ECTROCHIMICA
`ACTA
`
`
`—\
`
`‘PART B: ATOMIC SPECTROSCOPi?“%
`'
`Vol. 40B, No. 4, 1985
`
`'
`
`“fax
`
`Energetiq Ex. 2081, page 1 - |PR2015-01368
`
`Energetiq Ex. 2081, page 1 - IPR2015-01368
`
`

`
`01 LJ\./ 1 .[\\Jl._/1'lllV.llK41‘3t [Kb 1 A
`
`'
`
`PART B: ATOMIC SPECTROSCOPY
`EDITOR—IN-CHIEF
`Dr. P. W. J. M. Boumans, Philips Research Laboratories, 5600 JA Eindhoven, The
`Netherlands
`EDITORS '
`Dr. Walter Slavin, Perkin—Elmer Corporation, Main Avenue, Norwalk, CT 06856, USA
`Dr. H. W. ’ Werner, Philips Research Laboratories, 5600 JA Eindhoven, The Netherlands
`ASSISTANT EDITORS
`Dr. J. A. C.’ Broekaert, Institute of Spectrochemistry and Applied Spectroscopy ([5915),
`P.O. Box 778, 4600 Dortmund, Federal Republic of Germany
`Dr. R. J. Declcer, Rhodesia University, Chemistry Department, P.0. Box MP 169, Harare
`Zimbabwe
`EDITORIAL ADVISORY BOARD
`Chairman: J. D. Winefordner, Gainesville, FL
`C. Th. J. Alkemade, Utrecht, The Netherlands
`S. F. A. Kettle, Norwich,‘ UK
`R. M. Barnes, Amherst, MA
`S. R. Koirtyohann, Columbia, Mo
`L. S. Birks, Washington, DC
`K. Laqua, Dortmund, GFR
`R. F. Browner, Atlanta, GA
`B. V. L’vov, Leningrad, USSR
`C. L. Chakrabarti, Ottawa, Canada
`F. J. M. J. Maessen, Amsterdam,
`K. Dittrich, Leipzig, GDR
`The Netherlands
`H. Falk, Berlin, GDR
`S. L. Mandelshtam, Moscow, USSR
`V. A. Fassel, Ames, [A
`J. M. Mermet, Vernaison, France
`K. Fuwa, Tokyo, Japan
`N. Omenetto, Ispra, Italy
`L. de Galan, Delft, The Netherlands
`J. M. Ottaway, Glasgow, UK
`Y.,Gohshi, Tokyo, Japan
`E. Plsko, Bratislava, CSSR
`D. Hercules, Pittsburgh, PA
`J. Robin, ‘Villeurbanne, France
`G. M. Hieftje, Bloomington, [N
`F. Riidenauer, Vienna, Austria
`G. Horlick, Alberta, Canada
`M. Someno, Nagoaka City, Japan
`A. M. Huber, Orsay, France
`A. Strasheim, Pretoria, RSA
`H. Kawaguchi, Nagoya, Japan
`A. Walsh, Brighton, Australia
`M. Kent Wilson Washington, DC
`K. Zimmer, Budapest, Hungary
`Publishing, Subscription and Advertising Ofiices: Pergamon Press Inc., Maxwell House, Fairview Park, Elmsford,
`NY 10523, U.S.A.; Pergamon Press Ltd., Headington Hill Hall, Oxford OX3 OBW, U.K. (Oxford 64881).
`Payments must be made in advance. Subscription enquiries from customers in North America should be sent
`to: Pergamon Press Inc., Maxwell House, Fairview Park, Elmsford, NY 10523, U.S.A., and for the remainder
`of the world to: Pergamon Press Ltd., Headington Hill Hall, Oxford OX3 OBW, U.K.
`One volume per annum. Published monthly,
`Subscription rates: For libraries, university departments, government laboratories, industrial and other mulli-
`reader institutions (1985), U.S. $325.00; two-year rate (1985/86) U.S. $617.50 (including postage and insurance).
`Specially Reduced Rates to Individuals
`In the interests of maximizing the dissemination of the research results published in this important international
`journal we have established the following price structure. Any individual, whose institution takes out a library
`subscription, may purchase a second or additional subscription for personal use at a much reduced rate of U.S.
`$75.00 per annum.
`Microforttl Subscriptions and Back Issues
`year.
`Back issues of all previously published volumes are available in the regular editions and un microfilm and microfiche. Current subscriptions are
`available on microfiche simultaneously with the paper edition and on microfilm on completion of the annual index at the end at‘ the subscription
`Copyright © 1985 Pergamon Press Ltd.
