained only by the interaction of a laser pulse with a samplesurface. During this time, significant amounts of material are removed from thesample and ionized, providing a source of ions for mass analysis.Experimentally, the irradiance of the laser can be adjusted by eitherchanging the flux of radiation from the laser focused upon the surface, or bychanging the focus of a given laser flux upon the sample surface. Theseexperimental methods are examined in this chapter to further our knowledgeabout LAITMS for direct solid analysis. Because certified stainless steelsamples were analyzed by the LAITMS spectrometer, detection limits were alsodetermined in order to quantify the results obtained.Sample surface preparation is a very important consideration for amicrosampling technique such as laser ablation. The method removesnanogram quantities of material from a localized surface area, on the order of100 tm diameter spot size. Surface quality will affect the laser solid interactionin several ways, including not only the energy transfer to the solid but alsothrough impurities present on the sample surface. Musselman and co-workers98[107] investigated the effect of different surface preparations upon quantificationof the signals obtained with a laser ablation time of flight mass spectrometer.Their results showed that planar sample films gave the best precision. Non-planar sample geometry resulted in variations of laser focus, power density andanalytical volume, which was shown to hinder quantification. Clearly, surfacepreparation is important for micro-sampling methods, such as laser ablationmass spectrometry. For this thesis, the surfaces of NIST stainless steel sampleswere prepared by both abrasive paper and diamond polishing methods. Byobserving the ion signals obtained, and by examining the sample surfaces withelectron microscopy, the effect of surface preparation upon LAITMS of solidmetal samples was determined.Laser induced plasmas are formed during laser solid interactionwhenever the plasma formation threshold is exceeded. The effect ofwavelength upon the plasma and its physical properties is important, yet therehas been little investigation of the effects of different ablation laser wavelengthsupon direct solid analysis methods such as LAITMS. This chapter addressesthis concern by investigating LAITMS of different samples using two differentablation laser wavelengths.5.2 Experimental5.2.1 EquipmentThe LAITMS spectrometer used for these investigations has beendescribed previously [52] in Chapter 4. Experimental parameters for thespectrometer are given in Table IV. Sample preparation using diamondpolishing was accomplished using a model DU172 planetary lapping/polishingsystem (Canadian Thin Film Ltd., Toronto, Ont.). The electron micrographs ofthe sample surfaces were obtained by using a Hitachi (Tokyo, Japan) model S992300 electron microscope (Metallurgical Engineering, U. B. C.) and a Hitachimodel S-4100 electron microscope (Biosciences Electron Microscope Facility,U. B. C.).Table IV: Experimental parameters for LAITMS studies.Laser ablation---Wavelengths 532 or 266 nmPulse width 10 nsBeam waist at sample surface 50 iimIrradiance 5 X 1010 Wcm2Storage and ejection---Storage time 5 msRE. 1.05MHzRE. storage potential * 250 VInitial RE. scan potential * 250 VFinal RE. scan potential * 3600 VRE. scan rate 2.54 X i0 Vs1Ion detection--CEM voltage bias -1800 VCEM current amplification 1 X iO V•A-1*Measured (0 - peak)1005.2.2 SamplesFor the LAITMS experiments, many different solid samples were used,including NIST stainless steels (Cl 151, 11 55),type 308 stainless steel, silversolder, MACORTM (a sintered ceramic) and VESPELTM (a polyimide). Samplecompositions for analytes of interest in the NIST samples are given in Table V.Table V: Composition for NIST stainless steel samplesSRM Chromium Manganese Nickel iron(0/ \ 10/ \ (0/ \ (0/ \*t /0) I /0) i /0) I /0)C1151 22.70±0.08 2.50±0.08 7.29±0.05 65.7±0.51155 18.4±0.1 1.63±0.01 12.18±0.05 64.5±0.5*CalculatedThe sample surface was prepared [52] by 600 gauge abrasive paper (wettedwith methanol) polishing followed by a methanol (Omnisolve grade, BDHChemicals, Toronto, Ont.), de-ionized water then methanol rinse, finally airdrying. For the experiments involving diamond polished surfaces, sampleswere further prepared by mounting them in acrylic resin (“Quickmount”, FultonMetallurgical Products Corp., U.S.A.). The mounted samples weresubsequently polished using planetary lapping / polishing methods with thefollowing diamond pastes and times: 60 im for 36 hours, 6 j,m for 3 hours, 3im for 1.5 hours and 1 im for 30 mm. Final polishing was done by hand withalumina paste: 0.3 im for 10 mm. and 0.05 im for 5 mm. The mountedsamples were then immersed in warm acetone (Omnisolve grade, BDH101Chemicals, Toronto, Ont.) to dissolve the acrylic resin. The free polishedsamples were then washed as described above. For LAITMS, samples werealigned flush with the ring electrode inner surface using a helium-neonalignment laser.5.3 Results5.3.1 Laser IrradianceIn Chapters 2 and 3 [53] it was reported that the LAITMS spectraobtained exhibited significant peak broadening and a degraded signal-to-noiseratio resulting from space charge effects [40] in the trapping volume. Thisproblem has been addressed by a previous publication [52] as well as in thischapter by adjusting the laser pulse energy to reduce the amount of samplebeing ablated and ionized per laser pulse. For the earlier investigations [53],laser pulse energy was adjusted by reducing the flash lamp voltage. The pulse-to-pulse energy reproducibility problems, encountered previously, were greatlyreduced by sampling 10 % of the laser output beam for ablation purposes,allowing flash lamp operation at higher voltages. Figures 5.la through 5.ld areLAITMS spectra obtained for silver solder and type 308 stainless steel.Experimental parameters used for these spectra are given in Table IV. Thespectra represent the sum of spectra collected for 100 laser shots. Clearly, thenew spectra obtained in this investigation exhibit superior resolution when thelaser pulse irradiance is properly adjusted (Figures 5.la and 5.lc), and inferior,broad peaks for higher laser pulse irradiances (Figures 5.lb and 5.ld). In fact,Figures 5.lb and 5.ld show marked similarities with the spectra published inthe earlier study [53] in which it was difficult to vary the irradiance in areproducible manner. The new spectra in Figure 5.1 also exhibit less signal-to-noise degradation (note the abscissa scales for each spectrum), attributed toC2500-113Cd+114Cd+112cd+—2000-N111Cd++(U1500-°CdC’)C109A1o7Ag+(U116Cd+.51000-/65CU+500-64Zfl+Znlo6Cd+jCu\N•_i4J0-111111111111111liiiIII11IIIIIIIIIII11II11l111IIIIIIIJI11111IIIII1I11IIIIII1111I1I1lIIIII5060708090100110120130140masslchargeIamuFigure5.la:LAITMSspectrumobtainedforasampleofsilversolderusinglaserpulseirradiancësof3.2X1010W•cm2.Co08000-113Cd+114Cd+6000-112Cd+lllCd+(0C11OCd+’116Cd+.4000-lo9Ag+\>(U1O7Ag+02000-p1.ir--.————----—i._—l,_fi_—--—-n...L.-tr*-IdtF4.I.u,0—111111111111111111111111111111111111111111iiii.iiiiIIIIIIIIIIIIIIIIIIIIIIlIIIIIIIIIIIIIIIIIIIIIIIIII405060708090100110120130140mass/chargeIamuFigure5.lb:LAITMSspectrumobtainedforasampleofsilversolderusinglaserpulseirradianesof4.5XlOW•cm2.0&5000-56Fe+4000-(U52Cr+3000-w>2000-1000-\\58N1+5oCr488Sr4/LrLjrJ_iinJ.wii0—IIII1111111III1111111liii111111111III111111IIIliiiIII1111111111111111111111111IlII.I...•I2030405060708090100110mass/charge/amuFigure5.lc:LAITMSspectrumobtainedforasampleoftype308stainlesssteelusinglaserpulseirradiancesof5.2X1010W•cm2.LX)0115x103-56Fel10-(0a,4.’52Cr+>J58N1+a,5./-—.-________________-——---_-0—1111111111111111111111111111111111111111111111111IIllillIllIlIIIIII1II1I1IIIII1IIIIIIII1I2030405060708090100110mass/chargeIamuFigure5.ld:LAITMSspectrumobtainedforasampleoftype308stainlesssteelusinglaserpulseirradiancesof8.1X1010W•cm-2.106better pulse-to-pulse reproducibility of the ablation laser beam, resulting fromthe improved optical configuration. A diagram of the improved opticalconfiguration is given in Figure 4.3. Since the vacuum system used for theseexperiments was free of residual pump oil and had a high throughput pumpingspeed (170 Lpm.), trace neutral contaminants observed previously (i.e. CCI3+,m/z = 117, 119, 121) are not seen in the Figure 5.1 spectra.5.3.2 Sample Surface PreparationThe nature of the surface of a solid exposed to laser radiation will affectthe resulting interaction. Smooth, planar sample surfaces are more desirable[107] in many respects because the surface is of known geometry (planar) andmay be readily reproduced for subsequent analyses. In practical applications,the preparation of planar surfaces for microanalysis is not a simple task,involving laborious and time consuming polishing steps which may alsointroduce contaminants to the sample surface. The other extreme is no samplesurface preparation, desirable because no contamination from preparationsteps will occur. Surface contaminants already present on the sample may,however, interfere with the analysis. This section presents results obtained [52]which discuss sample surfaces prepared by diamond polished surfaces andthose prepared by polishing with a fine (600 grit) abrasive paper. The studyalso examines the effects of cumulative laser shots upon the same samplelocation.Sample pins of SRM C1151 were prepared with finely polished surfaces(diamond paste and alumina) and with fine abrasive paper polished surfaces inorder to investigate the effects of surface preparation upon LAITMS spectra.The certified composition of the elements of interest for SRM Cl 151 stainlesssteel are given in Table V. Experimental parameters used are given in Table IV.1—0tOZO.8(U>b(0C(UC0>.1-i(U0.4(U0.20.0Figure5.2a:Plotsofrelativesignalintensities(error=±2a)for52Cr,55Mnand56Feionsresultingfromlaserablationofastainlesssteel(SRMC1151)sampleversusnumberoflasershotsforafineabrasivepaperpolishedsurface.SpectrumNumber(10Shot)co01.21.00.8U,Ca)Ca)>(U0.60.40.20.0Figure5.2b:Plotsofrelativesignalintensities(error=±2a)for52Cr,55Mnand56Feionsresultingfromlaserablationofastainlesssteel(SRMC1151)sampleversusnumberoflasershotsforadiamondpasteandaluminapolishedsurface.12SpectrumNumber(25Shot)4109The results of these studies are given by Figure 5.2 and electron micrographs ofsurfaces prepared by the methods described above are shown in Figures 5.3and 5.4. This investigation was conducted by obtaining spectra for a singlelocation upon the sample surface (le., the mass spectra for 10 laser pulses upona spot on the sample were summed to give a single spectrum, then another 10mass spectra for 10 more laser pulses upon the same spot summed to give thenext spectrum, etc.), the entire experiment for each type of surface performed intriplicate.The results of this study are presented in Figure 5.2. They show that forthe fine abrasive paper polished (APP) surface, maximum ion signal intensitiesare observed for the first laser ablation pulses whereas maximum ion signalintensities for the diamond paste and alumina polished (DPP) surface areobserved after 25 laser ablation pulses. This result may arise from the differentsample surface reflectivities: the initial ablation laser pulse energy depositedupon the DPP surface can be reduced/decreased as a result of reflection by thishighly polished surface. After several laser pulses, the surface becomes lessreflective as a result of crater formation, and can absorb more of the incidentlaser energy, suggesting an explanation for the observed experimental results.The APP surface is much less reflective than the DPP surface because it hasfine grooves and scratches arising from the preparation step. This may result ingreater initial ablation of these surface irregularities, followed by a reduction inobserved signal intensities after they are removed by the formation of a smooth,shallow ablation crater. For both APP and DPP surfaces, the observed signalsdecrease as the number of laser shots upon the same spot increases, likely theresult of deep ablation crater formation. The geometry of a deep crater reduceslaser irradiance by increasing the sample surface area exposed to the ablationpulse and makes material removal more difficult. Laser ablation craters for 1110(a)(b)-; \4 :—‘ -—.—- -:‘L ;d-‘—i -14?7d;-4q:TvFigure 5.3: Scanning electron micrograph (SEM) of a type 308 stainless steelsample which was polished with fine abrasive paper; evident in thepicture are laser ablation craters for pulse irradiances of 5.2 X 1010W•cm-2 resulting from (a) one laser pulse and (b) 100 laser pulses.111and 100 shots on a 308 stainless steel sample are shown in Figure 5.3; initiallya shallow surface crater forms which acts to smooth the APP surface, followedby deeper crater formation with an increasing number of laser pulses upon thesame spot. Laser ablation craters similar to those shown in Figure 5.3 and 5.4were observed for the samples used in the surface preparation experiments.5.3.3 Laser Focus for Irradiance ControlAs was presented in Section 5.3.1, laser irradiance regulation isessential for LAITMS experiments. Section 5.1 suggests that laser irradiancecan be adjusted by changing the laser fluence or by adjusting the size of thelaser focus upon a sample at a constant fluence. This chapter provides insightinto the effects of changing the laser irradiance by laser focus control upon thesignals obtained for direct solid analysis by LAITMS.In order to examine the effect of varying laser pulse irradiance uponLAITMS signal intensity, a study was conducted in which the ablation laserfocusing lens was translated towards the sample. At a constant laser pulseenergy, this results in a decrease in the laser irradiance by distributing it over alarger spot on the surface of the sample. For this investigation, a certifiedreference stainless steel (SRM 1155) sample was prepared with a highlypolished surface (diamond paste and alumina). Sample composition for theelements of interest is given in Table V. The lens was translated towards thesample surface from the position used for previous experiments by 0.5 mmincrements. LAITMS spectra were obtained for each lens position by summingthe spectra obtained for 25 laser pulses. Experimental parameters for thisinvestigation are given in Table IV. The results are summarized in Figure 5.5,which is the ion signal intensity maxima as a function of lens position. Thisresult may be attributed to moving the focal point of the ablation laser lens112Figure 5.4: Scanning electron micrograph of stainless steel (SRM 1 155) samplepolished with diamond paste and alumina which shows the laserablation craters formed by 100 laser shots but varying the irradianceby increasing the laser beam waist at the sample surface. This wasaccomplished by translating the focusing lens towards the samplesurface. Crater 1 was for ablation at the focal point of the lens;moving in a clockwise manner shows the ablation craters formed asthe lens was translated towards the sample. Spectrum for crater Awas not stored.U)0>Cr)1806040200Figure5.5:Plotofrelativesignalintensityfor 52Cr,55Mnand56Feionsresultingfromlaserablationofastainlesssteel (SRM1155)whichwaspolishedusingdiamondpasteandaluminaversusdistanceof lenstranslationtowardsthesamplesurface.Theseintensitiesresult fromsummingthespectraobtainedfor25laserpulses.012345LensPositionImm114towards, and eventually beyond the sample surface, and illustrates that LAITMSof SRM 1155 can be optimized by careful adjustment of laser irradiance at aconstant pulse energy.A SEM micrograph of the sample surface is given in Figure 5.4, in whichthe different laser ablation craters are numbered for clarity. The results of thisstudy may also be emphasized by considering the effect of different lenspositions (and thus, different irradiances) upon the mass resolution obtained inthis experiment. Figure 5.6 shows spectra obtained for several laser focuspositions. These spectra illustrate that an optimum laser focus exists, assuggested by Figure 5.5. By plotting both resolution and ion signal intensity forseveral isotopes (Fig. 5.7), it is evident that a trade off exists between resolutionand ion signal intensity. Judicious selection of laser irradiance by varying thefocus of the laser facilitates selection of either higher resolution for identificationof ions with similar mass-to-charge ratios, or higher ion signals with lowerresolution for quantitative determinations of analyte within a sample matrix.5.3.4 Detection Limits for NIST Stainless Steel SamplesIn order to quantify the analytical capability of the improved LAITMSspectrometer presented in this chapter, the detection limits for the method weredetermined. The analysis of certified NIST stainless steel samples in thischapter allowed calculation of detection limits for LAITMS of stainless steelsamples. According to Boumans [108], 3 sigma detection limit values forinstrumental methods can be determined by the following relationship0.03 (RSDB)cLOD= SBR 5.14-’0C04-’C0‘I4.’Figure5.6a:LAITMSspectraobtainedforstainlesssteel(SRM1155)fromthelenstranslationstudyforlensposition4,showingspacechargedpeaks.Refertofigure5.4forapictureoftheablationcratersforthedifferentlenspositions.10152Cr+56Fe25x10320-15-10-5-0...I••..I..