spectroscopic determination of rotational temperature in c_2h_4/c_2h_2/o_2 flames for diamond growth...

Spectroscopic determination of rotational temperaturein C2H4=C2H2=O2 flames for diamond growth with and

without tunable CO2 laser excitation

X. N. He,1 X. K. Shen,1 T. Gebre,1 Z. Q. Xie,1 L. Jiang,2 and Y. F. Lu1,*1Department of Electrical Engineering, University of Nebraska-Lincoln, Lincoln, Nebraska 68588-0511, USA

2Department of Mechanical and Automation Engineering, 3rd School,Beijing Institute of Technology, Beijing, 100081, China

*Corresponding author: [email protected]

Received 4 January 2010; revised 22 February 2010; accepted 22 February 2010;posted 23 February 2010 (Doc. ID 122046); published 11 March 2010

Optical emission spectroscopy (OES) and spectroscopic temperature determination were carried out tostudy C2H4=C2H2=O2 flames used for diamond deposition with and without an excitation by a wave-length-tunable CO2 laser. Strong emissions from C2 and CH radicals were observed in the visible rangein all the acquired OES spectra. When the flames were irradiated by using a continuous-wave (CW)CO2 laser at a wavelength of 10:591 μm, the emission intensities of the C2 and CH radicals in the flamesincreased owing to the laser excitation. The CO2 laser was also tuned to a wavelength of 10:532 μm topreciselymatch the resonant frequency of the CH2-wagging vibrational mode of the C2H4 molecules. OESspectroscopy of the C2 and CH radicals were performed at different laser powers. The rotationaltemperatures of CH radicals in the flames were determined by analyzing the spectra of the R branchof the A2Δ → X2Π ð0;0Þ electronic transition near 430nm. The deposited diamond thin-films werecharacterized by scanning electron microscopy, stylus profilometry, and Raman spectroscopy. The deposi-tion mechanism with and without the CO2 laser excitation was discussed based on the OES spectralresults. © 2010 Optical Society of America

OCIS codes: 120.1740, 300.2140, 300.6390.

1. Introduction

Chemical vapor deposition of diamond films has beenextensively studied since the discovery of simple low-pressure methods to deposit polycrystalline films ona variety of substrates [1,2]. The underlying reason isthat diamond coatings on different materials andequipments are very useful because of the superiorproperties of diamond, including outstanding optical,electrical, and mechanical properties that make itattractive for various applications [3]. As a result ofthese properties, diamond film research has contin-ued to attract more and more interest over the pastdecade. The combustion-flame deposition methodhas received considerable attention since its discov-

ery by Hirose et al. [4,5] because of its simplicity andhigh growth rate. In the combustion-flame method,C2H2 and O2 have been commonly used as precursorsto produce proper combustion flames for depositionof diamond films. However, other carbon-based pre-cursors, such as C2H4 and C3H6, can also be used todeposit diamond [6].

During the combustion-flame deposition ofdiamond, a number of excited chemical species arepresent in flames above substrate surfaces. Some ex-amples are CH, C2, OH, and CN free radicals [7–10].The presence and interactions of these species abovethe substrate surface is an important factor in thecombustion-flame deposition of diamond films. Sinceradicals or, equivalently, excited chemical species,can be detected and characterized by optical emis-sion spectroscopy (OES), it is widely employed to

0003-6935/10/091555-08$15.00/0© 2010 Optical Society of America

20 March 2010 / Vol. 49, No. 9 / APPLIED OPTICS 1555

identify the excited atomic and molecular speciesin diamond deposition flames and to study param-eters such as relative emission intensities andtemperatures.OES is a spectroscopic technique that examines