`
`
`
`’
`
`It is a condition of publication that manuscripts submitted to this journal have not been published and will not be simultaneously submitted UT
`published elsewhere. By submitting H manuscript, the authors agree that the copyright for their article is transferred to the publisher if and Wh€_fl
`
`microfon-n or any other reproductions of similar nature and translations. No part of this publication may be reproduced, stored in a retrieval system
`permission in writing from the copyright holder.
`or transmitted in any form or by any means, electronic, clectrusti-ilic, magnetic tape, mechanical, photocopying, recording or otherwise, without
`
`PERGAMON PRESS
`HEADINGTON HILL HALL, OXFORD OX3 OBW, ENGLAND
`MAXWELL HOUSE, FAIRVIEW PARK, ELMSFORD, NEW YORK 10523
`
`Energetiq Ex. 2081, page 2 - |PR2015-01368
`
`Energetiq Ex. 2081, page 2 - IPR2015-01368
`
`

`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Energetiq Ex. 2081, page 3 - IPR2015-01368
`
`

`
`
`
`Fig. 1. Photograph of the COD produced inside a small quartz tube containing Xe at 2000 torr. The
`cw—CO 2 laser beam entered the cell from the right. The plasma was photographed for 1 s, atf8.6, with
`a neutral density filter (ND) = 3 in front of the camera lens. Then the plasma was turned at!‘ and the
`cell photographed under room lights with the ND filter removed.
`
`[19] G. I. KozLov, V. A. Kuzmzrsov and V. A. MASYUKOV, Zh.1 'ekh. Fiz. 49, 2304 (1 979) [San Phys. Tech. Phys
`1283 (l979)].
`[20] D. C. SMITH and M. C. FOWLER, Appl. Phys. Len. 22, 500 (1973).
`
`[21] D. R. KEEFER, H. B. HENRIKSEN and W. 1-‘. BRAERMAN, J. Appl. Phys. 46, 1080 (1975).
`
`Energetiq Ex. 2081, page 4 - |PR2015-01368
`
`666
`
`DAVID A. CREMERS et al.
`
`atmospheric pressures or above. The discharge resides near the focus of the laser baa
`independent ofany physical support and does not require a gas flow to stabilize the plasmaipa
`some other sources. Because the discharge is maintained using optical frequencies (30 THZ
`the plasma is called a “continuous optical discharge” (COD).
`A photograph of the COD produced in this laboratory is shown in Fig. 1. The plasma
`small, about 1 mm in diameter, and appeared as a very bright white light. The plasma W,
`initiated by the spark produced by a focused Q-switched Nd : YAG laser pulse superimposg
`on the focal volume of the cw-beam because the cw powers used to maintain the plasma w '
`insuificient to induce optical breakdown. Typically, the focused powers of cw—CO; lasers
`106-107 W/cmz, several orders of magnitude smaller than the 108-109 W/cm2 breakdo
`threshold of atmospheric pressure gases. The pulsed laser spark plasma contains a h
`density of electrons (2 10‘ " cm”) which act as an absorbing center for the 10.6-um bea
`The plasma can also be initiated using a conventional electrode spark [20—21] or the plasm
`produced on the surface of a material which is temporarily introduced into the focal volu 0
`of the cw-beam [15]. Once started, the plasma operates continuously as long as suffici
`intensity is supplied to the focal volume.
`The cw laser power needed to maintain the plasma depends upon the gas pressure, typeo
`gas, and whether the cw-beam is horizontal or vertical
`The maintenance thresho ‘i
`increases with the ionization potential (Ip) of the gas [7]: at 3192 torr, powers of 59, 93, 1
`and more than 480 W are needed to maintain the COD in Xe (Ip = 12.08 eV), Kr (13.93 eV)
`Ar (15.68 eV) and He (24.46 eV), respectively. About 2 kWare required to produce the plas
`in air at atmospheric pressure [20, 21]. The properties of the laser beam are also important as
`evidenced by the Widely different values for maintenance thresholds listed in the literature 1‘
`identical gases and pressures [4, 10, 17]. The temperature of the COD depends upon the
`and at 1520 torr has been measured spectroscopically to be a maximum of 14000 K in—_X
`
`Energetiq Ex. 2081, page 4 - IPR2015-01368
`
`

`
`Evaluation of the continuous optical discharge
`
`v
`
`667
`
`‘000 K in Ar, and 22000 K in N2 [16]. These are much higher than the temperatures
`an-acterizing other continuous excitation sources: arcs (4000-5000 K), ICP (6000—8000 K)
`d a microwave discharge (5000—7000 K). The temperature of the COD” can approach that
`gained by sparks (20 000 K or higher) and is due to penetration of the high frequency optical
`diation into the core of the plasma [16]. At laser frequencies, which are typically above the
`asma frequency [22], absorption occurs mainly via free—free transitions (inverse
`femsstrahlung) associated with e1ectron—ion collisions [23]. For comparison, at radio and
`‘ rowave frequencies, which are below the plasma frequency, plasma heating occurs
`1 ugh direct plasma-electric field interactions characterized by larger absorption coef-
`ients. Consequently, only the outer layers of these plasmas are heated. The higher
`[0 peraturc of the COD is also related to its greater input energy density (p) compared to
`5 re conventional discharges. For example, for a 0.1—cm dia. plasma, a laser power of 45 W,
`d assuming 50% absorption by the plasma [7], p = 5 x 103 W cm”, vs p = 188 W cm“ 3
`r an ICP of 1.5 kW and a volume of 8 cm3. The electron density of the Ar COD is
`"out 10” cm”, at least an order of magnitude greater than other common continuous
`urces [4].