•i11•i111111111111111020304050II...I.IIII,,.III,,,I60708090100mass/chargeIamua,0CCa,>a,a,CD52Cr56Fe8006004002000100Figure5.6b:LAITMSspectraobtainedforstainlesssteel(SRM1155)fromthelenstranslationstudyforlensposition8,exhibitinggoodresolution.Refertofigure5.4forapictureoftheablationcratersforthedifferentlenspositions.102030405060708090masslchargeIamuzCoS..C’,Ca,Ca,>(Ua,152Cr+6005004003002001000100Figure5.6c:LAITMSspectraobtainedforstainlesssteel(SRM1155)fromthelenstranslationstudyforlensposition9wherethesignalintensitiesarereduced.Refertofigure5.4forapictureoftheablationcratersforthedifferentlenspositions.56Fe+1060mass!chargeIamu708090>%4-.U)CCCCuC00(I)G)OD11—LI-- Resolution-o—Intensity(÷100)5040302010Figure5.7a:Plotofrelativesignalintensityandmassresolutionversuslenstranslationdistance(lensposition)towardsthesamplesurfacefor52Cr-i-ionsresultingfromlaserablationofastainlesssteel (SRM1155)whichwaspolishedusingdiamondpasteandalumina.1234567891011LensPositionIa.u.U)a)CC0*40a)0)—ci—Resolution-o—Intensity (÷70)605040302010Figure5.7b:Plotofrelativesignalintensityandmassresolutionversuslenstranslationdistance(lensposition)towardsthesamplesurfacefor 56Feionsresultingfromlaserablationofastainlesssteel (SRM1155)whichwaspolishedusingdiamondpasteandalumina.1234LensPositionIa.u.120where c is the analyte concentration, RSDB is the relative standard deviation ofthe background (%), SBR is the signal to background ratio (%) and LOD the limitof detection for the analytical method. In the previous lens translation study, ionsignals were maximized for LAITMS of NIST stainless steel 1155 using lensposition 4 on Figure 5.4. The elemental composition for the sample is given inTable V. Spectra obtained for lens position 4 (Fig. 5.6a) show significant peakbroadening as a result of space charge, whereas spectra from position 8 (Fig.5.6b) exhibited better mass resolution and were, thus, used for detection limitcalculations. For detection limit calculations, ion signals from isotopes withoutisobaric interference from other matrix species were used. Table VI lists theisotopes of interest for detection limit calculations, their natural abundances[104], the calculated isotopic concentrations for NIST stainless steel 1155 andthe ion signals obtained by LAITMS of this sample.Table VI: Calculated isotopic concentrations and their corresponding LAITMSsignals.Isotope Natural Sample Isotopic Ion SignalAbundance Corn position Composition fromLAITMSt(%) (%) (%) (arb. units)52Cr 83.79 18.45 15.46 82453Cr 9.50 18.45 1.75 12755Mn 100.00 1.63 1.63 30256Fe 91.8 64.5 59.17 82557Fe 2.1 64.46 1.35 8060Ni 26.1 12.18 3.18 67Background* -- 31.95±5.24* Ave of 100 data points from baseline of spectrum ± 2t Fig 5.6b121In an effort to quantify the LAITMS experiment, sample volumes removedfrom the solid by the ablation process were estimated for several craters. Thiswas crudely accomplished by approximating the crater to that of a cone shapeddepression, then estimating dimensions from the electron micrographs of thesample surface. Clearly, this estimate is only a conservative approximation,neglecting re-deposition of material at the crater rim (evident in the electronmicrographs); yet the result of the calculations in Table VII suggests about 5 ngof solid is removed per laser pulse, consistent with literature [109] values.Absolute sample volumes per laser pulse might also be measured [110] byusing a radioisotope spiked sample. By determining the activity before andafter laser ablation, the amount of material removed could be determinedprecisely. From Table VII, the sample volume removed by 25 laser pulsesshould, therefore, be about 125 ng, and by calculating the appropriate valuesfor Equation 5.1 in Table VIII, the absolute detection limits for atomic analytes inNIST stainless steel were determined.Table VII: Material removed by laser ablation, calculated by the coneapproximation.Laser Crater # Radius Depth Volume MassRemoved(cm) (cm) (cm3)* (g /100shots)f1 38.5 X io- 38.5 X 1O 5.98 X 10-8 4.8 X 1O2 65.45 X iO 15.4 X i0 6.91 X 10-8 5.6 X i03 69.3 X iO 7.7 X 1O 3.87 X 10-8 3.1 X iO* cone approximationt p = 8.03 g•cm3122Table VIII: Calculated limits of detection for atomic species in NIST 1155stainless steel.Isotope SBR in LCD Absolute LCD(%) (%) ()t52Cr 25.79 15.46 0.09423 11853Cr 3.98 1.75 0.06933 86.755Mn 9.45 1.63 0.02710 33.956Fe 25.82 59.17 0.3602 45057Fe 2.50 1.35 0.08500 1062.10 3.18 0.2383 298-125 ngt assumes 25 shotsThe results in Table VIII suggest that using conventional mass spectrometricdetection [96] with LAITMS, absolute detection limits of 34-450 pg could beobtained for atomic analyte species in a stainless steel sample.5.3.5 Laser Ablation by 532 nm and 266 nm RadiationThe final section of this chapter explores the effect of laser ablation usingdifferent wavelengths. Commercial laser ablation based microsampling massspectrometric methods, such as LAMMA [90] normally employ a fixedwavelength for the ablation laser. This may be due to the difficulty of convertingcommercial system optics etc. so that different wavelengths may be used. Thespectrometer used for these studies utilizes the frequency doubled output (532nm) of a Nd:YAG laser. By using a second harmonic generating crystal,frequency quadrupled output for the Nd:YAG laser (266 nm) can be readilyobtained. A discussion of laser harmonic generation is beyond the scope of this123thesis, but a treatment of lasers and non-linear optical phenomena such asfrequency doubling is given by Laud [67]. By using the different ablation laserwavelengths, their effect upon the ion signals obtained by LAITMS and uponthe laser produced plasma can be examined.For this investigation, a variety of different samples were studied. Theyincluded NIST C1151 stainless steel, silver solder, MACORTM (a sinteredceramic), and VESPELTM (a polyimide). Samples were prepared with fineabrasive paper as described in Section 5.2.2. The samples were then analyzedby LAITMS using laser ablation at both 532 nm and 266 nm. The resultingspectra are given in Figures 5.8 to 5.11. Through careful examination of thesespectra, several observations can be made with regard to the effect of laserablation at different wavelengths. Experimental parameters for theseinvestigations are give in Table IV. All spectra were collected by summing thespectra from 100 laser shots.The LAITMS spectra obtained for a silver solder sample using laserablation at both 532 nm and 266 nm are given in Figure 5.8. Cursory inspectionof these spectra shows ion signals which correspond to analytes known to bepresent in the sample. The differences between the two spectra manifestedthemselves in the relative ion signal intensities. For example, the ion signalsobtained by using 266 nm ablation exhibited much larger relative intensities forsilver, the dominant matrix species, and also for trace components such ascadmium, copper and zinc. These differences in relative ion signal intensitiessuggested that there were differences in the ionization conditions present in thelaser produced plasma used for sampling and ionization, which depend uponthe wavelength of the ablation laser.2500-113Cd+112cd+_114Cd+2000-‘NllOCd+(U1500-0C109AAgN\116cd+1000107/a)500-1/0—111111111111111111111111111111111111111111111111111111111111111111111111111111111111111115060708090100110120130140mass/chargeFigure5.8a:LAITMSspectraobtainedforasilversoldersampleusing532nmlaserablation.Lt)11O7Ag+1O9Ag+4000-110Cdt3000-(0C0112Cd+____/__2000-=(UCu+63+\/Zn+114cd+1000Cu0-111111111111IIIIlIlIllIllIll111111IIIIIII11111111111111IIIIIIIIIIIIIIIIIIIIIIIllIllIllIll5060708090100110120130140mass!chargeIamuFigure5.8b:LAITMSspectraobtainedforasilversoldersampleusing266nmlaserablation.(0c’J&15x103-10-C)AISiN2>1!AISiC2cc27A1+28S1+\AIN-.L-0-2030405060708090100110masslchargeIamuFigure5.9a:LAITMSspectraobtainedforasampleofMACORTMusing532nmlaserablation.C’]8000-&6000-(0004000-0(U02000-27A1+“2AIN-—1F-Jj__-—-—————2030405060708090100110masslchargeIamuFigure5.9b:LAITMSspectraobtainedforasampleofMACORTMusing266nmlaserablation.128Further experiments were then conducted in which non-conductingsample matrices were examined by LAITMS at different ablation laserwavelengths. The samples analyzed included both MACORTM and VESPELTM.Spectra obtained by LAITMS at both 532 nm and 266 nm for these samples aregiven in Figures 5.9 and 5.10. Figure 5.9a shows various matrix species arisingfrom laser ablation of MACORTM using 532 nm laser light pulses. Based uponelements known to be present in the sample matrix, these species aretentatively assigned the structures indicated in the figures. Although thisexperiment was not capable of structural identification because of electroniclimitations, species could be verified by using an ion trap mass spectrometer[101] capable of MS/MS type experiments. Upon changing the ablation laserwavelength to 266 nm, the spectra obtained by LAITMS changed somewhat toshow fewer fragment species and suggesting increased fragmentation in thelaser micro plasma. Figure 5.9b shows the spectrum obtained for LAITMS ofMACORTM at 266 nm. A similar result was observed for LAITMS of VESPELTMat the two wavelengths. Figure 5.10 shows spectra for both wavelengths; fewerfragment species are observed for 266 nm ablation, suggesting again thatincreased fragmentation of the polymer backbone occurs for laser pyrolysis withthe shorter wavelength, more energetic (4.67 eV instead of 2.33 eV) photonsfrom the frequency quadrupled output of the Nd:YAG laser.The final spectra obtained for this study are given in Figure 5.11 whichshows the ion signal intensities obtained by LAITMS of NIST C1151 stainlesssteel for both 532 nm and 266 nm laser ablation. These spectra have ionsignals corresponding to the isotopes known to be present in the sample, but asobserved for Figure 5.8, certain ion signals (le. 55Mn+ ) are more intense with266 nm ablation. These differences in ion signal intensities reflect the differentionization conditions present in the two laser produced plasmas. The large0)c’j1&C2HN8000-6000-0C4Ø-41K+&CHN+0>1!0/C2H5NC2HNOC2HNO200023Na+/C”N”2’’3“2N”+/2’‘5J3L0—2030405060708090100110mass/chargeIamuFigure5.lOa:LAITMSspectraobtainedforsamplesof VESPELTMusing532nmlaserablation0C’)8000-++39K&C2HN6000->(0C.4000-C‘I41K+&CHN2000-23+INa/C2H8NC2H3NO___________IC2HNO+/,/C2H5NO3_____________—2030405060708090100110mass/chargeIamuFigure5.lOb:LAITMSspectraobtainedforsamplesof VESPELTMusing266nmlaserablation.0CC0>4..Cu01C•j52Cr4000300020001000056Fe+58N1+mass/chargeIamuFigure5.lla:LAITMSspectraobtainedforsamplesof C1151stainlesssteelusing532nmlaserablation.C)Cu>%CoCuCCu>(UCu52Cr+8000600040002000056Fe+58Nj+2030405060708090100mass/chargeIamu110Figure5.llb:LAITMSspectraobtainedforsamplesofC1151stainlesssteelusing266nmlaserablation.133signal enhancements observed for specific analytes (such as copper, zinc andcadmium in silver or manganese in stainless steel) when the ablation laserwavelength is changed suggests that selective analyte ionization schemes at afixed laser ablation wavelength can afford enhanced sensitivity. The concept ofselective analyte ionization is addressed later (Chapter 7) in this thesis.5.4 SummaryThis chapter has been very important for the development of LAITMS aspresented in this thesis. Important experimental parameters such as ablationlaser wavelength and irradiance, as well as sample surface quality wereinvestigated to elucidate their effect upon LAITMS. Clearly, laser irradiancecontrol is essential for any laser sampling method. For these experiments,irradiance control allows the analyst to ablate varying amounts of material fromsolid sample matrices, thus storing more or less analyte in the ion trap. Thedetrimental effects of space charge [40] can be eliminated by controlling thequantity of ions to be confined in the ion trap. Ablation laser irradiance controlcan be implemented by either adjusting the fluence from the ablation laser, orby adjusting the focus of the laser radiation upon the sample surface. Theseinvestigations suggested that varying the laser focus was superior to varying theflashlamp output of the laser for irradiance control.Sample surface quality is an important consideration for LAITMS.Previous investigations by Musselman [107] suggested that planar samplesurfaces gave the best quantitation for laser ablation coupled with a time-offlight mass spectrometer. The experiments presented in this chapter show thatsurfaces prepared by fine abrasive paper polishing give initially high analytesignals, which diminish upon further irradiation on the same location.Conversely, surfaces prepared by diamond paste polishing methods exhibit134initially low ion signals which increase then decrease as the number of lasershots upon the same spot increases. These observed phenomena areattributed to sample surface reflectivity. The surface of the abrasive polishedsamples was covered by irregular scratches and grooves, the removal of whichby the initial laser shots gave rise to the large ion signals observed. Furtherirradiation on the abrasive prepared surface results in a shallow, smooth crater,which is more reflective to the incident laser radiation, giving the reduced ionsignals observed. Deep crater formation occurs for extended irradiance uponthe same location on the sample, producing an eventual reduction of the ionsignals. The diamond polish prepared surface was free from the surfaceirregularities of the abrasive paper prepared sample, resulting in a very highinitial reflectivity to the ablation laser pulses. This high initial reflectivity explainsthe reduced initial ion intensities observed for this sample. The increase in ionsignal intensity which occurred upon increasing the number of laser pulsesupon the same location was the result of shallow crater formation (less reflectivethan the highly polished surface)as well as the further decrease in signalintensity attributed to deep crater formation. Thus, the condition of the samplesurface affects the ion yield and ultimately the sensitivity of analysis for solidsamples using LAITMS. By