Diode Laser Diagnostics of Laser-Induced Plasmas and Atomic Vapor Cells

Material Information

Diode Laser Diagnostics of Laser-Induced Plasmas and Atomic Vapor Cells
Lauly, Benoit
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (137 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Committee Chair:
Winefordner, James D.
Committee Members:
Vala, Martin T.
Powell, David H.
Omenetto, Nicolo
Reitze, David H.
Graduation Date:


Subjects / Keywords:
Chemistry -- Dissertations, Academic -- UF
absorption, atomic, laser, libs, lip
Lasers ( jstor )
Plasmas ( jstor )
Cesium ( jstor )
Electronic Thesis or Dissertation
born-digital ( sobekcm )
Chemistry thesis, Ph.D.


DIODE LASER DIAGNOSTICS OF LASER-INDUCED PLASMAS AND ATOMIC VAPOR CELLS The main goal of the research is focused on the exploitation of diode lasers for several applications involving photon detection, high resolution spectroscopy and imaging of selected species in laser induced plasmas. Laser Induced Breakdown Spectroscopy (LIBS) is commanding much attention as an atomic emission spectroscopy technique due to its multiple attractive features. Much effort in the LIBS community has been, and still is directed toward the understanding of plasma fundamentals. Understandably, much information remains to be gathered in order to fully comprehend the laser-sample interaction. Of all the diagnostic techniques applied to plasmas and extensively described in the literature, absorption spectroscopy seems to be receiving comparatively less attention. In this work, we describe the use of selective absorption methods to follow the evolution of the plasma in time, and as a consequence, to better understand the temporal and spatial evolution of the different populations involved. The temporal behavior of a specific transition can be followed by measurements with a Photomultiplier Tube (PMT) and line shapes can be evaluated by scanning the diode laser. In spectrochemical analysis, line shapes plays a major role in the understanding of spectral interferences, plasma conditions and behavior of analytical applications. By spatially expanding the laser probe beam, the temporal and spatial evolution can be followed with a gated Intensified Charge-Coupled Device (ICCD), consequently assessing the studied species homogeneity within the plasma plume. Cesium atomic vapor filters or detectors have been a primary focus of this work as they demonstrate the potential to excel both in terms of spectral resolution and sensitivity. Atomic vapor detectors have a spectral resolution that is governed by the properties of the atomic vapor used as the sensing element, while maintaining the same value of the luminosity. Cesium vapor cells have been extensively investigated because of cesium s high number density at low temperature and its strong resonance transition in the near-infrared at 852nm (62S1/2 to 62P3/2). A promising fluorescence scheme for cesium has been demonstrated here that includes a single transition at 852nm and fluorescence detection at 894nm (62P1/2 to 62S1/2). For efficient detection, a rapid fine-structure mixing (62P3/2 , 62P1/2) is required and is provided by the presence of ethane in the cell. The absorption properties of this cell are reported as well as its potential application to a selected analytical problem such as the detection of Raman photons. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2008.
Adviser: Winefordner, James D.
Electronic Access:
Statement of Responsibility:
by Benoit Lauly.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Lauly, Benoit. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2008 ( lcc )


This item has the following downloads:

Full Text




2008 Benot Lauly 2


A ma famille et mes amis, 3


ACKNOWLEDGMENTS I would like first to thank my graduate research advisor, Ji m Winefordner, for his support and friendship during the past f our years. Jim provides an environment that supports research freedom, innovation, independent thinking, teamwork, and personal development. I will always remember Jims philosophies on research, teachi ng, and leadership and aspire to incorporate these ideologies into my personality. Nicolo Omenetto has also played a large role in my scientific and personal development. His insight and knowledge ex tend to a variety of fields, and his contributions to my research are numerous. His enthusiasm and excitement toward science are unequaled, and even the most rigorous of scientific discussions were always enjoyable. I would also like to thank Ben Smith for his friendship and contributi ons to my research. Ben has given me practical advice on numerous occasions to approach a given problem or develop an idea. Most importan tly, Bens appreciation of scien ce, history, culture and life has helped me realize that the world is full of ex citing and rewarding experiences and opportunities. I would also like to thank Nic holas Taylor for his friendshi p and for the long nights spent together in the dark laboratory. Jonathan Mert en has also made many contributions to this research, and I would like to thank him for the important role he played in my life outside chemistry. Jamshid Temirov and Igor Gornushki n have provided me with advice on numerous occasions, and I am thankful to have had such amazing postdocs in our laboratory. I would also like to thank the supporting staff, especially J eanne Karably for efficiently handling a variety of needs from purchase orders to travel reimbursements. My appreciation also goes to the chemistry departments electronic and machine shops for good advice and top-notch work. 4


I would also like to thank the past and present members of the Wineforner and Omenetto groups for their friendship and support. Perhaps most of all, I would like to th ank all my friends and family for their encouragement, love and support. 5


TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................11 CHAPTER 1 INTENT AND SCOPE OF SDUDY......................................................................................13 2 INTRODUCTION TO LASER-INDUC ED PLASMA SPECTROSCOPY..........................19 History and Fundamentals......................................................................................................1 9 Laser Properties Used in LIP..................................................................................................20 Laser Wavelength............................................................................................................21 Laser Pulse Width and Energy........................................................................................21 LIBS Setup..............................................................................................................................22 Applications of LIBS..............................................................................................................23 Conclusions.............................................................................................................................25 3 RESONANT ABSORPTION IMAGING IN LASER-INDUCED PLASMAS....................29 Introduction................................................................................................................... ..........29 Experimental................................................................................................................... ........33 Experimental Setup.........................................................................................................33 Sample Preparation..........................................................................................................35 Data Acquisition..............................................................................................................3 6 Results and Discussion......................................................................................................... ..38 Time-Resolved Atmospheric Plasma Evolution..............................................................38 Effect of the Pressure on the Plasma Evolution..............................................................40 Effect of the Laser Focusing Distance.............................................................................41 Shot-to-Shot Reproducibility..........................................................................................42 Conclusions and Remarks.......................................................................................................4 3 4 HIGH RESOLUTION TIME-RESOLVED ABSORPTION SPECTROSCOPY IN LASER-INDUCED PLASMA...............................................................................................59 Introduction................................................................................................................... ..........59 Theory Considerations.....................................................................................................61 A Review of Broadening Mechanisms............................................................................63 Experimental................................................................................................................... ........65 Frequency Calibration by a Confocal Fabry-Perot Interferometer..................................65 6


Experimental Setup.........................................................................................................68 Results and Discussion......................................................................................................... ..70 TOF Measurements for a Pressure of 0.01mbar..............................................................70 TOF Measurements for Pressures of 0.5 mbar, 5 mbar, 1 atm........................................74 Conclusions.............................................................................................................................76 5 INTRODUCTION TO ATOMIC VAPOR DETECTION.....................................................89 6 ATOMIC VAPOR DETECTORS: TO WARD THE DETECTION OF RAMAN SCATTERING PHOTONS....................................................................................................97 Introduction................................................................................................................... ..........97 Improvement of the Fl uorescence Efficiency, F.................................................................101 Investigation of Ethane-Induced Collisional Energy Mixing in Cesium.......................102 Fluorescence Intensity...................................................................................................103 Rate equation Approach................................................................................................104 Experimental Setup.......................................................................................................107 Results........................................................................................................................ ...108 Improvement of the Absorption Efficiency..........................................................................108 Experimental..................................................................................................................1 09 Results........................................................................................................................ ...110 Overall Performance of an AVD for Raman Detection........................................................111 Conclusions and Remarks.....................................................................................................112 7 FUTURE WORK.................................................................................................................. 120 LIST OF REFERENCES......................................................................................................124 BIOGRAPHICAL SKETCH.......................................................................................................137 7


LIST OF TABLES Table page 6-1 Values for several detectors of the lu minosity, Resolving power and their product.......119 6-2 Fluorescence ratios at different temperature....................................................................119 6-3 Calculated mixing rates a nd cross sections of Cs 62P3/2 62P1/2 induced by 100mTorr of ethane..........................................................................................................119 8


LIST OF FIGURES Figure page 1-1 Elements that can be determined by applying the fundamental or second harmonic wavelengths of laser diodes...............................................................................................17 1-2 External cavity diode lasers............................................................................................... 18 2-1 Different plasma featur es within the shockwave...............................................................26 2-2 Important time periods after plasma form ation during which emi ssions from different species predominate...........................................................................................................2 7 2-3 General laboratory bench-top setup for Laser Induced Breakdown Spectroscopy............28 3-1 Experimental setup used for shadowgra phic imaging of plasma evolution and the resulting images............................................................................................................... ..45 3-2 Experimental setup for time-resolved la ser-induced plasma absorption images at different pressure with a high resolution diode laser.........................................................46 3-3 Schematic diagram of gated 2-D Inte nsified Charge-Coupled Device (ICCD)................47 3-4 Microscope images of the laser craters..............................................................................48 3-7 Temporal evolution of the cesium ground state population. Images taken for single shots on 10 ppm CSCl cellulose pellets.............................................................................51 3-8 Time propagation of the shockwave crea ted from a LIP. Samples of 10 ppm CsCl cellulose sample were used................................................................................................52 3-10 Cesium ground state absorption images for several pressures...........................................54 3-12 Effect of the focusing distant on the cesium ground state distribution for several delay times.........................................................................................................................56 3-13 Successive LIP on the same ta rget spot for two matrices..................................................57 3-14 Transmittance values for successive laser shots on the same target spot..........................58 4-1 Partial energy level diagram of cesium..............................................................................78 4-3 Schematic of the experimental setup for time-resolved high-resolution absorption of the cesium ground state line at 852nm...............................................................................79 4-4 Records of the absorption profile of the cesium reference cell and the Q-switched output trigger from the pulsed-laser...................................................................................80 9


4-5 Contour plot of time-resolved spectra of the cesium absorption at a pressure of 0.01 mbar...................................................................................................................................81 4-6 Typical TOF profile recorded by absorption of Cs atoms in vacuum and the shifted Maxwellian fitting............................................................................................................. .82 4-7 Calculated contributions of the spectral line width, and shifts, in GHz, for the Cs 852nm line, as a function of plasma temperature..............................................................83 4-8 Absorption spectra and time-resolved ground state number density.................................84 4-9 Contour plot of time-resolved spectra of the cesium absorption at a pressure of 0.5 mbar and 5 mbar................................................................................................................85 4-10 Contour plot of time-resolv ed spectra of the cesium absorp tion at a pressure of 1 atm....86 4-11 Effect of the distance from the probe beam to the sample surface on absorption spectra at 0.5mbar and 5mbar............................................................................................87 4-12 Effect of the distance from the probe beam to the sample surface on absorption spectra at 1 atm............................................................................................................... ...88 5-1 Schematic of a resonance fluorescence dete ctor and resonance i onization detector.........94 5-2 The principle of operation of an atomic vapor photon detector, the atomic vapor absorbs at fixed frequency.................................................................................................95 5-3 Applications of a cesium RFD...........................................................................................96 6-1 Schematic of the different rates involved in the 62S1/262P3/2 detection process...........113 6-2 Experimental setup to measure the fluor escence response of an atomic vapor cell........114 6-3 Fluorescence spectra of a cesium-ethane absorption cell................................................115 6-4 Effect of the laser power on the fluorescence intensity...................................................115 6-5 Calculated Raman linewidth and corresponding signal absorbed...................................116 6-6 Experimental setup for measuring the sp ectral absorption linesha pe of the RFD and evaluate its imaging performance....................................................................................117 6-7 Spectral profile and image of the Cs/ethane RFD at room temperature..........................118 10


Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DIODE LASER DIAGNOSTICS OF LASERINDUCED PLASMAS AND ATOMIC VAPOR CELLS Benot Lauly August 2008 Chair: James D. Winefordner Major: Chemistry The main goal of the research is focused on the exploitation of diode lasers for several applications involving photon detection, high resolution spectroscopy and imaging of selected species in laser induced plasmas. Laser Induced Breakdown Spectroscopy (LIBS) is commanding much attention as an atomic emi ssion spectroscopy technique due to its multiple attractive features. Much effort in the LIBS comm unity has been, and still is directed toward the understanding of plasma fundame ntals. Understandably, much information remains to be gathered in order to fully co mprehend the laser-sample intera ction. Of all the diagnostic techniques applied to plasmas and extensively described in the literature, absorption spectroscopy seems to be receiving comparatively less attenti on. In this work, we describe the use of selective absorption methods to follow the evolution of the plasma in time, and as a consequence, to better understand the temporal and spatial evolu tion of the different populations involved. The temporal behavior of a specific tr ansition can be followed by measurements with a Photomultiplier Tube (PMT) and line shapes can be evaluated by scanning the diode laser. In spectrochemical analysis, line sh apes plays a major role in the understanding of spectral interferences, plasma conditions and behavior of analytical applications. By spatially expanding the laser probe beam, the temporal and spatial ev olution can be followed with a gated Intensified 11


Charge-Coupled Device (ICCD), consequently assessing the studied species homogeneity within the plasma plume. Cesium atomic vapor filters or detectors have been a primary focus of this work as they demonstrate the potential to excel both in terms of spectral resolution and sensitivity. Atomic vapor detectors have a spectral resolution that is governed by the properties of the atomic vapor used as the sensing element, while maintaining the same value of the luminosity. Cesium vapor cells have been extensively investigated be cause of cesiums high nu mber density at low temperature and its strong resonance transition in th e near-infrared at 852nm (62S1/2 62P3/2). A promising fluorescence scheme for cesium has been demonstrated here that includes a single transition at 852nm and fluorescence detection at 894nm (62P1/2 62S1/2). For efficient detection, a rapid fine-structure mixing (62P3/2 62P1/2) is required and is provided by the presence of ethane in the cell. The absorption pr operties of this cell are reported as well as its potential application to a select ed analytical problem such as the detection of Raman photons. 12


CHAPTER 1 INTENT AND SCOPE OF STUDY The main goal of the research is focused on the exploitation of diode lasers for several applications involving photon detection, high resolution spectroscopy and imaging of selected species in laser induced plasmas. Laser Induced Breakdown Spectroscopy (LIBS) is commanding much attention as an atomic emi ssion spectroscopy technique due to its multiple attractive features. Much effort in the LIBS comm unity has been, and still is directed toward the understanding of plasma fundame ntals. Understandably, much information remains to be gathered in order to fully co mprehend the laser-sample intera ction. Of all the diagnostic techniques applied to plasmas and extensively described in the literature, absorption spectroscopy seems to be receiving comparatively less attenti on. In this work, we describe the use of selective absorption methods to follow the evolution of the plasma in time, and as a consequence, to better understand the temporal and spatial evolu tion of the different populations involved. The temporal behavior of a specific tr ansition can be followed by measurements with a Photomultiplier Tube (PMT) and line shapes can be evaluated by scanning the diode laser. In spectrochemical analysis, line sh apes plays a major role in the understanding of spectral interferences, plasma conditions and behavior of analytical applications. By spatially expanding the laser probe beam, the temporal and spatial ev olution can be followed with a gated Intensified Charge-Coupled Device (ICCD), consequently assessing the studied species homogeneity within the plasma plume. Cesium atomic vapor filters or detectors have been a primary focus of this work as they demonstrate the potential to excel both in terms of spectral resolution and sensitivity. Atomic vapor detectors have a spectral resolution that is governed by the properties of the atomic vapor used as the sensing element, while maintaining the same value of the luminosity. Cesium vapor 13


cells have been extensively investigated be cause of cesiums high nu mber density at low temperature and its strong resonance transition in th e near-infrared at 852nm (62S1/2 62P3/2). A promising fluorescence scheme for cesium has been demonstrated here that includes a single transition at 852nm and fluorescence detection at 894nm (62P1/2 62S1/2). For efficient detection, a rapid fine-structure mixing (62P3/2 62P1/2) is required and is provided by the presence of ethane in the cell. The absorption pr operties of this cell are reported as well as its potential application to a selected analyti cal problem such as the de tection of Raman photons. An example could be the Surface-Enhanced Raman sp ectroscopy detection of dipicolinic acid in bacterial spores. Diode Laser Spectroscopy: Semiconductor lasers or diode lasers have proven to be valuable tools in atomic spectroscopy. The active medium is a pn semiconductor material, doped by a huge concentration of donors and acceptors, thus the Fermi level is in the valence band of the p region and in the conduction band for the n region. The population inversion, for laser action, is achieved by applying a voltage difference at the direct polarized junction. As in other lasers, the gain region is surrounded with an optical cavity to form a laser. The two ends of the crystal are cleaved to form perfectly smooth, pa rallel edges, forming a Fabry-Perot resonator. Because of their electrical efficiency, low cost diode lasers provide the spectroscopic advantages of laser diagnostic in a compact, transportable and easy-to-use system. Compared to gas lasers, no external pumping is needed whic h significantly reduces the size and cost. A diode laser head can be smaller than a one milli meter cube; the laser reaches few centimeters dimension with an external cav ity configuration. The temperature controller and power supply unit defines, the actual size of the entire laser device. 14


Such sources, in certain conditions, allo w for single mode opera tion and a continuous single mode tuning range. Diode lasers are tunable source mostly for the infrared and red region of the spectrum, different wavelength regi ons are achievable by changing the semiconductor crystal. With the continuous de velopment of laser technology, th e emitted wavelength of diode lasers is now available in the blue region of the spectra. However, they are currently quite expensive compared to other diode lasers. Figure 1-1 shows all th e elements as of 2005 that can be studied by applying the fundame ntal and second harmonic wavele ngths of diode laser. Thirty five elements are directly accessible, nineteen by second harmonic generation. The wavelength emitted by a single diode laser can be slightly tuned by changing the temperature and the supply current of the semiconduc tor. If the diode lase r is working in multi mode, an External Cavity configuration can be us ed to force the laser to oscillate in a single longitudinal mode. Diode lasers with linewidth in the order of MHz are commonly achieved. In the Littrow configuration, the light emitted by a diode laser is dire cted to a tunable grating as displayed on Figure 1-2. The 0th diffraction order is coupled out as laser beam and the1st order diffraction is coupled back into the laser diode chip. If the feedback of the gratings is much larger than the feedback of the frontal edge of the laser, an optical cavity is formed between the back edge of the laser and the tunable grating. In this case the wavelength varies also by modulating the voltage applied to the piezoelectr ic element which rotate s the grating of the external cavity. Other configurations exist co mmercially and a Littman/Metcalf configuration can be found on Figure 1-2. Exte rnal cavity diode lasers features like narrow linewidth (<5 MHz), high passive stability (long term drift, typically of 300 MHz), fine and accurate tuning ability, are required for high resolution spectroscopy and have been commonly used for subDoppler spectroscopy. In atomic spectroscopy, th e diode laser bandwidth s are often narrower 15


than line profiles, specifically in laser-in duced plasmas and therefore peak absorption measurements are possible. Because diode lasers are used in continuous wave mode, transient events in the plasma such as the absorpti on of a specific populati on can be continuously observed. Time-dependent detectors, photomultiplier tubes (PMT), photodiodes or Intensified Couple Charged Devices (ICCD) re cord the temporal evolution. Several restrictions apply when using diode lasers for spectroscopic measurements. The coarse tuning range for those diode lasers are ty pically from 10 nm to 100 nm but a fine tuning, modehop free, is only possible over a maximum of 100 GHz (0.24 nm at 850 nm) and more often over a range of about 20 GHz. Therefore, those di ode lasers have to be used for scanning over profiles with full width half maxima narrower than approximately 20 pm at 850 nm. The output power is often in the order of tens of miliwa tts, even if higher powers are achievable, which limits the type of application. In atomic abso rption spectroscopy, where low fluences are usually used to avoid saturating transitions, the power of those lasers is sufficient. External measurements with high spectral resolution are required to accurately monitor the output wavelength of diode lasers wh ich can significantly compli cate ones setup. Fabry-Perot interferometric techniques are of ten used for wavelength calibration. 16


Figure 1-1. Elements that can be determined by applying the fundamental or second harmonic wavelengths of laser diodes. Elements in purple boxes can be measured from the ground state or low-excited state, elements in yellow must be measured from highly excited metastable levels in a plasma source. Appropriate diodes are not commercially available for the elements in white boxes.1 17


a) b) Littrow diode grating d) c) Littman-Metcalf diode grating tuning mirror Figure 1-2. External cavity diode lasers. Diagram, a), and picture, b) of an external cavity diode laser in Littrow configurati on. Schematic, c), and picture of an external cavity diode laser in Littman/Metcalf configuration.2 18


CHAPTER 2 INTRODUCTION TO LASER INDUCED PLASMA SPECTROSCOPY History and Fundamentals Since the first report of laser operation in a ruby crystal in 1960, researchers have not ceased to ponder the mysteries of laser-light and matter interaction. With the development of Qswitched lasers, Brech and Cross, in 1962,3 demonstrated the capabili ty to produce high focused laser fluences sufficient to initiate a breakdown a nd to produce an analytical plasma from solid materials In 1963, Debras-Guedon and Liodec repo rted the analytical use of laser-induced plasmas for the spectrochemical analysis of surfaces,4 and in 1964, Maker, Tehrune, and Savage reported the first optical breakdown in a gas.5 Over the years, this technique has been extensively reviewed in the literature.6-9 Several names and acronyms have been applied to Laser-Induced Plasma Spectroscopy (LIPS). These include Laser-Induced Breakdown spectroscopy (LIBS), Laser Ablation Optical Emission Spectroscopy (LA-OES), Laser Spark Spectroscopy (LASS). Laser-Induced Plasmas (LIPs) can be defined by the generation of a vaporizing a nd exciting plasma from a high energy density focused laser pulse. A glow or flash is gene rally observed in the focal region. One major characteristic of a LIP is its transient naturethe plasma evolves extensively both in space and time. Much effort has been made and has to be made in understanding the fundamental mechanisms governing LIP. The laser-material inte raction leading to the breakdown is composed of two steps.10 The first involves absorption of the radi ation, creating free el ectrons through three body collisions of electrons, photons and neutrals. The second step is avalanche ionization. The high velocity of the free electrons induces an in crease in the thermal energy density that causes the underlying materials to reach critical temp eratures and pressures and then explode. The materials from the explosion contain a high numbe r density of electrons, ionized species from 19


collisions, and atoms that expand at superson ic speeds. A shockwave with the surrounding atmosphere is formed and is responsible for the loud sound usually associated with LIBS. A general feature of the interac tion between the vapor plasma a nd the ambient gas is shown on Figure 2-1. After some time, several microseco nds, the plasma plume slows down and cools. Due to radiative processes, que nching and recombination, a dense neutral region forms. A temporal evolution with the dominant emitting species during the plasma evolution is clearly summarized on Figure 2-2.6 Several factors have significant influence on the plasma characteristics. The behavior of the plasma is mainly dictated by the ablation lase r, which supplies the required fluence to reach a breakdown. The type of laser, wavelength, pulse characteristics an d beam quality, have different effects on plasma formation as does the focusing el ement and its position relative to the target. Laser Properties Used in LIP Almost all available types of pulsed-laser have been used as an excitation source for LIBS, from gas lasers to solid state lasers. The most popular lasers to create the breakdown are solidstate lasers and especially, la sers with an amplification me dium doped with neodymium (Nd): Nd-YAG (yttrium-aluminum-garnet), Nd-YLF (yttrium-lithium-fluoride), Nd-YVO4 (yttriumorthvanadate) for instance. Nowadays, solid state Nd-YAG lasers (1064 nm) are the most commonly used in LIP studies and applications, and are generally equipped with an active QSwitched and a flashlamp pump. Despite their highe r cost, the number of diode-pumped lasers is continuously increasing due to their compactne ss and better shot-to-shot reproducibility. The fundamental wavelength can eas ily be converted to shorter wavelengths (532 nm, 256 nm and 355 nm) with harmonic generation crystals. 20