`Studies of the COD to date have concentrated mainly on measurements of its physical
`_operties in various gases under different conditions. Several theoretical models of the COD
`ave been developed to account for these properties [24~29] and a few applications have
`een suggested [30—32]. An almost complete listing of the previous studies is presented in the
`ferences cited herein. To our knowledge, the COD has not previously been investigated as
`n excitation source for atomic emission spectroscopy. The high temperatures and electron
`ensity of this plasma should provide improved excitation for species difficult to detect with
`oler conventional cw sources. In addition, the COD is a single source which combines the
`igh temperature of a spark with the continuous operation of a dc arc, the goal behind the
`evelopment of many other types of spectroscopic sources [33]. In this paper we present the
`‘sults of a preliminary examination of this plasma for spectrochemical analysis. Particular
`inphasis is placed on characteristics of COD generation, variation of plasma properties with
`perating parameters, and analytical performance.
`
`
`
`2. EXPERIMENTAL
`
`Apparatus
`schematic of the apparatus used to generate the COD and record the emission spectrum is shown in
`1g. 2. The experimental conditions are listed in Table 1. The gas cell was evacuated to a pressure
`50 um before adding plasma gas. The cw—CO2 and pulsed Nd : YAG laser radiations were focused
`to the same volume of the cell at right angles as shown. The focal volume of the pulsed beam was
`djusted to overlap that of the cw-CO2 beam by moving the glass imaging lens slightly using an X YZ
`
`22] The plasma frequency (vp) is given by v; = (ezne)/‘ (41rZs0me), where e, nc, and me are the electronic charge,
`density, and mass, respectively, and so is the vacuum permittivity. At frequencies below vp, EM waves interact
`directly with the plasma to induce heating. At frequencies above vp, the waves only interact with individual
`electrons and ions to heat the plasma. For an electron density of8 X 10”‘ cm”, vp : 2.5 X 10” Hz, which is
`about an order of magnitude below the C0; laser frequency of 30 THz.
`‘
`.] T. P. HUGHES, Plasmas and Laser Light. John Wiley, New York (1975).
`A. A. KURBATOV, T. YA. POPOVA and N. G. PRF,()Bl(AZENSKY, Prac. of the XIII Int. Conf. on Phenomena in
`_ Ionized Gases, 1977, Part II, p. 899, Berlin, GDR (1977).
`25] G. I. KOZLOV and I. K. SELEZNEVA, Zh. Teklt. Fiz. 48, 386 (1978) [Son Phys. Tech. Phys. 23, 227 (1978)].
`] M. V. GERASIMENI-(O, G. I. KozLov, V. A. Kuzmarsov, Pis’ma Z11. Tekh. F£2. 6, 485 (1980) [Sov. Tech. Phys.
`Len. 6, 208 (l980)].
`_
`27] YU. P. RAZIER, Pis’ma Zh. Tekh. Fiz. 7, 938 (1981) [S002 Tech. Phys. Lelt. 7, 404 (198l)].
`23] S. MULLER and J. UHLENBUSCH, Physica 112C, 259 (1982).
`19] G. G. GLADUSH and A. N. YAVOKHIN, Kvunlovaya Elektron. (Moscow) 10, 1399 (1983) [Sov. J. Quantum
`I‘
`~ Electron. 13, 908 (l983)].
`] R. W. THOMPSON, E. J. MANISTA and D. L. ALGER, Appl. Phys. Len, 32, 610 (1978).
`Laser Focus, Dec. l977. p. 20.
`-32] N. H. KEMP and R. G. R001’, J. Spacecraft I6, 65 (1979).
