Basics Of Plasma Spectroscopy - AAPT
HomeSearchCollectionsJournalsAboutContact usMy IOPscienceBasics of plasma spectroscopyThis content has been downloaded from IOPscience. Please scroll down to see the full text.2006 Plasma Sources Sci. Technol. 15 View the table of contents for this issue, or go to the journal homepage for moreDownload details:IP Address: 198.35.1.48This content was downloaded on 20/06/2014 at 16:07Please note that terms and conditions apply.
INSTITUTE OF PHYSICS PUBLISHINGPLASMA SOURCES SCIENCE AND TECHNOLOGYPlasma Sources Sci. Technol. 15 (2006) S137–S147doi:10.1088/0963-0252/15/4/S01Basics of plasma spectroscopyU FantzMax-Planck-Institut für Plasmaphysik, EURATOM Association Boltzmannstr. 2, D-85748Garching, GermanyE-mail: fantz@ipp.mpg.deReceived 11 November 2005, in final form 23 March 2006Published 6 October 2006Online at stacks.iop.org/PSST/15/S137AbstractThese lecture notes are intended to give an introductory course on plasmaspectroscopy. Focusing on emission spectroscopy, the underlying principlesof atomic and molecular spectroscopy in low temperature plasmas areexplained. This includes choice of the proper equipment and the calibrationprocedure. Based on population models, the evaluation of spectra and theirinformation content is described. Several common diagnostic methods arepresented, ready for direct application by the reader, to obtain a multitude ofplasma parameters by plasma spectroscopy.spectroscopy for purposes of chemical analysis are describedin [11–14].1. IntroductionPlasma spectroscopy is one of the most established andoldest diagnostic tools in astrophysics and plasma physics(see for example [1, 2]). Radiating atoms, molecules andtheir ions provide an insight into plasma processes and plasmaparameters and offer the possibility of real-time observation.Emission spectra in the visible spectral range are easy toobtain with a quite simple and robust experimental set-up.The method itself is non-invasive, which means that theplasma is not affected. In addition, the presence of rf fields,magnetic fields, high potentials etc. does not disturb therecording of spectra. Also the set-up at the experimentis very simple: only diagnostic ports are necessary whichprovide a line-of-sight through the plasma. Thus plasmaspectroscopy is an indispensable diagnostic technique inplasma processing and technology as well as in fundamentalresearch. Although spectra are easily obtained, interpretationcan be fairly complex, in particular, in low temperature, lowpressure plasmas which are far from thermal equilibrium, i.e.non-equilibrium plasmas.These notes give an introduction to plasma spectroscopyof low temperature plasmas for beginners. For furtherreading the following selection of books is recommended.Principles and fundamental techniques of plasma spectroscopyare very well described in [3, 4]. Elementary processes thatdetermine the radiation of atoms and molecules in plasmasare discussed in detail in [5]. An introduction to lowtemperature plasma physics and common diagnostic methodswith focus on applications to plasma processing is given in[6, 7]. An overview of plasma diagnostic methods for variousapplications can be found in [8–10]. Applications of plasma0963-0252/06/040137 11 30.00 2006 IOP Publishing Ltd2. Radiation in the visible spectral rangeElectromagnetic waves extend over a wide wavelength range,from radio waves (kilometre) down to γ -rays (picometer). Thevisible range is only a very small part ranging from 380 to780 nm by definition. However, common extensions are to theultraviolet and the infrared resulting roughly in a range from200 nm to 1 µm. From the experimental point of view thiswavelength region is the first choice in plasma spectroscopy:air is transparent, quartz windows can be used and a variety ofdetectors and light sources are available. Below 200 nm quartzglass is no longer transparent and the oxygen in the air starts toabsorb light resulting in the requirement of an evacuated lightpath. Above 1 µm the thermal background noise becomesstronger which can only be compensated for by the use ofexpensive detection equipment.Radiation in the visible spectral range originates fromatomic and molecular electronic transitions. Thus, the heavyparticles of low temperature plasmas, the neutrals and theirions basically characterize the colour of a plasma: typically ahelium plasma is pink, neon plasmas are red, nitrogen plasmasare orange and hydrogen are purple—these are first results ofspectroscopic diagnostics using the human eye.2.1. Emission and absorptionIn general, plasma spectroscopy is subdivided into two types ofmeasurements: the passive method of emission spectroscopyand the active method of absorption spectroscopy. In the casePrinted in the UKS137