Other noteworthy types of laser are the CO2 laser (10.6m) 11 and excimer lasers (XeCl, 308 nm, KrF, 248 nm, ArF, 193 nm).12-14 Several publications can be found using these lasers; however, gas lasers require high maintenan ce and are generally bulky and expensive. Laser Wavelength The laser wavelength-matter in teraction has a direct influence on the plasma generation, 15 it has been used as an advantage in terms of in creased energy coupling with a particular sample. UV laser-induced breakdown leads, generally, to higher ablation efficienc y, lower fractionation and background emission, but requires a higher flue nce threshold and is a ssociated with greater matrix effects. Several papers directly compared the effect of laser wavelengths on specific samples. The ArF laser and KrF laser were comp ared while removing coatings from historical objects. 16 Within the same conditions, it was conclude d that the ArF gives better results due to a specific absorption from the coating at the lase r wavelength 248 nm. The different harmonics of a Nd-YAG laser were studied by Pinni ck et al. on different aerosols 17 and Costela et al. on spray paints.18 A higher spatial resolution can also be reached by UV lasers, an advantage that can be exploited for surface mapping in microanalysis.19 Laser Pulse Width and Energy One can easily understand that the pulse energy, as well as the fluence, affects directly the plasma evolution.20, 21 Pulses with high energy result in larger and more energetic plasmas. Temperature, ionization, mass ablated from the surface and emission intensity; parameters conventionally used to describe a plasma are all increasing with the pulse energy. The laser pulse width also influences the plas ma formation and its properties. Conventional LIBS has mainly focused on several nanoseconds (ns) to tens of nanoseconds pulse widths, though a more recent interest in picosecond (ps) and femtosecond (fs) has grown. The ablation process and the morphology of the craters have been studied for those regimes.22, 23 On the 21


femtosecond time scale, the mechanisms leading to plasma formation are dominated by multiphoton ionization over thermal proce sses. This results, compared to ns scale lasers, in higher ablation efficiencies. The material is removed without any thermal effect or melting beyond the ablated region. The temperature of the plasma also decays at a faster ra te. The result is the decrease of background emission from the pl asma and more specifically the continuum emission.24 A shorter plasma lifetime and a better resolution of spectral lines were also observed.25 Despite some appealing advantages, the cost, complexity and power demand of femtosecond lasers are high and limit its use to a small scientific comm unity. An alternative approach to improve analytical fi gures of merit that have received significant interest is the use of dualor multiple laser pulses. In a double-puls e experiment, a second la ser is applied after a certain delay from the first one. Several delay times and geometry configurations (orthogonal or parallel, for instance) have been reported.26-28 Although the mechanisms behind the improved figures of merits are not completely clear, th ere is an agreement in the possible sources of enhancements, higher ablated mass, re-excitation of the material ablated in the first pulse and energetic and physical effect s from the first pulse. 29, 30 LIBS Setup The common LIBS setup records the emission in tensity of atomic a nd ionic lines versus the wavelength during the entire plasma lifetime. Figure 2-3 shows the di agram of a generalized LIBS instrument. It includes a pulsed laser, a me thod of spectrally select ing one or more narrow regions of the emission lines and a detector. Seve ral factors are taken into consideration when choosing the components for the apparatus: the elements to be monitored (wavelength, number of elements), the nature of the sample and the ty pe of analysis (quantitat ive or qualitative). Those factors will influence the colle ction elements (lenses or fibe r optics), the nature of the 22


spectrometer (Echelle spectrometer,31 Czerny-Turner type monochr omator ) and the detector (Photomultiplier tube, CCD, gated-CCD, ICCD, phot odiode). For instance, the price for high spectral resolution is a higher cost and narrow spectral window. Depending on the number of lines desired, the window should be carefully chosen. Niemax et al. 32 reported the use of an Echelle spectrometer coupled w ith an ICCD. They improved th e precision and detection limits by observing multiple lines from single laser shots. Time-resolved LIBS is used to determine the time-evolution of line intensities emitted from different species. Baudelet et al. 33 presented the characteristics of the spectra obtained from orga nic species and particularly molecular species such as CN, which exists at longer delay time an d result from recombination. Gated detectors are commonly used to eliminate the conti nuum background emission at early times. Other parameters or conditions th at greatly affect the plasma evolution have been studied, including temperature, pressure, buffer gas and the matrix composition of the sample. Applications of LIBS Hundreds of potential and act ual applications for laser i nduced plasmas can be found in the literature. Presenting an extensive overview is not the scope of this work. However, several applications which take advantage of the simplicity, remote capabilities and low-destructive properties of LIBS are worth noting. Laser ablation is an attractiv e tool for alloy and metallurg ic industries because samples can be analyzed without pretreatment. On the indus trial scale, this saving of cost and time is of primary importance and several of efforts have been resulted in the implementation of LIBS. LIBS has been used to monitor industrial proce sses, especially in the steel industry. Brass, gold and aluminum alloys have also been studied. Some examples i nvolve microanalysis of surfaces such as the detection of de fects and elemental analysis, 34, 35 the analysis of molten alloys with high temperature resistance probes 36 and the analysis of scrap meta ls in industria l applications 23


for sample quality processing.37 Other industrial applications of laser ablation include welding 38, cutting and micromachining.39 Another interest in using LIBS is attributed to its potential in-situ field analysis, notably in environmental analyses. The reports of portable LIBS devices have increased over the years with the constant miniaturization of lasers, spectrometers and computers.40 Research has been done in and outside the laboratory on organic ma terials like plants, wood,41 soil.42 Capitelli et al. 43 evaluated the performance of quantitative and qualitative analyses of he avy metals in soils. The performance of using LIBS was compared w ith Inductively Coupled Plasma (ICP) emission analysis. As generally found in the literature, lower limits of detection were found with ICP based techniques, mainly when coupled with a mass spectrometric detector. However, ICP methods usually require digestion procedur es and are not portabl e. Cremers et al. 44 measured total soil carbon by placing the sample in a sm all quartz tube. LIBS measurements give satisfactory results when compared with meas urements obtained using the conventional dry combustion method. Other biological matrixes like blood 45, hair 46 and bacterial spores 47 and samples including pharmaceuticals 48, plastics 49, 50, oil 51and ceramic 52 products have been investigated. The main problem faced by those who want to employ LIBS with complex samples is in the interference caused by matrix effects. Seve ral methods can be used for improvements like internal standards or other calib ration methods. One interesting appr oach is calibration-free LIBS 53 which overcomes the matrix effect without the use of calibration curves. Statistical methods can also be used to improve the interpretation of spectroscopic data. No wadays, more people are combining LIBS with chemometrics, such as partial least-squares regression,54 pattern recognition (principal component analysis and cluster analysis) 55 and neural network analysis.56 24


The depth profiling ability of LIBS is anothe r advantage for different applications. By recording the distinctive plasma emission from successive single laser shots on the same spot, the local composition of the material at a partic ular depth can be studied. This approach has proven useful for investigating inhomogeneous or multi-layer samples, as reported by Anderson et al. using zinc coated steel.57 Some fields are in need of non-(or minimally) destructive techniques. This is the case in the conversation and restoration of cultural inheritance in art and archeology. LIBS have been used to determine the chemical composition of paints 58 and more precisely the identification of different pigments and additives in paintings to enable a better restoration or for testing their authenticity.59 Remote-LIBS can be used when the analysis requires large distances from the target. Lopez-Moreno et al. 60 demonstrated stand-off detection of rocks, soil and vegetation with the use of a telescope at a di stance greater than 10 m. Conclusions LIBS has found several niches. Its simplicity for elemental analysis allow to this technique to achieve acceptable results over a large range of applications. In combination with other techniques, more particularly molecular techniques such as Raman spectroscopy, the range of applications of LIBS are likely to increase. 25


TARGET Contact front Ionization front Shock front Shocked ambient gas Shocked and ionized ambient gas Vapor/plasma Plasma core Figure 2-1. Different plasma f eatures within the shockwave. 26


0510152025303540 -4 -2 0 PMT voltagetime (s) (b) Figure 2-2. Important time periods after plasma formation during which emissions from different species predominate. The box in a) represents the time which the plasma light is monitored using a gatable detector.6 Time-resolved plasma emission detected with a photomultiplier tube (PMT) at a spectral window containing an ionic and atomic line, b). 27


LIP Pulsed Laser Laser Focusing Optics Target Wavelength Selector Focusing Optics Radiant Power DetectorFigure 2-3. General laboratory bench-top setup for Laser Induced Breakdown Spectroscopy 28


CHAPTER 3 RESONANT ABSORPTION IMAGING IN LASER-INDUCED PLASMAS Introduction Over the years, a continuous effort has been and is still directed toward the understanding of the laser-sample interaction and plasma evolution. Understandably, much information can still be gathered to fully comprehend laser ablation processes. Different complementary diagnostic techniques have helped to improve the analytic al performance of laser-i nduced plasmas. Several theoretical models 61, 62 have been developed to fully describe the evolution of different species, as well as the major plasma parameters: temper atures, electron number densities, broadening parameters, velocities, etc. Laser-aided diagnostic methods ar e used extensively in a wide variety of fields in almost all areas of science. Laser spectroscopic methods are often characterized by high resolution and accuracy and have several advantages for plas ma diagnostics. Tunable lasers can provide spectral information and high spatial and temporal resolution can be achieved from lasers beam properties, such as beam size and coherence. An other key element is the possibility of imaging with the help of a beam expande r and a two dimensional detector. In addition, the laser source and the detector can be located far from the object to allow measurements in a the presence of a high luminosity background. Optical diagnostic tech niques are, in general, non-intrusive, allowing the study of non-perturbed plasmas. For this reason, they are invaluable components in experimental plasma studies. Of all the diagnostic techniques applied to plasmas (ICP, microwave) and extensively described in the literature, laser absorption spec troscopy seems to be the one having received comparatively less attention.6, 63-69 This maybe due to the fact th at the laser plasma is highly luminous and therefore a high intens ity radiation source is necessa ry in order to overcome the 29


strong background continuum which lim its the attainable signal signa l-to-noise ratio. Diagnostics techniques include laser-induced fluorescen ce, emission spectroscopy, shadowgraphy imaging, laser-enhanced ionization, X-ray sp ectroscopy and scattering processe s. From an analytical point of view, absorption spectroscopy has the adva ntage over the other te chniques (emission and fluorescence) that it is easily amenable to quantif ication. In other words, an absorption scan over the line profile or a peak absorption measuremen t can directly provide th e total number density of absorbing species, if the oscillator strength of the transition is known.70 Another advantageous feature of the absorp tion technique would be the possibility of measuring absorption line profiles and therefore ga in a better insight on the main broadening mechanisms operating in the plasma. This will be possible if a narrow excitation source, such as a diode laser, can be frequency scanned across the profile. Lastly, if a very narrow laser beam is direct ed into the plasma, vertical line of sight measurements can be obtained with the corr esponding spatial distribution of the absorbing species. Alternatively, the laser beam can be ma de larger than the plasma and one can there obtain spatially resolved distribution of the speci es with the use of a bidimensional detector (CCD). This chapter will focus on the use of diode laser abso rption spectroscopy to obtain information about the plasma morphology and its relation to laser-induced plasma analysis. Spectral considerations are the focu s of the following chapter. Resonant-Absorption Imaging for Morphology Studies: The plasma morphology or spatial structure of the plasma produced by lase r pulses provides valuable information on the distribution of atomic species in the plasma and their dynamics. Morphological data can be applied to understand and to improve LIP analytic al figures of merit in several ways. Signal 30


fluctuations in LIBS analysis has been a major pr oblem, especially laser shot to shot instability, variation of the sample surface and the non-linear dependence of several plasma parameters. Integrating the signal in a LIP ove r several laser shots can, then l ead to rather poor results. On the other hand, morphological information from a single laser-shot can be used for proper analysis of spectra or for selecting the part of the plasma characterized by the best analytical results. Xu et al. 71 used such data to re-normalize th eir emission spectra according to the conditions of specific plasmas. The renormalizati on approach has also been used in aerosol analysis to compensate for individual variations.72 Plasma morphology can al so be used to find the best sampling conditions by optimizing optical geometry.73 Most of the techniques used to study the mor phology of a LIP can be found in the literature of inductively-coupled, microwave, gl ow-discharged and arc plasmas. Perhaps the most widely used method for obtaining spectral information from various locations in the plasma is spectrometer slit imaging.74, 75 Simultaneous two-dimensional information can be achieved by using an imagi ng spectrometer (a lens and a monochromator) and a CCD camera. However, for each individua l plasma formation, the spatial vertical dimension is limited by the direction along the slit a nd a translation of the plasma is required to obtain the full image of the plasma. With a CCD coupled with an intensifier, time-resolved images are produced by a proper gating of the high voltage applied to the microchannel plate. If an interference filter is placed in front of the CCD detector, mi nimal spectral resolution of the plasma images can be obtained. A tunable filter offers the advantage of a llowing easy spectral acquisition. The tunable filter may be of the liquid crystal type 76 or of the acousto-optic type 40. In both cases, however, a limited spectral resoluti on is obtained, ~5 nm for liquid crystals and few nanometers for acousto-optic filters. Over such a wide spectral window, several lines, 31


atomic or ionic, are likely to be present, ma king it difficult to observe the population of single selected levels. Another approach to spectral imaging is to use spectrally resolved sources. Previous absorption measurements in ICP plasmas have attempted to use m odulated electrodeless discharge and glow discharge lamps as spectral sources.77 The radiation of these lamps is usually not intense enough to overcome the luminosity of th e plasma, especially at earlier delay times when the continuum is still present. Absorptio n measurements using fixed-frequency sources provide a measure of absorbance for a narrow sp ectral region under a line profile and cannot be used to determine the overall absorption feature. Near-resonance or resonance absorption and sha dow imaging is another tool widely used for plasma morphology.78-81 In most cases, the plasma plume is created and the absorption measurement is obtained by recording the transmitted light of a secondary pulsed laser. The beam is enlarged and spatially filtered to illumi nate as uniformly as possible the plasma over its whole extent. The high directionality of the b eam over long distances allows line-of-sight absorption measurements in high ly luminous plasma environments The plasma images in Figure 3-1,81 obtained by absorption shadowgraphy show two distinct features: a white, darkfringed line external to the laser spark body, which is pro duced by the refraction of the laser light at the shock-wave propagating in the medium, and da rk areas which are produced by absorption of laser light by the ionized gas. If a tunable lase r is used as the excitation source, resonance shadowgraph images can be related to single species in the plasma. Among the many possible techniques available to obser ve the morphological properties of the plasma, the resonant absorption imaging with a tunable diode laser deve loped in this work combines many interesting features. Resonant absorption imaging offers the possibility of following the temporal evolution 32


of the morphology of an individual population in a single laser shot with the inherent spectral advantages of diode lasers. Atomic and ionic species have been thoroughly studied in the literature but little attention has been paid to the ground state population of atoms. Whereas the most common diagnostic approach, emission spectroscopy, is normally unable to give direct info rmation about the ground state population, absorption spectroscopy provides complementary measurement and will further improve our understanding of the processes of plasma creation and e volution. Due to the availability of diode lasers at wavelengths of the ground state atomic population of cesium, cesium plasmas have been investigated in this work. Experimental Experimental Setup A detail schematic of the experimental se tup for resonant absorption imaging of ground state cesium is shown in Figure 3-2. The excitation laser source used to generate the breakdown is a Q-Switched Big Sky Ultra Nd-YAG laser wi th maximum pulse ener gy of 50 mJ, though a power of 25 mJ is used in all experiments. The laser produced 9 ns pulses at the fundamental wavelength of 1064 nm and is fired at a repetition rate of 1 Hz. If not mentioned specifically, the pulse is focused slightly below th e sample surface, at distance of 0.2 inches from the focal plane of the lens. The focal length is 2 inches. The samp le is placed in a vacuum chamber pumped by a mechanical pump and controlled by a needle valv e to allow a range of pressures from ambient pressure, 1 bar to 0.01 mbar. The chamber is made of brass and has multiple accessible windows for line-of-sight measurements in the plasma. The chamber is placed on an XYZ stage to adjust for difference of heights between samples and to assure fresh ablation areas for successive laser shots. 33


The probe radiation at 852.34 nm, corresponding to the D2 ground state 6S1/2 6P3/2 transition of atomic cesium, is provided by an external cavity diode laser operated in LittmanMetcalf configuration (Model TEC 500, Sacher Lasertechnik, Marburg, Germany) with a manufacturer specified linewidth of 5 MHz. It is a safe assumption that such linewidth is much narrower than the absorption line in the plasma. An iris is used to restrict the radiation diameter to 1mm, providing a homogeneous radial inte nsity distribution. The probe beam is then expanded by a beam expander, which consists of a short focal length dive rging lens followed by a converging lens. The distance betw een the two lenses is adjusted to provide a well collimated beam of several centimeters in diameter. The size of the beam exceeds the plasma dimensions to ensure a full illumination of the entire plasma volume. A small fraction of the beam is sent through a low pressure cylindrical cesium sealed cell (Opthos Instruments, Inc., Rockville, Maryland) and detected by a near-infrared photomultiplier tube (R636, Hamamatsu, Japan). The PMT out put is amplified (Model 427, Keithley Instruments, Cleveland, OH) and recorded by an oscilloscope (TDS3000 Series, Tektronix, Willsonville, OR). This fraction of the beam permits tuning the diode laser to the center of the strong hyperfine component of the ground state transition (F=4). This assures that all measurements are done at the peak absorption and that no mode hopping occurs; the laser emits in the single mode configuration. Resonance radiation transmitted through the plasma is directed into an 852 nm interference filter (Optometrics, LLC, Ayer, MA) to spectrally eliminate background light and protect the detector. The detector is placed several feet aw ay from the sample to further minimize the isotropic continuum emission, presen t at early delay times, that co uld saturate the detector. The detector is a Princeton Instrument (PI Mode l# ICCD-576S) gated 2-D Intensified Charge34