`] P. W. J. M. BOUMANS, Analytical Emission Spectroscopy, Ed. E. L. GROVE, Part II, Chap. 6. Marcel Dekker.
`New York (1972).
`
`
`
`Energetiq Ex. 2081, page 5 - |PR2015-01368
`
`Energetiq Ex. 2081, page 5 - IPR2015-01368
`
`

`
`
`
`668
`
`DAVID A. CREMERS er al.
`
`
`
`5‘/GA/Al.
`
`Fig. 2. Top view diagram of the apparatus used to generate the COD and monitor the emission
`spectrum. The large gas cell is shown here. The cw-CO2 laser beam was focused using a mirror (dashed
`line) or lens (shown here). The COD was sometimes started by replacing the side window by a pair of
`electrodes shown in the insert. PB: pulsed laser beam to start the COD; TS: translation stage to
`position the lens focusing the pulsed beam; W: quartz window; SW: salt window; C: chopper; QL:
`quartz lens; P: photomultiplier tube; M: monochromator; LIA: lock-in amplifier; R: chart recorder; E;
`electrode; B: vacuum bellows; G: pressure gage; V: vacuum pump; Xe: Xe gas supply. The tesla coil was
`connected to wire HV.
`
`translation stage. The mode of the cw-CO2 laser was adjusted to be TEM00. The laser power was not of
`stabilized via feedback control but it varied by less than 1 "0 over a 4-h period. The laser was operated
`broadband with a maximum available power of about 70 W. The pulsed laser was operated at about i
`1 Hz and was turned off following ignition of the COD. The cell pressure reached a constant valu '
`within ~ 1 min after ignition.
`-
`For measurements of line intensities and line widths, the plasma light was mechanically chopped V
`imaged on the entrance slit of a scanning monochromator with unit magnification, and detected with
`photomultiplier tube (PMT). The PMT signal was processed by a loclc—in amplifier (LIA). Th
`monochromator wavelength was scanned over spectral features and the LIA output recorded on a chart
`recorder. The spectral response of the detection system from 350 to 960 nm was determined using 1
`standard lamp calibration. Phase sensitive detection with the LIA minimized noise introduced by a
`slight regular fluctuation of the plasma intensity discussed below.
`;
`Some measurements of the stability of the COD were made by tuning the monochromator T
`wavelength to a spectral feature (emission line or background), storing the PMT current on a capacitor
`over an “exposure” period of 5 s, and then digitizing the accumulated voltage. This method was used:
`because intensity data could be obtained over a much shorter period compared to scanning over the '
`spectral line. Measurements were repeated 16 times to obtain statistical information.
`To spectrally resolve the intensity profile of the COD, the light at the monochromator exit slit was
`imaged with unit magnification on a vertically oriented photodiode array. The voltages from each ofthe 3
`512 photodiodes of the array were digitized and read into the computer memory for averaging.
`2.2. Gas cells
`
`v
`
`Two gas cells were used in this work. The large cell, shown in Fig. 2, was designed to withstand up I0 .
`7600 torr. The cw-laser beam entered the cell through a thick salt window and was focused using either 3
`lens or spherical mirror, as shown. The side windows were u.v. grade fused silica. These windows and the i
`rear flange of the cell were often replaced by various devices to test different analyte introduction
`techniques or nonlaser methods of starting the COD, some of which are discussed below. The front
`brass flange ofthe cell became very hot during prolonged operation and was water cooled to prevent the
`salt window from cracking. It was not necessary to cool the rest of the cell.
`.
`The small gas cell is shown in detail in Fig. 3. The plasma was generated inside a quartz tube 3.3~Cm
`long by 1.5-cm dia. to provide a small volume (~ 7 cm’) and to eliminate the need for separate windows
`to ignite the plasma and view the spectral emissions. These were accomplished by imaging directly’
`through the tube walls. The 0.18-cm wall thickness of the quartz tube limited the gas pressure to
`S 2800 torr. Analyte gases were introduced into the COD by injection either into a flowing gas stream.