U Fantzin units of W (m2 sr) 1 , where 4π represents the solid angled (isotropic radiation), measured in steradian (sr). The lineprofile Pλ correlates the line emission coefficient with thespectral line emission coefficient ελ : Pλ dλ 1 .(3)ελ εpk Pλ with1.0ImaxIntensity [a. u.]0.8 λFWHM0.6lineA characteristic of the line profile is the full width at halfmaximum (FWHM) of the intensity, λFWHM , as indicatedin figure 1. The line profile depends on the broadeningmechanisms [4]. In the case of Doppler broadening the profileis a Gaussian profile; the line width correlates with the particletemperature (see section 4.2). A convenient alternative to theline emission coefficient (equation (2)) is the absolute lineintensity in units of photons (m3 s) 1 :0.40.2λ00.0499500Ipk n(p)Apk .501Wavelength [nm]Figure 1. Line radiation and its characteristics.of emission spectroscopy, light emitted from the plasma itselfis recorded. Here, one of the basic underlying processes isthe excitation of particles (atoms, molecules, ions) by electronimpact from level q to level p and the decay into level kby spontaneous emission with the transition probability Apkresulting in line emission εpk . In the case of absorptionspectroscopy, the excitation from level q to level p takesplace by a radiation field (i.e. by absorption with the transitionprobability Bqp ) resulting in a weakening of the appliedradiation field which is recorded. The intensity of emissionis correlated with the particle density in the excited staten(p), whereas the absorption signal correlates with the particledensity in the lower state n(q), which is in most cases theground state. Thus, ground state particle densities are directlyaccessible by absorption spectroscopy; however, absorptiontechniques need much more experimental effort than emissionspectroscopy. Since some principles of emission spectroscopyapply also to absorption and, since emission spectroscopyprovides a variety of plasma parameters and is a passive andvery convenient diagnostic tool these lecture notes are focusedon emission spectroscopy. Further information on absorptiontechniques and analysis methods can be found in [6, 7, 15, 16].The two axes of a spectrum are the wavelength axis and theintensity axis as shown in figure 1. The central wavelength ofline emission λ0 is given by the photon energy E Ep Ekcorresponding to the energy gap of the transition from levelp with energy Ep to the energetically lower level k (Planckconstant h, speed of light c):λ0 h c/(Ep Ek ) .(1)Since the energy of a transition is a characteristic of the particlespecies, the central wavelength is an identifier for the radiatingparticle, unless the wavelength is shifted by the Doppler effect.As a principle, the wavelength axis λ is easy to measure, tocalibrate and to analyse. This changes to the opposite forthe intensity axis. The line intensity is quantified by the lineemission coefficient: hc ελ dλ(2)εpk n(p) Apk4π λ0lineS138(4)This relationship reveals that the line intensity dependsonly on the population density of the excited level n(p)which, in turn, depends strongly on the plasma parametersn(p) f (Te , ne , Tn , nn , .). This dependence is describedby population models and will be discussed in section 3.2.2. Atomic and molecular spectraThe atomic structure of atoms and molecules is commonlyrepresented in an energy level diagram and is stronglyrelated to emission (and absorption) spectra. The electronicenergy levels of atoms and diatomic molecules have theirspectroscopic notation:n w2S 1LL Sandn w2S 1 , g,u,(5)respectively. n is the main quantum number, the angularmomentum, w the number of electrons in the shell, S the spin,2S 1 the multiplicity, L S J the total angular momentum.This represents the LS coupling which is valid for light atoms.Details of atomic structure can be found in the standard books,e.g. [17–20]. In case of diatomic molecules the projection ofthe corresponding vectors onto the molecular axis is important,indicated by Greek letters. , and g, u denote the symmetryof the electronic wave function (for details see [4,20–22]).Optically allowed transitions follow the selection rules fordipole transitions which can be summarized as: L 0, 1, J 0, 1, S 0 for atoms and 