Coupled Device (ICCD). The ICCD photosensitive face or photocathode contains a grid, 576 x 384, of 2.2 m pixels. Incident photons striking th e photocathode release electrons that are then accelerated towards a micro-channel plate (MCP). A fluorescent screen absorbs the amplified electrical signal and the photons emitted are de tected by a CCD. A schematic of the ICCD detector is shown in Figure 3-3. Different exposur e times with different delays can be achieved by controlling the gate voltage between the photocathode and MCP. The gate is adjusted from the PG-200 (Princeton Instruments) pulse generator and triggered with the output Q-Switch of the Nd-YAG laser. A signal from the pulse generator is sent to the ST-138 camera controller that synchronizes the CCD with the gate and sends the output image to the computer. Sample Preparation For this study, two different t ypes of sample have been pr epared. Cesium pellets in a cellulose matrix are prepared by soaking a known amount of cellulose into a solution of known concentration of CsCl. The mixture is then dried overnight and carefully ball-milled before being pressed at a specific pressure to form a homogeneous pellet. The final concentration of the pellet can then be calculated, assuming th at all the water has been eva porated and that the CsCl has been retained by the cellulose matrix. Different pellet concentrations can easily be achieved by changing the solution concentration and a 10 ppm pellet was chosen for the following study. The last pellet provided a number density of cesium in the plasma which resulted in a transmittance of the order of 50%. In absorpti on, the concentration of the an alyte is based upon the difference between two large signals: the detector signal in the presence and absence of the analyte, which, in this case, is synonymous w ith detecting the signal on a nd off resonance, respectively. However, a CCD detector is known to have a limited dynamic range, lower concentrations resulted in a very weak absorption and higher con centrations resulted in such strong absorption that the study of the morphology of the plasma b ecomes difficult. Suitable contrast in the images 35


was not obtained with transmittance lower than a few percent. A second type of sample was provided (courtesy of Dr. Galan Moore from Corning Inc). CsCl fused pellets were prepared using standard additions. The CsCl and the Li2B4O7 flux were added together and ball-milled. SiO2 is then added and the standard is ball-milled again. SiO2 was added to the CsCl flux mix in order to make the pellets opaque for better absorption of 1064nm radiation. Several concentrations have been provided and a 0.1% of cesium pellet was chosen for this work. Not only do the fused silica and cellulose pellets have a very different matrix in terms of chemical composition, but their roughness and rigidity also cause strong variation in mass removal. Cellulose pellets are similar to a fine powder that binds together while fused silica pellets are extremely rigid with a smooth and reflective surface. Th e microscope pictures of the laser craters after one and ten lase r shots on the same sample spot can be seen in Figure 3-4. The obvious difference in crater diamet ers, approximately 500 m in di ameter for the cellulose pellet and 200 m in diameter for the fused silica is easily explained by their rheology. The difference in mass ablated understandably results in dispar ate number densities of atomic cesium in the plasma plume. This explains the use of differe nt concentrations, 10 pp m and 0.1%, with in the cellulose and the fused silica matrices, respec tively, that provided su itable contrasts for the plasma morphology study. Data Acquisition All the ICCD images were captured and ma thematically processed by the WinSpec32 software v2.5.18.2 (Princeton Instruments, NJ). Monochromatic radiation traversing several opticalsurfaces such as lenses and filters creates inference patte rns due to etalon effects. Those patterns, which are highly dependent on the la ser beam direction and on the wavelength, are recorded by the detector as well as plasma ab sorption. Optical elements can also present some defects that will a lter the perceived morphology of the ground state population. To correctly 36


account for these phenomena, a reference image had to be taken for each set of measurements or every time a parameter was changed. The reference image served as the true I0. The plasma transmittance was then plotted using the well known relation dark darkII II T 0 (3-1) where Idark is the dark image. The dark image was recorded when the MCP voltage was turned off. Figure 3-5 shows a typica l absorption image, the reference image and the resulting transmittance image. The color scale represents the transmittance intensity. It is important to note that those images represent the overall transmittance of the plasma by species absorbing at the probe laser wavelength and not only by the ces ium ground state population. For the particular image shown on Figure 3-5, the absorption by the shockwave can be observed from the concentric half circles. The blue circle repr esent values of transmittance lower than one while the red circle indicates values of transmittance higher than one: this result can be easily explained by the fact that the shockwave is defl ecting a part of laser radiation, accounting for an apparent absorption. A similar halo can be ob served around the mushr oom shaped absorption because of the deflection of the beam caused by the change in index of refraction. To assess which portion of the plasma represents the ce sium ground state absorption, images were taken with the laser tuned to the maximum absorption of the cesium transition and detuned by several tenths of nanometers. This is illustrated in Figure 3-6. Throughout the entire duration of the plasma, no transmittance lower than 70% is observ ed when the laser is tuned away from the transition, except for a very small layer close to th e sample surface (i.e. at grazing incidence) and at the location of the shockwave. The lower va lues of the transmittance (darker scale on the images) are therefore only attr ibuted to cesium ground state populat ions. A linear scale from 0 to 0.8 in transmittance is chosen for all the images from this work. The dimensions of the plasma 37


are calibrated with reference to a vision target that displayed precision test patterns printed on a glass substrate. The spacing betw een two lines of the dimension scale is displayed in every Figure and is always set to 1 mm. Results and Discussion Time-resolved Atmospheric Plasma Evolution Time-resolved resonant images produced on cellulose pellets at atmospheric pressure are shown in Figure 3-7. The gate width was set to 0.5 s and the delay varied from 1 s to 150 s. It should be noted that the transmittance values are independent of the laser probe energy. No differences were observed when neutral density f ilters attenuated the intensity of the radiation. Thus, all imaging experiments we re performed under linear interac tion between the laser and the atoms. Several papers 78, 82-86 have described time evolution of diffe rent species in the plasma plume and typical morphologies similar to those obtaine d here were reported. Once the electron number density of the vapor plume is greater than a cr itical number density, the vapor plume reaches a critical temperature and rapidly undergoes an almost one-dimensional expansion along a preferential axis, i.e., along the focusing le ns axis. The rapidly expanding vapor further compresses the background gas as it expands, gene rating a strong external shockwave in the background gas region. For high background pressures, a blast wave model has been used to describe the shock front caused by th e expansion of laser ablation plasmas.87 This model describes the propagation of a shock wave, caused by an expl osive release of energy through a background gas and is applied to this work (cf. Figure 3-8). The propaga tion of the shock front by the background gas follows the laser supported detonation wave model and is described by the following distance-time relation 38


ntR ][2 (3-2) where n depends on the symmetry of the s hock front ( for a cylindrical geometry).87, 88 This model provides an approximate initial veloci ty of the shockwave, on the order of 105 m/s which is coherent with measurements from other works.89 After several microseconds, the speed of the external shockwave decays to a sound wave as the pressure of the sh ockwave approaches the background gas pressure. The absorption of ground state cesium increases with time as the plasma temperature decreases, as the atoms recombine and the electr ons decay to the ground state. At early delay times, most of the atomic population appears to occur at ~2 mm above the surface, where the front of the plasma core contracts. The compre ssion from the background gas in the leading edge of the plume corresponds to a snowplow effect.90 The large velocity and temperature gradient between the vapor plume and both the sample and the background gases generates a boundary layer during expansion of the plum e. This boundary layer is the sour ce of vorticity in the flow of the plume and strong vortex rings provide strong velocities that lift the vapor plume.91, 92 Additional outward velocity creates the m ushroom shape, clearly observable for few microseconds and the plume changes to a more spherical shape at a 10 s delay. The high absorbing region moves towards the sample surf ace tens of microseconds after the laser pulse terminates when the interacti on with the background sufficiently reduces the strength of the vortex rings. Even after very long delay time, hundreds of microseconds, a strong absorption from the ground state is s till clearly visible. Figure 3-9 shows the same time-resolved plasma evolution for fused silica pellets. It is difficult to directly compare the transmittance va lues with the cellulose matrix, due to their distinctive nature, but the time evolution show s a similar behavior. Lower transmittances are 39


located in the plume front, lifted by ~2 mm from the surface during the first microseconds and the population collapses towards the surface at la ter delays. Ground state cesium seems not to be absorbing at early delay times in the region betw een the front of the plume and the surface. One could speculate that higher laser energy is absorbed by the surface, resulting to a faster expansion of the plume, generating a stronger vortex ri ng. The plasma would likely be more ionized, consequently decreasing the absorption of th e ground state atomic cesium outside the front plume. Effect of the Pressure on the Plasma Evolution The ambient gas pressure influences in the plume morphology and has been studied by several groups.82, 92, 93 Plume images obtained at different pressures for two delay times are shown in Figure 3-10. The effect of the pressure on the plasma evolution can be described by three pressure regimes. Below 1 mbar the ambient gas pressure is such that the plasma expands without collisions from the ambient gas molecu les: as a consequence, no sharp boundary is formed. The plume is expanding freely and th e distribution of the cesium ground state is associated with time-of-flight measurements. For higher pressure, the plasma becomes more collisional. Narrowing of the plume is clearly visible as the pressure is increased. Note that at pressures of 5 mbar and higher, a narrow stream of dense absorp tion is observed along the center line of the plume, indicating a si gnificant concentration of absorbers there. The gas molecules have compressed the plume to form a symmetrical elliptical shape. The time-resolved absorption at 5 mbar is presented in Figure 3-11. It has to be noted that the absorption reaches two apparent maxima, around 0.3 s and 2 s. These two maxima w ill be discussed in more details in the next chapter. When the pressure is increased to 25 mbar and more, a compressed region is formed between the expanding cesium plasma and ambien t gas. Here the transition to a hydrodynamic 40


regime takes place, with the plume acting as a piston on the surrounding ambient environment.91 This compressed region moves ahead of the plume and generates a shock wave. It also leads to enhance absorption from the plume front as a re sult of transfer of plume kinetic energy into thermal energy plume heating. When a shoc k boundary is formed between the plume and ambient gas medium, it effectively shields the diffusion of the ambient gas species from the plasma species. A strong vorticity occurs, lift s the plume and diffuses the species radially.92 Effect of the Laser Focusing Distance The ground state distribution for different focusing distances at a few delay times for cellulose pellets at atmospheric pressure is presen ted in Figure 3-12. The foca l length used in this work is 50.8 mm and the sample was positioned at different focusing distances below the sample surface. The position of the sample relative to the focal plane of the lens plays an important role in the analytical use of LIP and has been studied by several groups.94, 95 A significant variation is observed, both in the shape of the distribution and the transmittance values. When the laser is focused well below the surface, the absorbed ir radiance by the surface decreases and a smaller plasma is observed. At an intermediate focusi ng distance, absorption and plasma stability are obtained and the shape assumes an increased sp herical symmetry, which makes this distance the optimum when the plasma is used as an analyt ical spectroscopic tool. The optimum focusing is produced when the sample is placed at approxi mately 10%, or 5 mm in this experiment, below the focal plane of the lens, where the laser irradiance corresponds to approximately 10 GW/cm2. At focusing distances closer to the lens focal plane, the plasma become s elongated along the lens axis. An extreme case results in the breakdown of the air above the sample. An air breakdown not only affects the plasma shape but also the plasma reproducibility, and needs to be avoided in analytical applications. 41


Shot-to-shot Reproducibility Several parameters influence the LIBS m easurement and account for the well-known relatively poor precision of the technique. Intrinsic noises, which can be shot noise, detector noise or instrumental noise do not account for a ll of the poor repeatability of LIBS. The plasma time evolution as described previously, dramatica lly affects the plasma spatial structure. As a consequence, the probed or collected region that produces reproducible analytical measurement has to be chosen carefully. For a specific dela y time, the plasma morphology may also vary due to sample heterogeneities. Heterogeneities arise within disparities in the sample concentrations, surface profiles but also from the successive laser shots hitting the same crater. One of the most common approach to improve the precision of LIBS measurements is to accumulate several (n) shots to reduce the shot noise as n but the number of shots in the same crater is generally limited by modification of the sampled surfaces and formation of craters.96 Figure 3-13 shows the transmittance of successive laser shots on the sa me crater of the sample. A deeper and deeper crater is created after every shot, changing the focusing distance and the sample surface characteristics due to melting effect. For cellulose pellets, Figure 3-13 a), in which the ablation dimensions are greater, the ground state distribution and concentra tion varies significantly. The plume is extended along the lens axis as the number of shots increas es and a considerable variation in the transmittance is observed. Fused silica pellets generate more reproducible plumes, as shown in Figure 3-13 b). The first shots are often, consider ed as cleaning or preparatory shots, as the laser interacts with a fresh and rough surface that can contain dust, an oxidation layer or other contaminants not representative of the sample. A significant difference in the morphology and a lower absorption is observed for the first shots. For quantitative measurements, the first few shots are often discarded to improve repeatability. The set of images from Figure 3-13 are a good example on how careful the LIBS user must be when performing 42


quantitative measurements; the plasma evolves in time and in space and the choice of the region and time-scale observed is essential. It can vary from sample-to-sample cf. cellulose versus silica matrices, but it can also be a function of several other parame ters that were not investigated here, such as laser energy and sample temperat ure. The average transmittance, binned from a 1 mm to 1 mm square centered in the shock front of the plume where the transmittance is minimal and where the best repeatability was found, is plotted in Figure 3-14 for two delay times. There is evidence of a drift in the transmittance, which is more pr onounced with the cel lulose pellets. This is likely due to laser heating and change in the focusing distance. As expected, an improvement in the relative sta ndard deviation (RSD) is observe d when the first two shots are not included. The experimental RSD of the transmittance found are of the order of 10%. Conclusions and Remarks Spatial and time evolution of the cesium gr ound state population in laser-induced plasma are studied for different pressures and for two significantly different sa mples, a powder bonded cellulose sample and a fused silica sample. The morphology of laser-induced plasmas is greatly affected by several parameters, namely the samp le matrix, the focusing ablation laser and the shot-shot reproducibility. The plasma evolves rapi dly in time and space and different pressure regimes drastically change this evolution. Imagi ng measurements with an ICCD reveal spatial variation within the plasma formation as well as its structure. By evaluating the plasma homogeneity, the area of the plume at different delay times where th e best signal-to-noise ratio is obtained, can be determined. As a practical outcome, the knowledge of these different morphologies will help designing and implementi ng the optical collection and detection system in order to achieve the optim um analytical performance. This chapter focused on two specific concentrations, according to the nature of the sample. These concentrations provided th e best contrast images for structure studies. In terms of 43


analytical performance, these ICCD images wo uld give poor figures of merit. The low dynamic range of the ICCD greatly limits the use of these absorption images for quantitative measurements. For concentrations ten times higher, the plasma is optically thick, resulting in two absorption regions, while for concentrations ten times lower, no absorption can be measured. It should be noted that different possibilities exist to improve th e dynamic range. Several groups described an extended range of the detection with diode lasers by scanning the laser wavelength to the wings of the absorption profile.97 If the profile is mainly due to collisiona l broadening, especially at longer delay times, the absorptio n in the wings can be obtained at much higher concentrations of the absorbers while maintain ing a linear relation with the concentration. However, an accurate knowledge of the absorpti on profile and the possibility of adjusting the diode laser to known frequencies is required to successfully improve the dynamic range. As already said earlier, absorption measurements have been used in diagnostic measurement not only for morphology studies but also as direct measurement of plasma number densities. However, in this case, spectral information is required. The following chapter will focus on spectrally-resolved absorption spect ra of the cesium ground state population. 44


45 Figure 3-1. Experimental setup used for shadow graphic imaging of plasma evolution and the resulting images. Shadowgraphy of the laser-produced plasma and shock wave evolution is displayed on b). The photograph (right bo ttom) has a delay time of 375 ns with respect to the laser pulse, while the temporal delay between photographs is 500 ns. The shock wave position is highlighted with dashed semicircles.81 a) b)


46 Max power =50mJ mbarNeedle vale + Pressure gageFigure 3-2. Experimental setup for time-reso lved laser-induced plasma absorption images at different pressure with a high reso lution diode laser.


Figure 3-3. Schematic diagram of gated 2D Intensified Charge-Coupled Device (ICCD) 47


48 Figure 3-4. Microscope images of the laser craters. Images are taken after a) c) one and b) d) ten laser shots at 25 mJ. The top images ar e taken from a powde r-type sample, 10 ppm CsCl in a cellulose binder. The bottom images are taken from a fused silica pellet that contains 0.1% CsCl. 0 1mm0 1mm 0 1mm0 1mm


49 I I0 T 02 1 02 1 Figure 3-5. Absorption images at 852.4nm from laser-induced plas ma on cesium based pellets. The transmitted signal, I and the reference signal, I0, combine to display a transmittance image, T of all the species absorption at this wavelength. The scale of the transmittance image is shown at the bottom.


50 Figure 3-6. Diode laser images of laser-i nduced plasmas for the ground state resonance frequency of cesium. Images are shown o n resonance on the left and for a wavelength detuned from the transition on the right. The color intensity represents the transmittance and the scale is shown on top.


51Figure 3-7. Temporal evolution of the ce sium ground state population. Images taken for single shots on 10 ppm CSCl cellulose pellets. The transmittance scale and dimens ion scale is displayed at the bottom. 1 s 3 s 2 s 6 s 4 s 5 s 7 s 8 s 9 s 13 s 10 s 11 s 15 s 19 s 17 s 21 s 25 s 30 s 35 s 50 s 75 s 100 s 150 s


52 Figure 3-8. Time propagation of the shockwave cr eated from a LIP. Samples of 10 ppm CsCl cellulose sample were used. The evolution curve in red is fitted to a laser supported detonation wave model for cylindrical geometry.


53 1 s3 s 2 s 6 s 4 s5 s 7 s8 s9 s 13 s 10 s 11 s1 5 s1 9 s 17 s 21 s 25 s3 0 s3 5 s5 0 s 75 s100 s 150 s Figure 3-9. Time evolution transmittance of the cesium resonan ce ground-state. Measurements are for 0.1% CsCl fused silica pel lets at atmospheric pressure and single shots.


54 0.3mbar 10mbar 5mbar 0.5mbar 50mbar 200mbar 1mbar 25mbar 1bar b ) 5 s 0.3mbar 10mbar 5mbar 1mbar 0.5mbar 25mbar 1bar a ) 2 s 50mbar 200mbarFigure 3-10. Cesium ground state absorption images for several pressures. Images taken at two delay times, a) 2 s and b) 5 s.


55 0.05 s 0.4 s 0.2s 0.6 s 0.8 s 1 s 2 s 1.5s 2.5 s 3 s 4 s 10 s 5s 15 s Figure 3-11. Time evolution transmittance of the cesium resonan ce ground-state. Images taken for 0.1% CsCl fused silica pellet s at a pressure of 5mbar for single shots.


56 Figure 3-12. Effect of the focusing distant on the cesium ground state distribution for several delay times. The focusing values are the di stances of the surface of the sample from the focal plane of the lens. 5s 10s 25s 100s 5mm 4.2mm 3.4mm 5.8mm 6.6mm 7.4mm 1.5mmFocusing below the surface f = 50.8mm


57 12345 67891 0 15 20 25 35 40 Shot n: Shot n1 2 3 4 5 6789 10 12 15 20 25 22 Figure 3-13. Successive LIP on the same target spot for two matrices. a) cellulose and b) fused silica. b) a)


RSD 7 s = 21% (16% without shot 1&2) RSD 5 s = 11% (8% without shot 1&2) 0 510152025 0.2 0.3 0.4 0.5 0.6 TransmittanceShot number 19s delay 7s delay RSD 19 s = 9.3% (8.7% without shot 1&2)0510152025 0.3 0.4 0.5 0.6 0.7 0.8 TransmittanceShot number 5s 15sRSD 15 s = 11% (8% without shot 1&2)a) Cellulose matrices b) Fused silica matricesFigure 3-14. Transmittance values for successive laser shots on the same target spot. In a) cellulose matrices and in b) fused silica matrices were used. The relative standard deviations are given on the top of the graphics. 58


CHAPTER 4 HIGH-RESOLUTION TIME-RESOLVED AB SORPTION SPECTROSCOPY IN LASERINDUCED PLASMA Introduction In the previous chapter, it was shown that spatially resolved absorption images of laserinduced plasma provided valuable information on the plasma spatial structure and the population distribution. Time-resolved morphology was measur ed using a gated detector, an intensified charge-coupled device (ICCD). Time evolution was discretely plotted for successive laser shots by changing the gate delay on the ICCD and therefore non-gated detectors with temporal discrimination are more suitable for time-resolv ed measurements. Photomultiplier tube (PMT) detection provided complementary measur ements to spatial absorption images. The shape of spectral lines in plasmas is a topi c of strong interest a nd has been the subject of study for many years.98-101 In spectrochemical analysis, line shapes plays a major role since it helps understanding spectral interf erences, verifying plasma conditions (optically thin or thick) and explains the behavior of anal ytical calibration curves. The role of spectral line shapes as a diagnostic tool is magnified because of the abr upt spatial and temporal gradients which exist during the formation and evolution of a LIP. The derivation of atom ic number densities from an absorption measurement requires high spectral reso lution. Here, the narr ow bandwidth of external cavity diode lasers is used to obtain the spectral be havior in the cesium ground state populations in the ablation plume. Spectral line shapes have been investigat ed since the first observations of LIP 102 and have been more and more thoroughly stud ied in view of the possibility of obtaining high spectral and temporal resolution simultaneously. Numerous examples of measurements that follow the LIP evolution of spectral line shapes have been commonly used and reported in the literature. Emission spectroscopic detection is the most common method of linewidth measurement. For 59


example, Nmet et al. 103 studied the structure of spectral line s of Au, Ag and Cu lines at various delay times from several alloys. Hermann et al. 104 made a detailed study of the temporal and spatial development of LI P spectra of Ti in a N2 atmosphere. Although emission spectroscopy has the advantage of not requiring a second ra diation source, several properties make other diagnostic methods advantageous in line shape determination. At early times, the continuum emission produces a strong backgr ound and the temperature and num ber density are so high that the lines are generally self absorbed and/or self reversed. These effects complicate the determination of the true linewidth and make it difficult to extract plasma parameters. Even at later delay times, resonance lines or lines with strong transition probabilities are likely to be affected by self-absorption. On the other hand, in absorption spectrosco py, line shapes can be extracted directly if the laser power is low enough so that the transi tion is not saturated.65 Furthermore, the determination of the ground st ate lineshape is not possible in emission spectroscopy. Saturation spectrosco py requires a different treatment than the simple absorption theory.105 Measurements of 87Rb/86Rb have been made using a continuous wave Ti-sapphire laser for atomic absorption measurements with laser sampling us ing a 1064 nm Nd-YAG laser.106 Dominant line broadening mechanisms for Ca and Rb in low pressure LIP have been determined by Gornushkin et al.99. Another similar approach to measure line shapes is fluorescence spectroscopy and diode laser excited atomic fluorescence was used by Smith et al. 107 for isotopically selective measurements in uranium plasmas. In absorption spectroscopy, the concentration of the species is based upon measuring the difference between two large signals obtained in th e presence or in the absence of the analyte. Such measurement is accomplished by either tuning the source on and off resonance or by continuously tuning the source wavelength over th e resonance. A common approach to improve 60


the dynamic range in absorption spectroscopy is to use wavelength or frequency modulation.97 A diode laser is driven to smoot hly modulate the wavelength over the spectral line at a certain frequency in the kHz-MHz range. Rather stable plasma generated in furnaces 108 or in ICP can benefit from modulation spectrosc opy; however, with the steep temporal gradients occurring in LIP, modulation frequencies of the order of 100 MHz are required and are not easily reachable with modern diode lasers. Theory Considerations The absorbed intensity, IA, of a probe beam of light propagating through a linearly absorbing medium may be described by the Beer-Lambert relationship, given by line k t AdeIIII )1)(()( 0 0 (4-1) where I0( ) and It ( ) are the spectral distributions of the incident and transmitted radiant intensities (W cm-2), k( ) (cm-1) the absorption coefficient depe nding of the atomic spectral line at frequency and (cm) the absorption path length. Sp ectral lines are never strictly monochromatic and a spectral di stribution exists near the central wavelength. The atomic profiles are a function of the numbe r of absorbing partic les present in the plasma and can give distinct information about elect ron number densities and broa dening mechanisms. The Full Width Half Maximum (FWHM) is often used to define the line profile and is the wavelength interval of a spectral line between two points whose intensity is equal to half of the maximum intensity.70 The overall experimental line shape is a convolu tion of the line profile in the plasma and the linewidth of the probe beam. In the case of high-resolution diode lase rs, equivalent to the ones utilized in this work, the external-cavity provided such high resolution that the probe beam bandwidth was much narrower than th e probed line profile. Therefore, k( ) is constant over the 61