`
`
`
`T
`
`Energetiq Ex. 2081, page 6 - |PR2015-01368
`
`Energetiq Ex. 2081, page 6 - IPR2015-01368
`
`

`
`Evaluation of the continuous optical discharge
`
`Table 1. Experimental apparatus and conditions
`
`669
`
`A. COD generation apparatus
`(1) cw-CO2—1aser
`wavelength
`power
`beam diameter '
`beam divergence
`(2) Nd’:YAG pulsed laser (ignition)
`wavelength
`pulse duration
`pulse energy
`repetition rate
`(3) Sample cells
`material
`size
`large cell (Fig. 2)
`small cell (Fig. 3)
`gas pressure measurements
`
`(4) Optics
`106 pm focusing lens
`
`1.06 pm focusing lens
`cw-CO2 input window
`(both cells)
`side windows (large cell)
`B. Detection system
`(1) Spectrometer
`gratings
`
`slit width
`slit height
`second order filter
`
`(2) Photomultiplier tubes
`
`(3) Chart recorder
`(4) Chopper
`(5) Photodiode array
`‘
`evaluation board
`C. Solid sample introduction
`(1) Nd:YAG pulsed laser
`wavelength
`pulse duration
`pulse energy
`repetition rate
`(2) Lens
`
`Advanced Kinetics MSL-100
`10.6 um
`45 W typical
`6 mm
`6 mrad
`
`Quanta-Ray DCR-lA
`1.06 pm
`15 113
`250 ml/pulse
`1 Hz to initiate COD
`
`brass
`
`2OL><10H ><1()Wcm
`9 L x 9 H X 9 Wcm
`MKS 254 flow controller/Barntron,
`0—l0000 torr head
`
`2.54 cm dia., 2.54 cm ll, positive meniscus
`ZnSe, anti-reflection coated for 10.6 llm
`2.54 cm dia., 7.5 cm ll
`5 cm dia., 1.25 cm thick NaC1
`
`1cm thick, Scm dia. quartz
`
`Spex industries 05 m Czerny—Turner
`1200 lines mm”, 500 nm blaze, A 2 350 nm
`3600 lines mm“, 300nm blaze, A < 350nm
`2 10 pm
`100 ;im~2 cm
`Corning CS 3-70, A 2 600 nm
`RCA C31034, 1400-1600 V, J. 2 350 nm
`Hamamatsu R955, 800~1 200 V, A $ 350 nm
`Houston Instruments, Omniseribe
`Laser Precision CTX-534
`Reticon RLSl2S
`RC1024S-2
`
`ILS NT-90H
`1.06 um
`15 ns
`75 mJ/pulse (attenuated)
`10 Hz
`10 cm 11
`
`4uzemim;e,n««mmmuaggmu:rmmum§:«w.w~anmmaw=mm.,_,
`
`
`_._%*
`
`‘ctssing through the cell from front to rear or into a static gas fill using the rear latex seal shown in Fig. 3.
`because it was melted by the large fraction of l0.6‘um
`he seal was positioned out of view of the COD
`iation transmitted through the plasma volume. The front salt window flange and the rear analyte
`ection port were water cooled. The quartz tube was air cooled using convection and only a slight
`
`
`
`Used Nd:YAG pulses from striking and damaging the tube walls.
`
`3. RESULTS
`
`1. Plasma generation
`In this section, results are presented describing some characteristics of plasma generation
`Xe using combined lower laser powers and lower gas pressures than previously reported.
`
`Energetiq EX. 2081, page 7 - IPRZO15-01368
`
`Energetiq Ex. 2081, page 7 - IPR2015-01368
`
`

`
`670
`
`DAVID A. CREMERS er al.
`
`YGAS FLOW PORT
`
`
`
`Fig. 3. Diagram of the small gas cell, photographed in Fig. 1, used in analytical experiments with a
`static Xe fill. Gas samples were introduced through the injection port using a syringe. Solid samples
`were laser ablated from a metal strip positioned below the COD (see Section 3.3). WCC: water cooling
`coil.
`
`Xenon was used because of the difficulty of producing the plasma in Ar at the low pressures’:
`(<3 atm) and low laser powers (< 70 W) available to this study. A minimum of 80Wis;
`needed to generate the COD in Ar at 3040 torr
`
`requirements. The maintenance threshold of Xe is lowest of all stable inert gases because ofits *
`relatively low ionization potential and thermal conductivity. The cw-CO2 laser beam was;
`imaged i11to the gas cell horizontally because of the significantly greater cw maintenanc
`thresholds required to generate the COD with a vertical beam
`Although Xe is expensiv
`and would not normally be employed in a routine analysis, the results obtained here shoul
`indicate general trends applicable to a COD generated in other inert gases.
`Initial attempts at generating the COD used a 2.8-cm focal length spherical copper mirro
`similar to those used in other studies [3, 4, 10]. The resulting plasma was unstable, at
`indicated by large irregular intensity fluctuations, and remained operating for at most 30
`This behavior was attributed to the small diameter (6 mm) of the CO2 laser beam. The CO
`intercepted a significant portion of the central intense region of the incoming laser beam
`before it reached the mirror. This not only reduced the beam intensit
`
`'
`
`
`
`In addition, with mirror focusing strong convection currents and density gradients from th
`COD distort the incoming laser beam [17] and this effect will be more pronounced with
`smaller beam diameters. Instability in the COD noted here using mirror focusing wa
`probably not as apparent in previous work because of the larger laser beam diameters, whic
`were typically 2 cm or more [3, 7, 19]. In this study, a stable plasma was generated at lower X
`pressures and laser powers than
`
`
`
`
`
`material because of its low absorption at 10.6 um and absence of “thermal runaway’
`exhibited by germanium at high cw laser powers.