the wavelengths of photons emitted from atoms ormolecules during their transitions from an excitedstate to a lower energy state. In a combustion flame,the energy of a molecule can change via rotational,vibrational, and vibronic (combined vibrational andelectronic) transitions. More specifically, moleculesare heated and excited by a high temperature thatresults from a highly exothermic combustion reac-tion. When the molecule falls back down from theexcited state, energy is re-emitted in the form of aphoton. The wavelength of the photon is determinedby the energy difference between the two states.These energy transitions of molecules often lead toclosely spaced groups of many different spectrallines, known as spectral bands.OES spectra [11] can provide information in a

temperature range of 2000–8000K, where atomicspectra are not strong enough to ensure a goodsensitivity [12]. Emission spectra from various mole-cules have been used for temperature determina-tions in combustion-flame diagnostics. The UV OHspectrum was employed by de Izarra [13] as a mole-cular pyrometer to determine flame temperatures.The C2 Swan band is also used for temperature de-termination in flames. Emission spectra of the C2Swan band (band head 516:611nm) were employedfor evaluation of the rotational temperature in differ-ent plasma sources [14]. The rotational temperature,Tr, of molecular species in a plasma is generally veryclose to the gas kinetic temperature. This is due tostrong coupling between translational and rotationalenergy states [9,14]. To determine the rotationaltemperature of a premixed C2H4=C2H2=O2 flame,one method is the well-known Boltzmann plot whena high-resolution optical spectrometer is available toresolve the rotational lines.In our previous studies, laser-assisted combustion-

flame deposition of diamond films was performed toimprove the deposition rate of diamond thin films. Awavelength-tunable CO2 laser was used to irradiatethe C2H4=C2H2=O2 combustion flames during the de-position of diamond films. Although C2H2=O2 flameis efficient in diamond growth, there is no suitablewavelength in the tuning range of the CO2 laserthat can excite molecules in the C2H2=O2 flame.C2H4, which is one of the suitable precursors fordiamond deposition [7,15], was added in the flameto resonantly match the laser wavelength by itsCH2-wagging vibrational mode to enable CO2 laserexcitation. Two different wavelengths of the tunableCO2 laser were used for comparison with the caseswithout laser excitation: 10:591 μm (equivalent to awavenumber of 944 cm−1, the common CO2 laserwavelength) and 10:532 μm (equivalent to a wave-number of 949 cm−1). The C2H4 molecule has an IRabsorption band corresponding to the CH2-wagging

vibration mode at a frequency of 949 cm−1. Since thisfrequency is very close to the common CO2 laser line,10:591 μm, the CO2 laser can be used to excite theC2H4 molecules at this wavelength because of thebroadening of the absorption band and can be usedto promote the production of species beneficial todiamond deposition. Furthermore, the wavelengthof 10:532 μmproduced by the CO2 laser was also usedto resonantly excite the CH2-wagging mode. Pre-vious results have shown that resonant excitationby a CO2 laser can improve both growth rate andfilm quality, indicated by sharper diamond peak inRaman spectra. However, the mechanism of the en-hancement in the diamond deposition is still notclear. Therefore, OES was performed in this studyto identify the excited species and determine the ro-tational temperatures in the deposition flames withand without CO2 laser excitations. OES spectra werealso used to calculate relative emission intensities.The electronic transition of CH [A2Δ − X2Π ð0; 0Þ,center near 430nm] was used to determine the rota-tional temperatures using the Boltzmann plot meth-od. The CH radicals were chosen because they haverelatively high emission intensity. Previous studiesby Welter [8] and Firchow [9] have shown that theyare one of the species that correlate best with dia-mond growth. The diamond films deposited on tung-sten carbide with cobalt binder (WC-Co) substrates(BS-6S, Basic Carbide Corp., containing 6% cobalt)were characterized by using scanning electronmicroscopy (SEM), stylus profilometry, and Ramanspectroscopy.