source profile, and equal to its maximum value, km. Thus the radiant intensity absorbed is given by S k AIeIm)1( (4-2) where IS the intensity of the probe light integrated over its emission line width. For low values of the product kml i.e., at low optical densities, equation 4-2 simplifies as follows SmAIkI (4-3) The fraction of radiation absorbed, defined as the ratio (IA/I0) is thus seen to be directly proportional to the peak value of the absorption coe fficient. One can also re late the absorbance to the fraction of radiation ab sorbed through the relation )1log()log(0tI I A (4-4) Finally, from the classical radiation theory,96 and in the limit of low intensity of the exciting radiation, the absorp tion coefficient k( ) (cm-1) can be expressed as l luNgB c h k )( )(0 (4-5) Where h (J s) is the Plank constant, 0 (Hz) is the central freque ncy of the transition, c (cm s-1) is the velocity of li ght in vacuum, Blu (J-1 Hz s-1 cm3) is the Einstein coefficient of stimulated absorption and Nl (cm-3) is the number density of the absorb ers in the lower energy level. The spectral distribution func tion is normalized, i.e. 1)(dg Integrating equation 4-5 over frequency, we have lluNB c h dk0)( (4-6) Or, by using the Einstein relations between Blu and Au; 62


l ulN A dk 8 )(2 0 (4-7) where Aul (Hz) represents the Einstein coefficien t for spontaneous emission between the upper level (u) and the lower level (l). The lineshape function g( ) reflects the various broadening mechanisms occurring in the plasma. A Review of Broadening Mechanisms In laser induced plasmas, an atomic spectra l line may be broadened due to Doppler and collisional broadening mechanisms Because of the low laser fluence used in the present work, laser induced saturation and power broadening, whic h could be caused by the probe laser, are not considered. Doppler broadening, due to the relative motion of an ab sorbing particle and the light source, is particularly significant at the temper ature common in atmospheric pressure plasmas. For a system of particles with a Maxwellian velocity distribution, the relation between the Doppler broadening FWHM, D, and the kinetic temperature, T(K), is given by 96 m TD 0 71016.7 (4-8) where T (K) is the temperature and m is the mass of the absorber (g). The line shape function exhibits a Gaussian shape described by: 2ln4)(2 02ln2 )( De gD D (4-9) where 0 is the central frequency and D is the FWHM. Collisional mechanisms include van der W aals, resonance and Stark broadening and result in Lorentzian profiles. Van der Waals broa dening is due to interac tion between two neutral species, cesium and all neutral spec ies, including cesium, not optica lly coupled to levels involved 63


in the probed transition. This is a short range force proportional to r-6 (attraction) and to r-12 (repulsion) described usually by the Lennard-Jones potential. This collision broadening can be diabatic or adiabatic.109 However, diabatic collision broadening is often overshadowed by adiabatic broadening which relates to collisions when the perturber does not induce a transition between atomic levels but causes a phase change of the emitting wavetrain. In the impact approximation, the collisional broadening FWHM a nd shift due to van der Waals interaction are given by nvCwidth Waalsdv 5/3 5/2 6 _..71.2 (4-10) nvCshift Waalsdv 5/3 5/2 6 _..98.0 (4-11) where v (cm/s) is the relative velocity, n (cm-3) is the concentration of perturbing particles, and C6 (cm6.s-1) is the constant determined by the interaction potential. Resonance broadening, which is 1/r3 dependent, is due to interactions with similar atoms and may have effective cross sections orders of magnitude larger than foreign gas kinetic cross sections.109 This mechanism may be significant when an excited atom (state p) interacts with an identical atom in a significantly populated state (state q) and when a strong electric di pole transition couples the two levels. The Lorentzian FWHM, due to resonance br oadening is given by the relation 110 N m fe g ge q p res 00 2 2/12 2 3 (4-12) where gp and gq are the degeneracies of th e two states involved, f is th e oscillator strength of the resonance transition, e the electron charge (C), me is the mass of an electron (kg), 0 is the vacuum permittivity (J-1C2m-1), 0 is the central frequency of the transition and N (m-3) is the population number density of state q. Note that that equation 4-10 does not depend on the relative velocities of the colliding at oms and hence on the plasma temperature. 64


Stark Broadening, due to Coulombic interactions with charged particles, is a function of ne and Te and is often the most significant collisiona l broadening mechanism, especially at early delay times in the plasma evolution. Heavy elem ents, such as cesium, are generally influenced by the quadratic Stark effect which yields an appr oximately Lorentzian line shape. In addition to broadening, an absorption band may be shifted in wavelength (Stark shift) due to interactions with electric fields. Pressure sh ift and Doppler shift may also be present but can be negligible compared to Stark shifts when ne>1015 cm-3, a condition satisfied in atmospheric laser-induced plasmas. Theoretical Stark parameters, tabulated by Griem 110, relate the broadening and shift of the spectral lines in plasmas to electron numbe r density and temperatur e. For singly-ionized plasmas, the total theoretical FWHM due to quadratic Stark broadening Starkwidth (), and the total theoretical Stark shift, Starkshift (), are given by the formulas 110 e e e Starkwidthwn Tn n16 2/16/1 4/1410) 068.01( 1075.112 (4-13) e e e Starkshiftwn Tn n w d16 2/16/1 4/1410) 068.01( 100.2 (4-14) where ne is the electron number density (cm-3), Te is the electron temperature (K), w is the electron impact parameter (), is the ion broade ning parameter, and d/w is the ratio of shift to width (dimensionless). The true lie function is then the convol ution of the Lorentzian and Doppler contributions and is known as the Voigt profile. Experimental Frequency Calibration by a Confocal Fabry-Perot Interferometer Interferometers serve a variety of purposes in laser spectroscopy, they aid in the measurement of absorption or emission line profiles as well as in the diagnosis of laser mode structure, linewidth, and performance.111 Perhaps more importantly, interferometers provide a high resolution frequency ruler to appropriately analyze atomic sp ectra. In this work, the probe 65


laser frequency is monitored by sending a fraction the beam into a confocal Fabry-Perot interferometer (FPI). The varying transmission function of a FPI is caused by interference between the multiple reflections of the incident light between the two reflecting surf aces. Constructive interference occurs if the transmitted beams are in phase, and this corresponds to a high-transmission peak of the etalon. If the transmitted beams are out-of-phase, destructive interference occurs and this corresponds to a transmission minimum. The cond ition for constructive interference within a FPI is that the optical path is equal to an integr al number of wavelengths of the incident light.112 Specifically, a confocal FPI consists of two partially transmitting spherical mirrors precisely aligned to form a reflective cavity. The mirrors have equal radii of curvature ( r ) and are separated by a distance d where d = r In this work, spherical mirro rs with a 50 cm radius of curvature (CVI Laser Corporation, Albuquer que, NM) 98% reflectivity from 850-920 nm, and /10 optical quality is used. The two mirrors we re permanently fixed by a 50 cm quartz tube to provide a rigid FPI which, afte r initial mirror alignment, c ould be moved as the whole. Given the mirror charact eristics and neglecting spherical ab erration, the free spectral range ( ), or frequency separation between successive inference maxima, can be determined from the following relation: d c 2 (4-15) where c is the speed of light and d is the distance between the mirrors. The theoretical free spectral range of the devi ce was therefore 150 MHz. The overall performance of an interferometer is quantified by its finesse ( F* ). A higher finesse value is indicative of a higher resolu tion measurement. The total finesse of an interferometer is influenced by the optical and alignment imperfections of the device, stemming 66


from non-ideal reflectivity, surface quality and we dge angle. In a confocal device, the total finesse is mainly governed by the mirror reflectivity, R which is given by the following relation, and to a lesser extent, by the surface quality, respectively.113 R R Frefl 1* (4-16) The theoretical finesse of the device was a pproximately 9.9. The device was calibrated by comparison to the fine structure splitting of the cesium 6S1/2 6P3/2 transition at 852.34nm from a sealed cell containing cesium at low pressure. In a pure cesium vapor at room temper ature, the hyperfine component of the 6S1/2 (F = 3, 4) 6P3/2 (F = 2, 3, 4, 5) transition has a Doppler shape (cf. equation 4-9). Six hyperfine transitions exist between the cesium 6S1/2 and 6P3/2 states, labeled a, b, c, d, e and f in Figure 4-1 114. For each component i, k( ) D and k0 are given by the following equations, respectively: 2 0)( 0)(D i iekki i (4-17) m kT ci i D20 (4-18) i i D ifN mc e k 12 0 (4-19) The hyperfine shift ( o i) and oscillator strength (fi) for each individual hyperfine component can be found in Figure 4-1. The total D oppler envelope of the weak component (F = 3 F = 2, 3, 4) and the st rong component ( F = 4 F = 3, 4, 5) was then calculated by adding each individual profile with their respective we ights derived from the selection rules. The calculated Doppler envelopes are shown in Figure 4-2. The spacing between the centers of the weak and strong components is measured to be 8.77 GHz and it can be compared to the spacing from the experimental spectra with its fringes. The theoretical free spec tral range of 150 MHz is 67


confirmed by the number of fringes, 58.6, corresponding to a spacing of 8.77 GHz. It should be noted that the non-linear dispersion of the fringes from the confocal FPI at the far edge of a single scan is minimal and is neglected in this study. Experimental Setup The experimental setup used to acquire resonant ab sorption spectrum in a LIP is shown in Figure 4-3. The radiation at 852 nm was provided by an external cavity diode laser with nominal (manufacturer specified) linewidths of 5 MHz, as reported in the prev ious chapter. An iris of 200 m restricted the dimension of the probe beam to allow spatia lly resolved measurement. The radiation is sent through the pl asma generated by a Nd-YAG laser and detected several feet away by a spectrometer composed of a classic C zerny-Turner monochromator (0.5 m,1200 gr/mm grating blazed at 400 nm, Acton Resear ch) and a PMT (R5108, Hamamatsu, Japan). The spectrometer is computer controlled and set at 852 nm. As opposed to an interference filter, a monochromator is chosen in experiments i nvolving the PMT detection, to provide a higher spectral background rejection, mainly from the strong early continuum emission. The latter would saturate the non -gated PMT and the recovery tim e would prevent early delay time observations. The PMT output is amplified by a pr oprietary module from Jobin Yvon inc. with an approximately 10 ns rise time and record ed by a 500 MHz oscilloscope (TDS 520D series, Tektronix, Wilsonville, OR) connected to a com puter by a GPIB contro ller (GPIB-USB-HS, National Instruments, Austin, TX). Single seque nce oscilloscope traces were imported into Microcal Origin Pro 7.5 for data analysis. The recorded time-resolved transmission signals are converted to absorban ce by the usual relation (4-4), where I0 is determined by averaging 200 pretrigger data points. The sample is mounted on a XYZ stage and pla ced in a vacuum chamber. For this work, 0.1% CsCl fused SiO2 pellets were solely used as they provided good reproducibility and fewer 68


particles ejection. It is importa nt to note that both the vacuum chamber and the ablation laser are mounted on a secondary XYZ stage so that the entir e plasma can be translated without changing the ablation parameters (i.e. focal length), wh ile keeping the iris a nd detecting optics fixed. To scan the diode laser, a waveform pr ovided by a function gene rator (Model FG3C, Wavetek Meterman, Everett, WA) is applied to a piezoelectric element placed behind the tuning mirror in the laser cavity. While scanning, the 852.34 nm diode laser is operated in the current coupling mode. The current coupling function couples the piezo voltage and the injection current of the laser diode. The benefit of this coupling is to extend the mode-hop-free tuning range of the laser system. Typically, a 1V peak -to-peak triangular wave at a frequency of 10 Hz was used for frequency scanning. The triangle function provides a linear frequency scan. It was found that, if a lower frequency was driving the piezo element, the laser would suffer from a small frequency drift over several scans. Relative frequency calibration is achieved by directing a portion of the scanning beam into the previously described co nfocal FPI. Interference fringes are detected by a photodiode. Absolute frequency calibration is obtained by scanning over th e absorption profile of the previously mentioned ces ium low pressure sealed cell. The 10 Hz scanning frequency of the diode la ser is considered small compared to the reciprocal of the plasma lifetime in order to justify the assumption the diode laser frequency remained constant for the duration of the plasma absorption. During a scan of 100 ms, the radiation frequency of the laser is scanned ov er 20 GHz. If a plasma lifetime of 100 s is considered, the laser frequency has varied only by 20 MHz, just four times the laser bandwidth. As a quasi-fixed frequency is generated within the occurrence of the observed phenomena, the TTL output of the function generato r triggered the output of an ad justable pulse/delay generator (Model DG535, Stanford Research Systems, Inc) that finally triggered the pulsed laser. The 69


delay generator controlled the frequency at which the probe beam is absorbed in the plasma. An absorption profile is then obtaine d by discretely cha nging the pulse delay af ter every single laser shot. Every time the Nd-YAG laser is fired, oscill oscope traces from the low pressure cesium cell scan and the Q-switched outpu t pulse are recorded to monito r the probe laser frequency, as shown on Figure 4-4. Finally, a set of neutral density filters was used to insure that the interaction between the laser and the atoms was linear. To verify this assumption, absorption measurements were taken with and without neutral density filters and found to be the same, within the uncertainty of our measurement. Results and Discussion To determine the influence of the ambien t gas environment on the cesium ground state population in the plasma plume, absorption measurem ents were performed at different pressures, i.e. 1 atm, 5 mbar, 0.5 mbar and 0.01 mbar. Different experimental regimes can be expected here. At the lowest pressure, due to the absence of ambient gas particles, the PMT measurements taken at different distances from the target surf ace are equivalent to Time-Of-Flight (TOF) measurements of the atoms released from the target and reaching the probe volume. As the pressure increases, the kinetic distribution of the species in the plume will be affected because of the interactions between the plume and the surrounding gas. TOF Measurements for a Pressure of 0.01mbar From the TOF measurements, relevant info rmation regarding the time taken by a given atomic population to develop in a given energy state after the pl asma creation can be obtained. Time-resolved absorbance spectra taken at two heights from the plasma surface, 1 mm and 2 mm are displayed in Figure 4-5. A height of 1mm means that the laser probe beam diameter is centered at 1.1 mm from the surfa ce. The beam diameter is 200 m. The graphs represents a 70


contour plot, where the x-axis is the time in s the y-axis the frequency in GHz and the color intensity represents the absorbance. The intensity scale is displayed next to the contour plot. The time evolution gives an insight on the velocity, hence the kinetic en ergy of the emitted particles. In addition, relevant parameters which help the interpretation of the mechanisms responsible for the plume absorption are obtained as well. Fr om Figure 4-5, the absorption is maximal at approximately 0.4s and decreases with time as the plasma temperature decreases. TOF absorption data can be re presented by a shifted Maxwel l-Boltzmann distribution with a center of mass shifted as expected with th e formation of a thermalized Knudsen layer according to the relation.115 ) 2 )( exp(.)()(2 2 3tRT tvzm t z NtAs cm (4-20) where N is an amplitude normalization factor that depends on the ground state number density, z is the height of the probe beam above the surface, vcm is the center-of-mass velocity, Ts is the effective internal temperat ure of the directed beam of pa rticles or sample temperature, m is the mass of the species, t the time. The Knudsen la yer is the region within a few mean free paths of the target surface in which th e particles vaporized, sputtered due to the laser collides due to high-number densities.115 In this region the Ma xwellian formulation assumes that the speed distribution has equilibrated afte r propagating a short distance from the target surface, and that the signal is a direct measure of the concentrati on of the indicated specie s. Figure 4-6 gives a typical TOF profile recorded for Cs atoms in vacu um. The fit to the data is shown by the red line in the Figure 4-6 and is in relatively good agr eement with the experimental TOF curve. The calculated parameters, center of mass velocity and temperature are calculated to be 2.3x105 cm s1 and 22000 K respectively. Those values are in ag reement with similar m easurements performed on other samples.76, 116 71


The spectral structure which can be deduced from Figure 4-5, coincides with the ground state absorption of cesium, the strong (F=4 F = 3, 4,5) and weak (F = 3 F = 2, 3, 4) components of the 6S1/2 state are clearly visible but not resolved (cf Figure4-8 a)). The significance of the Doppler, Star k and van der Waals broadenings mechanisms can be estimated from the theoretical dependence of the ces ium 852 nm line width and shift upon plasma temperature as shown in Figure 4-7. The temper ature dependence of the Doppler line width is given by equation 4-8 and is assumed to be domi nant; calculations of van der Waals broadening and shift (equations 4-10, 4-11) are carri ed out using a typi cal value of ~10-42 m6.s-1 for the parameter C6, 117 and calculations of the Stark width and shift were made using equations 4-13, 414. The parameters w d/w and were not available for the Cs 852 nm line but were assumed to be identical to the parameters for the Cs 894 nm line. As one can see from Figure 4-7, van der Waals collisional broadening only is important for large number densities of perturbing atoms (over 4x1017 cm-3), and so van der Waals broadening is negligible except for pressures near atmosphere. The other possible broadening to be considered is resonanc e line broadening due to collisions of like species (Cs and Cs). Resona nce broadening occurs wh en a high number density is present for a coupled state. There are several states coupled to the ground states and from Figure 4-7, one can see that the number densities of the excited states have to be of the order of 5x1016 cm-3 to contribute significantl y. At a delay time of about 0.4 s, when most cesium is likely to be excited, this broadeni ng might therefore be important. Figure 4-8 a) shows the spectral behavior of the absorbance obs erved at a height of 1 mm above the surface and at a delay time of 0.4 s, t ogether with that obtained from the reference cell. In the latter, the linewidth is purely broade ned and has no spectral shift. A small shift, less than 0.5GHz, is observed but this value is lim ited by the precision of the measurement. As one 72


can see from Figure 4-7, for St ark broadening to be dominan t over the Doppler broadening (ne>1015 cm-3), a larger spectral shift would have b een observed. Van der Waals broadening is not dominant for the same reason. The noise in the data makes an accurate fit to a Voigt profile difficult due to the complexity of the line, cf. six hype rfine components. The frequency scanning range of the diode laser barely resolves th e spectral profile and c ontribute also to the uncertainties of the fitting. Several combinations of Doppler linewidth and Lorentzian line width can gave satisfactory results in the range of the experiment unc ertainty, approximately 10%. The relative standard deviati on was obtained from the data discussed in the previous chapter. For that reason, both Lorentzian and Doppler shapes were f itted to obtain the limiting case values. As can be seen from the Figure 4-8 a), Doppler fits better the tail of the ab sorbance profile. Pure Lorentzian lines resulted in a FWHM of 6 GHz and pure Doppler lines in a FWHM of 2.5 GHz. Both fittings give similar values for the numbe r density of ground state that can be extracted from the line peak absorbance. The expression us ed for the peak absorbance in our case can be derived from equation 4-3, 4-5 and 4-7, and is given by the relationship eff l ul mNA kA 82 0 (4-21) where eff (Hz) is the effective line width of the absorption profile and is defined as m effk dk )( (4-22) The effective line width is determined as the width of a rectangular line profile which has the same height and area as the re al line. The absorption pathlength, l, was estimated from absorption images to be 1 mm. The temporal evolution of the number density is shown in Figure 4-8 b). Number densities of 1016 cm-3 are calculated A Doppler FWHM corresponds to a kinetic 73


temperature of 13000 K 1500 K (cf. equation 4-6) a value commonly repor ted in the literature for similar plasmas at early delay times.99 Therefore, on the basis of th e set of data for Cs, Doppl er broadening is the main contributor to the cesium 852 nm line width. Alth ough Stark broadening exists, cf. the small shift in the lineshape, it is indisti nguishable from resonance broadeni ng as all the popula tion of all the states involved are no t known and have not been measured here. TOF Measurements for Pressures of 0.5 mbar, 5 mbar, 1 atm As the pressure is increased, gas phase coll isions dramatically change the evaluation dynamics of the ablation plume and a double peak structure in the temporal profiles of the cesium ground state species is obse rved. Absorbance profiles recorded at various locations in the plume, 0 mm, 1 mm and 2 mm in a direction perp endicular to the target surface are given in Figure 4-9 and Figure 4-10. The time evolution as well as the spectral absorption profile obtained at 0.5 mbar, 5mbar and 1atm is plotted in a contour plot. In Figure 4-10, the time-resolved absorbance at the peak of the strong hyperfine co mponent (at 1 mm for 0.5 mbar) clearly reveals that the cesium ejected by the plume exhibits a double peak TOF distribution. The first peak in time appears at approximately 0.4 s for every pre ssure. The velocity of the particles sputtered and vaporized from the first peak is sufficiently high that the ambient gas pressure has a shifted Maxwellian distribution. The center-of-mass velocity is calculated as 1.4 105 cm s-1 and the surface temperature as 20000 K. However, the broade ning of the peak is smaller. The kinetic temperature was calculated as (done previously ) and found to be 8000 K 1000 K. This lower value is indicative of the surroundi ng gases slowing down the TOF particles. It should be noted that the Doppler lineshape fits better than the Lorentzian, supporting the idea of Doppler broadening as the dominant broadening. 74