using sample surfaces prepared by diamond pastepolishing, absolute detection limits for atomic analytes in a stainless steelsample were determined. Although the sample volume was approximated, itwas done so that the final result was overestimated, giving detection limits of34-450 pg for atomic analytes in a certified stainless steel sample.The wavelength of the laser used for sample ablation was found to affectthe results obtained by LAITMS. For molecular analytes, such as for ceramicand polymer samples, increased fragmentation of matrix species was observedfor higher energy 266 nm photon radiation (4.67 eV) over lower energy (2.33135eV) 532 nm radiation when used for ablation purposes. This suggests that theionization conditions of the laser produced plasma from 266 nm laser ablationare more energetic than those for 532 nm laser ablation produced plasma.When atomic analytes were analyzed by LAITMS of metal samples, differentrelative ion intensities were observed for certain analytes. In general, increasedionization for certain analytes was observed when 266 nm laser radiation wasused for ablation purposes. These differences in ionization between the twoablation laser wavelengths further suggests that shorter wavelength laserablation is more energetic than that for 532 nm radiation. Researchers [111-113] realized that laser ablation / desorption using longer wavelengths resultedin less molecular fragmentation, a desirable phenomenum for molecularanalysis. Thus, depending upon the ablation wavelength chosen for LAITMSexperiments, the analyst may extract atomic and/or molecular information abouta solid sample.136CHAPTER 6COMPARISON OF NORMAL AND STRETCHED HYPERBOLICELECTRODE GEOMETRY6.1 IntroductionThe preceding chapter was devoted to investigations regarding physicalparameters external to the mass analyzer. The effects of sample surfacepreparation as well as ablation laser irradiance, wavelength and number ofshots upon the sample were shown to have dramatic effects upon the directanalysis of solid matrices by Laser Ablation Ion Trap Mass Spectrometry(LAITMS). This chapter is devoted to the evaluation of the operation of the iontrap mass spectrometer with “normal” electrode spacing as defined by previousauthors [33] in comparison with that of a modified or “stretched” electrodegeometry [65]. The commercially available Quadrupole Ion Trap MassSpectrometer (Finnigan ITMSTM) has been used for several years with astretched geometry (a trade secret). This stretched geometry was an empiricaladjustment developed to give improvements in the performance of thespectrometer by reducing mass defects observed for large mass organicfragments and provide more reproducible results for replicate analysis.To investigate the effect of a “stretched” electrode geometry with the iontrap mass spectrometer used for LAITMS, it was necessary to evaluate theperformance of the ion trap with “normal” electrode spacing, as defined by Toddet. a!. [33] through Equations 1.6 to 1.8. Because the ring electrode has aninternal radius of 10.00 mm, the equations require the distance between theend cap electrodes to be 14.14 mm. All experiments conducted in this137laboratory have utilized electrodes which satisfy these geometric requirements.The “stretched” trap geometry adopted commercially uses electrodes defined byequations 1.6 to 1.8, but which are arranged in a fashion such that the end capelectrode spacing is expanded eleven percent or:z0’=1.11z 6.1where z0’ is the “stretched” and z0 the “normal” end cap electrode spacing. Thedifferent electrode configurations were evaluated both for electron ionization ofneutral gas phase molecules and for LAITMS of a solid sample matrix.Parameters investigated for this study included peak shape, resolution, storagetime for analytes and magnitude of the radio frequency storage potential used toconfine analyte ions for mass spectrometric analysis.6.2 Experimental6.2.1 EquipmentThe LAITMS spectrometer used for these investigations has beenpreviously described in chapter 4. No changes were made to the system for the“normal” electrode geometry experiments. However, for the “stretched”electrode geometry experimental section, the distance between the end capelectrodes was increased by exactly eleven percent, as described above byEquation 6.1. Because the end cap spacing for the “normal” electrodes was14.14 mm, the stretched electrode geometry required 15.70 mm as the end capspacing in accordance with the above equation. To increase the end capelectrode spacing, VESPELTM spacers were inserted on the support shaftsused to mount the end cap and ring electrodes together in the spectrometer.These spacers were constructed so that the “stretched” geometry was138Table IX: Instrumental parameters for electron ionization experiments usedin the parametric investigations of the “normal1’and “stretched”electrode geometry.Ionization---Electron energyDuration of electron beamStorage and ejection---Storage timeRE.RE. storage potential *Initial RE. scan potential *Final RE. scan potential *RE. scan rateIon detection--CEM voltage biasCEM current amplification5 - 5000 ms1.05 MHz250 V250 V3600 V2.54 x 1 4 Vs1-1800 V1 X io V•A1Number of scans per spectrum 50symmetric and the ring electrode was situated around the midpoint of the endcap separation; the spacers were half the required separation and werelocated on the support shafts both above and below the ring electrode.Instrumental parameters for the electron ionization and LAITMS experimentsare given in Tables IX and X.70 eV1 ms*Measured (0 - peak)139Table X: Instrumental parameters used for LAITMS parametric investigationsof the “normal” and “stretched” electrode geometry.Laser ablation---WavelengthPulse widthBeam waist at sample surfaceI rradianceStorage and ejection---Storage timeRE.RE. storage potential *Initial RE. scan potential *Einal RE. scan potential *RE. scan rateIon detection--CEM voltage biasCEM current amplificationNumber of scanst per spectrum*Measured (0 - peak)tone scan per laser shot532 nm10 ns50 tm5 X 1010 Wcm-25-35 ms1.05 MHz200 - 700 V200 - 700 V3600 V2.54 X 1O V•s1-1800 V1 X 10 VA-1506.2.2 SamplesCarbon tetrachloride was chosen as a model compound for analysis byelectron ionization (El) because it was used as a mass scale calibrant inprevious studies [52, 53]. The carbon tetrachloride (Omnisolve grade, BDHChemicals, Toronto, Ont.) used was not purified further prior to analysis.Gaseous analyte molecules were introduced to the mass spectrometer via avariable leak valve. All El of neutral molecules were performed directly in thestorage volume of the ion trap.140For the LAITMS experiments, NIST stainless steel (C1151) and a sampleof Inconel 747 (Dr. A. Mitchell, Metallurgical Engineering, U.B.C.) were used assolid sample matrices. Sample compositions are given in Table Xl.Table Xl: Elemental composition for metal alloy samples analyzed by theLAITMS investigations.NIST stainless steel SRM C1151Element Percentage Composition (mass)Iron 65.7±0.5Chromium 22.70 ± 0.08Nickel 7.29 ± 0.05Manganese 2.50 ± 0.08Inconel 747Element Percentage Composition (mass) *Nickel 63Chromium 18Iron 10Zinc <1*values given without error limitsThe sample surface was prepared [53] by 600 gauge abrasive paper (wettedwith methanol) polishing followed by a methanol rinse (Omnisolve grade, BDHchemicals, Toronto, Ont.), de-ionized water then methanol rinse, then air drying.Samples were aligned flush with the ring electrode inner surface using ahelium-neon alignment laser.1416.3 Results and Discussion6.3.1 Storage TimeIon storage time, the period in which an ion can be confined within thestorage volume of a quadrupole ion trap between creation and detection, isdependent upon many factors [32]. Factors considered include the nature of theanalyte ions (atomic, molecular, mass, charge, etc.), chemical reactivity to otherionic or neutral species present in the bath gas, trajectory within the storagefield, the depth of the storage potential well and homogeneity of the storagefields. For these studies, identical experimental parameters were used tocompare the performances of both “normal” and “stretched” quadrupole ion trapelectrode geometry. This eliminated interference from many of the parameterslisted above and compared performance for the two configurations on the basisof their differing potential well shapes.Figures 6.1 and 6.2 show integrated signal intensities for CCl3 ionsversus storage time for electron ionization of carbon tetrachloride in both normaland stretched mode operation. These figures plot the average values from 6replicate experiments; each experimental spectrum represents the summedspectra of 50 individual ion trap scan sequences. Comparison of these figuresshows ion signals are low for short storage times, increasing to maxima atapproximately 100 ms then slowly decreasing in intensity for longer storageperiods of up to five seconds. The same trends are noticed for both “normal”and “stretched” geometry. There are differences between Figures 6.1 and 6.2;the “normal” electrode geometry gives less reproducible spectra, as reflected inthe larger error bars. Also, the “stretched” electrode geometry reaches asharper maximum.Similar comparisons of integrated signal intensity versus storage time foratomic ions from LAITMS of C1151 stainless steel (Figures 6.3 and 6.4)z0U,C0CV0)0Cc’J-0-123-.0-- 121—ti—119-0-117200x103150100505000Figure6.1:Plotofintegratedionsignalintensityforelectronionizationofcarbontetrachloride(CCI3+ions,error=2a)versusstoragetimeobtainedwithaniontrapwith“normal”electrodespacing,obtainedfor6replicateexperimentsof50scanseach.01000200030004000StorageTimeImsPlotof integratedionsignalintensityforelectronionizationofcarbontetrachloride(CCI&ions,error=2a)versusstoragetimeobtainedwithaniontrapwith“stretched”electrodespacing,obtainedfor6replicateexperimentsof50scanseach.Cl)1200x115011::Figure6.2:[-o--121I-c’—1211l—t—119101000StorageTimeIms 30005000>1(0C04.’C04.’I4.’C80x1036040200Figure6.3:PlotofintegratedionsignalintensityforLAITMS(532nm)ofC1151stainlesssteelatomicionsversusstoragetimeobtainedwithaniontrapwith“normal”electrodespacing,obtainedfor6replicateexperimentsof50shotseach,usingastoragepotentialof250V(O-p).510152025StorageTimeIms(UU,Ca,4.’CVa,4.’(UIa,4-CIf)40x1302010Figure6.4:PlotofintegratedionsignalintensityforLAITMS(532nm)ofCl151stainlesssteelatomicionsversusstoragetimeobtainedwithaniontrapwith“stretched”electrodespacing,obtainedfor6replicateexperimentsof50shotseach,usingastoragepotentialof400V(0-p).5101520253035StorageTimeIms146revealed an analogous trend: Ion signals decreased with increasing storagetimes. No minima/maxima were observed for short storage times; this mayresult from 5 ms being the shortest storage time available with the spectrometerhardware/software used for the experiment. Ion intensity signals decayed farmore rapidly for LAITMS atomic ions, possibly due to a higher reactivity for theatomic ions with neutrals in the vacuum system; as well, the atomic speciesmay have less stable trajectories within the storage fields because of theirinitially larger kinetic energy (creation by the laser produced plasma) whencompared to the 70 eV electron ionized molecules of the previous experiment.Additionally, the ion signals for LAITMS using the ‘stretched” geometry showslightly larger signals for longer storage times, which may be attributed toslightly larger radio frequency storage potentials (400 V versus 250 V (0-p))used for this investigation.Ion trajectory simulation studies were conducted using MacSimionTMversion 1.0 to model the effect which different electrode configurations haveupon the nature of the storage field and ion motion within the trapping volume.Parameters used for these simulation studies are given in Table XII.Table XII: Parameters used for the MacSimionTM 1.0 simulation studies*.Storage RF 1.05 MHzStorage potential 100 V (0-peak)Storage time 63 isIon charge +1lonmass 109 amuInitial Ion energy 0 eV* refer to APPENDIX II for program datails.147(a)(b)Figure 6.5: Cross sectional diagram through an ion trap with (a) “normal”electrode geometry and (b) “stretched” electrode geometry thatshow equipotential contour lines for the storage fields created bysimulation using MacSimionTM 1.0. The contour lines shownrepresent 10 V increments for the phase when the ring electrodeis 100 V and the endcaps at ground.(a)8 0%1 000%I 000080I a§§1 8C8O aO 00O 0a aa 0a B0 0O DO0 00 08 00080000ODD0 000 0a a8 DOaa000a8a0008aaa00aa000000a0a880Ba0a0000 0000an-0000000 ao00• 00000000000000B0DD0000a0OO00D000000000000000 000000000000 000000000000000000DO0000000 CD0000000000000000000000000000000000000000000000D000000000000000D00148Figure 6.6: MacSimionTM 1.0 ion trajectory simulation for a single ion ofmlz = 109 amu positioned initially at the center of the storage fieldfor (a) “normal” and (b) “stretched” electrode geometry. Param.used for this simulation are given in Table 6.4. MacSimionTM dàesnot compensate for the effects of helium buffer gas upon the iontrajectories developed by the simulation.(b)0a0D0aO000000DQ0000Oa000000DOa00O0O0Oa0CDDDCD0D00 0000 00000000 DOD000 CD0000 CCCCC 00000 0000 00 00000000 00000000a 80% 000 00 800 0000 B0 000 00000 000 R0C —0 0o8o 0g8 00 -C CB0 000 0O 000 00O 00O 0 R0 00Caa 00000 000D0 —O0 00C 00C00 CODBaO 00000000 000000000 0000 0000000 000 000 00 C00 DC000 COO 8DO aDDO00 0000000000000000000000D000000000000000000000000000000000000000C149Figure 6.5 shows equipotential surfaces for a vertical cross section through theelectrodes as well as the storage volume for “normal” and “stretched” geometry.Notice that the “stretched” geometry has similar radial field gradients to the“normal” geometry, but has a reduced axial gradient due to the expandeddistance between the end caps. This simulation result suggests that stored ionswill have trajectories which range closer to the endcaps for the “stretched”electrode geometry. Ion trajectory studies were conducted with MacSimionTMfor single ions in the center of the trapping volume of both electrodeconfigurations. Figure 6.6 shows plots of the trajectories obtained for singlycharged ions with a mass of 109 amu. This simulation does not account for theeffects of a buffer gas, shown to improve ion trap performance [96] becauseVersion 1.0 MacSimionTM is unable to include this variable. Other researchershave generated experimental stability diagrams for the stretched trap geometry[114] and concluded that the electric fields created by the “stretched” geometrywere slightly different from those of a “normal” electrode geometry. Thesimplistic trajectory study presented in this thesis showed the ion trajectory for“normal” electrode geometry to be more symmetric and confined within thecenter of the storage volume, whereas the “stretched” geometry gave a slightlyexpanded trajectory, the result of the modified storage fields created by thealtered electrode geometry.6.3.2 Storage PotentialIon storage time experiments suggested that the magnitude of the RadioFrequency storage potential may affect the signals obtained for LAITMS of solidsample matrices. A sample of Inconel 747 was chosen for this investigationbecause it has isotopes with mass-to-charge ratios over the range of 52 to 64amu, the range of interest for previous LAITMS investigations of steel matrices.The instrumental parameters used were those described previously (Table X),(U0Ca,CVa,mIa)a,C0I0250x1200150100500700Figure6.7:PlotofintegratedionsignalintensityobtainedbyLAITMS(532nm)ofatomicionsfromaninconel 747sampleversusstoragepotentialobtainedusinganiontrapwith“stretched”electrodespacing,obtainedfor5replicateexperimentsof50shotseach,usingaconstantstoragetimeof5ms.300400500600StoragePotential