2. Experiments

Figure 1 shows a schematic diagram of the CO2 laser-assisted C2H4=C2H2=O2 combustion-flame deposi-tion system used in this study. A CW tunable CO2laser (STS 1000T, PRC Company) was used to irradi-ate the flames. The initial diameter of the laser beamwas 14mm. A ZnSe convex lens (f ¼ 25:4 cm) wasplaced 29:2 cm in front of the flame to focus the laserbeam to a diameter of approximately 1mm. A weld-ing torch with a nozzle diameter of 1:5mm was usedto produce a C2H4=C2H2=O2 flame [with a gas ratio0:23=0:49=1:21 slpm (standard liters per minute)].These precursor gases were mixed in the torchthrough three mass flow controllers (Smart-Trak ser-ies 100, SIERRA Instruments, Inc). The CO2 laserbeam was directed perpendicular to the flame feath-er and parallel to the substrate. In this study, we onlyused unseeded substrates for diamond film deposi-tion. There was no diamond powder dispersed ontothe WC-Co substrate surfaces, which has a surfaceroughness (Ra) of 0:4 μm. The substrates werecleaned in a supersonic bath of acetone for 15 minbefore deposition. During the deposition, the sub-strates were placed on a hollow brass block withwater cooling. The substrate was placed in the feath-er region for diamond growth [16]. The distancebetween the substrate surface and the inner coneof the flame was around 1mm in all experiments.

1556 APPLIED OPTICS / Vol. 49, No. 9 / 20 March 2010

The surface temperature of the substrates duringdeposition was monitored by a pyrometer (OS3752,Omega Engineering, Inc.) to ensure a steady sub-strate temperature at approximately 700–800°C.The diamond depositions were carried out for15 min under several different laser powers. Thethicknesses of the deposited films were characterizedby a stylus profiler (XP-2, AMbios Technology).Raman spectroscopy (inVia Raman microscope,Renishaw Inc.) was used to characterize the depos-ited diamond films.Optical emission spectra of flames were collected

during diamond deposition in a direction perpendicu-lar to the flame feather. The flame was imagedthrough one UV grade quartz lens (f ¼ 10 cm) andintroduced into a spectrometer (Andor ShamrockSR-303i-A) via a slit 0:5mm × 10 μm in size. Thespectral information was collected by an intensifiedcharge-coupled-device (ICCD, Andor iStar DH-712)and analyzed by a personal computer. The wide-range OES spectra for studying the overall emissionintensities were obtained using a grating of150 lines=mm (spectral resolution: 0:52nm), whereasthe high-resolution spectra for rotational tempera-ture calculations were obtained by using a gratingof 2400 lines=mm (spectral resolution 0:02nm).Micrographs of the deposited diamond films wereobtained by using an SEM (S-4700, Hitachi High-Technologies Corporation).

3. Method for Determining Rotational Temperature

Figure 2(a) shows a typical OES spectrum of the Rbranch of the A2Δ − X2Π ð0; 0Þ band of the CH radi-cals. The population density distribution of radicalsin a plasma can be described by the Boltzmannequation [14,17]. The emission intensity I of a spec-tral emission line within a rotational band due to atransition from an upper state J0 to a lower state J″

can be described by using Eq. (1). The method usedfor calculations are described in detail by Kim and

Cappelli [7], Pellerin et al. [14], and Moon andChoe [17]:

I ¼ CSJ0J″λ−4 exp�−

EJ0

kBTr

�: ð1Þ

In the above equation, I is the relative emission in-tensity of a rotational line obtained from the experi-mental spectra, C is a proportionality constant thatis the same for all rotational transitions within aband, SJ0J″ is the rotational intensity factor orHöln–London factor, λ is the wavelength of theemitted spectral line, EJ0 is the rotational energyof the initial level, kB is the Boltzmann constant,and Tr is the rotational temperature [7,14,17].