The second absorption occurs at a different tim e, depending on the pressure and the height of observation, as shown in Figure 4-9 and 4-10. The time scale displayed is adjusted and is thus different for each pressure in order to follow the full time evolution. The time spacing between the two peaks, decreases from 0.5 mbar and 5 mbar. This is in agreement with morphology observations from the previous chapter. As the pressure increases, the plume is characterized by a strong interpenetration of the plasma species and the background gas leading to a thermalization of the species with their surroundings. Recombination reactions result in the formation of atoms and consequently induced th e absorption observed. At 0.5 mbar, due to the low collisional environment, thermalization occurs from approximately 2 s to 10 s. After 10 s, the plasma has cooled and no absorption is observed. At 5 mbar, the surrounding gases force the absorbing species to be c ontained into a smaller volume. Recombination occurs sooner and the absorbance is greatly enhanced. At atmos pheric pressure, the apparent delay in the recombination process at 0 mm seems to indicate the occurrence of strong vorticity effects. The plume front is lifted from the surface and only after a long delay time, for example 20 s in this case, the species closed to the surface absorb. Th e contour plot at 2 mm for 1 atm is not shown because the absorption was not resolved and abso rbances greater than 1.5 are obtained. In this spatial region of the plasma, optically thick co ndition exists and a spectroscopic analysis could not be made. The spectral broadening at 0.5 mbar and 5 mbar is dominated by Doppler broadening after the first peak occurs. According to Figure 4-8, wh ich compares the differe nt widths, one can see that Stark Broadening is not dominant with respect to Doppl er broadening because of the absence of the Stark shift. As one can see from Figure 4-8, at high electron number densities, on the order of 1016 cm-3, the Stark shift of the Cs 852 nm line is approximately 2 GHz. Therefore, 75


the upper limit for the electron num ber density, for delay >5 s, can be estimated to be no more than approximately 1015 cm-3. The same reasoning can be applied to indicate the absence of van der Waals broadening. A remaining possibility is resonance line broadening; however, the number density of resonant excited states has to be greater than 1016 cm-3 for resonance broadening to be important. If we assume that Local Thermodynamic Equilibrium (LTE) conditions exist at a time delay of 2 s and later, the states are in thermal equilibrium following a Boltzmann distribution and high temperature (>10000 K) are required to possibly achieved such high number densities for even the first excited state. Figure 4-11 shows Doppler fitting for different heights, the calculated kinetic te mperature and the calculated cesium ground state number density. At 1 atm, the confinement of the plasma speci es in the front part of the plume and the shielding generated by the steep gradient of pr essures enhance the absorption of the cesium ground state. Temperature and number densities cal culated are much highe r in the confinement region. Doppler broadening is likely the main br oadening mechanisms but Stark, van der Waals and resonance can all contribute to a Lorentzian broadening. A published work 98 at atmospheric pressures reported contributi ons of 0.85 and 0.25 for Doppler and Lorentzian fractions, respectively, at atmospheric pressure. The auth ors made use of curve of growth methodology associated with absorption measurements to cl ose estimate the ratio of the two broadening mechanisms. Doppler line shapes are fitted to atmospheric data and displayed in Figure 4-12. Conclusions Laser atomic absorption with a narrow band scanni ng diode laser has been used to evaluate laser plasma diagnostics at different pressures. So it was shown that, in the case of a trace element (cesium in a fused silica matrix), Doppler broa dening is the dominant broadening mechanism. The number density of cesium atoms in the ground state, as well as the kinetic temperature is 76


evaluated by measuring line prof iles and peak absorptions at 852 nm. Overall, high resolution measurements of the hyperfine structure of ces ium under different experimental conditions have been illustrated. It is felt that the combination of time-resolv ed morphology measurements with high resolution spectral profiles data as demonstrated in this work, represents an attractive tool for the diagnostics of laser induced plasmas. 77


78 89Figure 4-1. Partial energy level diagram of cesium. The hyperfine structure of the 6S1/26P3/2 transition are detailed. The diagram shows frequency detuning from the line center of each hyperfine component involves in the tran sition as well as relevant wavelengths and oscillator strengths.114 Figure 4-2. Calibrated disper sion of the confocal FPI. -101234567 0 1 2 k*l 58.5 fringes = 8.77GHz v


79Figure 4-3. Schematic of the experimental setup for time-resolv ed high-resolution absorption of the cesium ground state line a t 852nm. Nd-YAG (Big Sky Laser 50mJ) 852nm Diode laser low pressure Cs cell PMT photodiode function generator pulse generator delay External triggerQ-switch out Piezovoltage TTL pulse Cs pellet300miris vacuum chamber Monochromator + PMT detector Beam expander


0.00 0.02 0.04 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 Voltagetime (s) Figure 4-4. Records of the abso rption profile of the cesium reference cell and the Q-switched output trigger from the pulsed-laser. The Q-switched trigger signal allows scanning through the profile by adju sting the gate on the pulsed-delay generator. 80


a) 5 10 15 20 25 time (s)frequency (GHz) -0.10 0.020 0.14 0.26 0.38 0.50b) 012345 5 10 15 20 25 time (s)frequency (GHz) -0.10 0.020 0.14 0.26 0.38 0.50 Figure 4-5. Contour plot of time-resolved spectr a of the cesium absorptio n at a pressure of 0.01 mbar. The diode laser beam probes a 200 m in diameter region at a) 1 mm and b) 2 mm from the target surface. 81


82 Ts= 22000K Vcm= 2.3x105cm/s012 0.0 0.1 0.2 0.3 Chi^2/DoF= 0.00022 R^2= 0.94948 P10.01552.00054 vcm2.27088.05797 M/2RTs0.66003.03898Absorbancetime ((mm2/ s2) (mm/s) s) ( s) Figure 4-6. Typical TOF profile recorded by abso rption of Cs atoms in vacuum and the shifted Maxwellian fitting. gure 4-6. Typical TOF profile recorded by abso rption of Cs atoms in vacuum and the shifted Maxwellian fitting.


83Figure 4-7. Calculated contributi ons of the spectral line width, and shifts, in GHz, for the Cs 852nm line, as a function of p lasma temperature. 0500010000150002 00002500030000 0 2 4 6 width and shift (GHz)Temperature (K)DopplerStark width, ne=1016cm-3Stark width, ne=5*1015cm-3Stark width, ne=1015cm-3Stark shift, ne=5*1015cm-3Stark shift, ne=1016cm-3Stark shift, ne=1015cm-3V.dWaals width, n=3*1017cm-3 V.dWaals width, n=1018cm-3 V.dWaals shift, n=1018cm-3 V.dWaals shift, n=3*1017cm-3 res, nCsexcited state=8.1016cm-3


84 0.0 5.0x10151.0x10161.5x10162.0x1016Figure 4-8. Absorption spectra and time-resolv ed ground state number density. Comparison of the absorption spectrum of Cs at 0.4 s for a pressure of 0.01 mbar and the absorption spectrum of the reference cell in a). The frequency between the two peaks corresponds to the hyperf ine structure. Gaussian and Lorentzian line shapes are fitted to the spectra. Time-resolved number densit ies in b) of the cesium ground state for two heights of observation calculate d from peak absorption values. 0.5 1.0 1.5 1mm 2mmGround state number density (cm-3) time (051015202530 0.0 0.1 0.2 0.3 0.4 0.5s) Absorbancefrequency (GHz) Gaussian fit Lorenztian fit 0.4s reference cell b) a) (s)


85 02468101214 5 30 15 25 20 10 time (s)frequency (GHz) 02468101214 5 10 15 20 25 time (s)frequency (GHz) 02468101214 5 10 15 20 25 time (s)frequency (GHz) 02468101214 5 10 15 20 25 30 time (s)frequency (GHz) 02468101214 5 10 15 20 25 30 time (s)frequency (GHz) 02468101214 5 10 15 20 25 time (s)frequency (GHz)PRESSURE OF 0.5mbar: PRESSURE OF 5mbar: 0mm 0mm 1mm 1mm 2mm 2mm Figure 4-9. Contour plot of time-resolved spectra of the cesium absorptio n at a pressure of 0.5 mbar and 5 mbar. The diode las er beam probes a 200m in diameter region at 0m m, 1mm and 2mm from the target surface.


86Figure 4-10. Contour plot of time-resolved spectra of the cesium absorption at a pres sure of 1 atm. The diode laser beam probe s a 200m in diameter region at 0 mm and 1mm from the target surface. 20406080100120140160 5 10 15 20 25 30 time (s)frequency (GHz) 20406080100120140160 5 10 15 20 25 30 time (s)frequency (GHz) -0.10 0.080 0.26 0.44 0.62 0.80PRESSURE OF 1atm: 0mm 1mm


0.5mbar: 4s 5mbar: 3s 51015202530 0.00 0.05 0.10 0.15 4s, 0.5mbar, 1mm Gaussian fitAbsorbancefrequency (GHz)51015202530 0.00 0.02 0.04 0.06 0.08 Absorbacnefrequency (GHz) 4s, 0.5mbar, 0mm Gaussian fit51015202530 0.00 0.02 0.04 0.06 0.08 Absorbancefrequency (GHz) 4s, 0.5mbar, 2mm Gaussian fit51015202530 0.0 0.2 0.4 0.6 0.8 1.0 1.2 4s, 5mbar, 1mm Gaussian fitAbsorbancefrequency (GHz)51015202530 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Absorbancefrequency (GHz) 3s, 5mbar, 0mm Gaussian fit51 01 52 02 53 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Asorbancefrequency (GHz) 3s, 5mbar, 2mm Gaussian fit T = 4000 K N = 6 1015 cm-3 T = 4000 K N = 5 1015 cm-3 T = 2000 K N = 2 1015 cm-3 T = 1500 K N = 8 1014 cm-3 T = 2000 K N = 1 1015 cm-3 T = 1300 K N = 4 1014 cm-3 Figure 4-11. Effect of the distance from the pr obe beam to the sample surface on absorption spectra at 0.5mbar and 5mbar. The Gaussian fitting is shown on red at 4 s and 3 s delay, respectively. 87


1atm: 30s 01020304050 0.0 0.2 0.4 0.6 0.8 1.0 Absorbancefrequency (GHz) 1atm, 0mm, 30s Gaussian fit51 01 52 02 53 0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Absorbancefrequency (GHz) 1atm, 1mm, 30s Gaussian fit T = 8000 K N = 1 1016 cm-3 T = 8000 K N = 1 1016 cm-3 Figure 4-12. Effect of the distance from the pr obe beam to the sample surface on absorption spectra at 1 atm. The Gaussian fi tting a 30 s delay is shown on red. 88


CHAPTER 5 INTRODUCTION TO ATOMIC VAPOR DETECTION There is a growing need towards the development of photon detectors with high spectral resolution. Many applications such as laser Do ppler velocimetry, chemical imaging and longrange optical communications require or can be enhanced by superi or spectrally-resolved detectors. Atomic vapor detection is characterized by spectral discrimination in the range of MHzGHz, or approximately 10-4 nm, in the near-infrared regi on. Several devices can detect photons with high spectral resolution,118, 119 however, these detectors often lack other critical figures of merit such as high throughput, sensitivity, spatial resolution and image quality. High spectral resolution is achieved by ex ploiting the absorption features of atomic lines in their gas phase. Atomic vapor cells have been extensively used as ultra-narrow Notch filters,120 but their application as photon detectors is still limited. The pot ential applications of resonance detectors was described by Matveev et al. 121. The primary absorption proce ss occurring in the atomic vapor can be monitored by either detecting the fluorescence or the ionization resulting from a selected state, usually excited with a laser. Imaging can be accomplished by spatially expanding the excitation laser into two dimensions. Single-point photon detectors ar e resonance ionization dete ctors (RID) and resonance fluorescence detectors (RFD). Severa l versions of these detectors have been described based on the atomic vapors, mercury 122, 123 and cesium 124, 125 for both single-point photon and imaging detection. However, the different types of dete ction are based on the sa me operating principle. Figure 5-1 depicts the general ope ration of a resonance fluorescen ce detector and a resonance ionization detector. Fluorescence photons or el ectrical charges are created and only detected when the interrogated object emits or scatter radi ation of an energy matching exactly that of the narrow absorption linewidth of the atomic transition chosen and is summari zed in Figure 5-2. In 89


a fluorescence detecting scheme, the signal phot ons at the resonance absorption wavelength, 1, which are absorbed, excites ground state atoms to th e first excited energy level. Photons outside the narrow absorption band pass through the ce ll unaffected. Excited atoms can further be excited by a pump laser tuned to a second transition at the wavelength, 2. For imaging detection, 2 is expanded into a sheet of li ght and directed into the side of the cell behind the input window. Excited atoms decay to the fluorescent level and emit at the wavelength, F. This fluorescence can be detected using a CCD camera for imagi ng or PMT or photodiode fo r single point. Due to the difference in wavelength between 1 and F, a standard optical filter or a dichroic mirror for back-fluorescence, can be used to spectrally separate F from the unabsorbed photons. The ionization detector works in si milar fashion; however, the phot oionization of excited atoms requires a multistage excitation from two lasers, 2, 3. Electrons created in the plane of these two lasers are accelerated to ward a microchannel plate (MCP ) by the high voltage applied between the input window and MCP. Electrons striking the MCP are amplified and transfer from the electric field to phosphorous screen from which a image is captured by a CCD. Because all atomic vapor photon detectors have a fixed frequency of detection, a tunable source of light is required to investigate the phenomenon. Photon detection occurs by tuning in wavelength the incident light such that that upon interaction with an object or phenomenon, the scatter light will be shifted into resonance with the transition of the atomic vapor. Any phenomenon that generates wavelength shifts upon interaction with in cident radiation, Mie scattering, Raman scattering, Doppler velocime try, can benefit from these detectors. Different transition schemes in the case of multiphoton excitation have been tested. A mercury based RID was first de monstrated by Matveev et al. 122. A three-step ionization scheme was used with the first transition, 1 at 253.7 nm. Electrons crea ted from photoionization were 90


directly directed toward a lumi nescent screen by a voltage app lied between the screen and input window. The mercury RID was further improve d by the addition of a microchannel plate 126. A compact sealed-cell Hg RID was also reported 127 where an image composed of 1000 photons was detected by image summation. The spectral re solution obtained was on the order of 25 GHz due to the combination of Doppler-broadened profiles of seven mercury isotopes. The use of cesium as the atomic vapor has several advantages. Both cesium and mercury possess high vapor pressures at room temperatur e but cesium has only one stable isotope so its absorption profile is not complicated by isotope spectral shifts. The ground state transition for cesium has a large oscillator streng th (f=0.72) and occurs in the infrared whic h is beneficial for several ways. Wavelengths in the near infrared are accessible by external cavity diode laser which are spectrally narrow, tunable, inexpensive and small; as opposed to UV radiations used to reach Hg resonance transition. Large excimeror Nd-YAG-pumpe d, pulsed dye laser systems are used to produce a tunable UV source. In add ition, near-infrared can effectively penetrate biological tissues with minimum p hotodegradation, enabling potential applications in biological and medical fields. The near abse nce of native fluorescence in this region is another advantage in detecting Raman scattering. Temirov et al. 125 described a cesium based RID in a six-way cross vacuum chamber. The detector consisted of a metal coated vapor input window, a commercially available alkali metal vapor dispenser and a MCP-phosphorous scr een. The excitation/ionization scheme (62S1/262P3/282D3/2ionization continuum) was employi ng radiation at 852.34 nm, 917.47 nm and 1064 nm. A portion of the Nd-YAG laser wa s used to pump a Raman wavelength shifter and the resulting radiation was directed into a second harmonic ge neration crystal. The second harmonic at 599 nm pumped a color center crys tal which could simultaneously produce 852.34 91


and 917.47 nm. The spectral resolution of the de tector was limited by the Doppler-broadened linewidth of the signal transition and determined as 540 MH z. A major limitation of the cesium RID is the reactivity of cesium. Despite consid erable effort in constructing the device to overcome the reactivity, the detector had only a limited lifetime of about a week before the electrodes and windows required cleaning. The co ating left by cesium on the surface caused a decrease the threshold voltage before an arc was created between the atom chamber and the electrodes, thus limiting its operation. A cesium based RFD has been described by Korevaar et al.128. This RFD was based on a two-step excitation scheme (62S1/262P3/282S1/2) followed by wavelength -shifted fluorescence at 455.35 nm (72P3/262S1/2). A spectral resolution of 600 MHz was reported at the working temperature of 100C. The Cesium RFD uses comm ercially available abso rption sealed-cells as the atomic vapor reservoirs. RFDs provide the advantage of being compact, inexpensive and do not require any maintenance. The dimension of a sealed Cs RFD cell varies from few millimeters to tens of centimeters depending on the applica tion desired. The most recent RFD was reported by Pappas et al.,124 who used the same first resonance excitation as the detector radiation but further excited the 62P3/2 state by pumping to the 62D5/2 state by laser excitation at 917.47 nm, 2. The excited atoms radiatively decay to the 7P3/2 and subsequently generate fluorescence photons at 455.65 nm. A limitation of the cesium RFD is th e relatively low sensitivity when compared to high-gain devices with poor spectral resolution su ch as acousto-optic t unable filters or liquid crystal tunable filters.129 Correll et al. 130 devoted a significant effort in increasing the quantum efficiency of the process by collisional excitation energy transfer. Pixley et al. 131 used the latter cesium resonance fluorescence detector for m oving-object detection where they measured Doppler-shifted radiation (cf. Figure 5-3 a)). Velo city differences of 15 m/s could be resolved 92


and working detection range of up to 1500 m/s wa s available given the tu ning characteristics of the first excitation laser. Pappas et al. 132 used the same RFD to detect Mie scattering from silica particles by deconvolution of Mie sc atter from the instrumental response (cf. Figure 5-3 b)). The Mie scattering linewidth (140 MHz) correlated well to the literatu re value of 100 MHz in air. Another type of scattering, namely Raman scattering, has been investigated, mainly because of the versatility and quantitative ability of the Raman technique. However, while Mie scattering benefits from the high spectral resolution of the detector and the high number of signal photons available, Raman scattering does not. In f act, the Raman linewidth is much larger than the absorption linewidth and the scattering cross section is very small. The number of Raman photons detected is limited and thus, an excellent limit of detection in terms of number of photons is required. The goal of this work is to further deve lop cesium-based vapor detectors by simplifying the detection process for practical use and to focus towards the detection of Raman photons. 93


atomic cell a) 12 fl detector Filter b) Ionization Re g ion Phosphor Screen MCP to CCD Ion Atom ElectronInput Window Figure 5-1. Schematic of a reso nance fluorescence detect or and resonance ionization detector. In a), when the radiation 1 is absorbed, the atoms excited are further excited to a fluorescing state by a radiation 2. The fluorescence photons can be detected at the back of the cell after going th rough a standard optical filter In b), when the radiation 1 is absorbed, 2 and 3 ionized the excited atom. The ions or electrons are accelerated and amplified in the MCP by a voltage between the input window and MCP. The electrons hit the phosphorous sc reen and are converted to photons. The latter are detected by a CCD. 94


ground state excited state laser shift = exc Laser source laserObject / Phenomenon Atomic vapor reservoir Detection Figure 5-2. The principle of operation of an atomic vapor photon detector, the atomic vapor absorbs at fixed frequency. A tunable source is required to interrogate the frequency shift from the moving object or species that scatter light inelastically. 95


-2000-1500-1000-5000500 0 1 2 3 4 5 6 7 96 m/s 121 m/s ReferenceFluorescence Intensity (a.u.)Frequency (MHz) a) b) Deconvolution RFD Frequency response Convoluted Mie scatter Figure 5-3. Applications of a cesium RFD. Fluorescence excita tion profiles of the reference beam and scatter representing different disk velocities in a), adapted from 131.The measurement of Mie scattering from a partic ulate solutionin b): spectral response of the RFD; scattering signal de tected from a suspension of particles; deconvoluted Mie spectrum. 132 96


CHAPTER 6 ATOMIC VAPOR DETECTORS: TOWARD THE DETECTION OF RAMAN SCATTERING PHOTONS Introduction Raman spectroscopy is a powerful tool capa ble of providing information about the structure, properties and concentration of molecu les. Its major disadvantage is that of having a very weak signal due to the small cross section of the interaction process. In fact, the Raman cross section is approximately 12-14 orders of magnitude lower than the competitive effect, fluorescence.133 The fundamentals and applications of Raman spectroscopy has been reported extensively in the literature and ca n be found in several reviews and books 133-135. Most of the scientific studies in the field of Raman research were and still are, performed as single point nonimaging methods.136, 137 Compact lasers, notch filters, and CCD cameras revolutionized the field in the late 1990s. From a single point scanning method, Raman spectroscopy has developed into a two dimensional chemical and biological imaging techni que. Several instruments are available for a variet y of applications.138 Imaging Raman spectroscopy is advantageous when high fidelity images at a limited number of wave lengths (in the simplest case one wavelength) provide sufficient chemical and spatial information. Reducing th e number of spectral channel decreases the time necessary for 2-D experiments without losing spatial di stributions and is of high interest for fixed wavelength de tectors as atomic vapor detectors. Although several versions of Raman imaging devi ces are commercially available, there is still considerable room for improvement in the field, which can be addressed with the further development of atomic vapor detectors (AVD). It is important to note that the spectrally selective atomic vapor detectors do not require filtering of stray radiation that might occur at the excitation wavelength, since only Raman scattered radiatio n corresponding to the narrow absorption profile of atoms will be detected.121 97


The detection of Raman photons by mean s of laser-enhanced ionization was experimentally demonstrated for the first time by Smith at al. 139. Magnesium was chosen as the atomic vapor due to its efficient atomization in an air-acetylene flame, and Raman scatter from carbon tetrachloride molecules were detecte d. A tunable dye laser was scanned around the resonance transition of Mg at 285.213 nm. Further excita tion with a second la ser and subsequent collisional ionization in the flam e resulted in the creation of ion/ electron pairs. Raman spectra of dimethyl sulfoxide, carbon tetrachloride, and chloroform were also successfully recorded with a Mg flame resonance ionization detector.140 The use of atomic vapors as spectrally selective filters has also been demonstrated to reject Rayleigh scattering.141 Rb and Hg vapors formed in a furn ace were used to filter Raleigh scattered radiations from carbon te trachloride samples. Continuous wave Ti-Sapphire laser at the 794.76 nm resonance transition of Rb and XeCl ex cimer pumped dye laser at the Hg resonance line at 253.64 nm interrogated the Raman active samples. One point of concern in the detection of Ra man signals with low pressure atomic vapor detector, such as a resonance ionization detector (RID), is that the typical spectral bandwidth of Raman scatter is much larger than the Doppler -limited absorption profile of atomic vapor at room temperature. Typically, the Raman band half width, R, is in the range of 30-50 cm-1 while the atomic linewidth, 1, is about 400 MHz (~0.01 cm-1). Because of this spectral discrepancy, the absorption efficiency w ill decrease proportionally with the ratio 1/ R and so will the background-limited signalto-noise ratio, which is propor tional to the square root of ( 1/ R ) To better illustrate this point, the theore tical background that in cludes a comparison of Raman signal levels expected from an atomic va por detector (AVD) (fluorescence or ionization) 98


as compared to a conventional Ra man spectrometer is useful to our approach. The Raman signal, SR (in counts) for the spectrometer is given by the following relation: tTAN d d ESD Da a p R)( (6-1) Here, Ep is the laser photon irradiance (photons s-1 cm-2), ad d is the Raman differential cross section of the analyte molecules (cm2 sr-1), Na is the number of scatteri ng molecules per unit area (cm-2), AD is the useful area of the detector (cm2), is the collection solid angle (sr), T is the transmission factor of the optical system, D is the detector effici ency (counts/photons) and t is the integration (measurement) tim e (s). The first product (within braces in the above equation) corresponds to the excitation effi ciency and the second represents the optic al conductance or luminosity of the spectrometer. In this approach, the Raman signal will be given by: t TAN d d ESAVDDAVD AVD Da a p AVD R)( (6-2) The terms and symbols are the same as in e quation 6-1 and the suffix AVD has been added. By comparing the two equations, and assuming equal values of the optical transmission, as well as equal integration times, we can see that D AVDD D AVD D R AVD RA A S S )( (6-3) The first ratio on the right hand side of equation 6-3 represents the ratio of the optical conductances, en. This ratio can be much larger than unity, indicating that the optical conductance of AVD can be much larger (typically 104 times higher) than of a conventional Raman spectrometer, for the same spectral reso lution. This clear advantage of AVD has been 99