I V(0-peak)z>1.1-i0)I0140x103120100806040200700Figure6.8:PlotofintegratedionsignalintensityobtainedbyLAITMS(532nm)ofatomicionsfromaninconel747sampleversusstoragepotentialobtainedusinganiontrapwith“normal”electrodespacing,obtainedfor5replicateexperimentsof50shotseach,usingaconstantstoragetimeof5ms.200300400500600StoragePotentialI V(0-peak)152with a constant storage time of 5 ms used throughout these investigations.Figures 6.7 and 6.8 show plots of integrated peak intensities versus storagepotential for stretched” and “normal” electrode configurations. Bothconfigurations exhibit similar performance trends, in that initially low ion signalsare obtained for lower storage potentials, followed by a region of storagepotentials where the ion signal response remains relatively constant. Furtherincreases in storage potential give rise to larger signal intensities, although thesignal intensity ratios begin to vary as well. In other words, an optimal range ofstorage potentials is observed, with lower and higher storage potentialsresulting in inferior results for LAITMS. The smaller ion signals observed forlower storage potentials can be attributed to less effective ion storage in thecorrespondingly weaker trapping field. At higher storage potentials ions aremore efficiently confined within the trapping volume, reflected by the results inFigures 6.7 and 6.8.Comparison of integrated ion signal intensities for “stretched” and“normal” electrode geometry in this study reflects the results of Section 6.3.1:the “normal” geometry gives less reproducible results than the “stretched”electrode geometry. These studies of storage potential for both configurationssuggest that a storage potential of 300-500 V (0-p) is desirable for LAITMS ofsolid stainless steel and Inconel metal samples. This result proves that earlierinvestigations of LAITMS in this laboratory were conducted at a less thanoptimal storage potential [52, 53] which resulted in signals that may have beenas much as one fourth of those possible at an optimum storage potential(estimated from Figures 6.7 and 6.8). Thus far, the use of quadrupole ion trapelectrodes in a “stretched” geometry appears to provide more reproducibleresults than can be achieved with electrodes in the standard “normal” geometrydefined by Todd et. a!. [33]. The final investigation presented in this study153describes the effects which “normal” and “stretched” electrode configurationshave upon the mass resolution and peak shape achieved in LAITMSexperiments.6.3.3 Peak Shape and ResolutionTypical spectra were obtained for electron ionization of carbontetrachloride and LAITMS of Inconel 747 solid samples from experiments inSections 6.3.1 and 6.3.2. Experimental parameters were as given in thepreceding sections. Figure 6.9 shows mass spectral peaks corresponding toCCI3+ ions resulting from electron ionization of a carbon tetrachloride sample.Peaks for both “normal” and “stretched” electrode configurations are given tofacilitate direct comparison. The abscissa is given in data points rather thanmass numbers to facilitate resolution calculations; mass-to-charge ratio isdirectly proportional to data point number. The first peak, corresponding tomass-to-charge ratio 119, was superimposed for both electrode configurations.It is evident that the “stretched” geometry results in broadened peaks which tailmore gradually to lower mass numbers; the number of data points betweenpeak centers is larger for the “stretched” electrode geometry. This effect wasnoted by March [115] when comparing data obtained from different ion trapresearchers (some using “stretched” and some using “normal” electrodespacing). Resolution calculations [98] for mass to charge 119 and 121 arelisted in Table XIII and show that the “normal” electrode geometry affords 30%better mass resolution than “stretched” geometry. “Normal” electrode geometryalso results in more symmetric peak shapes than the “stretched” geometry.Similar phenomena were observed for LAITMS of Inconel 747 samples.Figure 6.10 shows mass spectral peaks for 56Fe+, 58Ni+ and 6ONi+ resultingfrom LAITMS of a solid Inconel 747 sample. As previously stated, the abscissa(0CC0)>0)—Stretchedelectrodegeometry—NormalelectrodegeometryJ30x12520_____151050120Figure6.9:MassspectralpeaksforCCl3(119,121,123amu)ionsobtainedbyelectronionizationofcarbontetrachlorideforboth“normal”and“stretched”electrodegeometry.Theabscissaisgivenindatapointstofacilitateresolutioncalculations,withm/zdirectlyproportionaltodatanumber.020406080100DataNumber!a.u.Cu>‘0Ca)Ca)4-Cua,10U,1—Stretchedelectrodegeometry—Normalelectrodegeometry30002500_______________________________2000150010005000120Figure6.10:Massspectralpeaksfor56Fe,58Niand60NiionsobtainedbyLAITMS(532nm) for bothhlnormaluand“stretche&electrodegeometry.Theabscissaisgivenindatapointstofacilitateresolutioncalculations,withmlzdirectlyproportionaltodatanumber.020406080100DataNumberIa.u.156is plotted in data points and the 56Fe peaks for both electrode configurationswere superimposed to allow direct comparison and facilitate resolutioncalculations. Again, broadened peaks were observed for LAITMS (similar to theelectron ionization experiments), with peak resolutions for 56Fe+ and 58Ni+given in Table XIII.Table XIII: Resolution calculations for “normal” and “stretched” electrodeconfigurations.Electron ionization of carbontetrach brideMass Resolution *CCl3 (119 amu) “normal” geometry 127CCI+(119 amu ) “stretched” geometry 93Percent difference -26.4 %CCl3 (121 amu) “normal” geometry 167CCI+(121 amu ) “stretched” geometry 107Percent difference-35.9 %LAITMS of Inconel 74756Fe+ “normal” geometry 61.656Fe+ “stretched” geometry 33.5Percent difference -46%58Ni+ “normal” geometry 79.858Ni+ “stretched” geometry 44.9Percent difference-44%* Reference [98]157These resolution values demonstrate that for LAITMS of Inconel 747, “normal”electrode geometry affords 40% better resolution and more symmetric peakshapes than the “stretched” geometry, consistent with the conclusion from theelectron ionization studies of carbon tetrachloride.6.3 SummaryThis chapter has investigated an empirical modification of thequadrupole ion trap electrode geometry obtained by increasing the end capseparation by eleven percent, similar to the commercially available FinniganITMSTM quadrupole ion trap mass spectrometer. The results showed thatincreasing the end cap electrode separation improved signal reproducibility, butat the expense of the 30-40% better resolution possible with the end capseparation used previously [52]. Moreover, it was determined that normal endcap electrode spacing results in more uniform mass peak shapes than the“stretched” geometry, likely the result of different ion ejection characteristicsduring detection which arise from the modified quadrupole electric fields. Inmass selective instability ion trap operation [96], the ions confined within thetrapping volume are detected by ejection through one end cap, thus physicalalteration of the electrode geometry results in not only modifying the storagefields but also acts to change the dynamics of ion detection. During the courseof this study, it was determined that a storage potential value of 300-500 V 0-pwas desirable for LAITMS; this storage potential range enhances the signalsobtained in this laboratory by up to four times more over those obtained fromprevious investigations [52, 53].158CHAPTER 7EVALUATION OF TWO COLOR LASER ABLATIONION TRAP MASS SPECTROMETRY7.1 IntroductionThe preceding chapters described the development of Laser Ablation IonTrap Mass Spectrometry (LAITMS) and the improvements on the initialexperiments by delineating how the various parameters influenced the signalsobtained. This chapter is devoted to development of a method of LAITMS whichselectively enhances the sensitivity for a desired analyte. Several methods ofanalysis are possible, such as selective analyte storage, selective detection ofanalyte species of interest and enhanced ionization of the desired analyte.Selective analyte storage in ion trap mass spectrometry can beaccomplished by either ejecting all species below a cutoff mass [116, 117](provided the analyte is of higher mass-to-charge ratio than the undesirablecomponents), or by creating storage fields which are selective for one analytespecies. These selective electric fields can be created by superimposing a DCcomponent upon the RE storage potential [33, 118, 119], creating quadrupoleelectric fields which lie at an apex of the stability diagram of a desired ion. Thisis schematically represented in Figure 7.1, which shows two stability diagrams.By using electric fields corresponding to location “a”, selective storage of themass-to-charge 18 ion would occur. Another approach to selective ion storage[120] involves combining an RE storage voltage and a frequency synthesizer.Any of these methods for selective analyte storage may be used to enhancesensitivity by either storing and integrating analyte ions from several laser159Figure 7.1: Typical stability diagrams for ion storage in an ion trap for ions ofmass/charge 18 and 28 amu.U(V)v(V)160ablation events [121] or eliminating space charge effects [40] arising from matrixspecies.Analyte ions stored in an ion trap can also be selectively detected byoptical methods such as atomic fluorescence. The ultimate limit of detection (le.single atom detection) was demonstrated over a decade ago [122] for a bariumion in a quadrupole ion trap. Selective detection schemes are advantageous inthat specific analyte sensitivity can be dramatically improved, but this selectivitymay be difficult, costly and often impractical for analysis of several analytespecies.Methods such as Resonance Ionization Mass Spectrometry (RIMS) [123]use lasers to selectively ionize analyte species of interest for massspectrometry. Analysis by RIMS requires that the sample be gaseous andconfined to a relatively small volume where the laser beam(s) can ionize thedesired analyte species for mass analysis. Some RIMS experiments [124-126]analyze solid materials by desorbing or ablating neutral material directly fromsolid sample matrices. During these experiments, the ablation laser normallyacts only as a sampling tool to vaporize and atomize analyte from a solidsample, which is subsequently ionized by the RIMS laser(s). Because theablation laser pulse generates ions directly from solid samples for LAITMS, asecond laser beam of desired wavelength passed through the neutrals createdduring the laser ablation event may enhance ionization of specific analytespecies, resulting in enhanced sensitivity. Experimental obstacles to thesuccess of this methodology include spatial and temporal alignment of thesecond laser beam with respect to the ablation laser beam. Because a nitrogenlaser pumped dye laser was available in the laboratory, this experiment wasattempted for LAITMS of Mn in a stainless steel sample.161Manganese was the analyte of choice for several reasons. It is acomponent of readily available NIST stainless steel (C1151) and has atomicground state transitions with favorable transition probabilities (see Figure 7.2[127] and Table XIV [1281). Moreover, a dye was available for the laser whichcould be tuned to the required wavelength (403.07 nm) to populate the 4p levelfor Mn(l). The dye laser was used to populate an excited state of neutral atomicanalyte, in order to investigate the enhanced ionization for Mn(l) in the ablationlaser plasma and the resulting effects upon sensitivity.Table XIV: Transition probabilities for manganese(l).Wavelength Spectrum Lower Energy Upper Energy(nm) Level (K) Level (K) (108/sec)403.076 Mn (I) 0 24802 1.4403.307 Mn (I) 0 24788 0.95403.449 Mn (I) 0 24779 0.54c.’JCD155450UzUi35—725IoT60e---z4.j(’-)Figure7.2:Grotriandiagram(adaptedfromStoner&Bashkin,1975)forMn(l).Thediagramrepresentstransitionsfroma6sande6stoSextetlevelsbelow56500cm’forMn(l) (1s2sp63d54).6P4‘01637.2 Experimental7.2.1 EquipmentThe LAITMS spectrometer used for these experiments is described in aprevious publication [52]. The spectrometer was modified slightly byintroducing a second laser assembly. The lasers used were a PhotochemicalResearch Associates (PRA, London, Ontario) model LN1000 pulsed nitrogenlaser, pumping a PRA model LN1O7 dye laser. To enhance the ionization ofneutral manganese, a PRA 7A400 dye solution (82.7 mg of 2-[1,1’-biphenyl]-4-yl-6-phenyl-benzoxazole {PBBO} in 50 mL 70/30 v:v Toluene Ethanol) wasused in the dye laser to provide laser radiation of 403.07 nm. The dye laseroutput wavelength was adjusted by controlling the laser grating with a PRAmodel DD1790 digital drive unit. Wavelength calibration was achieved bymethods similar to those used by LeBlanc [129], whereby a 0.35 m CzernyTurner monochromator (Model 270, Schoeffel-MacPherson, MA, USA) wasused to measure the output wavelength of the dye laser. The nitrogen and dyelaser were mounted on an optical rail placed near the LAITMS spectrometertable so that the dye laser beam could be directed coaxially with the ablationand alignment laser beams. An schematic of this arrangement is given inFigure 7.3. For these experiments, Q-switched, frequency doubled Nd:YAGlaser radiation (10 ns, 532 nm) was used. When the ablation laser pulse isfocused upon the sample surface, the coaxially aligned dye laser (403.07 nm)pulse reaches its focal waist at a point slightly in front of the sample surface.This behavior results from the effects of chromatic aberration [130] arising fromfocusing both beams with a simple piano convex lens. This discrepancy inlaser foci acts to favor the two color LAITMS experiment, because the neutralgas phase analyte atoms of interest for this investigation are located above the164Figure 7.3: Schematic diagram of the optical configuration used for the2 color LAITMS experiment.DifferentiallyPumpedSample ProbeA = Spatial Filter30cm LensAHigh EnergyMirrorBeamA DumpBeam Steering ColumnRight Angle PrismHe-Ne Alignment Laser165sample surface (in the laser plasma) where the dye laser power density ismaximized.Even though the lasers used for this investigation can be alignedspatially, taking advantage of chromatic aberration, temporal considerations forthe laser pulses are crucial for the success of this experiment. The ablationlaser has a 10 ns duration, which results in the formation of a microplasma atthe sample surface as described in Chapter 1. The lifetime of this transientplasma can vary, but ion production terminates by approximately 100 ns [85]according to literature sources. This means that the arrival of the dye laserpulse (width = 700 ps) must not only coincide with the ablation laser pulsearrival at the sample surface, but must be delayed slightly after its arrival (within100 ns) in order to excite neutral manganese atoms so that ionization in theplasma can occur. For the experiment, this delay was implemented byemploying a variable delay pulse generator constructed in house (TheElectronics Shop, Chemistry Department, U. B. C.). This delay generator is thesame as the one described previously for gating the electron gun for electronionization experiments. The delay generator was implemented by sampling thetrigger pulse sent to the ablation laser, delaying for a variable period, thensending a trigger pulse to the nitrogen and dye lasers. A pictorialrepresentation of the electronic arrangement used is given in Figure 7.4. Inorder to get the two laser pulse arrival times approximately the same, a quartzbeam sampler was placed in front of the laser window on the vacuum manifold.The lasers were triggered by the ion trap data acquisition computer, whilearrival times for