0, u g formolecules. L 0 or J 0 transitions are not allowed ifthe angular momenta of both states involved are zero.Figure 2 shows the energy level diagram for heliumwhich is a two electron system. The levels are separatedinto two multiplet systems: a singlet and a triplet system.Following Pauli’s principle the spin of two electrons in theground state is arranged anti-parallel resulting in the 1s 11 Sstate. The fine structure is indicated only for the 23 P state.Electronic states which cannot decay via radiative transitionshave a long lifetime and are called metastable states (23 S and21 S). Transitions which are linked directly to the ground stateare called resonant transitions. The corresponding transitionprobability is high, hence the radiation is very intense. Dueto the large energy gap these transitions are often in thevacuum ultraviolet (vuv) wavelength range. Optically allowed
Basics of plasma spectroscopysinglet systemE [eV]11S24.581Ptriplet system1D3F3S3PD3F23.673 1P3 1S22.923 1D667nm501nm728 nm388 nm706nm2 1P587nm2 3P2058nm20.613 3D3 3P3 3S2 1S1083 nm53.7 nm19.822p 3P3/2,1/22 3S58.4 nmMetastable stateLargeenergygapResonant transitionHelium1 1S configuration 1s20Figure 2. Example of an atomic energy level diagram.E [eV]16Hydrogen H2singlettriplet141210United atomapproximationn 3E,FCaBcn 2B 1Σu 8b 3Σu b6Repulsive state420Figure 4. Molecular excitation and radiation according to theFranck–Condon principle for the ground state and two excited statesof molecular hydrogen.···Vibrational levels vv 3v 2v 1v 0Rotational levels JX 1Σg n 1Figure 3. Example of an energy level diagram of a diatomicmolecule.transitions with upper quantum number n 3 are indicatedby an arrow and labelled with the corresponding wavelength.Radiation in the visible spectral range mostly originates fromtransitions between excited states.The energy level diagram of a diatomic molecule withtwo electrons, i.e. molecular hydrogen, is shown schematicallyin figure 3. Again the two electrons cause a splitting into asinglet and a triplet system. In the united atom approximationfor molecules a main quantum number can be assigned. Inmolecules the energy levels are usually abbreviated by upperand lower case letters, where X is the ground state (as arule). The corresponding spectroscopic notation is shownin figure 3 for the X and b states. Due to the additionaldegrees of freedom, each electronic state has vibrational levels(quantum number v) and each vibrational level has rotationallevels (quantum number J ) which appear with decreasingenergy distances. The vibrational levels in the ground stateare indicated in figure 3. A special feature in the energylevel diagram of the hydrogen molecule is the repulsive stateb 3 u . Due to the repulsive potential curve, the energyrange covers a few electronvolts. Molecules in this stateeventually dissociate. Radiation in the visible spectral rangecorrespond to an electronic transition without restrictions forthe change in the vibrational quantum number and appear inthe spectrum as vibrational bands. Each vibrational band has arotational structure, where the rotational lines must follow theselection rules: J 0, 1, forming so-called P , Q andR branches.An additional feature of diatomic molecules is theinternuclear distance of the two nuclei. Thus, potential curvesdefine the energy levels. This is shown in figure 4 for theground state and two electronically excited states of hydrogentogether with vibrational levels, i.e. the vibrational eigenvaluesand the corresponding vibrational wave functions. Under theassumption that the internuclear distance does not change inthe electron impact excitation process and during the decayby spontaneous emission, excitation and de-excitation followa vertical line in figure 4, which means that the Franck–Condon principle is valid. In other words, a vibrationalpopulation in the ground state is projected via the Franck–Condon factors into the electronically excited state. TheFranck–Condon factor is defined as the overlap integral ofS139