PAGE 100

reported by Matveev et al.121. Most type of spectrometers ca n be designed to provide a high resolving power, R (R=106-108). However, for most spectrometers, increasing the resolving power inherently comes at the expense of the luminosity. Table 6-1 shows some values for the luminosity, resolving power and their product. It is clear that, in theory, atomic vapor detect ors largely surpass the performance of the other detection systems. Atomic vapor detectors are quantum devices and atoms inside the cell are isotropically sensitive to resonant photons. The luminosity is then constant and independent of R. From equation 6-3, one can see that if D for a conventional spectro meter detector system is assumed unity, ( D)AVD must equal to ( en)-1 for AVD to detect the same number of Raman photons as the traditional Raman spectrometer. The overall detection efficiency of an AVD system is given by the product of several efficiency factors, namely the absorp tion efficiency or fraction absorbed, abs, the fluorescence efficiency, F (or ionization efficiency, ion for RID), and the collection efficiency, c (of photons for RFD or electrons for RID). In an AVD working under vacuum, the absorption efficiency is bound to be low due to the large difference in the spectral profiles of Raman emission and atomic absorption. It is clear that in this case both the ionization (or fluorescence) efficiency and collection efficiency need to be closed to unity; otherwise the gain du e to high luminosity is completely counter-balanced. The improvement of th ese efficiency factors dictates our approach in this work. First, a new cesium fluorescence dete ction scheme with the presence of an ethane buffer gases improves the fluorescence efficienc y. Second, the absorption efficiency is improved by broadening the absorption bandwidth with helium as a buffer gas. It should be reported that some efforts we re first dedicated to improve the cesium resonance ionization detector previ ously developed in our laboratory.125 Although theoretically 100

PAGE 101

superior to RFD, the practical and successful implementation of cesium ionization imaging remains a task facing several engineering challenge s. High ion collection efficiency in a Cs-RID detector requires a low working pressure to avoid recombination of the formed ions while they are accelerated to the multichannel plate. The consequently low (~10-4) absorption efficiency require the ionization efficiency to be close to unity. The ioni zation efficiency depends upon the excitation-ionization scheme chosen and the lase r power available. A 100% efficiency can be reached if the second excitation step (917 nm) and the photoionization step (wavelengths >744 nm) are both saturated. If a commercially available Alexandrite/LiF:F2 + laser matched the requirement and improved the efficiency, ultimatel y, the high reactivity of cesium with all the different components of the detector prevente d the RID from working in proper conditions. A major change in the materials used or design ha d to be performed and therefore our focus was shifted to RFD using sealed cells. Improvement of the Fluorescence Efficiency, F Fluorescence efficiency of 10-3 have been reported for RF D described by Pappas et al. 124 based on the two-step detection scheme, (62S1/262P3/262D5/2): absorption at 852 nm, excitation at 917 nm followed by fluorescence dete ction at 455 nm. The fluorescence efficiency was limited by the radiative coupling from the 62D5/2 to 7P3/2 states required from fluorescing at 455 nm and all the other coupled state that decr eases the quantum efficiency. A novel cesium fluorescent detecting scheme is developed in this work that does not require a second step excitation. Photons matching the gr ound state transition at 852 nm (62S1/262P3/2) are collisionally coupled to the 62P1/2 state that is resonant with the ground state, and fluorescence occurs at 894 nm. The transition 6p2P1/2 -6p2P3/2 is forbidden by the sel ection rules, hence the coupling has to be induced by collisi ons. For this coupling to be effi cient, a very fast mixing rate is required to overcome the large differen ce in energy between th e two states, 554.1 cm-1 and 101

PAGE 102

ethane was chosen for this purpose. Recently, Beach et al. 142 developed an end-pumped cesium vapor laser based on the same approach with high average power that potentially competes with diode-pumped solid-states lasers in app lications that requ ired cw operation. Single step detection significantly decreases th e complexity of the atomic vapor detector, while several strong lasers to saturate the res onant transitions are no longer required. Many applications that required small and compact instrument can greatly benefit from the development of this simple fluorescence detector. Investigation of Ethane-induced Collisional Energy Mixing in Cesium Several studies have focused on collisional energy mixing in laser excited alkali vapors using noble gas perturbers, involving Cs-He and Cs-Ar.114 In a low pressure cesium cell that does not contain ethane, the relaxation rate from the 62P3/2 level to the 62P1/2 is slow compared with the 3.3 x 107 s-1 natural decay rate out of the 62P3/2 level (30.5 ns lifetime), which decays radiatively back to the ground state 62S1/2 level by means of an electric dipole allowed transition. The requirement for this rapid fine-structure mi xing is the reason for the presence of ethane in the cell. Krause et al. 143 measured the cross-sectional values for the Cs 62P3/2 62P1/2 population transfer by collision with various hydrocarbons to be in the order of 10-15 cm2. The participation of the molecular rotational degrees of freedom of ethane in the energy transfer accounts for the very large magnitudes of the mixi ng cross sections. The cross section was found 4 to 6 orders of magnitude larg er than similar cross sections for collisions with noble gases 143. In this work, the cross-section value of the cesium-ethane mixing from 62P3/2 62P1/2 is determined by measuring the fl uorescence intensity of both the radiation at 894.6nm and 852.6nm when excited by a laser at 894.6 nm and at 852.6 nm. 102

PAGE 103

Fluorescence Intensity Fluorescence photons are direc tly related to the fraction of radiation absorbed. The fluorescence intensity from a specific state, IF excited by a monochromatic source of intensity, I0 is described according to the relation below a line source assuming excitation and no preor post-filter effects )1(0 mk FeIYI (6-4) where km is the peak absorption coefficient (cm-1), l (cm) is the absorption pathlength, I0 the intensity of the input radiation ( and Y is the fluorescence power yield, given by the ratio between the spontaneous emission from state j to state i, Aji and the sum of all the deactivation channels from level j, including collisions, Rjq. q jqjq jiRA A Y (6-5) If the product is much smaller than 1, which is the case for low number densities, the absorption system is optically thin and the prev ious expression can be simplified as followed mkm FkIYI0 (6-6) In the above equation, km is given by the expression eff l ul mnA k 82 0 (6-7) where nl is the number density of perturbers (cm-3), eff (Hz)is the effective width of the absorption. The measured fluorescence signal finally depends on the solid angle of detection, c, the collection and efficiency of the detector (counts/photon), and the probed volume (cm3), V and is given by the expression 103

PAGE 104

F C FIV S 4 (6-8) Rate equation Approach The steady-state and temporal behavior of laser interaction with an atomic system are often described by either the rate equation formalism or the density matrix formalism.144 The density matrix formalism properly provides the more co mplete description of laser excited atomic systems. It properly treats coherence effects resul ting from the finite bandwidth of a laser source interacting with a broader abso rption profile. It can also desc ribe certain effects caused by narrowband light at high intensities, such as dynamic Stark effects (broadening and splitting of levels) and two-photon excitation.145 In general, those effects are significant whenever high intensity, narrowband light excites atoms in a weakly collisional media. For collisional media, with external cavity diode lase r, the rate equation approach gives a reasonable description. The principle of detailed balance states that the number of atoms per second leaving the considered quantum state, a, through any particular process must equal the number of atoms per second entering that state, i.e. 0 dt dNa, where Na is the population of the state a. This balance holds for collisional processes and is individually applicable to each pair of levels. For each state an equation can be written to describe the ch ange on the corresponding population due to all the radiative and non radiative processes whic h are present. For a considered state a, aaii i ia aNDNP dt dN (6-9) where is the radiative and non radiative rate of populati on of the state from state i, is the radiative and non radiative probab ility of depopulation from the state a to state i. For the non radiative or collisionally induced transition, the corresponding tran sition probability is given by iaPaaiD 104

PAGE 105

the rate Rmm, where m and m are the initial and final state from the collisions between A and B. This rate is directly proportional to the number density of perturber atoms, Nk. In analogy with gas kinetics, the rate Rmm, is also proportional to the mean re lative velocity of the colliding atoms A and B: k rel mm mmNV R' (6-10) Under typical experimental conditions i.e., in cesium absorption at room temperature, the distribution of atomic velocitie s is Maxwellian and the temper ature-dependent mean relative velocity of the Maxwellian dist ribution is given by the relation BA B relTk V8 (6-11) where kB is the Boltzmann constant, T is the temperature of the system and A-B is the reduced mass of A-B quasimolecule formed by a collision between the two atoms. In equation 6-7, the proportionality constant mm with the dimension of area is the cross section of the process. In general, cross sections for elec tron excitation energy transfer depend on the relative velocity and are therefore functions of the system temperature. In this work, as the cross sections from cesiu m-ethane is orders of magnitude higher than from cesium-helium, the latter are neglected. Another assumption made is to neglect the quenching from helium and ethane. Figure 6-1 summarizes the rates involved in the level scheme when the cesium sealed is pumped by the 62S1/262P3/2 transition at 852.6 nm. Bij is the stimulated emission coefficient and the energy density of the laser. The steady state rate equation for the level 1 (6p2P1/2 ) when level 2 is pumped by the radiation at 852 nm is 01121102 21 1 NRNANR dt dN (6-12) 105

PAGE 106

Thus the number density ratio is equal to 21 1210 852 1 2)( R RA N N (6-13) If we now look at the fluorescence ratio from equation 6-8, we obtain 1 1 2 2 1 2Fc F c F FIV IV S S (6-14) If the two input ra diations at 852.4 nm and 894. 6 nm are co-linearly prop agating in the cell, the fluorescence will be detected by the same collection optics. The efficiency of the detection system is assumed constant over the range of th e two transitions. Furthermore the power of the two lasers are monitored and adjusted to the same value with a set of neutral density filters. Consequently, the previous equation can be simplified, and by substituting from equations 66 and 6-7, the detected fl uorescence ratio is obtained FI 2 110 1 2 20 10 20 10 1210 2120 20 1 2 eff eff F FNA NA A RA RA A S S (6-15) Since the effective width of the two transitions is derived from the same Doppler and Lorentzian broadening parameters, they can be considered id entical. Finally, the following equations enable a direct calculation of th e cesium-ethane rate, R21 and R12, and therefore thei r cross sections assessing the efficiency of the coupling, by solving the ratios of intensity fluorescence 20 21 10 12 101021 20201012 852 1 21 1 )( A R A R AR AAR S SF F (6-16) 20 21 10 12 10102021 202012 894 1 21 1 )( A R A R AAR AR S SF F (6-17) 106

PAGE 107

The experimental tabulated values for A20 and A10 are respectively 2.86x107 s-1 and 3.28x107 s-1. Experimental Setup The cross sections for cesium 62P3/2 62P1/2 mixing in cesium vapor were investigated in a series of fluorescent experiments in a sealed cell as shown in Figure 6-2. The absorption sealed cell in this experime nt, measuring 25 mm in diameter and 40 mm in length was commercially procured (Opthos Instru ments, Inc, Rockville, MD) with uncoated optical windows on each end and filled at room temperature with 100 Torr of ethane and 350 Torr of He, in addition to a small quantity of cesium metal prior to be sealed off. The temperature of the cell varied from room temperature to 150 C by a current-controlled heating tape wrapped around the cell and is further mo nitored by a thermocouple. Higher temperature were not attempted as the inside pressure of the cell would exceed atmospheric pressures and could compromise the integrity of the cell. The cesium atoms were excited from ground state to, alternatively, 6P1/2 and 6P3/2 state by two diode lasers. A flipping mirror enables alterna ting from the 894.59 nm laser diode -operated in Littrow external cavity (model TEC-100, Sacher Lasertechnik, LLC, Marburg, Germany) to a 852.34 nm diode laser operated in a Littman-Me tcalf external cavity (Model TEC-500, Sacher Lasertechnik, LLC, Marburg, Germany). Both lase r beams were collimat ed and superimposed. An iris was used to restrict the radiation di ameter to a common valu e of 2mm, providing a homogeneous radial intensity dist ribution over the entire cell.The stability of the laser radiation was monitored by PMT detector (R636, Hamama tsu, Japan). The PMT output was amplified (Model 427, Keithley Instruments, Cleveland, OH) and recorded by an oscilloscope (TDS3000 Series, Tektronix, Willsonville, OR). The fluorescence was collected from the input window at an angle by a fiber optic and sent to an Ocean Optics spectrometer (crossed Czer ny-turner monochromator with a CCD detector 107

PAGE 108

calibrated for 770 nm to 1030 nm). The spectra were recorded with different integration times and the results shown were normalized in Counts/second for comparison of different spectra. The resonance fluorescence signal was corrected for the scattering of the incident light which mainly occurred from input window of the cell. An optimum collection angle was experimentally found where the fl uorescence/ scatter ratio is ma ximal, due to the different distributions of the fluorescence and the scatte ring. The remaining scattering was corrected by subtracting the recorded spectrum with the laser detuned from th e cesium transition. Results Figure 6-3 shows the fluorescence spectra meas ured from 852 nm and 894 nm excitation radiations. The fluorescence intensities are derive d from the two spectra and the measurement is repeated several times with different collecti on angles and different wavelengths. The total pressure of the cell is 450 Torr, the main br oadening mechanism is collisional broadening, more precisely van der Waals broadening (cf. chapter 4) When the laser is scanned over the transition, the fluorescence intensities vary but the ratio rema ins constant. The effect of the laser power is shown in Figure 6-4 and confirms that the atomic system was not saturated. The ratio for several temperatures is displayed in Table 6-2 and the calculated values of R12, R21 and the cross section 21 (cm2) are summarized in Table 6-3. The calculate d cross section are in the same order of magnitude as the values obtained by Krause et al. 143 confirming the fast mixing induced by ethane molecules. For the cesium-ethane cell, th e fluorescence efficiency is calculated to be 0.09 (cf. equation 6-5) and is 2 orders of magnitude higher compared to previous RFD. Improvement of the Absorption Efficiency As stressed before, the absorption efficiency of the Raman photons is bound to be low for RFD operating at low pressure. When the cesium absorption profile is pressure broadened, a significant improvement is expected. Figure 6-5 shows the Raman signal and the expected signal 108

PAGE 109

absorbed for a low pressure cell dominated by Doppler broadening at 25 C. Raman lineshape can be described by a Lorentzian profile 133 and a 30 GHz bandwidth (1 cm-1) was chosen for the sake of the argument. The fraction absorb ed was calculated to be approximately 10-4. The pressure of the different buffer gases was selected to give an absorption linewidth similar to that of a Raman signal. In addition of 100 Torr of ethane, 350 Torr of He was added to further broaden the absorption linewidth. The broadening processes already described in chapter 4 are used here to calculate the theo retical linewidth. It should be not ed that, in our experimental conditions, the main source of collision broadeni ng results from the inter action of cesium atoms with helium, ethane and other cesium atoms, i.e. heliumCs ethanetCs CsCs Lv vvv (6-18) The individual contributions to the total Lorentzian width, L, were calculated using the values for the reduced broadening constant = L /n where n is the number density of perturbers. These values are listed below: Cs-Cs = 6.7-7 cm3 s-1 146 Cs-He = 3.610-10 cm3 s-1 147 Cs-ethane = 50 MHz/Torr 147 The values used for the perturber number densities were 2.410 cm-3 148 for the cesium vapor cesium and 3.1319 cm-3 for 350 Torr of helium. Experimental The experimental setup is shown in Figure 66 and the measurements were carried out at room temperature, 25C.The radiation at 852 nm was provided by an external cavity diode laser with manufacturer-specified linewidths of 5 MH z as described in the previous chapter. A waveform provided by a function generator (M ODEL FG3C, Wavetek Meterman) was applied 109

PAGE 110

to a piezoelectric element placed behind the tuning grating in the laser cavity. A portion of the scanned laser beam was directed to a custombuilt confocal Fabry Perot interferometer to monitor the wavelength emitted by the laser. The ra diant intensity transmitted through the cell is directed into a 852 nm interference filter (Opt ometrics, LLC) and detected by a Photomultiplier tube (R636, Hamamatsu) sensitive in the near-infrared. The potential use of this RFD as an image de tector is measured by expanding the beam to the dimension of the cell, then by passing though a transmission target (USAF 1951 3-Bar Resolving Power Test Target). The resulting im age was then passed through a dielectric mirror positioned at 45 that transmit at 852 nm and reflect a fraction of the 894 nm radiation which is sent into the RFD cell previously describe d. The fluorescence photons passing through an interference filter at 894 nm (Optometrics, LLC) were imaged onto the photocathode surface of a single-stage, proximity focused image in tensifier (MODELV807OU-64-N132, Hamamatsu Corporation, Japan). Photons emitted from the phosphor-coated back surface were imaged onto a cooled, digital CCD camera (Mod el Penguin 150CL, Pixera Corpor ation, Los Gatos, CA) with integration times from 0.1 ms to 60 s. Results Because of collisional broadening, the RFD ab sorption bandwidth is estimated to be approximately 15 GHz and verified experimentally; however, as shown in Figure 6-7 a), the mode-hop free range is limited to 25 GHz and a prope r correction for the baseline is difficult as the profile is not resolved. Difficulty with baselin e correction is enhanced by the distortion of the baseline from etalon effects every time the la ser radiation passes thr ough a surface. With a bandwidth of 15 GHz, the fraction absorbed for a Raman signal of 1cm-1 was estimated to be 0.4. This is three orders of magnitude higher than fr om a low pressure RFD. The fraction absorbed is improved at the cost of the imaging performance as shown in Figure 6-7 b) The high collisional 110

PAGE 111

environment blurs the fluorescence images. It should be noted that measurements at higher temperature increases the fraction absorbed as the number density of cesium in the vapor phase increases. Overall Performance of an AVD for Raman Detection The different efficiencies of the cesium/eth ane RFD have been estimated and the number of photons to be detected can be calculated from equation 6-2: t TAN d d ESAVDDAVD AVD Da a p AVD R)( (6-2) We assume a laser photo irradiance, P of 1016 photons s-1 cm-2 (10 mW cw laser), a differential cross section, ad d of 10-29 cm2 sr-1, a number of scattering molecules per unit area, Na, of 1019 molecule cm-2 a luminosity, AD of 0.12 and an optical system transmission factor, T of unity. The detector efficiency for our particular RFD is the product of the measured fluorescence efficiency, 0.09, the fraction absorbed, 0.4 and th e secondary collection/detection efficiency estimated to be 0.08. By combining these values a photon rate of ~100 photon/s needs to be detected by the secondary detector. Several detectors, mainly phot odiodes have the potential to measure such low signal in the ne ar infrared at 894 nm but were not available in our laboratory. Moreover, the estimated photon rate represents the higher ideal limit of detection and an order of magnitude higher should be required to hopefully detect Raman photons. The optical transmission factor was assumed equal to unity and the spectral bandwidth of Raman scatter is more, typically for biological systems, in the order of hundreds of GHz further reducing the absorption efficiency. 111

PAGE 112

Conclusions and Remarks A simple, compact cesium resonant fluorescent detector has been developed. The detection scheme is based on the absorption of the ground state atoms at 852 nm (62S1/262P3/2) and fluorescence decay from the 62P1/2 state. A buffer mixture of ethane and helium improve the detector efficiency by several fold compared to previously reported RFD. Ethane collisionally coupled the 62P3/2-62P1/2 states, very fast mixing rate were measured (mixing across sections in the order of 10-15cm2, and fluorescence efficiency was improved by two orders of magnitude. Helium collisionally broadens the absorption bandwidth and fraction ab sorbed was improved by three orders of magnitude. However, the primar y goal, detecting Raman scattering photons have not been successfully reached. Raman scattering is characterized by a low cross section and large emission spectral linewidth compared to AVD. 112

PAGE 113

Figure 6-1. Schematic of the different rates involved in the 62S1/262P3/2 detection process. The cesium atomic vapor is pumped by a 852 nm laser radiation. In this figure, the radiative absorption and emission rates due to stimulated processes are indicated as Bji where Bji is the Einstein coefficient for stimulated absorption (emission) and the energy density of the laser. 6p2P3/26s2S1/2 852 nm 894 nm 6p2P1/2 R12 R21 2 1 B02 R20 B20 R10 0 113

PAGE 114

Figure 6-2. Experimental setup to measure the fluorescence respons e of an atomic vapor cell. The cell contains 350 Torr of helium and 100 Torr of ethane from a 852 nm and 894 nm diode laser radiation. 114

PAGE 115

a) b) 8408608809000 500 1000 1500 2000 Intensity (a.u)wavelength (nm) 894 nm excitation 8408608809000 500 1000 1500 2000 2500 3000 intensity (a.u)wavelength (nm) 852 nm excitation Figure 6-3. Fluorescence spectra of a cesium-ethane absorption cell. The cell is excited by a) a 852 nm laser radiation and b) a 894 nm laser radiation. The spectra were corrected for scattering. Fluorescence from 852nm excitation (70C)1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E-061.E-041.E-021.E+00Laser Power (mW)Intensity (counts/ms) 852fluo 894fluo Figure 6-4. Effect of the laser power on the fluorescence intensity (Log-Log scale) 115

PAGE 116

116 Figure 6-5. Calculated Raman linewidth and co rresponding signal absorbed. Spectra are shown for a) a low pressure cesium cell and b) a cesium cell containing 100 Torr of ethane and 350 Torr of helium at 50C. 11011 51010051010110110 510 12110 111.510 11210 112.510 11 3010120Rf () glf ()1101111011 f Signal intensity (a.u)frequency -5 5 01 0 10Raman signal Signal absorbed(GhZ) a) 11011 5101005 (GHz)1010110110 510 12110 111.510 11210 112.510 11 3010120Rf () glf ()11011 f11011 Signal intensity (a.u)frequency (GhZ) -5 10Ram Sign (GHz)b) 5 01 0an signal al absorbed