the ablation and dye laser pulses measured with an InstrumentTechnology Ltd. model TF1850 high current 50 2 photodiode (The TechnologyShop Inc., Sudbury, MA) which was placed in the sampled beam. Laser pulsearrival times were synchronized to be approximately the same time, then varied166Figure 7.4: Box diagram of the electronics and instrumentation used forthe two color LAITMS experiment.167in a systematic manner with the variable delay on the pulse generator.Experimental parameters for these two color LAITMS experiments are given inTable XV.Table XV: Instrumental and laser parameters for the two color LAITMSexperiment.Laser ablation---Wavelength 532 nmPulse width 10 nsBeam waist at sample surface 50 tmIrradiance 5 X 1010 W•cm2Dye Laser--Wavelength 403.07 ±0.05 nmPulse width 700 PsPulse energy -2.5 jiJStorage and ejection---Storage time 5 msRF. 1.05MHzRF. storage potential * 250 VInitial RE. scan potential * 250 VFinal RE. scan potential * 3600 VRE. scan rate 2.54 X iO Vs1Ion detection--GEM voltage bias -1800 VCEM current amplification 1 X 1o V•A-1Number of scanst per spectrum 50*Measured (0 - peak)fone scan per laser shot1687.2.2 SamplesA NBS stainless steel sample (C1151) was used for these investigations.The sample surface was prepared [52] by polishing with 600 gauge abrasivepaper (wetted with methanol) polishing followed by a methanol rinse(Omnisolvegrade, BDH Chemicals, Toronto, Ont.), de-ionized water then methanol rinse,then air drying. Samples were aligned flush with the ring electrode innersurface using the helium-neon alignment laser.7.3 Results and DiscussionAfter careful alignment of the lasers used for this experiment, two colorLAITMS spectra were obtained for Cl 151 stainless steel at different delayvalues. Figure 7.5 represents the intensity ratio for 55Mn+ / 56Fe+ at variousdelay values. Changes in the relative intensity ratio reflect enhancements in thesignal for 55Mn+, produced by the supplementary dye laser radiation. Eachpoint is the average of five spectra (see Table XV). The timing delay valueshave errors of -±20 ns. Even though the timing delay error was very high, thisinvestigation showed enhancements for the ion signal obtained for manganeseusing dye laser delay values of 50-100 ns after the ablation laser pulse. Theseresults are consistent with the Balazs investigations [85], where ion generationby the plasma terminated after approximately 100 ns. Examination of the pulsedelay generator output showed that the device had 250 ns pulse delay shiftsoccurring at irregular intervals. Although the symptoms were observable,attempts (The Electronics Shop, Chemistry Department, U.B.C.) aimed atremoving this irregularity proved futile. These pulse shifts are showndiagramatically in Figure 7.6 and may explain the relatively large error barsobserved in Figure 7.5. The dye laser pulses arriving before the ablation laserpulses do not contribute to 55Mn+ signal enhancements, but the shifted pulses0)Co1200Figure7.5:Plotof55Mn/ 56Fe-’-ionsignalintensityratio(±2a)versusdyelaserdelayrelativetolaserablation,averagedfor5separateexperiments.0.90.5DyeLaserDelayIns-100-50050100150170______________________ _________Ablation Laser Pulseri 1200ns______ __ ___ __:..Dye Laser PulseNB: Not Drawn To ScaleFigure 7.6: Diagram of the dye laser pulse delay generator outputwhich shows the sporadic pulse shift “bug” produced bythe electronics._ HAblation Laser Pulsen200nsI 1 Dye Laser PulseNB: Not Drawn To ScaleFigure 7.7: Diagram of the dye laser pulse delay generator whichshows the shorter delays used for the second experiment.110.90>C0.7C+0LL.%0.60.5200Figure7.8:Plotof55Mn/56Feionsignalintensityratio(±2a)versusdyelaserdelayrelativetolaserablationforthesecondinvestigation,averagedfor5separateexperiments.-300-200-1000100DyeLaserDelayIns172arriving later in time coincide with the ablation event and result in the observedmaxima.A second experiment was performed in an attempt to experimentallyverify the effect of these sporadic pulse shifts upon the observed analyte signal.This experiment used shorter time delays (summarized by Figure 7.7) and theresults of this investigation are combined with the previous one in Figure 7.8.Each point in the figure corresponds to the average of five spectra (see TableXV). The timing delay values have errors of ± 20 ns. Figure 7.8 exhibits twomaxima, one for dye laser delays of about -150 ns and one at 50-100 ns. Thenegative delay values signify that the dye laser pulse arrives before the ablationevent (which affords no enhancement in ionization), verifying the presence ofthe random pulse delay observed from the pulse delay generator. Thistemporal jitter manifests itself as large errors (present in figures 7.5 and 7.8),especially when the timing delay results in enhanced 55Mn+ signals.7.4 SummaryThis chapter investigates the potential application of a dye laser forLAITMS experiments in order to selectively enhance the ionization ofmanganese in a stainless steel sample. The methodology is simple in principle,but requires that timing for the laser pulses must be accurate on thenanosecond to picosecond time scale. For these investigations, an inadequatedelay generator was used for timing laser pulse arrivals, giving promising yetinferior results. This study demonstrated selective ionization enhancement of asingle analyte can be implemented by applying a supplementary excitationlaser to the laser ablation plasma used for sampling and ionization. Thesupplementary radiation populated an excited state for the neutral atomicanalyte (manganese), facilitating thermal ionization of the excited atoms andproducing enhancements in observed signals. The results obtained werepromising and further study is warranted when a more suitable delay generatorbecomes available.173174CHAPTER 8INVESTIGATION OF THE USE OF AN ION TRAP WITHCYLINDRICAL ELECTRODE GEOMETRY8.1 IntroductionThe quadrupole ion traps described thus far have been constructed witha ring electrode described by a hyperboloid of one sheet and end capelectrodes given by two complimentary hyperboloids. The cross section of thiselectrode arrangement is described by Equations 1.6 to 1.8. To achieve idealquadrupole storage fields with electrodes described by these equations impliesthe use of infinite electrode surfaces. Obviously, this is not possible for anexperimental situation, where finite electrode geometry must be used; thus,only approximately ideal quadrupole storage fields are created insideexperimental ion traps. Preliminary ion trap experiments in this laboratory [53]used electrodes with truncated hyperbolic surfaces. This electrode geometrywas replaced in later investigations by electrodes with extended hyperbolicsurfaces carried out to a radius of 30.00 mm [52] in an effort to create more idealstorage fields. The results of these design modifications were discussed inChapter 4.Hyperbolic electrodes are not easily constructed, requiring precisemachining and polishing, making their manufacture both time consuming andcostly. Moreover, apertures are often made through the electrode surfaces forelectron, ion or photon transmission, which undoubtedly alter the ideality of thestorage fields created by the electrodes. Even though hyperbolic electrodegeometry has been emphasized for quadrupole ion trap electrodes, early work175by P. H. Dawson and co-workers [118] as well as J. F. J. Todd’s research group[35] showed that ion storage devices could function with non-hyperbolicelectrodes. In fact, a patent was issued to Langmuir et. a!. in 1962 [131] for anion trap with cylindrical electrode geometry two years after Paul andSteinwedel’s original [24] quadrupole ion trap patent.Cylindrical quadrupole ion trap electrodes have a barrel-shaped cylinderas the analog of the ring electrode of a “conventional” ion trap and two circularplates above and below to serve as endcap electrodes. In 1973, Benilan andAudoin [37] reported a simulation study for two forms of cylindrical ion traps. Intheir simulation studies, the radio frequency (RE) potential was applied to theendcaps with the ring electrode grounded, a situation which presentsexperimental difficulties for ion ejection through the endcaps for mass analysis.Bonner and co-workers [132] further discussed theoretical aspects of cylindricalion trap operation and demonstrated the first evidence of ion storage with acylindrical device in 1977. Additional theoretical and experimental evidence ofthe potential use of cylindrical ion traps was put forth in 1980 by Nassiopouloset. a!. [133].In order to evaluate cylindrical ion trap operation, a rather extensivestudy [134] was conducted in which the experimental stability diagrams for ionmotion and the mass spectrometric performances of several cylinderconfigurations were directly compared to that obtained with an ion trap havinghyperbolic electrodes. Quadrupole ion trap electrode dimensions investigatedby this study are given in Table XVI. The study showed cylinder 2 was the bestfor both ion storage (based upon experimental stability diagrams) and alsogave the best mass spectra for stored ions.176Table XVI: Dimensions of the cylindrical ion traps used by Mather et. a!. [134].Ion Trap Zi LiLZ-i(cm) (cm)Cylinder 1 0.678 0.556 1.215 1.476Cylinder2a 1.001 0.712 1.406 1.976Cylinder 3b 0.999 0.751 1.330 1.769QUISTOR 1.000 0.707 1.414 2.000aThese dimensions are for the inscribed cylinder of the QUISTOR.bThese dimensions are for the cylinder whose field at the centerresembles the QUISTOR most closely.As Table XVI indicates, cylinder 2 most closely represents the geometry of aquadrupole ion trap with hyperbolic electrodes. All cylinders worked as ionstorage devices and gave mass spectra, but the results were inferior for cylinder1 and 3 compared to cylinder 2. Their study concluded [134] that although theperformance of cylindrical ion traps as mass spectrometers was not as good asthat for hyperbolic ion traps, for their size and simplicity, cylindrical ion trapsgave remarkably good mass spectra. Additionally, the research of Lagadec etal. [64] demonstrated that cylindrical ion traps achieve greater ion densities andtotal numbers in the storage volume than their hyperbolic analogs.The objective of this chapter is to investigate some of the operationalcharacteristics of a simple cylindrical ion trap and compare it’s performance withan ion trap having hyperbolic electrodes. This direct comparison will determinethe feasibility of using simple, inexpensive cylindrical electrodes forexperiments such as Laser Ablation Ion Trap Mass Spectrometry (LAITMS) aswell as for experiments involving optical interrogation of stored ions.-4--14.14[‘:HHri/iRingElectrode-( VESPEL’ Spacersr End Cap Electrode 2177L _j20.00 Immft60.00 mmi End Cap Electrode 1Ring450400Figure 8.1: Detailed schematic diagram of the cylindrical ion trap electrodes.178Table XVII: Operational parameters used for the cylindrical ion trapexperiments.Ionization---Electron energyFilament currentDuration of electron beamStorage and ejection---Storage timeRE.RE. storage potential *Initial RE. scan potential *Final RE. scan potential *RE. scan rateIon detection--CEM voltage biasCEM current amplification70 eV2.5- 3.23 Alms5 - 1000 ms1.05 MHz50 - 550 V50- 550 V3600 V2.54 X io V•s1-1800 V1 X i0 VA1Number of scans per spectrum 100*Measured (0 - peak)1798.2 Experimental8.2.1 EquipmentThe spectrometer used for these investigations was described in aprevious publication [52], but in this case employed electrodes with cylindricalgeometry. The electrodes were constructed so that the endcap and ringelectrode spacings were identical to those for the hyperbolic electrodesdescribed in Chapter 4. These dimensions were similar to those used byMather et al. [134] in their investigations of cylindrical ion trap performance. Adetailed schematic diagram of the cylindrical ion trap electrodes is given inFigure 8.1. Table XVII lists operational parameters used for the cylindrical iontrap experiments.8.2.2 SamplesFor these investigations a gaseous sample of 10 % xenon in helium(Spectra Gases Inc., Newark, NJ) was introduced to the trap via the variableleak valve.8.3 Results and Discussion8.3.1 Comparison of Cylindrical and Hyperbolic Electrode PerformanceInitial investigations with the cylindrical electrodes gave large signals,suggesting that analyte was being ionized and stored by the new electrodes.After careful adjustment of sample introduction via the variable leak valve, massspectra were obtained for xenon isotopes. The mass spectra obtained wereinferior to those obtained in earlier experiments involving hyperbolic electrodes:They were poorly resolved, showing the entire Xe isotopic cluster as anunresolved broad peak. Figure 8.2 shows a typical mass spectrum obtained forXe isotopes with the cylindrical electrodes. Operational parameters are given in(U>10CC)CC)CaC)CQDXe16001400120010008006004002000200Figure8.2:Typicalelectronionizationmassspectrumobtainedforxenonisotopesusinganiontrapwithcylindricalelectrodes.Experimentalparametersusedaregivenintable8.2.50100150mass/chargeIamuco6000-1291305000-131132Xe34000->..1U)3000-1342000-136‘0_I••••IIIIIIIIIIIIIIIIIIIIIIIIJIIIIIII100105110115120125130135140145150mass/chargeIamuFigure8.3:Typicalelectronionizationmassspectrumobtainedforxenonisotopesusinganiontrapwithhyperbolicelectrodegeometry.182Table XVII, using 2.72 A filament current, 1.6 pBar He and 25 ms storage time.In comparison, Figure 8.3 shows a mass spectrum for Xe obtained using ahyperbolic electrode geometry. Because of the large discrepancy inperformance between the two electrode configurations, parametricinvestigations were conducted with the cylindrical electrodes in order toevaluate their ability to function as a simple mass spectrometer.8.3.2 Electron Ionization and Buffer Gas PressureTo evaluate the degree of ionization of neutral analyte within the storagevolume of the cylindrical electrodes, the effect of varying the flux of electrons forionization was examined. This was accomplished by varying the currentpassed through the filament of the electron gun. Experimental parameters aregiven in Table XVII , using a 250 V and 25 ms storage period, as well as 1.6p. Bar He buffer gas pressure. Figure 8.4 shows the effect of varying the electrongun filament current upon the signal obtained for the Xe isotope cluster. Eachpoint on the curve represents the sum of 100 separate scans. The resultsclearly show that after a threshold value of 2.5 A, the ion signal obtainedincreases until a maximum value is reached at about 3 A. This can be attributedto either ionization at maximum efficiency of the spectrometer and/or storage ofthe maximum number of ions within the trap. Because the signal intensitiesobtained were relatively low, it can be assumed that the electron ionizationefficiency for Xe by 70 eV electrons was maximized for a filament current of 3 Ausing these cylindrical ion trap experimental conditions.The presence of a buffer gas remarkably improves the operation of iontraps with hyperbolic electrodes [96]. To this end, the operation of thecylindrical ion trap was investigated at various helium buffer gas pressures.Experimental parameters are given in Table XVII, using a 250 V and 25 msC)a:,>4-.0CC,4-’CC,>4-’C)PlotofXeionintensityversuselectrongunfilamentcurrent.Eachpointrepresentstheaverageof100scans.2500200015001000-0-Xe-Background02.52.62.72.82.93.03.13.2ElFilamentCurrent/ AFigure8.4:>‘(00C0>(U0OD1-0—Xe.Background20001500100050000.51.0HeliumPressureI Bar1.5PlotofXeionintensityversusheliumbuffer gaspressure.Eachpointrepresentstheaverageof100Figure8.5:scans.185storage period with a 2.72 A filament current. The results are summarized inFigure 8.5, which shows an increase in signal intensity upon increasing the Hegas pressure within the vacuum system. All pressures are corrected [97] heliumpressure values. The upper pressure of the investigation was limited to 2.33tBar so that the channel electron multiplier would not be damaged by arcing athigher operational pressures. Clearly, the use of a helium buffer gas improvesthe operation of the cylindrical ion trap in that more ions are effectively storedand detected, giving enhanced ion signals from the spectrometer. Unlike thehyperbolic electrode geometry, vastly improved resolution was not observed formass spectrometry of the stored ions, suggesting inferior operation of thecylindrical electrodes for mass spectrometry by mass selective instability mode.Thus far, the cylindrical electrodes have shown similar operational trends(excepting mass resolution) with respect to electron ionization and buffer gaspressure to those of hyperbolic electrodes. Further investigation of operationalparameters such as storage potential and time provide additional insight intothe operation of simple cylindrical electrodes.83.3 Storage Time and PotentialSection 8.3.2 dealt with operational parameters external to thequadrupole electric fields for ion storage and analysis. This section investigatesboth storage time and potential, both of which are respectfully regulated by theduration and magnitude of the quadrupole electric field used for ion storage.Figure 8.6 shows the effect of storage time upon the ion signals obtained for Xeisotopes. Instrumental parameters for this study are given in Table XVII, with a250 V storage potential, 2.72 A filament current and pressure of 1.6 j..tBar for thisexperiment. The result of this investigation suggests that increasing the storagetime decreases the observed stored ion signals. These conclusions differ from(0114001200ci1000U,c800C)Ca)>6004-(UC)rr40020001000—0--Xe’.Background200400600800StorageTimeImsFigure8.6:Plotof Xeonintensityversusionstoragetime.Eachpointrepresentstheaverageof100scans.F.