U FantzIntensity [a.u.]1.2Exit slitNa D-line1.03 2P3/2 - 3 2S1/20.83 2P1/2- 3 587588589590591Collimatingmirror592Wavelength [nm]Entrance slitIntensity [a.u.]1.0Nitrogen N20.8C 3Πu - B 3Πg0.6 v'-v''0.40.20-21-3Figure 6. Monochromator in Czerny–Turner Wavelength [nm]Figure 5. Atomic and molecular spectra: NaD-lines and vibrationalbands of the second positive system of N2 .two vibrational wave functions. In addition for radiativetransitions the electronic dipole transition momentum has tobe taken into account. Thus, the Franck–Condon factors arereplaced by vibrational transition probabilities and togetherwith the vibrational population of the excited state determinethe intensity of a vibrational band.Figure 5 shows the intense vibrational bands of molecularnitrogen. These vibrational bands v v with v 2 (v is the vibrational quantum number in the upper electronicstate, v is the vibrational quantum number in the lowerelectronic state) correspond to the electronic transition C 3 u –B 3 g , which is called the second positive system of nitrogen.The rotational structure of each vibrational band is observedclearly: however, the shape of the bands is determined by thespectral resolution of the spectroscopic system. The sameapplies to atomic spectra as shown in the upper part of figure 5for sodium. The two narrow lines correspond to a recordedspectrum where the fine structure is resolved, whereas thebroad line would be observed by a spectrometer with poorspectral resolution, a quantity which will be discussed in thenext section.2.3. Spectroscopic systemsThe choice of spectrographs, detectors and optics dependsstrongly on the purpose for which the diagnostic tool is tobe used. Details of various spectroscopic systems and theircomponents can be looked up in standard books about opticsor in [3, 4]. The basic components of a spectrometer are: theentrance and exit slit, the grating as the dispersive elementand the imaging mirrors, as illustrated in figure 6 for theCzerny–Turner configuration. The exit slit is equipped with adetector. The light source, i.e. the plasma, is either imaged byan imaging optics onto the entrance slit or coupled by fibres tothe slit. The latter is very convenient, particularly when directaccess to the plasma light is difficult. The individual parts ofS140the spectroscopic system determine the accessible wavelengthrange, the spectral resolution and the throughput of light.The choice of the grating, which is characterized bythe grooves per millimetre (lines/mm) is of importance forthe spectral resolution. Special types of gratings such asEchelle gratings are optimized for high order diffractionsresulting in a high spectral resolution. The blaze angleof a grating determines the wavelength range with highestreflection efficiency, i.e. the sensitivity of the grating.The focal length of the spectrometer influences the spectralresolution and together with the size of the grating defines theaperture and thus the throughput of light. The width of theentrance slit is also of importance for light throughput, whichmeans a larger entrance slit results in more intensity, with thedrawback that the spectral resolution decreases.At the exit either a photomultiplier is mounted behind theexit slit or a CCD (charge-coupled device) array is mounteddirectly at the image plane of the exit. In the first case, the widthof the exit slit or, in the second case, the pixel size influencethe spectral resolution of the system. The overall sensitivityof the system is strongly dominated by the type of detector:photomultipliers with different cathode coatings or CCD arrayswith different sensor types (intensified, back-illuminated,etc.). Spectroscopic systems which use photomultipliersare scanning systems whereas systems with CCD arrays arecapable of recording a specific wavelength range. Spatialresolution can be achieved by the choice of 2-dim detectors.Temporal resolution is completely determined by the detector:photomultipliers are usually very fast, whereas CCD arrays arelimited by the exposure time and read-out time.In the following some typical system set-ups are presentedin order to give an overview of the different types ofspectrometers and their typical application.For linemonitoring, which means following the temporal behaviourof an emission line, pocket size survey spectrometers are verysuitable. They have a poor spectral resolution λ 1–2 nmbut a good time resolution. The diagnostic technique itselfis a simple one, providing information on plasma stabilityor changes in particle densities. Another typical system is aspectrometer with a focal length of 0.5–1 m ( λ 40 pm) anda grating with 1200 lines mm 1 . The optical components andthe optics are very much improved; the time resolution dependson the detector. Using a 2–dim CCD camera reasonable spatialresolution can be achieved. This combination representsa flexible system and is a reasonable compromise betweenspectral resolution and temporal behaviour.