By taking the natural logarithm of the aboveexpression, the expression for the Boltzmann plotis obtained as shown in Fig. 2(b). Notice that thevalue −1=Tr is the slope of the Boltzmann plot oflnðIλ4=SJ0J″Þ versus EJ0=kB, which should be astraight line [7,14,17]. Since the values of SJ0J″

and EJ0 for the transitions in the CH molecular spec-tra are constants that can be calculated for eachrotational transition, the value of the rotational tem-perature can be found by calculating the slope of theBoltzmann plot:

lnðIλ4=SJ0J″Þ ¼ −

1Tr

EJ0

kBþ lnC: ð2Þ

For the R branch of the A2Δ − X2Π ð0; 0Þ band ofCH, which is shown in Fig. 2(a), centered near430nm used for the CH rotational temperature cal-culation, Kim and Cappelli [7] have given both theSJ0J″ and EJ0 values. The SJ0J″ andEJ0 values can alsobe calculated from formulas and rotational transitiondata given by Pellerin et al. [14], Herzberg [18], andKovacs [19].

It is important to define the quantities involved inthe use of OES as a diagnostic tool. A portion of the

Fig. 1. (Color online) Schematic diagram of the experimental setup for OES measurements. The inset image shows the structure of aflame in its unperturbed state and in the configuration used for diamond deposition. There are three distinct regions in a flame: (1) theinner flame, (2) the feather, and (3) the outer diffusion flame.


flame volume emission is focused onto the slit of aspectrometer via a lens or mirror optical system.The calibration of the detected signals are normallycarried out by substituting a surface emitting conti-nuum source of known spectral radiance, like a cali-brated tungsten strip lamp, in place of the volumeemitting source. This calibrated lamp is positionedat the focus of the emission optical system, andthe amount of its radiation per unit time per unit so-lid angle per unit wavelength band per unit area ele-ment is accurately known. Thus, the signal from thespectrometer and detection system is calibrated [20].

4. Results and Discussion

The shapes of the combustions under different laserpower excitations are shown in Fig. 3. In Fig. 3(a), theoptical emission of the flame went through the slitand then was reflected by a totally reflective gratingonto the ICCD. The flame was irradiated by a10:591 μm CO2 laser at different powers (from left toright) of 0 (no laser), 200, 400, 600, 800, and 1000W.From the images of the flames, we can see that theinner flame became shorter as the laser power in-

creased. The original (no laser) inner flame wasaround 5mm. However, it changed to ∼3:5mm whenirradiated by a CO2 laser power of 1000W. The flamechanged gradually as the laser power increased.However, as shown in Fig. 3(b), the flame becamemuch shorter when a 10:532 μm laser beam was usedto irradiate the flame. The inner flame changed from∼5 to ∼2mm as the laser power changed from 0 to1000W. The differences between the 10.591 and10:532 μm laser excitations were due to the differentlevels of wavelength matching with the 949 cm−1

CH2-wagging vibrational mode. 10:532 μm is moreprecisely in resonance with the CH2-wagging modethan 10:591 μm. When gas precursors left from thenozzle outlet and traveled into the inner flame, thelaser beam was aligned orthogonally to the flame in-ner feather, and the diameter laser beam of 1mmsize was positioned immediately under the flamenozzle. The flame shape variation proves that C2H4molecules were resonantly excited by the 10:532 μmlaser beam when C2H4 was mixed in the flame.Furthermore, more molecules were excited to higherexcited states under the 10:532 μmCO2 laser irradia-tion. Higher-energy-state molecules then relax tolower states, and thus produce much stronger opticalemission. Since more molecules from lower stateswere excited to higher states, combustion reactionsbecame faster, and the inner feather became shorterthan without laser excitation.

SEM microscopy was used to characterize themorphologies of the deposited diamond films.Figures 4(a)–4(c) show the diamond films depositedwithout laser excitation, with 800W 10:591 μm CO2laser excitation, and with 800W 10:532 μmCO2 laserexcitation, respectively. From Fig. 4, it can be seen

Fig. 2. (a) R branch of the CH band used in calculations;(b) Boltzmann plot to obtain the rotational temperatures.