—0:,€0>C,,CC0)€00)I1000800600400-0-XeIBackgroundRFStoragePotentialI VoltsFigure8.7:PEotofXeionintensityversusionstoragepotential.Eachpointrepresentstheaverageof100scans.188the results of Chapter 6 for the hyperbolic electrodes, which exhibit maxima forion signals at storage times of 100 ms, shown previously in Figures 6.1 and 6.2.This suggests that although similar, the quadrupole electric fields for thecylindrical electrodes differ markedly from those for hyperbolic electrodes. Thisis reflected by the observed inferior mass resolution obtained with the deviceconstructed for this chapter. Investigations of the effect which storage potentialhas upon ion signal intensity are shown in Figure 8.7. Operational parametersfor this experiment are given in Table XVII, keeping a constant 25 ms storagetime, 2.72 A filament current and 1.6 iiBar He buffer gas pressure. Theexperiment shows that a defined ion signal maxima exists between 150-300 V,again differing from the results in Chapter 6 (Figure 6.7 & 6.8) and furtherillustrating the dissimilar nature of the electric fields generated within the twoelectrode configurations.8.4 SummaryThe initial goal of this chapter was to evaluate the performance of simplecylindrical electrodes in an ion trap mass spectrometer, in an effort to simplifythe instrumentation required for experiments such as laser ablation ion trapmass spectrometry (LAITMS). Preliminary studies showed that although thecylindrical electrode arrangement did store ions, subsequent mass scanningusing mass selective instability operation [96] produced mass spectra of inferiorresolution to that possible with hyperbolic electrode geometry. Furtherparametric studies of the cylindrical trap indicated that it did behave in a similarmanner to the hyperbolic trap. However, the quadrupole electric fielddifferences between the two devices gave inferior operation for the cylindricalelectrode geometry when used as a simple mass spectrometer. Becausecylindrical ion trap electrodes are simple to construct and do store ions189effectively within the storage volume, these inexpensive devices could facilitateoptical ion trap experiments. The simple design of cylindrical ion traps makestheir fabrication inexpensive, and allows construction of traps with opticalprobing apertures through the ring and endcaps that are not practical withhyperbolic electrodes. Moreover, the electric fields of any electrode system arealtered by machining holes etc. through their surfaces, making inexpensiveelectrodes desirable. Permanent experimental modifications which corrupt iontrap operation (le.: machining apertures) are not as costly for cylindricalelectrodes as they are for expensive, precisely machined hyperbolic electrodes.The design simplicity and low cost of cylindrical ion trap electrodes caneffectively address the requirements for optical ion trap methods.190CHAPTER 9LASER ABLATION SYNCHRONIZATION WITH RESPECTTO THE RADIO FREQUENCY STORAGE FIELD9.1 IntroductionIon generation by impingement of high energy laser light upon a samplematrix has been demonstrated [135] to be a useful method of ionization formass spectrometry. The ion generation process occurs as a result of intenselocalized heating of the sample surface [74] through the formation of a laserinduced micro plasma. For laser irradiances of 1X109W•cm2 or greater, thelaser solid interaction removes significant amounts of material, forming a craterwith repeated shots from a pulsed laser system upon the same spot. Thisprocess of material removal is called laser ablation, and has been developed inthe previous chapters of this thesis as a means of sampling and ionizing solidmaterials for quadrupole ion trap mass spectrometry.Earlier in this thesis, investigations of laser ablation as a sampling andionization method were concerned with using the method for ion creation withinthe storage volume of the ion trap. Since ion generation by laser ablation hasbeen addressed in Section 1 .3 of this thesis, it will not be discussed in detailhere. Ion production by Q-switched laser radiation occurs on the 100 ns timescale [85], during the vaporization and plasma formation stages of the laserablation event. This makes laser ablation a pulsed transient ion source,suitable for use with mass spectrometric methods such as time of flight (TOE)and ion storage methods such as Ion Cyclotron Resonance (ICR) orQuadrupole Ion Trap (QIT) mass spectrometry, all of which require pulsed ion191sources for mass analysis. Ion storage methods are advantageous becausethey allow storage of analyte from several ablation events as well as selectiveanalyte storage to enhance the signals for trace species.Quadrupole ion traps utilize rotationally symmetric radio frequency (RE)electric fields for ion storage and analysis. The period of the oscillating electricfield is normally on the order of 1 ps for the RE fields in quadrupole ion trapmass spectrometry. Because ion production occurs on the 100 ns time scale[85], the phase of the electric field should be important during ion generation bylaser ablation. Eor LAITMS, direct solid sampling by laser ablation occurs at thesurface of a sample pin which is radially inserted flush with the inner surface ofthe ring electrode [52, 53]. Depending upon the phase of the RF field, the ringelectrode (and sample matrix) may be positive, negative or possess no chargeduring the laser ablation event. This chapter examines the effect ofsynchronization of the laser ablation event with the RE storage frequency in aneffort to understand the phenomena and enhance the ion signals obtained byLAITMS experiments.9.2 Experimental9.2.1 EquipmentThe spectrometer used for these investigations was described previously[52] in Chapter 4. Changes to the existing spectrometer include a 24 turninductance toroid mounted coaxially on the RE lead near the ring electrode inorder to sample the applied potential. The phase of this sampled RE storagefield is detected by a zero crossing detection device designed and built inhouse (Electronics Shop, Chemistry Department, U.B.C. ). This device isenabled by a signal pulse from the computer, and produces a trigger pulse forthe ablation laser at a variable delay after a positive zero crossing in the RE192NB: Not drawn to scale.RF StoragePotentialLaserAblationFigure 9.1: Schematic diagram which summarizes the function of theelectronics used for laser ablation synchronization withrespect to the phase of the ion storage field.xDetectzerocrossingX = Variable Delay193phase was detected. The function of the device is pictorially summarized inFigure 9.1. By varying the delay, the effect of laser ablation at different RFphases upon the ion signals obtained by LAITMS was investigated.Experimental parameters for this investigation are given in Table XVIII.Table XVIII: Instrumental parameters for synchronization of the ablation laserwith the storage potential for LAITMS experiments.Laser ablation---Wavelength 532 nmPulse width 10 nsBeam waist at sample surface 50 imlrradiance 5 X 1010 Wcm-2RF-phase synchronization for laser---Phase delay values 0.25 to 0.88 ± 0.05 isNd:YAG internal delay 242.3 ±0.1 tsStorage and ejection---Storage time 5 msecRE. 1.05MHzRE. storage potential * 400 VInitial RE. scan potential * 400 VFinal RE. scan potential * 3600 VRE. scan rate 2.54 X io Vs-1Ion detection--CEM voltage bias -1800 VCEM current amplification i x ioNumber of scansf per spectrum 50*Measured (0 - peak)tone scan per laser shot1949.2.2 SamplesFor these investigations, Inconel 747 (Dr. A. Mitchell, MetallurgicalEngineering, U.B.C.) was used as a solid sample matrix. Sample compositionis given in Table XIX.Table XIX: Elemental composition of inconel 747.Element Percentage Composition (mass) *Nickel 63Chromium 18Iron 10Zinc <1*values given without error limitsThe sample surface was prepared [53] by 600 gauge abrasive paper (wettedwith methanol) polishing followed by a methanol (Omnisolve grade, BDHChemicals, Toronto), de-ionized water then methanol rinse, then air drying.Sample pins were aligned flush with the ring electrode inner surface using ahelium-neon alignment laser.9.3 Results and DiscussionThe initial phase synchronization electronics design failed to perform atall. The problem was that the zero-crossing detector sensitivity was inadequatefor the signal obtained with the inductance toroid. After correcting this byincreasing the sensitivity of the zero-crossing detector, the effect of laserLI)0)CuC’,Ca)CCuL.0)a)CFigure9.2:Plotof56Fe-’-,58Niand60NiionsignalintensitiesforionsresultingfromLAITMSàfinconel 747versusphasedelayvaluesforlaserablationrelativetothephaseoftheionstoragefield.800060000.5PhaseDelayIs0.60.7CD0)0.350.300.2504.d(U>0.20(U0CCo0.0.10(UC)0.050.00Figure9.3:Plotof56Fe,58Niand60NiionsignalintensityRelativeStandardDeviation(RSD)valuesforionsresultingfromLAITMSofinconel747versusphasedelayvaluesforlaserablationrelativetothephaseoftheionstoragefield.0.30.4PhaseDelayIjts197ablation for various phase delays was examined. The results of these studiesare summarized in Figures 9.2 and 9.3. Figure 9.2 shows integrated ion signalintensities at various phase delay values, each point representing the averageof 4 separate experiments. The figure shows the average ion signal changesdepend upon the phase of the RE storage potential. The plot of ion signalrelative standard deviation (RSD) (for these ion signals) versus phase delayvalue is given in Figure 9.3. This figure also shows definite maxima andminima, depending upon the phase delay value. This data (Fig. 9.2 and Fig.9.3) suggests that there is a correlation between the phase of the RF storagepotential and ion generation by laser ablation of solid samples. Although theresults are promising, they are undoubtedly limited by the accuracy of the phasedelay (±0.O5is) and internal timing delay of the laser electronics (±0.1 is). In asimilar series of experiments conducted at Los Alamos National Laboratories,Nogar and Hemberger [136] independently investigated the effect ofsynchronizing laser desorption with the amplitude and phase of the storagepotential. They present a new method of storing laser desorbed ions (called“dynamic RE trapping”); ions are created when the RE storage potential is low,followed by a rapid storage potential increase for ion capture and storage. Thisexperiment utilized state of the art electronics and instrumentation; the jitter ofthe phase synchronization electronics was an order of magnitude less (10 ns)than that observed for the experiments presented in this chapter (0.1 jis). The“dynamic RE trapping” method gives an order of magnitude improvement of theion signals obtained by laser desorption ion trap mass spectrometry, and withbetter precision. Their results also conclude that the phase of the RE is animportant parameter when using a pulsed laser desorption/ablation ionizationscheme for ion trap mass spectrometry. This conclusion is further validated bythe earlier research of Cotter et. at who observed anomalies in mass198discrimination of large biomolecules by Matrix Assisted LaserDesorption/Ionization experiments in an ion trap [137]. Their results alsoindicated that for a laser desorption ion trap mass spectrometry, phase-lockingof the laser output with the storage potential improved the experiment. Theexperiments at U. B. C. and at Los Alamos verify that phase synchronization isessential to the success of LAITMS; ion signal reproducibility can besubstantially improved by laser ablation which is in phase with the storagepotential. The improvements in ion signal intensity that result from thesynchronization of the laser ablation event with the phase of the storage fieldare attributed to improved trapping efficiencies for ions created by laserablation.9.4 SummaryThis chapter investigates the significance of synchronizing laser ablationwith the phase of the RE storage potential. Ion generation by laser ablation witha Q-switched Nd:YAG laser occurs for a —1OO ns time window, whereas theperiod of the RE potential used for ion storage is —1 is. The findings of thischapter and those from independent experiments at Los Alamos NationalLaboratory prove that phase synchronization of laser ablation with the RFstorage potential is important for experiments such as LAITMS because ionsresulting from laser ablation are more efficiently trapped. Improved ion trappingefficiency gives rise to larger ion signals with better reproducibility, desirable forLAITMS analytical applications.199CHAPTER 10CONCLUSIONS10.1 GeneralitiesThis thesis presented the evolution of a mass spectrometer capable ofproviding direct mass spectrometry of solid samples, based upon laser ablationcoupled with quadrupole ion trap (ion trap) mass spectrometry. The newmethod is called Laser Ablation Ion Trap Mass Spectrometry (LAITMS).Analysis of solid samples which are both conductors and non-conductors ispossible with the spectrometer developed for this thesis; also, both atomic andmolecular information can be obtained for analytes of interest. Ion trap massspectrometers are simple, compact, robust and relatively inexpensive massspectrometers which are commercially available. The development of LAITMSby this work has provided yet another demonstration of the many applications ofion trap mass spectrometry for chemical analysis. This thesis has examined theLAITMS method to provide an understanding of some of the chemical andphysical processes involved. Understanding these processes provides insightfor the application of LAITMS to the analysis of “real” solid samples of interest toanalysts.The disadvantages of LAITMS include a