`For a C0, laser beam power of 45 W, the plasma could be reliably maintained in Xe 21
`pressures between 1335 and 3200 torr using the large cell. With the small cell the minimum
`pressure was about 1150 torr. The lower maintenance threshold results from insulficlell
`absorption of the laser beam by the plasma to compensate for energy losses due to thermal
`
`Energetiq Ex. 2081, page 8 - |PR2015-01368
`
`Energetiq Ex. 2081, page 8 - IPR2015-01368
`
`

`
`Evaluation of the continuous optical discharge
`
`671
`
`Q_..
`’ uctivity, convection, and radiation [16]. The upper threshold results from increased
`ative losses at the higher gas densities.
`he COD was ignited in most cases using the laser spark generated by focusing a 2 MW
`
`YAG pulse within the focal volume of the 10.6-pm beam. Careful alignment of the focal
`‘mes was necessary to produce ignition and movements of 0.03 cm were critical. Small
`ges in the alignment led to the observation of unique phenomena: one pulse would start
`
`5 e COD ignited but quickly went out. Trial and error proved to be the best method of
`jng the focal volumes. Although most convenient, plasma ignition need not be
`
`t the plasma using a tesla coil connected to the assembly shown in Fig. 2 consisting of a
`if of electrodes positioned inside a bellows and flange assembly. The flange was mounted in
`
`
`
`tivated, a bright series of sparks was formed between the two electrodes. As the spark was
`ually swept through the focal volume of the C0, laser beam, the COD was easily ignited.
`e‘ electrodes were then retracted out of the way.
`Unsuccessful attempts were made to generate the COD with up to 60W of 1.06-um
`diation from a multimode cw-Nd:YAG laser. Because laser heating of a plasma via inverse
`emsstrahlung varies as /12 [23], the failure to form the COD wasprobably due to the 100
`es lower absorption of the plasma at 1.06 pm compared to 10.6 pm.
`
`._ Plasma characteristics
`
`3.2.1. Eflects of pressure and laser power. Figure 4 shows the Xe COD spectrum, not
`forrected for detection system spectral response, over the 200-940 nm region at
`two
`ssures. Emissions from neutral and once-ionized Xe were observed having upper levels in
`he 9.6—l3.9 eV range. Because of the high background intensity, due to the high COD
`emperature and electron density, the signal-to—background ratio (SBR) was moderate for
`most lines. Relative intensities of some Xe lines, corrected for detection system response, are
`‘sted in Table 2. In general, XeI intensities of 100 or more on a scale in which the strongest Xel
`me is 10‘ [34], could be observed. Because of the high intensity of the COD, slit widths less
`han 10 pm were used to prevent saturation of the detection system.
`The high electron density of the COD broadens emission line widths via the Stark effect.
`
`Table 2. Relative intensities of some Xe lines observed in the COD*
`
`
`
`
`
`Rel. Int. Wavelength (nm)Wavelength (nm) Rel. Int.
`
`0.006
`529.22 (Xe II)
`0.030
`450.10
`0.044
`764.20
`0.018
`452.47
`0.839
`823.16
`0.013
`462.43
`0.029
`826.65
`0.033
`467.12
`0.259
`823.01
`0.037
`473.42
`0.146
`834.68
`0.013
`480.70
`0.070
`840.92
`0.008
`482.97
`1.000
`881.94
`0.016
`484.33
`0.278
`895.23
`0.010
`491.65
`
`
`
`0.008 904.55492.32 0.344
`
`.
`
`*Xe pressure 2 2000 torr; cw~CO2 laser power = 45 W. Lines are
`from Xe I unless otherwise indicated.
`
`[34] A. R. STRIGANOV and N. S. SVENTITSKII, Tables ofspeczral Lines ofNeutral and Ionized Atoms. IFI/Plenum,
`New York (1968).
`
`Energetiq Ex. 2081, page 9 - |PR2015-01368
`
`Energetiq Ex. 2081, page 9 - IPR2015-01368
`
`

`
`672
`
`DAVID A. CREMERS er al.