Basics of plasma spectroscopyThe next step is to deploy an Echelle spectrometer, whichprovides an excellent spectral resolution ( λ 1–2 pm)by making use of the high orders of diffraction providedby the special Echelle grating. Typical applications aremeasurements of line profiles and line shifts. An importantpoint in the choice of the spectroscopic system is theintensity of the light source. For example, measurementswith an Echelle spectrometer require much more light thanmeasurements with the survey spectrometer. However, this canbe partly compensated by the choice of detector and exposuretime.Another important issue is the calibration of thespectroscopic system. One part is the calibration of thewavelength axis, which is an easy task, done by using spectrallamps (or the plasma itself) in combination with wavelengthtables [23]. For example a mercury–cadmium lamp can beused as the cadmium and mercury lines extend over a widewavelength range. Groups of lines with various distancesbetween each other (wavelength axis) are very well suitedto determine the spectral resolution of the system and theapparatus profile. In this case, line broadening mechanismsmust be excluded, for example, by using low pressure lamps.Much more effort is needed for the calibration of theintensity axis, which can be either a relative or an absolutecalibration. A relative calibration takes into account onlythe spectral sensitivity of the spectroscopic system along thewavelength axis. An absolute calibration provides in additionthe conversion between measured signals (voltage or counts) inW/(m2 sr) or to Photons/(m3 s) according to equations (2) and(4). An absolute calibrated system provides calibrated spectra,which gives direct access to plasma parameters. Thus, theeffort is compensated by an increase in information. Forthe intensity calibration light sources are required for whichthe spectral radiance is known. One of the most criticalpoints in the calibration procedure is the imaging of the lightsource to the spectroscopic system. Here one must be verycareful to conserve the solid angle which is often adjusted byusing apertures. Calibration standards in the visible spectralrange, from 350 nm up to 900 nm, are tungsten ribbon lamps,providing black body radiation (grey emitter) and Ulbrichtspheres (diffusive sources). For extensions to the uv rangedown to 200 nm, the continuum radiation of deuterium lampsis commonly used. Since such light sources must have highaccuracy they are usually electrical stabilized but they alter intime. This means that their lifetime as calibration source islimited, which is less critical for relative calibration. Typicalcurves of a calibration procedure are shown in figure 7. Theupper part is the spectral radiance of an Ulbricht sphere(provided by an enclosed data sheet) which is typically usedfor the calibration of a spectroscopic system with fibre optics.The spectrum in the middle gives the measured intensity inunits depending on the detector (e.g. CCD detector, countsper second). The recorded spectrum is already normalizedto the exposure time. Dividing the spectral radiance curveby the recorded spectrum the conversion factor is obtainedrepresenting the spectral sensitivity of the system. As alreadymentioned, absolutely calibrated spectral systems are the mostpowerful tool in plasma diagnostics. Thus, the followingsections discuss the analysis methods based on absolutelycalibrated spectra. Nevertheless, some basic principles can beCalibrated spectrumspectral radiance[W/m2/sr/nm]Measurementintensity [counts/s]Conversion factorspectral sensitivityWm2sr nm (counts/s) 4π λphotons hcm2s nm 500600700Wavelength [nm]800Figure 7. Example for the three steps needed to obtain a calibrationcurve.applied even if only relatively calibrated systems or systemswithout calibration are available.3. Population modelsAccording to equation (4), the absolute intensity of a transitionis directly correlated to the population density in the excitedstate, the upper level. The population density of excited statesis described by a Boltzmann distribution provided that thelevels are in thermal equilibrium among each other. Sincelow temperature plasmas are non-equilibrium plasmas, whichmeans they are far from (local) thermal equilibrium, thepopulation density does not necessarily follow a Boltzmanndistribution. As a consequence, the population in an excitedstate depends not only on the electron temperature but ona variety of plasma parameters: temperature and density ofthe electrons and the heavy particles, radiation field, etc.The dominant parameters are determined by the dominantplasma processes. Thus, population models are requiredwhich consider populating and depopulating processes for eachindividual level of a particle. An