PAGE 117

117Figure 6-6. Experimental setup for measuri ng the spectral absorption lines hape of the RFD and evalua te its imaging performance. 852 nm diode laser Cs cell CCD Camera Beam expanderTransmission maskLens Image Intensifier 894 nm filter LensDichroic Mirror PMT Iris 852nm filter FPI/Photodiode Beam Splitter

PAGE 118

-5051015 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 K.l frequency (GHz) K.l a) b) Figure 6-7. Spectral profile and image of the Cs /ethane RFD at room temperature. Absorbtion spectrum, a) and image of the 852 nm lase r radiation through the transmission mask, b). No signal was measured when the laser was detuned from the transition. 118

PAGE 119

Table 6-1. Values for several detectors of th e luminosity, Resolving power and their product. Spectrometer LR Rmax L(cm2 sr) Interference filter 0.1 10410-5 Grating monochromator 1 5*1052*10-6Fabry-Perot interferometer 100 10610-4 Atomic vapor detector 8*10108*108100 Table 6-2. Fluorescence ratios at different temperature 25 C 45 C 55 C 65 C 852 pump ratio S2/S1 0.288 0.278 0.298 0.298 std dev 0.02 0.02 0.03 0.03 894 pump ratio S2/S1 0.227 0.253 0.279 0.280 std dev 0.02 0.02 0.03 0.03 Table 6-3. Calculated mixing rate s and cross sections of Cs 62P3/2 62P1/2 induced by 100mTorr of ethane T, C R12 (s-1) 0.5 108 R21 (s-1) 0.8 108 21 (cm2) 0.3 10-15 25 3.2 108 8.0 1083.3 10-15 45 4.5 108 9.8 108 4.2 10-15 55 6.4 108 1.2 1095.2 10-15 65 6,7 108 1.4 109 5.9 10-15 119

PAGE 120

CHAPTER 7 FUTURE WORK Future work in the field of atomic vapor photon detection and imaging can be done in the improving the sensitivity and expand ing the range of applications. Atomic vapors have been used as absorptio n or notch-like filters; for example, to discriminate against Rayleigh sc attering in Raman spectroscopy.120 These filters can provide less expensive alternatives to comm ercially available holographic notch filters. In this work, an atomic vapor detector based on cesium resona nce fluorescence has successfully improved the detection efficiency but has not yet reached the sensitivity needed to detect Raman scattered photons. A solution to increase the number of Raman photons is to use Raman enhancing techniques. In the past decades, Surface Enha nced Raman Spectroscopy (SERS) has shown promise to overcome the traditionally low sensitivity of normal Raman spectroscopy. Studies have been made to understand the si gnal enhancement observed when a molecule is attached to various metals such as silver, gold and copper. The signal enhancement has been increased by 14 orders of magnitude ma king single molecule detection possible.149, 150 This signal enhancement would translate into a Raman cross section of approximately 10-16 cm2/molecule. There are two operative mechanisms responsible for the SERS phenomenon, namely electromagnetic field enhancement and ch emical enhancement. The former contributes to the enhanced Raman signal when scattering takes place in the enhanced local optical fields of the metal and the latter is due to metal-molecule interaction. An example of SERS was explored in our laboratory, as a tool to study biological agents.151 Dipicolinic acid is an important biomarker used for detection of bacterial spores and the feasibility of SERS in a continuous flow mode has been evaluated. With the signal enha ncement of SERS, AVD can be exploited as Raman photon detectors. 120

PAGE 121

Bakker et al. have proposed another applica tion for the use of at omic vapor absorption filters in the field of plasma diagnostics, na mely in the detection of Thomson scattering.152 There is significance to the determination of number density (ne) and temperature (Te) of free electrons in a plasma as a function of time and space. These measurements can provide information concerning the details of energy transfer and departures from equilibrium. Electron number density also play a role in excitation and io nization mechanisms as well as plasma viscosity, electrical and th ermal conductivity.153 A common method of monitoring electron number density in an alytical plasmas uses Stark broadening of the absorption or em ission lines of the plasma gas (cf. Review of Broadening in chapter 4, equation 4-12). The method, as discussed in chapter 4, is complicated by the need to distinguish Doppler broadening and the instrume ntal response function. Electron number density and temperature have also been determined by ab solute line intensity m easurements of two or more of the plasma gas emission lines. In this method, the determination of ne and Te are dependent so errors in one results in error in the other. Emission in th is technique can only be laterally resolved and plasma symmetry must be assumed which is not always the case. Langmuir probes have been used in plasma diagnostics. In genera l, these probes are relatively inexpensive and provide adequate spatial re solution. In practice, however, they can only be applied to low temperature, low pressure plasmas and the probe is likely to disturb the plasma during measurements. Sheath formation around the probe, probe heating, and probe erosion have also hindered this technique.153 J.J Thomson presented the theory for the sc attering of electrom agnetic radiation by electrons in 1907.154 The phenomenon, however, was not stud ied in the laborat ory until decades later with the advent of intens e laser sources capable of produc ing appreciable scattering from 121

PAGE 122

high electron number density plasmas. Thomson s cattering was then used in the study of high temperature plasmas used in nuclear fusion res earch and later applied to processing plasmas. Hiefte et al. 155 have since applied the techniques to st udy analytically relevant plasmas including microwave plasmas and the inductiv ely coupled plasma or ICP. As J.J. Thomson theorized, when an incident la ser beam interacts with a charged particle in a plasma, the oscillating field of the radiation in duces the charge to oscillate in resonance with the field. The oscillating charge radiates light, a phenomenon al so known as Thomson scattering. The wavelength of this scattered ra diation that reaches a detector is double Doppler-shifted as the radiation is moving rapidly with respect to both the source and the detector. In general, the width of the Thomson scattering spectral profile is related to the velocity and therefore Te while the area of the spectrum is related to the ne.153 Detecting Thomson scattering requires exce llent stray light re jection and spectral discrimination; the small cross section of this event (6.7x10-25cm2) results in a weak signal as compared to the intensity of the source required to produce the scatter. Th e optical system used to collect the scattered light is usually a double or triple monochromator. However, the use of these low throughput systems is not useful in the detect ion of weak Thomson scattering signals from low density plasmas. An alternative method has been described for the detection of weak Thomson signals. Bakker et al. have demonstrated the use of a sodium va por absorption cell or atomic notch filter placed between the plasma and the detector.152 Scattered light was collected and directed through the vapor wh ich absorbed the stray light and Rayleigh scattering at the incident wavelength. This approach allowed broadened Thomson scatter to be transmitted and detected by conventional means. The main curren t constraints of this application are the high laser power necessary (hundreds of mJ) and plasma with relative high electron number density. 122

PAGE 123

Laser induced plasmas (ne>1016cm-3) can be investigated with Nd-YAG laser with 200mJ/pulse as excitation source. The developments of atom ic vapor-based devices will certainly aid the application of Thoms on scattering detection. 123

PAGE 124

LIST OF REFERENCES 1. Schnurer-Patschan, C.; Niemax, K., Plenary lecture, J. Anal. At. Spectrom 1993, 8, 1103. 2. Krupa, R. J.; Long, G. L.; Winefordner, J. D., An Icp-Excited Icp Resonance Monochromator and Fluorescence Spectrometer fo r the Analysis of Trace to Major Sample Constituents. Spectrochimica Acta Part B-Atomic Spectroscopy 1985, 40, (10-12), 1485-1494. 3. Brech, F.; Cross, L., Appl. Spectrosc 1962, 16, 59. 4. Debrasguedon, J.; Liodec, N., Spectroscopie De Lutilisation Du Faisceau Issu Dun Amplificateur a Ondes Lumineuses Par Emissi on Induite De Rayonnement (Laser a Rubis), Comme Source Energetique Pour Lexcitati on Des Spectres Demission Des Elements. Comptes Rendus Hebdomadaires Des Seances De L Academie Des Sciences 1963, 257, (22), 3336-&. 5. Maker, P. D.; Terhune, R. W.; Savage, C. M., Optical third harmonic generation Columbia University Press, New York: Paris, 1964; Vol. 2, p 1559. 6. Miziolek, A. W.; Palleschi, V.; Schechter, I., Laser-Induced Breakdown Spectroscopy Cambridge University Press: 2006. 7. Winefordner, J. D.; Gornushkin, I. B.; Correll, T.; Gibb, E.; Smith, B. W.; Omenetto, N., Comparing several atomic spectrometric methods to the super stars: special emphasis on laser induced breakdown spectrometry, LIBS, a future super star. Journal of Analytical Atomic Spectrometry 2004, 19, (9), 1061-1083. 8. Radziemski, L. J., Review of Selected An alytical Applications of Laser Plasmas and Laser-Ablation, 1987-1994. Microchemical Journal 1994, 50, (3), 218-234. 9. Moenke-Blankenberg, L., Laser Microanalysis John Wiley: New York, 1989. 10. Weyl, G. M., Laser-Induced Plasmas and Applications Marcel Dekker: New York, 1989. 11. Idris, N.; Terai, S.; Lie, T. J.; Kurniawan, H.; Kobayashi, T.; Maruyama, T.; Kagawa, K., Atomic hydrogen emission induced by TEA CO2 laser bombardment on solid samples at low pressure and its analytical application. Applied Spectroscopy 2005, 59, (1), 115-120. 12. Rieger, G. W.; Taschuk, M.; Tsui, Y. Y. ; Fedosejevs, R., Laser-induced breakdown spectroscopy for microanalysis using submillijoule UV laser pulses. Applied Spectroscopy 2002, 56, (6), 689-698. 13. Garcia, C. C.; Corral, M.; Vadillo, J. M.; Laserna, J. J., Angleresolved laser-induced breakdown spectrometry for depth profiling of coated materials. Applied Spectroscopy 2000, 54, (7), 1027-1031. 124

PAGE 125

14. Georgiou, S.; Zafiropulos, V.; Anglos, D.; Bala s, C.; Tornari, V.; Fotakis, C., Excimer laser restoration of painted artworks: procedures, mechanisms and effects. Applied Surface Science 1998, 129, 738-745. 15. Gunther, D.; Heinrich, C. A., Comparison of the ablation behaviour of 266 nm Nd : YAG and 193 nm ArF excimer lasers for LA-ICP-MS analysis. Journal of Analytical Atomic Spectrometry 1999, 14, (9), 1369-1374. 16. Pouli, P.; Melessanaki, K.; Giakoumaki, A.; Argyropoulos, V.; Anglos, D., Measuring the thickness of protective coatings on histor ic metal objects using nanosecond and femtosecond laser induced breakdown spectroscopy depth profiling. Spectrochimica Acta Part B-Atomic Spectroscopy 2005, 60, (7-8), 1163-1171. 17. Pinnick, R. G.; Chylek, P.; Jarzembski, M.; Creegan, E.; Srivastava, V.; Fernandez, G.; Pendleton, J. D.; Biswas, A., Aerosol-Induced Laser Breakdown Thresholds Wavelength Dependence. Applied Optics 1988, 27, (5), 987-996. 18. Costela, A.; Garcia-Moreno, I.; Gomez, C.; Caballero, O.; Sastre, R., Cleaning graffitis on urban buildings by use of second and third harmonic wavelength of a Nd : YAG laser: a comparative study. Applied Surface Science 2003, 207, (1-4), 86-99. 19. Menut, D.; Fichet, P.; Lacour, J. L.; Rivoa llan, A.; Mauchien, P., Micro-laser-induced breakdown spectroscopy technique: a powerful method for performing quantitative surface mapping on conductive and nonconductive samples. Applied Optics 2003, 42, (30), 6063-6071. 20. Gunaratne, T.; Kangas, M.; Singh, S.; Gross, A.; Dantus, M., Influence of bandwidth and phase shaping on laser induced breakdown sp ectroscopy with ultrashort laser pulses. Chemical Physics Letters 2006, 423, (1-3), 197-201. 21. Fernandez, A.; Mao, X. L.; Chan, W. T.; Shannon, M. A.; Russo, R. E., Correlation of Spectral Emission Intensity in the Inductively-C oupled Plasma and Laser-Induced Plasma during Laser-Ablation of Solid Samples. Analytical Chemistry 1995, 67, (14), 2444-2450. 22. Le Drogoff, B.; Margot, J.; Chaker, M.; Sa bsabi, M.; Barthelemy, O.; Johnston, T. W.; Laville, S.; Vidal, F.; von Kaenel, Y., Tempor al characterization of femtosecond laser pulses induced plasma for spectrochemical analysis of aluminum alloys. Spectrochimica Acta Part BAtomic Spectroscopy 2001, 56, (6), 987-1002. 23. Kumagai, H.; Midorikawa, K.; Toyoda, K.; Nakamura, S.; Okamoto, T.; Obara, M., Ablation of Polymer-Films by a Femtosecond HighPeak-Power Ti Sapphire Laser at 798-Nm. Applied Physics Letters 1994, 65, (14), 1850-1852. 24. Angel, S. M.; Stratis, D. N.; Eland, K. L.; Lai, T. S.; Berg, M. A.; Gold, D. M., LIBS using dualand ultrashort laser pulses. Fresenius Journal of Analytical Chemistry 2001, 369, (34), 320-327. 125

PAGE 126

25. De Giacomo, A.; Dell'Aglio, M.; Santagat a, A.; Teghil, R., Early stage emission spectroscopy study of meta llic titanium plasma induced in air by femtosecondand nano secondlaser pulses. Spectrochimica Acta Part B-Atomic Spectroscopy 2005, 60, (7-8), 935-947. 26. Galbacs, G.; Budavari, V.; Geretovszky, Z., Multi-pulse la ser-induced plasma spectroscopy using a single laser source and a compact spectrometer. Journal of Analytical Atomic Spectrometry 2005, 20, (9), 974-980. 27. Gautier, C.; Fichet, P.; Menut, D.; Lacour, J. L.; L'Hermite, D.; Dubessy, J., Study of the double-pulse setup with an orthogonal beam geometry for laser-induced breakdown spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, (7), 975-986. 28. Sobral, H.; Villagran-Muniz, M.; Navarro-Gonz alez, R.; Raga, A. C., Temporal evolution of the shock wave and hot core air in laser induced plasma. Applied Physics Letters 2000, 77, (20), 3158-3160. 29. Scaffidi, J.; Angel, S. M.; Cremers, D. A., Emission enhancement mechanisms in dualpulse LIBS. Analytical Chemistry 2006, 78, (1), 24-32. 30. Colao, F.; Lazic, V.; Fantoni, R.; Pershi n, S., A comparison of single and double pulse laser-induced breakdown spectrosc opy of aluminum samples. Spectrochimica Acta Part BAtomic Spectroscopy 2002, 57, (7), 1167-1179. 31. Lindblom, P., New compact Echelle spectro graphs with multichannel time-resolved recording capabilities. Analytica Chimica Acta 1999, 380, (2-3), 353-361. 32. Bauer, H. E.; Leis, F.; Niemax, K., La ser induced breakdown sp ectrometry with an echelle spectrometer and intensified charge coupled device detection. Spectrochimica Acta Part B-Atomic Spectroscopy 1998, 53, (13), 1815-1825. 33. Baudelet, M.; Boueri, M.; Yu, J.; Mao, S.; Piseltelli, V.; Mao, X.; Russo, R. E., Spectrochimica Acta Part B-Atomic Spectroscopy 2007, 62, (12), 1329-1334. 34. Cravetchi, I. V.; Taschuk, M.; Tsui, Y. Y.; Fedosejevs, R., Scanning microanalysis of Al alloys by laser-induced breakdown spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, (9), 1439-1450. 35. Loebe, K.; Uhl, A.; Lucht, H., Microanalysis of tool steel and gl ass with laser-induced breakdown spectroscopy. Applied Optics 2003, 42, (30), 6166-6173. 36. Gruber, J.; Heitz, J.; Strasser, H.; Bauerle, D.; Ramaseder, N., Rapi d in-situ analysis of liquid steel by laser-induced breakdown spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 2001, 56, (6), 685-693. 37. Vieitez, M. O.; Hedberg, J.; Launila, O.; Ber g, L. E., Elemental analysis of steel scrap metals and minerals by laserinduced breakdow n spectroscopy. Spectrochimica Acta Part BAtomic Spectroscopy 2005, 60, (7-8), 920-925. 126

PAGE 127

38. Lacroix, D.; Jeandel, G., Spectroscopic char acterization of laser-induced plasma created during welding with a pulsed Nd:YAG laser. Journal of Applied Physics 1997, 81, (10), 65996606. 39. Endert, H.; Patzel, R.; Powell, M.; Rebha n, U.; Basting, D., New Krf and Arf ExcimerLaser for Advanced Duv Lithography. Microelectronic Engineering 1995, 27, (1-4), 221-224. 40. Yamamoto, K. Y.; Cremers, D. A.; Ferris, M. J.; Foster, L. E., Detect ion of metals in the environment using a portable laser-in duced breakdown spectroscopy instrument. Applied Spectroscopy 1996, 50, (2), 222-233. 41. Moskal, T. M.; Hahn, D. W., On-line sort ing of wood treated with chromated copper arsenate using laser-induced breakdown spectroscopy. Applied Spectroscopy 2002, 56, (10), 1337-1344. 42. Eppler, A. S.; Cremers, D. A.; Hickmott, D. D.; Ferris, M. J.; Koskelo, A. C., Matrix effects in the detection of Pb and Ba in soils using lase r-induced breakdown spectroscopy. Applied Spectroscopy 1996, 50, (9), 1175-1181. 43. Capitelli, F.; Colao, F.; Provenzano, M. R. ; Fantoni, R.; Brunetti, G.; Senesi, N., Determination of heavy metals in so ils by laser induced breakdown spectroscopy. Geoderma 2002, 106, (1-2), 45-62. 44. Cremers, D. A.; Ebinger, M. H.; Breshears, D. D.; Unkefer, P. J.; Kammerdiener, S. A.; Ferris, M. J.; Catlett, K. M.; Brown, J. R., Measuring total soil carbon with laser-induced breakdown spectroscopy (LIBS). Journal of Environmental Quality 2001, 30, (6), 2202-2206. 45. Vanleeuwen, T. G.; Vanderveen, M. J.; Verdaasdonk, R. M.; Borst, C., Noncontact Tissue Ablation by Holmium Ysgg Laser-Pulses in Blood. Lasers in Surgery and Medicine 1991, 11, (1), 26-34. 46. Corsi, M.; Cristoforetti, G.; Hidalgo, M. ; Legnaioli, S.; Palleschi, V.; Salvetti, A.; Tognoni, E.; Vallebona, C., Application of laser-induced breakdown spect roscopy technique to hair tissue mineral analysis. Applied Optics 2003, 42, (30), 6133-6137. 47. Gieray, R. A.; Reilly, P. T. A.; Yang, M.; Whitten, W. B.; Ramsey, J. M., Real-time detection of individual airborne bacteria. Journal of Microbiological Methods 1997, 29, (3), 191199. 48. St-Onge, L.; Kwong, E.; Sabsabi, M.; Vada s, E. B., Quantitative analysis of pharmaceutical products by laser-induced breakdown spectroscopy. Spectrochimica Acta Part BAtomic Spectroscopy 2002, 57, (7), 1131-1140. 49. Anzano, J.; Casanova, M. E.; Bermudez, M. S.; Lasheras, R. J., Rapi d characterization of plastics using laser-induced plasma spectroscopy (LIPS). Polymer Testing 2006, 25, (5), 623627. 127

PAGE 128

50. Marshall, J.; Franks, J.; Abell, I.; Tye, C., Determination of Trace-Elements in Solid Plastic Materials by Laser Ablation Inductiv ely Coupled Plasma Mass-Spectrometry. Journal of Analytical Atomic Spectrometry 1991, 6, (2), 145-150. 51. Fichet, P.; Mauchien, P.; Wagner, J. F.; M oulin, C., Quantitative elemental determination in water and oil by laser induced breakdown spectroscopy. Analytica Chimica Acta 2001, 429, (2), 269-278. 52. Kuzuya, M.; Murakami, M.; Maruyama, N., Quantitative analysis of ceramics by laserinduced breakdown spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 2003, 58, (5), 957-965. 53. Ciucci, A.; Corsi, M.; Palleschi, V.; Rastel li, S.; Salvetti, A.; Tognoni, E., New procedure for quantitative elemental analysis by laser-induced plasma spectroscopy. Applied Spectroscopy 1999, 53, (8), 960-964. 54. Fink, H.; Gentsch, D.; Heimbach, M., Condensed metal vapor on alumina ceramic in vacuum interrupters. Ieee Transactions on Dielectrics and Electrical Insulation 2002, 9, (2), 201-206. 55. Amador-Hernandez, J.; Fernandez-Romero, J. M.; de Castro, M. D. L., In-depth characterization of screen-printed electrode s by laser-induced breakdown spectrometry and pattern recognition. Surface and Interface Analysis 2001, 31, (4), 313-320. 56. Sirven, J. B.; Bousquet, B.; Canioni, L.; Sa rger, L.; Tellier, S.; Potin-Gautier, M.; Le Hecho, I., Qualitative and quantit ative investigation of chromium -polluted soils by laser-induced breakdown spectroscopy combined with neural networks analysis. Analytical and Bioanalytical Chemistry 2006, 385, (2), 256-262. 57. Anderson, D. R.; Mcleod, C. W.; English, T. ; Smith, A. T., Depth Profile Studies Using Laser-Induced Plasma Emission-Spectrometry. Applied Spectroscopy 1995, 49, (6), 691-701. 58. Melessanaki, K.; Mateo, M.; Ferrence, S. C.; Betancourt, P. P.; Anglos, D., The application of LIBS for the analysis of archaeological ceramic and metal artifacts. Applied Surface Science 2002, 197, 156-163. 59. Brysbaert, A.; Melessanaki, K.; Anglos, D., Pigment analysis in Bronze Age Aegean and Eastern Mediterranean painte d plaster by laser-induced breakdown spectroscopy (LIBS). Journal of Archaeological Science 2006, 33, (8), 1095-1104. 60. Lopez-Moreno, C.; Palanco, S.; Laserna, J. J ., Remote laser-induced plasma spectrometry for elemental analysis of samples of environmental interest. Journal of Analytical Atomic Spectrometry 2004, 19, (11), 1479-1484. 61. Wen, S. B.; Mao, X. L.; Greif, R.; Russo, R. F., Radiative cooling of laser ablated vapor plumes: Experimental an d theoretical analyses. Journal of Applied Physics 2006, 100, (5), -. 128