Fig. 3. (Color online) Optical images of the C2H4=C2H2=O2

combustion flame with laser irradiations at (a) 10.591 and(b) 10:532 μm. Yellow bars represent WC substrate positions.


that there is an increase in the average size of thediamond grains deposited when the CO2 laser is in-troduced into the flame. In Figure 4(a), the average

grain size is less than 1 μm, whereas that in Fig. 4(b)is around 1:5 μm, and in Fig. 4(c) 2–3 μm. In all cases,the temperature of the WC-Co substrate was main-tained at 780–800°C. The deposition time was fixedat 15 min. The 10:532 μm CO2 laser excitation pro-vided more significant improvement in depositionrate than the 10:591 μm CO2 laser excitation, whichcan be seen from the stylus profiles in Fig. 5. Fromthe profiles, we can see that the diamond films de-posited without laser excitation, with the 800W10:591 μm CO2 laser excitation, and with the800W 10:532 μm CO2 laser excitation are 4, 8, and11:5 μm, respectively. Therefore, the increase in thedeposition rate was not due purely to thermal effectson the substrate, but more likely to the enhancementin the population of exited species that are beneficialto diamond formation.

The characteristic fingerprint of diamond is a sin-gle sharp Raman line at 1332 cm−1 [21]. As can beseen in Fig. 6, Raman peaks are slightly shifted ow-ing to the stress in the films. The Raman peak inten-sity increased significantly when 800W 10:591 μmand 10:532 μm CO2 laser excitations were appliedto the flame. The diamond line and the graphite (G)and disorder (D) peaks are all present [21]. In addi-tion, there is also a small, broad band at ∼1150 cm−1

that is evidence for the existence of nanosized dia-mond crystals [22,23]. This broad band becameweaker when the CO2 laser excitation was addedto the flame and disappeared when the 10:532 μmCO2 laser excitation was applied. Therefore, therewere fewer nanosized diamond crystals in the dia-mond films deposited with the 10:532 μm CO2 laserexcitation. From the three Raman spectra, it can beseen that the 10:532 μm CO2 laser excitation cangreatly improve the diamond film quality, which isalso made evident by the decreasing G band around1575 cm−1 [21]. 10:591 μm CO2 laser excitation canalso improve the diamond film quality, but not as sig-nificantly as the 10:532 μm excitation.

The C2H4=C2H2=O2 combustion flame had muchstronger optical emissions when the CO2 laser wasintroduced into the flame. The spectra in Fig. 7 showthe increases in the intensities of CH and C2 speciesas the laser power increased. All the spectra were

Fig. 4. SEM microimages of diamond films with (a) no laserexcitation, (b) 800W 10:591 μm CO2 laser excitation, and(c) 800W 10:532 μm CO2 laser excitation.

Fig. 5. Stylus profiler graphs of deposited diamond films: (a) without CO2 laser excitation, (b) with 800W 10:591 μmCO2 laser excitation,and (c) with 800W 10:532 μm CO2 laser excitation.


taken at the point 0:5mm above a WC substrate. InFig. 7(a), the inset spectrum is a highly resolved spec-trum of the CH band, including the R (pure rotation),

P, and Q branches. The OES spectra were takenunder the condition in which the C2H4=C2H2=O2combustion flame was excited by the 10:591 μmCO2 tunable laser excitation remotely matchingthe CH2-wagging mode. The emission intensity ofC2 (centered at 516nm) was enhanced by 176%and 272% at laser powers of 400 and 1000W, respec-tively. This increase in the population of excited spe-cies in the flame is due to the excitation of the flamethrough resonant absorption of laser energy by theC2H4 molecules. Figure 7(b) shows the optical emis-sion intensities at different powers from 0 to 1000Wof the 10:532 μm CO2 laser. The 10:532 μm CO2 lasercan excite the intensity of C2 by a factor of 292% atthe laser power of 400W, and to 658% at the power of1000W, which was much stronger than those with10:591 μm excitation.

Figure 8(a) shows the calculated CH rotationaltemperature distribution along the vertical axis of

Fig. 6. (Color online) Raman spectra of diamond films with nolaser excitation, 10:591 μm CO2 laser excitation, and 10:532 μmCO2 laser excitation.