excellent overview of thistopic is given in [5]. It is obvious that for molecules wherevibrational and rotational levels exist, the number of processesis incalculable and has to be reduced in some way.3.1. Populating and depopulating processesThe electron impact excitation process is one of the mostimportant processes. It increases the population of the upperlevel and decreases the population of the lower level. In turnelectron impact de-excitation depopulates the upper level andpopulates the lower level. In a similar way, this principleapplies to absorption and spontaneous emission for opticallyallowed transitions. Other population processes which couplewith the particle in the next ionization stage are radiative orthree-body recombination of the ion and the de-excitation byionization. The different types of processes and a detailedexplanation with examples are given in [4, 5, 7, 24].Each process is described by its probability. In thecase of spontaneous emission the probability is called theEinstein transition probability Aik , where i labels the upperlevel and k the lower level. Collisional processes are generallydescribed by cross sections or rate coefficients. The latterS141
U Fantzf(E)0.3Te 1.5 eV Maxwell EEDFσ(E)0.2Te 4.5 eV0.10.0051015Ethr10-172025Energy [eV]Xexc [m3/s]10-18steep dependenceon Te10-1910-20Te [eV]Figure 8. Convolution of a cross section with Maxwellian EEDFs(Te 1.5 eV and 4.5 eV) resulting in the excitation rate coefficient.can be obtained from the convolution of the cross sectionwith the corresponding energy distribution function of theimpact particle. For an electron impact process the electronenergy distribution is used, often described by a Maxwelldistribution function. However, it should be kept in mindthat this assumption is often not justified in low temperatureplasmas, as briefly discussed in section 4.3. The upper partof figure 8 shows Maxwellian electron energy distributionfunctions (EEDF) f (E) for two electron temperatures, Te 1.5 eV and Te 4.5 eV, and a typical cross section σ (E)(in arbitrary units) for an electron impact excitation process.The dashed area starting with the threshold energy of theexcitation process (Ethr ) indicates the part of the electronswhich contributes to the rate coefficient. The lower part offigure 8 shows the corresponding excitation rate coefficientXexc (Te ): Xexc (Te ) σ (E)(2/me )1/2 E f (E) dEEthrwith f (E) dE 1.(6)0It is obvious that the rate coefficient Xexc (Te ) shows a steepdependence on Te , in particular, at low electron temperatures,i.e. Te Ethr . Since the part of the cross section withenergies close to the threshold energy contributes most tothe convolution, the accuracy of the rate coefficient dependsstrongly on the quality of the cross section in this energy region.3.2. Corona modelA simple approach to population densities in non-equilibriumplasmas is presented by the so-called corona model. Thecorona equilibrium is deduced from the solar corona whereS142electron density is low ( 1012 m 3 ), electron temperature ishigh ( 100 eV) and where the radiation density is negligible.Due to the low electron density the probability of electronimpact de-excitation processes is much lower than deexcitation by spontaneous emission and can be neglected.The electron temperature guarantees that the plasma is anionizing plasma, i.e. recombination and thus populatingprocesses from the next ionization stage do not play arole. Due to the insignificant radiation field, absorption isnot important and excitation takes place only by electronimpact collisions. Since these conditions are often fulfilledin low pressure, low temperature plasmas the usage of thecorona model is a common method to deduce population (andionization) equilibrium. However, the applicability has to bechecked carefully in the individual case. These plasmas arecharacterized by a low degree of ionization. Each particlespecies (electrons, ions and neutrals) is characterized by itsown temperature (under the assumption that a MaxwellianEEDF can be applied) and a gradual decrease is obtained:Te Ti Tn .The corona model assumes that upward transitions areonly due to electron collisions while downward transitionsoccur only by radiative decay. Thus, in the simplest case,the population of an excited state p is balanced by electronimpact excitation from the ground state q 1 and decay byspontaneous emission (optically allowed transitions to level k): exc(Te ) n(p)Apk .