PAGE 129

62. Gornushkin, I. B.; Kazakov, A. Y.; Omenetto, N.; Smith, B. W.; Winefordner, J. D., Radiation dynamics of post-br eakdown laser induced plasma. Spectrochimica Acta Part BAtomic Spectroscopy 2004, 59, (4), 401-418. 63. Bakowski, B.; Hancock, G.; Peverall, R.; Ritchie, G. A. D., Number density and temperature measurements obtained using sensitiv e diode laser spectroscopy in an argon plasma. Applied Physics B-Lasers and Optics 2006, 82, (1), 123-131. 64. Beverini, N.; DelGobbo, G.; Genovesi, G. L.; Maccarrone, F.; Strumia, F.; Paganucci, F.; Turco, A.; Andrenucci, M., Time-resolved pl asma diagnostic by laser-diode spectroscopy. Ieee Journal of Quantum Electronics 1996, 32, (11), 1874-1881. 65. Baer, D. S.; Hanson, R. K., Tunable Diode-Laser Absorption Diagnostics for Atmospheric-Pressure Plasmas. Journal of Quantitative Spectroscopy & Radiative Transfer 1992, 47, (6), 455-475. 66. Baer, D. S.; Chang, H. A.; Hanson, R. K., Fluorescence Diagnostics for AtmosphericPressure Plasmas Using Semiconductor-Lasers. Journal of the Optical Society of America BOptical Physics 1992, 9, (11), 1968-1978. 67. Radziemski, L. J.; Cremers, D. A., :aserInduced Plasmas and Applications. In Marcel Dekker, Inc: New York and Basel, 1989. 68. Sdorra, W.; Niemax, K., Temporal and Spatia l-Distribution of Analyte Atoms and Ions in Microplasmas Produced by Laser Ablation of Solid Samples. Spectrochimica Acta Part BAtomic Spectroscopy 1990, 45, (8), 917-926. 69. Muraoka, K.; Maeda, M., Laser-Aided Diagnostics of Plasmas and Gases Institute of Physics Publishing: Bristo and Philadelphia, 1999. 70. Mitchell, A. C.; Zemanski, M. W., Resonance radiation and excited atoms Cambridge University Press: 1961. 71. Xu, L.; Bulatov, V.; Gridin, V. V.; Schechte r, I., Absolute analysis of particulate materials by laser-induced breakdown spectroscopy. Analytical Chemistry 1997, 69, (11), 21032108. 72. Schechter, I., Correction for Nonlinea r Fluctuating Background in Monovariable Analytical Systems. Analytical Chemistry 1995, 67, (15), 2580-2585. 73. Bulatov, V.; Xu, L.; Schechter, I., Spect roscopic imaging of laser-induced plasma. Analytical Chemistry 1996, 68, (17), 2966-2973. 74. Mateo, M. P.; Palanco, S.; Vadillo, J. M.; Laserna, J. J., Fast atomic mapping of heterogeneous surfaces using microline-im aging laser-induced breakdown spectrometry. Applied Spectroscopy 2000, 54, (10), 1429-1434. 129

PAGE 130

75. Scott, R. H.; Strashei.A, Laser Induced Plasmas for Analytical Spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 1970, B 25, (7), 311-&. 76. Geohegan, D. B.; Puretzky, A. A., Laser ablation plume thermalization dynamics in background gases: Combined imaging, optical absorption and emission spectroscopy, and ion probe measurements. Applied Surface Science 1996, 96-8, 131-138. 77. Hart, L. P.; Smith, B. W.; Omenetto, N., Laser-Induced Stepwise and 2-Photon Ionization Studies of Strontium in the Air Acetylene Flame. Spectrochimica Acta Part B-Atomic Spectroscopy 1985, 40, (10-12), 1637-1649. 78. Pichahchy, A. E.; Cremers, D. A.; Ferris, M. J., Elemental analysis of metals under water using laser-induced breakdown spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 1997, 52, (1), 25-39. 79. Gupta, A.; Braren, B.; Casey, K. G.; Husse y, B. W.; Kelly, R., Direct Imaging of the Fragments Produced during Excimer Laser Ablation of Yba2cu3o7-Delta. Applied Physics Letters 1991, 59, (11), 1302-1304. 80. AlWazzan, R. A.; Hendron, J. M.; Morrow, T., Line shape study of Ba ions in laser produced plasmas. Applied Surface Science 1996, 99, (4), 345-351. 81. Corsi, M.; Cristoforetti, G.; Hidalgo, M.; Iria rte, D.; Legnaioli, S.; Palleschi, V.; Salvetti, A.; Tognoni, E., Temporal and spatial evolution of a laser-induced plasma from a steel target. Applied Spectroscopy 2003, 57, (6), 715-721. 82. Wen, S. B.; Mao, X. L.; Greif, R.; Russo, R. E., Laser ablation induced vapor plume expansion into a background gas. II. Experimental analysis. Journal of Applied Physics 2007, 101, (2), -. 83. Aragon, C.; Penalba, F.; Aguilera, J. A., Spatial characterizati on of laser-induced plasmas: distributions of neut ral atom and ion densities. Applied Physics a-Materials Science & Processing 2004, 79, (4-6), 1145-1148. 84. AlWazzan, R. A.; Hendron, J. M.; Morrow, T., Spatially and temporally resolved emission intensities and number densities in lo w temperature laser-induced plasmas in vacuum and in ambient gases. Applied Surface Science 1996, 96-8, 170-174. 85. Thareja, R. K.; Misra, A.; Franklin, S. R., Investigation of la ser ablated metal and polymer plasmas in ambient gas using fast photography. Spectrochimica Acta Part B-Atomic Spectroscopy 1998, 53, (14), 1919-1930. 86. Yalcin, S.; Tsui, Y. Y.; Fedosejevs, R., Pr essure dependence of emission intensity in femtosecond laser-induced breakdown spectroscopy. Journal of Analytical Atomic Spectrometry 2004, 19, (10), 1295-1301. 130

PAGE 131

87. Ho, J. R.; Grigoropoulos, C. P.; Humphre y, J. A. C., Computational Study of HeatTransfer and Gas-Dynamics in the Pulsed-Laser Evaporation of Metals. Journal of Applied Physics 1995, 78, (7), 4696-4709. 88. Freiwald, D. A., Approximate Blast Wave Theory and Experimental-Data for Shock Trajectories in Linear Expl osive-Driven Shock-Tubes. Journal of Applied Physics 1972, 43, (5), 2224-&. 89. Hauer, M.; Funk, D. J.; Lippert, T.; Woka un, A., Time resolved study of the laser ablation induced shockwave. Thin Solid Films 2004, 453-54, 584-588. 90. Geohegan, D. B., Fast Intensified-Ccd Photography of Yba2cu3o7-X Laser Ablation in Vacuum and Ambient Oxygen. Applied Physics Letters 1992, 60, (22), 2732-2734. 91. Garrelie, F.; Catherinot, A., Monte Carlo simulation of the laser-induced plasma-plume expansion under vacuum and with a background gas. Applied Surface Science 1999, 139, 97101. 92. Garrelie, F.; Champeaux, C.; Catherinot, A., Study by a Monte Carlo simulation of the influence of a background gas on the expansion dynamics of a laser-induced plasma plume. Applied Physics a-Materials Science & Processing 1999, 69, (1), 45-50. 93. Harilal, S. S.; O'Shay, B.; Tao, Y. Z.; Tillack, M. S., Ambient gas effects on the dynamics of laser-produced tin plume expansion. Journal of Applied Physics 2006, 99, (8), -. 94. Aguilera, J. A.; Bengoechea, J.; Aragon, C., Spatial characterization of laser induced plasmas obtained in air and argon with different laser focusing distances. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, (4), 461-469. 95. Multari, R. A.; Foster, L. E.; Cremers, D. A. ; Ferris, M. J., Effect of sampling geometry on elemental emissions in laser-induced breakdown spectroscopy. Applied Spectroscopy 1996, 50, (12), 1483-1499. 96. Ingle, J. D.; Crouch, S. R., Spectrochemical Analysis Prentice-Hall, 1988. 97. Kluczynski, P.; Gustafsson, J.; Lindberg, A. M.; Axner, O., Wavelength modulation absorption spectrometry an extensive scrutiny of the generation of signals. Spectrochimica Acta Part B-Atomic Spectroscopy 2001, 56, (8), 1277-1354. 98. Gornushkin, I. B.; Anzano, J. M.; King, L. A. ; Smith, B. W.; Omenetto, N.; Winefordner, J. D., Curve of growth methodology applied to laser-induced plasma emission spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 1999, 54, (3-4), 491-503. 99. Gornushkin, I. B.; King, L. A.; Smith, B. W. ; Omenetto, N.; Winefordner, J. D., Line broadening mechanisms in the low pressure laser-induced plasma. Spectrochimica Acta Part BAtomic Spectroscopy 1999, 54, (8), 1207-1217. 131

PAGE 132

100. Aragon, C.; Bengoechea, J.; Aguilera, J. A., Influence of the optical depth on spectral line emission from laser-induced plasmas. Spectrochimica Acta Part B-Atomic Spectroscopy 2001, 56, (6), 619-628. 101. Castle, B. C.; Visser, K.; Smith, B. W.; Winefordner, J. D., Spatial and temporal dependence of lead emission in laser-induced breakdown spectroscopy. Applied Spectroscopy 1997, 51, (7), 1017-1024. 102. Allemand, C. D., Spectroscopy of Single-Spike Laser-Generated Plasmas. Spectrochimica Acta Part B-Atomic Spectroscopy 1972, B 27, (5), 185-&. 103. Nemet, B.; Kozma, L., Basic Investigat ions of Nanosecond Laser-Induced Plasma Emission Kinetics for Quantitative Elem ental Microanalysis of High Alloys. Journal of Analytical Atomic Spectrometry 1995, 10, (9), 631-636. 104. Hermann, J.; Boulmer-Leborgne, C.; Hong, D., Diagnostics of the early phase of an ultraviolet laser induced plasma by spectral line analysis consid ering self-absorption. Journal of Applied Physics 1998, 83, (2), 691-696. 105. Horvatic, V.; Corrella, T. L.; Omenetto, N.; Va dla, C.; Winefordner, J. D., The effects of saturation and velocity selectiv e population in two-step 6S(1 /2)-> 6P(3/2)-> 6D(5/2) laser excitation in cesium. Spectrochimica Acta Part B-Atomic Spectroscopy 2006, 61, (12), 12601269. 106. King, L. A.; Gornushkin, I. B.; Pappas, D.; Smith, B. W.; Winefordner, J. D., Rubidium isotope measurements in solid samples by lase r ablation-laser atomic absorption spectroscopy. Spectrochimica Acta Part B-Atomic Spectroscopy 1999, 54, (13), 1771-1781. 107. Smith, B. W.; Quentmeier, A.; Bolshov, M.; Niemax, K., Measurement of uranium isotope ratios in solid samples using laser ablation and diode la ser-excited atomic fluorescence spectrometry. Spectrochimica Acta Part B-Atomic Spectroscopy 1999, 54, (6), 943-958. 108. Schnurerpatschan, C.; Zybin, A.; Groll, H. ; Niemax, K., Improvement in Detection Limits in Graphite-Furnace Diode-Laser At omic-Absorption Spectrometry by Wavelength Modulation Technique Plenary Lecture. Journal of Analytical Atomic Spectrometry 1993, 8, (8), 1103-1107. 109. Alkamade, C. T.; Hollander, T. ; Snelleman, W.; Zeegers, T., Metal Vapours in Flame Oxford, 1982. 110. Griem, H. R., Plasma Spectroscopy McGraw-Hill: New-York, 1964. 111. Thorne, A., Spectrophysics Springler-Verlag: New-York, 1999. 112. Vaughan, J. M., The Fabry-Perot Interferometer Adam Hilger: Philadelphia, 1989. 113. Demtrder, W., Laser Spectroscopy. Springler-Verlag: New-York, 1996. 132

PAGE 133

114. Vadla, C.; Horvatic, V.; Niemax, K., Radia tive transport and coll isional transfer of excitation energy in Cs vapor s mixed with Ar or He. Spectrochimica Acta Part B-Atomic Spectroscopy 2003, 58, (7), 1235-1277. 115. Kelly, R.; Dreyfus, R. W., On the Effect of Knudsen-Layer Formation on Studies of Vaporization, Sputtering, and Desorption. Surface Science 1988, 198, (1-2), 263-276. 116. Bushaw, B. A.; Alexander, M. L., Inves tigation of laser abla tion plume dynamics by high-resolution time-re solved atomic absorption spectroscopy. Applied Surface Science 1998, 129, 935-940. 117. Wagenaar, H. C. The influence of spectral li ne profiles upon Analytic al curves in atomic absorption spectrometry. Technische Hogeschool, Delft, Netherland, 1976. 118. Morris, H. R.; Hoyt, C. C.; Miller, P.; Tr eado, P. J., Liquid crysta l tunable filter Raman chemical imaging. Applied Spectroscopy 1996, 50, (6), 805-811. 119. Goetz, A. F. H.; Vane, G.; Solomon, J. E.; Rock, B. N., Imaging Spectrometry for Earth Remote-Sensing. Science 1985, 228, (4704), 1147-1153. 120. Guenard, R. D.; King, L. A.; Smith, B. W.; Winefordner, J. D., Two-channel sequential single molecule measurement. Analytical Chemistry 1997, 69, (13), 2426-2433. 121. Matveev, O. I.; Smith, B. W.; Winefordner, J. D., Resonance ioniza tion image detectors: basic characteristics and potential applications. Applied Optics 1997, 36, (34), 8833-8843. 122. Matveev, O. I.; Smith, B. W.; Winefordner, J. D., Two-dimensional resonance ionization imaging with a low pressure mercury atom vapor cell. Optics Communications 1998, 156, (4-6), 259-263. 123. Podshivalov, A. A.; Shepard, M. R.; Matvee v, O. I.; Smith, B. W.; Winefordner, J. D., Ultrahigh-resolution, frequency-resolved resona nce fluorescence imaging with a monoisotopic mercury atom cell. Journal of Applied Physics 1999, 86, (10), 5337-5341. 124. Pappas, D.; Pixley, N. C.; Matveev, O. I.; Smith, B. W.; Winefordner, J. D., A cesium resonance fluorescence imaging monochromator. Optics Communications 2001, 191, (3-6), 263269. 125. Temirov, J. P.; Chigarev, N. V.; Matveev, O. I.; Omenetto, N.; Smith, B. W.; Winefordner, J. D., A resonance ionization im aging detector based on cesium atomic vapor. Spectrochimica Acta Part B-Atomic Spectroscopy 2004, 59, (5), 677-687. 126. Podshivalov, A. A.; Matveev, O. I.; Smith, B. W.; Winefordner, J. D., A novel and efficient excitationand ioni zation-scheme for laser reso nance ionization of mercury. Spectrochimica Acta Part B-Atomic Spectroscopy 1999, 54, (13), 1793-1799. 133

PAGE 134

127. Pappas, D.; Matveev, O. I.; Smith, B. W.; Shepard, M. R.; Podshivalov, A. A.; Winefordner, J. D., Sealed-cell mercur y resonance ionizati on imaging detector. Applied Optics 2000, 39, (27), 4911-4917. 128. Korevaar, E.; Rivers, M.; Liu, C. S., SPIE Space sensing, Communications and Networking 1989, 1059, 111-118. 129. Morris, H. R.; Hoyt, C. C.; Treado, P. J., Imaging Spectrometers for Fluorescence and Raman Microscopy Acoustooptic and Liquid-Crystal Tunable Filters. Applied Spectroscopy 1994, 48, (7), 857-866. 130. Correll, T. L.; Horvatic, V.; Omenetto, N. ; Vadla, C.; Winefordner, J. D., Quantum efficiency improvement of a cesium based re sonance fluorescence detector by helium-induced collisional excitation energy transfer. Spectrochimica Acta Part B-Atomic Spectroscopy 2005, 60, (6), 765-774. 131. Pixley, N. C.; Correll, T. L.; Pappas, D.; Om enetto, N.; Smith, B. W.; Winefordner, J. D., Moving object detection using a cesium resonance fluorescence monochromator. Optics Communications 2003, 219, (1-6), 27-31. 132. Pappas, D.; Correll, T. L.; Pixley, N. C.; Sm ith, B. W.; Winefordner, J. D., Detection of Mie scattering using a resonance fluorescence monochromator. Applied Spectroscopy 2002, 56, (9), 1237-1240. 133. Lewis, I. R.; Edwards, G. M., Handbook of Raman Spectroscopy Marcel Dekker Inc: 2001. 134. Laserna, J. J., Modern techniques in Raman spectroscopy John Wiley & Sons: 1996. 135. Lyon, L. A.; Keating, C. D.; Fox, A. P.; Baker, B. E.; He, L.; Nicewarner, S. R.; Mulvaney, S. P.; Natan, M. J., Raman spectroscopy. Analytical Chemistry 1998, 70, (12), 341r361r. 136. Petry, R.; Schmitt, M.; Popp, J., Raman Spect roscopy A prospective tool in the life sciences. Chemphyschem 2003, 4, (1), 14-30. 137. Kaminaka, S.; Ito, T.; Yamazaki, H.; Kohda, E.; Hamaguchi, H., Near-infrared multichannel Raman spectroscopy toward real-time in vivo cancer diagnosis. Journal of Raman Spectroscopy 2002, 33, (7), 498-502. 138. Harris, C., Anal. Chem. 2003, 75A-78A. 139. Smith, B. W.; Farnsworth, P. B.; Winefor dner, J. D.; Omenetto, N., Experimental Demonstration of a Raman-Scattering De tector Based on Laser-Enhanced Ionization. Optics Letters 1990, 15, (14), 823-825. 134

PAGE 135

140. Petrucci, G. A.; Badini, R. G.; Winefordne r, J. D., Photon Detection Based on Pulsed Laser-Enhanced Ionization and Photoionizati on of Magnesium Vapor Experimental Characterization. Journal of Analytical Atomic Spectrometry 1992, 7, (3), 481-491. 141. Indralingam, R.; Simeonsson, J. B.; Petrucci G. A.; Smith, B. W.; Winefordner, J. D., Raman-Spectrometry with Metal Vapor Filters. Analytical Chemistry 1992, 64, (8), 964-967. 142. Beach, R. J.; Krupke, W. E.; Kanz, V. K.; Payne S. A.; Dubinskii, M. A.; Merkle, L. D., End-pumped continuous-wave alka li vapor lasers: experiment model, and power scaling. Journal of the Optical Society of America B-Optical Physics 2004, 21, (12), 2151-2163. 143. Walentyn.E; Phaneuf, R. A.; Krause, L., In elastic-Collisions between Excited Alkali Atoms and Molecules .10. Temperature-Depe ndence of Cross-Sections for 2p1/2-]]-2p3/2 Mixing in Cesium, Induced in Collisions with Deuterated Hydrogens, Ethanes, and Propanes. Canadian Journal of Physics 1974, 52, (7), 589-591. 144. Axner, O.; Ljungberg, P., A Tutorial Review of the Rate-Equation and Density-Matrix Formalisms for 2-and 3-Level Atomic and Molecular-Systems in High Collisional Media Exposed to Pulsed-Laser Light wi th Arbitrary Laser Bandwidths. Spectrochimica Acta Reviews 1993, 15, (4), 181-287. 145. Travis, J. C.; Turk, G. C., Laser-Enhanced Ionization Spectrometry New York, 1996. 146. Jabbour, Z. J.; Sagle, J.; Namiotka, R. K. ; Huennekens, J., Measurement of the SelfBroadening Rate Coefficients of the Cesium Resonance Lines. Journal of Quantitative Spectroscopy & Radiative Transfer 1995, 54, (5), 767-778. 147. Siegling, F.; Niemax, K., Hi gh-Pressure Noble-Gas Broadening of the Cs Resonance Lines. Zeitschrift Fur Naturforschung Section a-a Journal of Physical Sciences 1984, 39, (5), 455-463. 148. Nesmeyanov, A. N., Vapor pressure of the chemical elements. ed. R.Gray, Elsevier: Amsterdam-London-New York, 1963. 149. Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S., Surface-enhanced Raman scattering and biophysics. Journal of Physics-Condensed Matter 2002, 14, (18), R597-R624. 150. Baker, G. A.; Moore, D. S., Progress in plasmonic engineering of surface-enhanced Raman-scattering substrates toward ultra-trace analysis. Analytical and Bioanalytical Chemistry 2005, 382, (8), 1751-1770. 151. Guingab, J. D.; Lauly, B.; Smith, B. W.; Orne netto, N.; Winefordner, J. D., Stability of silver colloids as substrate for surface enha nced Raman spectroscopy detection of dipicolinic acid. Talanta 2007, 74, (2), 271-274. 152. Bakker, L. P.; Freriks, J. M.; de Hoog, F. J.; Kroesen, G. M. W., Thomson scattering using an atomic notch filter. Review of Scientific Instruments 2000, 71, (5), 2007-2014. 135

PAGE 136

153. Warner, K.; Hieftje, G. M., Thomson scattering from analytical plasmas. Spectrochimica Acta Part B-Atomic Spectroscopy 2002, 57, (2), 201-241. 154. Thomson, J. J., The corpuscular theory pf Matter A. Constable & co., Ltd: London, 1907. 155. Marshall, K. A.; Hieftje, G. M., Thom son Scattering for Determining Electron Concentrations and Temperatures in an Inductively Coupled Plasma .1. Assessment of the Technique for a Low-Flow, Low-Power Plasma. Spectrochimica Acta Part B-Atomic Spectroscopy 1988, 43, (6-7), 841-849. 136

PAGE 137

BIOGRAPHICAL SKETCH Benot Lauly was born in Paris, France on November 16, 1979. He spent his early years in Reims, Champagne-Ardennes region, France. Benot graduated from the Ecole National Suprieur de Chimie et Physique de Bord eaux (ENSCPB) in 2003. While attending at the ENSCPB, Benot did 6 months internship in the analytical department of Agfa-Gevaert N.V, Antwerpen, Belgium. He enrolled at the Univers ity of Florida in August 2003 in the chemistry graduate program and received his PhD in August 2008 under the supervision of Dr James Winefordner. He accepted a postdoctoral position at Dukes biomedical engineering department under Dr Tuan Vo Dinhs supervision. 137