`
`Measurements of the half-width of the 450.10 and 834.68-nm Xe I and the 529.22-nm Xel
`lines were obtained over the pressure range of 1402 to 2204 torr. The average widths wet‘
`0.13, 0.17 and 0.13 nm, respectively. Only the 450.10-nm line showed a pressure dependencg
`decreasing from 0.14 to 0.12 as the pressure decreased. The instrumental resolution 0
`0.086 nm, determined using a low pressure Hg lamp, was not removed from thes
`experimental line widths. ,
`V
`The COD resides near the focus of the CO2 laser beam that supplies it with energy, at
`position dependent upon several parameters including laser energy and gas pressure
`Changes’ in these parameters moved the plasma along the axis of the focused laser beam. F0
`example, decreasing the Xe pressure from 3175 to 1336 torr moved the plasma 0.5 mm
`toward the focus. Similarly, at a pressure of 2000 torr, changing the laser power from 25 ti,
`60 W moved the plasma away from the focus by 0.3 mm. These results are understood by
`considering the motion of the plasma in a parallel laser beam: the plasma moves toward};
`away from the laser depending on whether the beam intensity is above or below the thresholfi
`to overcome losses, respectively [18]. The intensity gradient obtained by focusing represent‘
`a stable trap for the plasma. The angle of the converging beam and the absorption of the
`plasma determine how far the plasma resides from the focus. A significant fraction of thy
`1O.6—;4m beam is not absorbed by the COD
`At the steady state distance (x) from the focus
`the laser beam intensity absorbed by the plasma (AI ) is approximately given by AI = I0 (x) [1
`—exp(— pr)], where I0(x) is the beam intensity incident on the plasma, ;1 is the plasma
`absorption coefficient (cm‘ ‘), and r is the plasma diameter. Absorption is balanced by energy
`losses from thermal conduction, convection, and radiation. The absorption coefficient varies
`with pressure (p) as [1 z p2 [16] so that as the pressure increases, A1 also increases for a giv 0
`laser power, and the plasma moves further away from the focus, into a region of reduc ’
`intensity, until absorption is again balanced by the loss mechanisms. In the same way, as the
`laser power is increased at constant pressure, AI increases and the plasma moves toward t e
`focusing lens, as observed.
`
`2006- torr
`
`
`
`
`
`
`I404 torr
`
`200
`
`275
`
`350
`
`550
`
`750
`
`950
`
`Wavelength, nm
`Fig. 4. Spectrum from 200 to 950 nm of the Xe COD at two pressures, not corrected for detection
`system response. Different PMTs and gratings were used to monitor the regions above and below
`350 nm. The horizontal line under each spectrum indicates the baseline signal with the COD off.
`Emission lines marked by a triangle are off-scale. A filter was placed at the entrance slit beginning at
`the wavelength indicated by the arrows to eliminate the second order spectrum.
`
`
`
`Energetiq Ex. 2081, page 10 - |PR2015-01368
`
`Energetiq Ex. 2081, page 10 - IPR2015-01368
`
`

`
`Evaluation of the continuous optical discharge
`
`673
`
` (°.V) XeI,XeI,BG‘Intensities,V
`
`l6
`
`20
`
`24
`
`
`
`Ratios
`
`28
`
`32
`
`lO‘2- Pressure,
`
`torr
`
`Fig. 5. Variation of the intensities of (1) 450.10 nm Xe I and (1') adjacent background at 448.00 nm
`(BG 1), and (2) 529.22 rim Xe II and (2') adjacent background at 527.00 nm (BG II), and the ratios (3)
`[Xe I/BG I] and (4) [Xe II/BG II], as a function of xenon gas pressure. The laser power was 46 W.
`
`
`
`ns was adjusted at each pressure to maintain the most intense portion of the plasma over the
`ntrance slit. As the pressure was decreased from 3175 to about 1900 torr the intensities from
`11 three sources increased. As the pressure was reduced further, the Xe I intensity increased
`ntil about 1700 torr and the Xe II intensity peaked at 1600 torr. The net result was that the
`attic of the emission line to background intensity steadily increased as the pressure was
`educed from 3175 to about 1 500 torr. These observations agree with measurements reported
`elow, which indicate the temperature of the COD increases somewhat with decreasing
`ressure. Different behavior was observed by changing the laser power. As the power
`ecreased from 55 to 25 W, the intensities of Xe I (450.10 nm), Xe 11 (529.22 nm), and the
`djacent backgrounds all increased slightly, but the ratios of these lines to background
`ernained essentially constant.