restricted sample geometry, alimited dynamic range resulting from space charge effects and possible matrixinterferences such as the presence of easily ionized elements. Additionally,sampling and ionization are not independent because they result from the samelaser pulse. Advantages of LAITMS include: the powerful mass spectrometricmethods possible with quadrupole ion trap mass spectrometry (includingcollision induced dissociation and MS experiments for the identification ofsample matrix species), ablated neutrals mass spectrometry by using electron200ionization, integration of sample ions from several laser pulses to enhancesensitivity and optical detection schemes such as laser induced fluorescence toachieve very low detection limits.10.2 Understanding of LAITMS Developed by This ThesisThe initial goal of this thesis was to develop an ion trap massspectrometer which uses laser ablation directly inside the storage volume tosample, as well as ionize solid materials in order to obtain multielementalatomic mass spectrometric information about the sample. The first experimentsattempted to use a pulsed ruby laser for laser ablation of solid materials inside aretrofitted ion trap mass spectrometer (originally intended for use with a graphitefurnace atom source [41]). These experiments did not provide reliable orreproducible analysis of solid samples, yet demonstrated that ions created bylaser ablation within the storage volume of an ion trap could be successfullystored and then analyzed by the device. One of the problems encounteredduring this experiment was that the ruby laser system (including electronics andoptical configuration) was not suited to laser ablation for sampling andionization of solid samples for ion trap mass spectrometry. A desirable laser forthis experiment would require higher irradiance (Wcm2)for smaller laser pulseenergies. This was possible for a laser capable of shorter laser pulses,suggesting that a Q-Switched Nd:YAG laser would be more appropriate.Moreover, better temporal control over the timing of laser ablation with respectto the ion trap scan sequence was made possible with the use of this type oflaser.Further experiments were implemented with laser ablation ion trap massspectrometry, using a Q-Switched Nd:YAG laser for sampling and ionizing solidmaterials directly inside the storage volume of this ion trap. These experiments201[53] demonstrated that using the higher laser irradiances and better temporalcontrol provided by the Nd:YAG laser does improve the results obtained byLAITMS. Mass spectra were obtained for solid metal samples, with adequatemass discrimination to show signals for the isotopes known to be present in thesamples. This experiment provided sufficient evidence to warrant furtherinvestigation of laser ablation as a means for ion creation directly inside thestorage volume of an ion trap. To conduct systematic investigations of thechemical and physical processes involved with the method, called LaserAblation Ion Trap Mass Spectrometry (LAITMS), a spectrometer was designedand constructed to meet these needs.The improved spectrometer was designed to facilitate the investigation ofmany of the experimental parameters involved in LAITMS of solid samplematrices. Included in the design were many useful features, including adifferentially pumped sample introduction probe to facilitate samplemanipulation, an improved optical configuration for laser ablation, a cleanervacuum system for the ion trap experiment and paths through the storagevolume of the ion trap for optical interrogation of the stored ions. The design ofthis spectrometer has been published elsewhere [52]. This new spectrometermade possible the investigations presented by the chapters of this thesis.Early LAITMS experiments identified an “uncharacterized surface effect”[53] for the analysis of solid samples. This was examined by investigations ofthe effect of surface preparation. The findings of these studies suggest that thesurface quality directly affects the ion signals obtained by the experiment.Smooth (diamond paste polished) sample surfaces give rise to initially low ionsignals. Conversely, surfaces prepared by abrasive paper polishing methods(resulting in fine grooves and scratches) have initially large ion signalintensities when analyzed by LAITMS. The discrepancies in the observed202signals arise from the different reflectivity of the sample surfaces. After 10 - 20laser pulses impinge either sample surface, a shallow ablation crater forms;upon further ablation, a deep crater forms, which drastically reduces theintensity of ion signals that result from LAITMS of solid samples. These resultssuggest that a small number of laser pulses (10 - 20) directed upon the samplesurface prior to LAITMS analysis results in better ion signal reproducibility forplanar sample targets.Laser irradiance control is essential for successful direct solid massspectrometry by LAITMS. The results presented by this thesis show thatirradiance control allows the analyst to ablate varying amounts of material fromsolid sample matrices, thus storing more or less analyte in the ion trap. Thedetrimental effects of space charge [40] can be eliminated by controlling thequantity of ions to be confined in the ion trap. Ablation laser irradiance controlwas implemented by adjusting the fluence from the ablation laser and also byadjusting the focus of the laser radiation upon the sample surface. Theseinvestigations suggest that varying the laser focus is superior to varying theflashlamp output of the laser for irradiance control.The wavelength of the laser used for sample ablation was found to affectthe results obtained by LAITMS. For molecular analytes, such as for ceramicand polymer samples, increased fragmentation of matrix species was observedfor ablation with higher energy 266 nm photon radiation (4.67 eV) over lowerenergy (2.33 eV) 532 nm radiation. When atomic analytes were analyzed byLAITMS of metal samples, different relative ion intensities were observed forcertain analytes. Both of these observations suggest that the ionizationconditions of the laser produced plasma from 266 nm laser ablation are moreenergetic than those for 532 nm laser ablation produced plasma. In general,increased ionization for certain analytes is observed when 266 nm laser203radiation is used for ablation purposes. Other researchers [111-113] have alsoconcluded that laser ablation/desorption using longer wavelengths results inless molecular fragmentation, consistent with the findings of this thesis.The absolute geometry of the hyperbolic electrodes used for quadrupoleion trap mass spectrometry affects the nature of the results obtained for both Elmass spectrometry of neutral molecules and LAITMS of solid samples. Thisthesis investigates the empirical modification of the quadrupole ion trapelectrode geometry obtained by increasing the end cap separation by elevenpercent, similar to the commercially available Finnigan ITMSTM quadrupole iontrap mass spectrometer [65]. The results show that increasing the end capelectrode separation improves ion signal reproducibility, but at the expense ofthe 30-40% better resolution possible with the end cap separation predicted bytheory [33] and used for this thesis. For direct solid mass spectrometry byLAITMS, the optimum ion storage potential was found to be 300-500 V (0-p) forboth electrode configurations. Moreover, it was determined that “normal” endcap electrode spacing (as predicted by theory) results in more uniform masspeak shapes than the “stretched” geometry. This can be attributed to differention ejection characteristics during detection, which arise from the modifiedquadrupole electric fields.Selective analyte ionization for manganese in stainless steel wasdemonstrated by a two color LAITMS experiment. The success of theexperiment was dependent upon precise spatial and temporal alignment of thetwo lasers. Although the results obtained were limited by the experimentalapparatus used, they did show signal enhancements for manganese over thoseobserved for one color LAITMS.The use of a quadrupole ion trap with cylindrical electrode geometry wasexamined in an effort to simplify the instrumentation requirements for LAITMS.204Studies presented by this thesis show that although the cylindrical electrodesstore ions, the mass spectra obtained have inferior resolution to that possiblewith a hyperbolic electrode geometry. Further parametric studies of thecylindrical trap demonstrated that it behaves in an analogous manner to thehyperbolic trap. However, the quadrupole electric field differences between thetwo devices produce inferior operation for the cylindrical electrode geometrywhen used as a simple mass spectrometer. Because cylindrical ion trapelectrodes are easily constructed and store ions effectively within the storagevolume, these inexpensive devices can facilitate optical ion trap experiments.The simple design of cylindrical ion traps allows the fabrication of inexpensivetraps with optical probing apertures through the ring and endcaps that may notbe practical with hyperbolic electrodes. Because of their simplicity and low cost,cylindrical ion traps can provide an alternative to expensive hyperbolic iontraps, especially for optical (laser) ion probing experiments.The final experiments presented by this thesis examined thesynchronization of laser ablation with the phase of the ion storage potential.Because ion production by laser ablation is limited to an -100 ns time frame,the phase of the RE storage potential (period = -1 is) is significant forsuccessful ion trapping. Independent research conducted at Los AlamosNational Laboratories [1361 confirms this conclusion. By synchronizing laserablation with the phase of the storage potential, ion trapping efficiency isimproved, resulting in larger, more reproducible ion signals.10.3 Future Research DirectionsThis thesis has presents and examines Laser Ablation Ion Trap MassSpectrometry (LAITMS) as a new method for direct solid mass spectrometry. Tofully understand the method, their are many different research areas which205need to be examined. For LAITMS, fundamental research in two distinct fieldshas been coupled: laser plasma spectrometry and quadrupole ion trap massspectrometry. This section outlines research in both of these areas whichdeserves future investigation, in order to further the understanding of LAITMSpresented by this thesis.The nature of ion creation and storage by laser ablation within thequadrupole ion trap warrants further study. LAITMS relies upon theconfinement of ions created by laser ablation of a solid sample. It is evident thatat least some ions that result from this process are stored, yet it is unclearexactly how much of the solid analyte is ionized, and how much of the ionizedmaterial is actually stored by the ion trap. The laser plasma could be studiedwith time resolved spectroscopic methods, and could be probed by a secondlaser to determine spatial and kinetic information about analytes. Thisinformation could then be used to model the ion storage process, and byapplying newer ion storage methods such as “dynamic RE trapping” [136] toLAITMS, the process of ion generation and storage would be better understood.To fully develop the potential uses of LAITMS requires full exploitation ofthe many mass spectrometric methods available for quadrupole ion traps.Experiments such as MS, selective analyte manipulation (dissociation, storageand ejection), high mass resolution and high mass limit detection are readilyavailable features for quadrupole ion traps. These methods can provide awealth of information for LAITMS experiments, by not only positively identifyingsample matrix species, but enhancing the sensitivity possible. The powerfulanalytical features of ion trap mass spectrometry are implemented by controlover the ion confinement fields, making fundamental research regarding ionstability in modified quadrupole electric fields necessary. This research can be206Razor Blade lnteerence4Beam Dump FilterSample PinFor AblationMirrorElectronicSignal TuneableFrom CW“conventional” DyeL.AITMS LaserExperiment________ArgonIonLaserFigure 10.1: Experimental arrangement for laser inducedfluorescence detection of La and Gd by LAITMSof solid samples.207modeled by computer simulation and experimentally verified by ion tomographyexperiments [55, 56].Ultra-trace analyte detection is the ultimate goal of LAITMS. This thesispresents absolute detection limits in the picogram range. Although intended asa final chapter, LAITMS experiments which use laser fluorescence for iondetection were incomplete at the time of thesis submission and, thus, are notincluded in this thesis. Optical ion detection schemes based upon laserfluorescence have been demonstrated to achieve the ultimate limit of detection,single ion detection [57]. The spectrometer presented by this thesis wasconstructed so that optical fluorescence experiments were possible.The experimental arrangement for laser fluorescence detection of Gdand La in solid samples is summarized by Figure 10.1. Experimentally, a CWdye laser (Dr. A. Merer, Chemistry Department, U. B. C. ) will be used to excitethe stored ions by passing the laser beam through the ion storage volume. BothGd+ and La+ have transitions which can be excited by the wavelength tuningregion available by using Rhodamine-6-G laser dye. The transitions for thesefluorescence experiments were selected so that excitation was at a differentwavelength than detection (direct-line fluorescence) [69]. To isolate thedetection wavelength, narrow band width interference filters (10 nm FWHM) willbe used. The fluorescent signal will be detected using a CCD camera, whichwill provide not only a signal, but spatial and temporal mapping for thefluorescent species in the storage volume of the ion trap. These investigationsare currently underway at the University of British Columbia.208REFERENCES1. W. Wien, Verhand/ungen der Physika/ische Gesellschaft zu Berlin, 17,189 (1898).2. W. Wien, Anne/en der Physik und Chemie, Leipzig, 8, 224 (1902).3. J. J. Thompson, Philosophical Magazine V, 44, 293 (1897).4. J. J. Thompson, Philosophical Magazine VI, 24, 209 (1912).5. J. J. Thompson, Rays of Positive Electricity, , Longmans Green and Co.:London (1913).6. L. Kaufman, Physikalische Zeitschrift, 2, 602 (1901).7. F. W. Aston, Philosophical Magazine, 38, 709 (1919).8. F. W. Aston, Philosophical Magazine, 39, 449 (1920).R. Herzog, Zeitschrift fur Physik, 89, 447 (1934).P. Herzog and J. H. E. Mattauch, Anna/en der Physik, Liepzig, 19, 345(1934).A. 0. Nier, Physical Review, 48, 283 (1935).W. E. Stephens, Physical Review, 69, 691 (1946).A. J. Dempster,A. J. Dempster,A. J. Dempster,A. J. Dempster,K. T. Bainbridge,K. T. Bainbridge,K. T. Bainbridge,K. T. Bainbridge,K. T. Bainbridge,K. T. Bainbridge,9.10.11.12.13.14.15.16.17.18.19.20.21.22.Physical Review, 11, 316 (1918).Science, 52, 559 (1920).Physical Review, 18, 415 (1921).Physical Review, 20, 631 (1922).Physical Review, 39, 847 (1932).Physical Review, 39, 1201 (1932).Physical Review, 43, 103 (1933).Physical Review, 43, 1056 (1933).Journal of The Franklin Institute, 215, 509 (1933).Physical Review, 44, 123 (1933).20923. A. E. Cameron and D. F. Eggers, Rev.ofSci. Instrum., 19, 605 (1948).24. W. Paul and H. Steinwedel, “Apparatus for separating charged particlesof different specific charges,” German Patent 944,900 (1956), U.S.Patent 2,939,952 (1960).25. F. M. Penning, Physica, 3, 873 (1936).26. J. D. Baldeschweiler, H. Benz and P. M. Liewellyn, Advances in MassSpectrometry, , ed. E. Kendrick, Institute of Petroleum: London (1968).27. M. B. Comisarow, J. Chem. Phys., 