(7)n1 ne X1pkAs discussed above, inverse processes, such as electron impactde-excitation and self-absorption, are not of relevance. As aresult the population density is far below a population densityaccording the Boltzmann distribution and population densitiesof excited levels are orders of magnitude lower than thepopulation of the ground state. In this case, the ground statedensity n1 can be replaced by the overall particle density nn .The corona equation can be extended easily by furtherprocesses, for example, excitation out of metastable statesor cascading from higher excited states. For excited stateswithout an optically allowed transition, i.e. metastable states,the losses are often determined by diffusion. This can becharacterized by confinement times where the reciprocal valuereplaces the transition probability in equation (7). However,the selection of processes being important for the populationequilibrium is often very unclear in the individual case.3.3. Collisional radiative modelA more general approach to population densities is to set uprate equations for each state of the particle together with thecoupling to other particles, e.g. the next ionization stage. Sincesuch a model balances the collisional and radiative processesthe model is called a collisional radiative (CR) model. Thetime development of the population density of state p in aCR-model is given by: d n(p)excn(k) ne Xkp dtk p de excexcXpk Xpk Sp Apk n(p) nek pk pk p
Basics of plasma spectroscopyde excne Xkp Akpk p 10-2 ni ne ne αp βp .(8)The first term describes the electron impact excitation fromenergetically lower lying levels, k p. Loss processes areelectron impact de-excitation into energetically lower lyinglevels, k p, electron impact excitation of energeticallyhigher lying levels k p, electron impact ionization (ratecoefficient (S(p)), and spontaneous emission. Next, electronimpact processes and spontaneous emission from energeticallyhigher lying levels k p is taken into account. The last twoexpressions describe population by three-body recombinationand radiative recombination respectively. In addition, opacity,which means self-absorption of emission lines, may beimportant. A convenient method is to use the population escapefactors ( [25, 26]) which reduce the corresponding transitionprobabilities in equation (8).In the quasi-stationary treatment the time derivat
temperature plasma physics and common diagnostic methods with focus on applications to plasma processing is given in [6,7]. An overview of plasma diagnostic methods for various applications can be found in [8–10]. Applications of plasma spectroscopy for purposes of chemical analysis
1. Introduction to Spectroscopy, 3rd Edn, Pavia & Lampman 2. Organic Spectroscopy – P S Kalsi Department of Chemistry, IIT(ISM) Dhanbad Common types? Fluorescence Spectroscopy. X-ray spectroscopy and crystallography Flame spectroscopy a) Atomic emission spectroscopy b) Atomic absorption spectroscopy c) Atomic fluorescence spectroscopy
Visible spectroscopy Fluorescence spectroscopy Flame spectroscopy Ultraviolet spectroscopy Infrared spectroscopy X-ray spectroscopy Thermal radiation spectroscopy Detecting and analyzing spectroscopic outputs The goal of all spectroscopic systems is to receive and analyze the radiation absorbed, emitted, .
Plasma Etching Page 2 OUTLINE Introduction Plasma Etching Metrics – Isotropic, Anisotropic, Selectivity, Aspect Ratio, Etch Bias Plasma and Wet Etch Summary The Plasma State - Plasma composition, DC & RF Plasma Plasma Etching Processes - The principle of plasma etching, Etching Si and SiO2 with CF4
Plasma Fundamentals - Outline 1. What is a plasma ? Temperature Debye shielding Plasma frequency 2. The edge of a plasma Sheath physics 3. How to ignite a plasma Ignition, Paschen curve Streamer RF-ignition 4. Transport in a plasma Particle motion Plasma
Spectroscopy Beauchamp 1 y:\files\classes\Spectroscopy Book home\1 Spectroscopy Workbook, latest MS full chapter.doc Basics of Mass Spectroscopy The roots of mass spectroscopy (MS) trace back to the early part of the 20th century. In 1911 J.J. Thomson used a primitive form of MS to prove the existence of isotopes with neon-20 and neon-22.
Plasma Cleaner: Physics of Plasma Nature of Plasma A plasma is a partially ionized gas consisting of electrons, ions and neutral atoms or molecules The plasma electrons are at a much hi
2.0 MagMate Cut25 Plasma 6 2.1 Fundamentals of Plasma Cutting 6 2.2 Process operation for transferred arc applications 7 3.0 Plasma cutting components 8 3.1 Plasma cutting power sources 8 3.2 Plasma cutting capacity 8 3.3 Plasma cutting torches (general) 8 3.4 Air supply 9 3.5 Process comparisons 9 3.6 Work return cable assembly 9
the transactions are difficult to discern. This makes it difficult to determine the overall size of activity and to know what the fair price is for a particular technology. And, of course, in highly inefficient markets a good deal of potentially valuable trade in innovation does not occur. The costs are so high and the potential value so difficult to perceive that innovation often sits “on .