`3.2.2. Temperature. Detailed measurements of the COD temperature have been reported
`reviously [8, 9, ll, 17]. The intention here was to determine how the temperature and
`lectron density varied as certain experimental parameters were changed to ascertain the
`tability of this source. The COD temperature was measured using the Saha equation and a
`oltzmann plot. Because of the asymmetry of the horizontal plasma, the resulting data were
`or Abel inverted. However, only light from the center of the plasma image (l0—;tm wide by
`00-um high) was monitored by masking the monochromator slits. The resulting tempera-
`ures are therefore population averaged as discussed by BOUMANS [3 5]. The electron density
`as determined by seeding the Xe with about 1.3-2.0 ‘X, hydrogen gas by volume and
`measuring the width of the H, line at 656.28 nm. The H,, line (486.13 nm), which yields more
`ccurate densities [36], was not used because of several Xe I interferences and the about
`ixfold weaker intensity of this line compared to H,. To minimize perturbation of the COD,
`nly the smallest amounts of hydrogen needed to make a measurement were added. Above
`“0,
`the plasma noticeably decreased in intensity and was extinguished above 5 ‘,’/,’,
`oncentration. Using the formula of GRIEM [36], the electron density was deduced by an
`terative process with the Saha temperature. The 450.10-nm Xe I and 529.22-nm Xe 11 lines
`Were used for the Saha analysis. At the pressures of 2766,2130 and 1601 torr, the temperature
`Was measured to be 9630, 10 200 and 10 375 K, respectively, using 45 W of laser power. The
`lectron density also varied slightly: at the respective pressures it was 7.39, 8.95 and 7.31
`X 10' "‘ cm’ 3, which corresponds to about 0.12 "/0 ionization of the Xe. The width of the H,
`
`
`
`35] P. W. J. M. BoUMANs, Spectrochemicul Excitation. Hilger and Watts, London (1966).
`__ [36] H. R. GRIEM, Plasma Spectroscopy. McGraw~Hill, New York (1964). The formula used to calculate the
`electron density (nc) is given on p. 305: ne = C(nE, T)A;l3’z. Here, All is the full Stark line width and C(ne,T) a
`coefficient tabulated on pp. 538~539.
`
`=k<3)1«o:z.4r
`
`Energetiq Ex. 2081, page 11 - |PR2015-01368
`
`Energetiq Ex. 2081, page 11 - IPR2015-01368
`
`

`
`
`
`674
`
`DAVID A. CREMERS et at.
`
`line varied in the range 0.754185 nm as the pressure varied. At 2130 torr, the COD
`temperature changed by less than 2 ‘Z, and the electron density by less than 10 "/3 as the [age
`power increased from 32 to 48 W.
`.
`To obtain a good measure of the Boltzmann temperature of the plasma, the spread in (11
`upper energies. of the emitting levels should be at least 2 eV. Unfortunately, the majority O
`Xel lines for which the oscillator strengths have been measured correspond to upper level
`grouped between 11.4 and 12.0 eV [37]. However, when about 66 ‘X, Kr was added to the X
`COD at least 13 lines of Kr I were observed corresponding to upper levels spanning the range
`from 11.44 to 13.49 eV for which the oscillator strengths are known to i 50 E’/.; (Table 3). Th '
`relative intensities of these lines were measured to produce the Boltzmann plot shown in Fig
`6. The integrated intensities were determined by tracing the lines on paper, cutting them on
`Table 3. Data for Kr I lines used to construct Boltzmann
`plot*
`
`Wavelength (nm)
`427.40
`435.14
`440.00
`445.39
`446.37
`587.09
`760.15
`768.52
`769.45
`785.48
`810.44
`811.29
`850.89
`
`g1
`5
`1
`5
`5
`3
`5
`5
`1
`3
`3
`5
`7
`3
`
`10‘ 9 x A (s‘ 1)?
`0.026
`0.032
`0.020
`0.0078
`0.023
`0.018
`0.31
`0.49
`0.056
`0.23
`0.13
`0.36
`0.24
`
`Eu(e\/)1
`12.82
`13.49
`13.46
`12.82
`12.81
`12.14
`11.55
`12.26
`11.53
`12.14
`11.44
`11.44
`12.10
`
`‘Data from: Wavelengths and Transition Probabilities for
`Atoms and Atomic Ions Part II, Transition Probabilities,
`NSRDS-NBS 68, US. Government Printing Office,
`Washington, DC (1980).
`1‘Here, g and E“ are the statistical weight and energy,
`respectively, of the upper emitting level corresponding to the
`transition having transition probability A. The estimated
`
`uncertainty of A is 150“/0.
`
`
`I
`
`:2
`
`I3
`
`Eu, eV
`
`Fig. 6. Boltzmann plot of Kr 1 lines in the Xe/Kr COD used to measure the plasma temperature.
`Only t