69, 4097 (1978).28. “Press Release,” Royal Swedish Academy of Sciences, October 12,1989.29. P. E. Kelley, J. G. C. Stafford and D. R. Stephens, “U.S. Patent 4 540884 10 (1985).30. P. E. Kelley, J. G. C. Stafford and D. R. Stephens, Can. Patent 1 207918 (1986).31. R. F. March, mt. J. Mass Spectrom. Ion Proc., 119, 71(1992).32. R. E. March and R. J. Hughes, Quadrupole Storage Mass Spectrometry,1 ed. Chemical Analysis: A series of monographs on analyticalchemistry and its applications, ed. J. D. Winefordner, Wileylnterscience: Toronto (1989).33. R. F. Mather, A. M. Waldren and J. F. J. Todd, Dynamic MassSpectrom., 5, 71(1978).34. P. H. Dawson, Quadrupole Mass Spectrometry and Its Applications,Elsevier: Amsterdam (1976).35. G. Lawson, R. Bonner and J. Todd, J. Phys. E: Sciinstr., 6,357(1973).36. F. Mathieu, J. Math. Pure Appi. (J. LiouvilIe), 13, 137 (1868).37. M. Benilan and C. Audoin, mt. J. Mass Spectrom. Ion Phys., 11, 421210(1973).38. G. C. Stafford, P. E. Kelley and D. C. Bradford, American Laboratory,June, 51(1983).39. J. C. Schwartz, R. G. Cooks, M. Weber-Grabau and P. E. Kelley, 36thASMS Conf. Mass Spectrom. Allied Top. (1988).40. G. C. Stafford, D. M. Taylor, J. E. P. Bradshaw and M. Uhrich, 35thASMS Conf. Mass Spectrom. Allied Top. (1987).41. J. A. B. Daigle, Ph. D. Dissertation, University of British Columbia(1990).42. S. A. McLuckey, G. J. V. Berkel, G. L. Glish, E. C. Huang and J. D.Henion, Anal. Chem., 63, 375 (1991).43. G. J. V. Berkel, G. L. Glish and S. A. McLuckey, Anal. Chem., 62, 1284(1990).44. S. J. Bauer and R. G. Cooks, Talanta, 40, 1031 (1993).45. L. E. Slivon and M. R. Bauer, Anal. Chem., 63, 1335 (1991).46. S. Zhao, K. V. Somayajula, A. G. Sharkey, D. M. Hercules, F.Hillenkamp, M. Karas and A. lgendoh, Anal. Chem., 63, 450 (1991).47. R. C. Beavis and B. T. Chait, Chem. Phys. Lett., 181, 479 (1991).48. M. A. Posthumus, P. G. Kistemaker, H. L. C. Meuzelaar and M. C. TenNoever de Brauw, Anal. Chem., 50, 985 (1978).49. M. Karas, U. Bahr and F. Hillenkamp, mt. J. Mass Spectrom. IonProc., 92, 231 (1989).50. J. N. Louris, J. S. Brodbelt-Lustig, R. E. Kaiser and R. G. Cooks, 36thASMS Conf. on Mass Spectrom. and Allied Top., San Francisco, CA(1988).51. S. A. McLuckey, G. L. Glish, D. C. Duckworth and R. K. Marcus, Anal.Chem., 64, 1606 (1992).21152. C. G. Gill and M. W. Blades, Journal of Analytical Atomic Spectrometry,8, 261 (1993).53. C. G. Gill, B. Daigle and M. W. Blades, Spectrochimica Acta, 46B,1227 (1991).54. M. Weber-Grabau, P. E. Kelley, J. E. P. Syka, S. C. Bradshaw and J. C.Brodbelt, 35th ASMS Conf. Mass Spectrom. and Allied Top., Denver,CC, p 1114(1987).55. P. H. Hemberger, N. S. Nogar, J. D. Williams, A. G. Cooks and J. E. P.Syka, Chem. Phys. Lett., 191, 405 (1992).56. J. D. Williams, R. G. Cooks, J. E. P. Syka, P. H. Hemberger and N. S.Nogar, J. Am. Soc. Mass Spectrom., 4, 792 (1993).57. P. E. Toschek, ed., New Trends in Atomic Physics, Amsterdam: NorthHolland Physics Publishing (1990).58. W. Paul and H. Steinwedel, Z. Naturforsch., 8a, 448 (1953).59. E. Fischer, Z. Phys., 156, 1 (1959).60. V. v. G. Rettinghaus, Z Angew. Phys., 22, 321 (1967).61. J. E. P. Syka and W. J. Fies, 35th Ann. Conf. Amer. Soc. MassSpectrom., Denver, CC, p. 767 (1987).62. M. B. Comisarow and A. G. Marshall, Fourier Transform Ion CyclotronResonance Spectroscopy, Interactions Between Ions and Molecules.NATO Adv. Study Inst. Ser. B (Physics), ed. P. Auloos, Plenum: NewYork (1975).63. M. B. Comisarow and A. G. Marshall, Chem. Phys. Lett., 25, 282(1974).64. H. Lagadec, C. Meis and M. Jardino, mt. J. Mass Spectrom. Ion Proc.,85, 287 (1988).21265. J. Louris, G. C. Stafford, J. Syka and D. Taylor, 40th ASMS Conf. MassSpectrom. and Allied Top., Washington, DC, East Lansing, p. 1003(1992).66. T. H. Maiman, Phys. Rev. Lett., 4, 564 (1960).67. B. B. Laud, Lasers and Non-Linear Optics, John Wiley & Sons:Toronto (1985).68. D. O’Shea, W. R. Callen and W. T. Rhodes, Introduction to Lasers,Addison-Wesley: London (1976).69. N. Omenetto, Analytical Laser Spectroscopy, John Wiley & Sons: NewYork (1979).70. G. M. Hieftje, J. C. Travis and F. E. Lytle, Lasers in Chemical Analysis,Humana Press (1981).71. W. Demtroder, Laser Spectroscopy, Basic Concepts andInstrumentation,, Springer-Verlag: New York (1981).72. T. P. Evans, Applications of Lasers to Chemical Problems, John Wiley &Sons: New York (1982).73. F. Bretch and L. Cross, Applied Spectrosc., 59 (1962).74. L. J. Radziemski and D. A. Cremers, ed., Laser Induced Plasmas andApplications, New York: Marcel Dekker (1989).75. L. Moenke-Blankenburg, Prog. Analyt. Spectrosc., 9, 335 (1986).76. P. S. Adrain and J. Watson, J. Phys. D: Appi. Phys, 17, 1915 (1984).77. N. Fürstenau, F. Hillenkamp and P. Nitsche, tnt. J. Mass Spectrom. IonPhys., 31, 85 (1979).78. N. Fürstenau and F. Hillenkamp, mt. J. Mass Spectrom. Ion Phys., 17,135 (1981).79. P. Hughes, Plasmas and Laser Light, , Adam Hilger: Bristol,MA (1975).80. 0. N. Krokhin, “Generation of high temperature vapor and plasmas by213laser radiation,” Laser Handbook, ed. F. T. Arrechi and E. 0. S. d. Bois.North Holland: Amsterdam, 1972, P. 1371.81. C. DeMichelis, IEEEJ. Quant. Electron., QE-6, 630 (1970).82. J. F. Ready, App!. Phys., 36, 462 (1965).83. J. F. Ready, Effects of High Power Laser Radiation, , Academic: NewYork (1971).84. J. F. Ready, Industrial Applications of Lasers, , Academic: Toronto(1978).85. L. Balazs, R. Gijbels and A. Vertes, Anal. Chem., 63, 314 (1991).86. M. W. Blades, P. Banks, C. Gill, D. Huang, C. L. Blanc and D. Huang,IEEE Trans. on Plasma Sd., 19, 1090 (1991).87. D. W. Koppenaal, Anal. Chem., 62, 1284 (1990).88. R. B. Cody, R. C. Burnier, W. D. P. (Jr.), T. J. Carlin, D. A. McCrery, R. K.Lengel and B. S. Freiser, Int. I Mass Spectrom. Ion Phys., 33, 37(1980).89. B. S. Freiser, Talanta, 32, 697 (1985).90. R. W. Odom, “Laser Microprobe Mass Spectrometry-lon ans NeutralAnalysis,” Lasers and Mass Spectrometry, ed. D. Lubman. OxfordUniversity Press: Oxford, 1990.91. R. G. Strife and J. R. Simms, Anal. Chem., 61, 2316 (1989).92. D. N. Holler, I. Lys, P. J. Cotter and 0. M. Uy, Anal. Chem., 61, 1083(1989).93. B. D. Nourse and R. G. Cooks, Anal. Chim. Acta., 228, 1 (1990).94. P. F. Bonner, mt. J. Mass Spectrom. Ion Phys., 23, 249 (1977).95. G. Lawson, J. F. J. Todd and P. F. Banner, Dynamic Mass Spec., 4, 39(1975).96. G. C. Stafford, P. E. K. Jr. , J. E. P. Syka, W. E. Reynolds and J. F. J.214Todd, mt. J. Mass Spectrom. Ion Proc., 60, 85 (1984).97. T. A. Flaim and P. D. Owenby, J. Vac. Sd. Technol., 8, 661 (1971).98. J. F. O’Hanlon, A Users Guide To Vacuum Technology, 2 ed. , Wileylnterscience: Toronto (1989).99. D. J. Wineland, W. M. Itano and J. C. Bergquist, Opt. Lett., 12, 389(1987).100. N. Bradshaw, E. F. H. Hall and N. E. Sanderson, J. Analyt. AtomicSpectrom., 4, 801 (1989).101. J. N. Louris, R. G. Cooks, J. E. P. Syka, P. E. Kelley, G. C. Stafford andJ. F. J. Todd, Anal. Chem., 58, 1677 (1987).102. G. M. Weyl, “Physics of laser-induced breakdown: an update,” LaserInduced Plasmas and Applications, ed. L. J. Radziemski and D. A.Cremers., Dekker: New York, 1989, p. 1.103. W. D. Kingery, H. K. Bowen and D. R. Uhlmann, Introduction toCeramics, Wiley: New York (1976).104. R. C. Weast, ed., CRC Handbook of Chemistry and Physics, Florida:CRC Press (1979).105. F. W. Billmeyer, ed., Textbook of Polymer Science, New York: Wiley(1971).106. 0. F. Folmer and L. V. Azarraga, 5th mt. Symp. Adv. in Chromatog,Las Vegas, NV, Preston Technical Abstracts, p. 216 (1969).107. I. H. Musselman, D. S. Simons and R. W. Linton, mt. J. Mass Spectrom.Ion Proc., 112, 19 (1992).108. P. W. J. M. Boumans, J. C. Ivaldi and W. Slavin, Spectrochimica Acta,46B, 641 (1991).109. J. Perez and B. R. Weiner, Appl.Surf. Sc!., 62, 281 (1992).110. N. S. Nogar, Personal Communication, November (1993).2151 11. C. H. Becker, L. E. Jusinski and L. Moro, mt. J. Mass Spectrom. IonProc., 95, Ri (1990).112. F. Engelke, J. H. Hahn, W. Henke and R. N. Zare, Anal. Chem., 59, 909(1987).113. D. M. Lubman, Anal. Chem., 59, 31(1987).114. J. V. Johnson, R. E. Pedder and R. A. Yost, Rapid Commun. MassSpectrom., 6, 760 (1992).115. R. E. March, F. A. Londry and R. L. Alfred, Org. Mass Spectrom., 27,1151 (1992).116. P. E. Kelley, G. C. Stafford, J. E. P. Syka, W. E. Reynolds, J. N. Louris, J.W. Amy and J. F. J. Todd, 33rd ASMS Conf. on Mass Spectrom. andAllied Top., San Diego, CA, p 307 (1985).117. P. E. Kelley, G. C. Stafford, J. E. P. Syka, W. E. Reynolds, J. N. Lourisand J. F. J. Todd, 10th mt. Conf. on Adv. in Mass Spectrom., Swansea,Wales, p. 869 (1985).118. P. H. Dawson, J. Hedman and N. P. Whetten, Rev. Sd. Instrum., 40,1444 (1969).119. J. E. Fulford and R. E. March, mt. J. Mass Spectrom. Ion Phys., 26, 155(1978).120. P. E. Kelley, J. E. P. Syka, P. E. Ceja, G. C. Stafford, J. N. Louris, H.F. Grutzmacher, D. Kuck and J. F. J. Todd, 34th Ann. Conf. Amer. Soc.Mass Spectrom., Cincinnati, OH, p. 963 (1986).121. 5. M. Michael, M. Chein and D. M. Lubman, Rev. Sd. Instrum., 63,4277 (1992).122. W. Newhauser, M. Hohenstatt, P. E. Toschek and H. Dehmelt, Phys.Rev. A, 22, (1980).123. D. H. Smith, J. P. Young and P. W. Shaw, Mass Spectrom. Rev., 8, 345216(1989).124. N. S. Nogar and R. C. Estler, “Laser desorption I laser ablation withdetection by resonance ionization mass spectrometry,” Lasers andMass Spectrometry, ed. D. M. Lubman. Oxford University Press:Oxford, 1990, P. 65.125. S. Mayo, T. B. Lucatorto and G. G. Luther, AnaL Chem., 54, 553 (1982).126. N. Nogar and R. Estler, Anal. Chem., 64, 465 (1992).127. J. 0. Stoner and S. Bashkin, Atomic Energy-Level and GrotrianDiagrams, Amsterdam: North Holland Publishing (1975).128. C. H. Corliss and W. R. Bozman, Experimental Transition Probabilitiesfor Spectral Lines of Seventy Elements, , National Bureau ofStandards: Washington D.C. (1962).129. C. W. LeBlanc, M. Sc. Dissertation, University of British Columbia(1989).130. F. W. Sears, Optics, 3 ed. ,Addison Wesley: Don Mills, Ontario (1949).131. D. B. Langmuir, P. V. Langmuir, H. Shelton and Wuerker, “R. F.Containment device,” U.S. Patent 3,065,640 (1962)132. R. F. Bonner, J. E. Fulford, R. E. March and G. F. Hamilton, mt. J. MassSpectrom. Ion Phys., 24, 255 (1977).133. A. G. Nassiopoulos, P. A. Moller and A. Septier, Rev. Phys. App!., 15,1529 (1980).134. R. E. Mather, R. M. Waldren, J. F. J. Todd and R. E. March, mt. J. MassSpectrom. Ion Phys., 33, 201 (1980).135. D. M. Lubman, ed., Lasers and Mass Spectrometry, New York: OxfordUniversity Press (1990).136. G. C. Eiden, A. W. Garrett, M. E. Cisper, N. S. Nogar and P. Hemberger,mt. J. Mass Spectrom. Ion Proc., submitted December (1993).137. V. M. Doroshenko, T. J. Cornish and R. J. Cotter, Rapid Commun. MassSpectrom., 6, 753 (1992).217218APPENDIX IThese programs were written by the author for control of the ion trapexperiment using Borland Turbo Pascal Version 3.0, in collaboration with Dr. B.Daigle (formerly U. B. C.) and Dr. Michael. W. Blades (U. B. C.). The followingsubroutines were used in conjunction with a main program initially developed toacquire and manipulate data from a 4096 pixel linear photo diode array. Themain program was written by Dr. B. Browne; further information about the codefor the main program can be obtained upon request. Address inquiries to Dr. M.W. Blades, Chemistry Department, University of British Columbia.The program creates 4096 point data files which are ASCII format.Further processing of these files was done using Apple MacintoshTMcomputers, with Version 1.11 IgorTM software. The following subroutines(called procedures in Turbo Pascal 3.0) are copies of those used for the ion trapexperiments presented by this thesis.PROCEDURE ScanSpectrum (VAR darray : diodearraytype; scansinteger);PROCEDURE SetupISCl6; {initialize the ISC16 data acquisit. board}VAR i : integer;BEGINPORT[$030B] := 0; {disable board for initialization}FOR i:= 1 to 16 DOPORT[$0310J 0; {load muxRAM to sample ch. 1 only}PORT[$0311] := 0; {external clock input}PORT[$0307J := $74;PORT[$0305] := LO(CLOCKTIME);PORT[$0305] := HI(CLOCKTIME);219PORT{$0308] := 11; {external, + slope trigger}PORT[$0307] := 50; {set burst to sample 1 channel/clock pulse}PORT[$0304] := 1;PORT[$0304] := 0;PORT[$0307] := 178; {set post trigger delay for 4096 samples}PORT[$0306] := 0;PORT[$0306] := 16;END; {SetupISCl6}PROCEDURE GetAverageData (VAR specarray : diodearraytype,numscans : integer);TYPE byteArrayType ARRAY[1 ..MaxArraySize] OF INTEGER;VARlobyte, hibyte byteArrayType;i,j,A,B,C,D,E,S,V : iNTEGER;hibytefactor, lobytefactor : REAL;ch :CHAR;BEGINPORT[$21B] := $80; {set up triggers on DM12-10}A := $218;B := $219;C $21A;FOR i 1 TO 4096 DO {clear data buffer}BEGINlobyte[i] := 0;hibyte[i] := 0;ENlGoToXY(20,12);ch :=WRITE(’Scanning Spectrum’);FOR i := 1 TO numscans DO {enter the scan loop}BEGINIF Keypressed THENBEGIN220READ(Kbd,ch);IF Keypressed AND (ch := #27) THENREAD(Kbd ,ch);END;IF NOT (ch := #27) THENBEGINGoToXY(3 9,12);CirEol;GoToXY(3 9,12);WRITE(i);PORT[$0315] 0; {bank B man., save data bnk. A}PORT[$031B] := 04; {reset bnk. A I set 4K buffer}PORT[$0308] : $OB; {+slope external trigger}PORT[$030B] := 0;PORT[$030D] 0; {enable trigger logic/data acq.}{wait for trigger status flag indicating external trigger then}{begin the ion trap scan sequence}V := 4095- D; {set storage potential}OUTDAC12_S (0,V);DELAY(5); {delay to allow storage potential to stabilize}PORT[A] := $0; {1 msec electron gun pulse)PORT[A] := $FF;DELAY( 1);PORT[A] := $0;PORT[C] $0; {1 msec ablation laser trigger pulse)PORT{C] := $FF;DELAY( 1);PORT[C] $0;PORT[B] $0; {set data acquisition trigger on DM12-lO to zero}DELAY(STORETIME- 3); {store time for ions in trap}PORT[B] := $FF; {trigger ISC16 to begin data acquisition}DELAY( 1);PORT[B] := $0;DELAY( 1);FOR J := 1 TO FINVOLT DO {ramp the stor. pot. to eject ions for detect.}221BEGINV := 4095 - (INITVOLT + J);OUTDAC12_S (0,V);ENLFOR I := 10 TO 40 DO {quick ramp to max. pot. to eject remaining ions}BEGINV := I * 100;OUTDAC12_S (0,V);ENIV := 4095; {set storage pot. to zero to ensure no ions are stored}OUTDAC12_S (0,V);DELAY(50);PORT[$0314] 0; {set bank A manual, and read the data buffers}FOR j := 1 TO 4096 DOBEGINlobyte[j] := mem[$D000:0j + lobyte[j];hibyte[j] : mem[$D000:0] + hibyte[j];ENEEND {if ESC (ch := #27) is pressed}END, {the scan loop)hibytefactor 256/numscans; {convert the buffer to a spectrum}lobytefactor := 1/numscans;FORk := 1 TO 4096 DOBEGINspecarray [k] : = hibyte [k] * hibytefactor + lobyte [k] *lobytefactoi.;END;END; { GetAverageData }BEGIN { Scanspectrum, called by main program for the ion trap expt. }setupISCl6; {call subroutine to initialize data acq. board)GetAverageData(darray, SCANS); {call sub. to run IT expt. }END; {Scanspectrum}222APPENDIX IIMacSimionTM Version 1.0 is distributed by Montech Pty. Ltd., MonashUniversity, Melbourne, Australia. The program was developed by DonaldMcGilvery and Richard Morrison to be used as an aid for understanding ion andelectron optical devices. The program is a simulation, and the users ofMacSimion are cautioned not to place undue faith in exact numerical resultsfrom the simulations it provides. The program has three sections: A specializeddrafting portion for electrode construction, a refine section that calculates thefield potentials created by the electrodes using an iterative application ofLaplace’s equation, and finally trajectories of charged particles are calculatedbased upon the field conditions near the particle. Further information regardingthe program and it’s application can be obtained by contacting Montech Pty. Ltdat Email address CHE2O1N @ VAXC.CC.MONASH.ED.AU.

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059528/manifest

Comment

Related Items