Fig. 7. (Color online) Series of optical emission spectra ofC2H4=C2H2=O2 combustion-flame excited by 0, 200, 400, 600,800, and 1000W CO2 laser emission at a wavelength of(a) 10.591 and (b) 10:532 μm.

Fig. 8. (Color online) (a) CH rotational temperature distributionsalong the flame feather vertical direction with and without CO2

laser excitation; (b) optical emission intensity distributions ofCH radical (centered at 431nm) along the flame feather verticaldirection.


the feather flame. The three curves show the caseswith no laser excitation, 800W 10:591 μm CO2 laserexcitation, and 800W 10:532 μm CO2 laser excita-tion, respectively. The figure shows the CH rotationaltemperature distribution as a function of distancefrom the nozzle. The nozzle was placed at a pointof x ¼ 0mm. The substrate was positioned at differ-ent places because of the different inner flamelengths under various excitation conditions. The lastpoint of each curve was taken at 0:5mm above thesubstrate surface under that condition. From thetrend of the temperature distribution, it can be seenthat the CH rotational temperature gradually in-creased from the nozzle to the substrate surface asthe distance increased. The CH rotational tempera-ture at 1:5mm below the nozzle was as low as 2200Kbecause combustion reactions were not fully com-pleted, resulting in less heat generation. The tem-perature at 0:5mm above the substrate increasedwith the laser excitation. With the 10:532 μm CO2laser excitation, the temperature increased to thehighest point. Apparently, molecules were excitedto higher rotational energy states, and thus a higherrotational temperature was obtained. Figure 8(b)shows optical emission intensities at 431nm (CHband head) as a function of distances from the torch

without and with the 10.532 and 10:591 μm laser ex-citations. The flame with the 10:532 μmCO2 laser ex-citation is the shortest but has the highest intensity.Therefore, the intensity was highly increased by the10:532 μm CO2 laser excitation. The C2 emissionintensities have almost the same trend as the CHemission.

Figure 9 shows the calculated CH rotationaltemperatures as a function of excitation laser powerranging from 0 to 1000W with two wavelengths of10.591 and 10:532 μm. All spectra used for tempera-ture calculation were taken at 0:5mm above thesample surface with the presence of a substrate.Figure 9(a) shows the rotational temperature inthe flame excited by the 10:591 μmCO2 laser. The in-crease in the rotational temperature is not signifi-cant as the laser power increases. In Fig. 9(b),however, the CH rotational temperature obviouslyincreased when the 10:532 μm CO2 laser power in-creased from 0 to 1000W, owing to the much higherenergy coupling efficiency between the 10:532 μmCO2 laser and the flames.

Excited intermediate species in the flame and ahigher collision frequency could affect diamond nu-cleation. In this study, the film quality and growthrate were greatly improved by the CO2 laser excita-tion. More significant effects can be seen from the10:532 μm excitation, which has a perfect match withthe CH2-wagging mode. Both a higher growth rateand a higher flame temperature were observed.

5. Conclusions

The optical emission from the excited species in aC2H4=C2H2=O2 combustion flame was enhanced byintroducing CO2 laser excitation, and the CH rota-tional temperatures have a rising trend with increas-ing laser powers. The increase in the emissionintensity from the flames was found to be propor-tional to the laser power used. The laser excitationof the flame resulted in the deposition of diamondfilms with larger grains on the substrate. The CH ro-tational temperatures calculations showed that the10:591 μm laser excitation had relatively smallereffects on the CH rotational temperature. Great tem-perature change appeared when the flame wasexcited by the 10:532 μm laser beam. From the in-creasing OES intensities and temperatures, it is con-cluded that the CO2 laser can excite the C2H4molecules by IR absorption corresponding to theCH2-wagging vibrational mode of the C2H4 mole-cules centered at 949 cm−1, and this excitation is ben-eficial for diamond film depositions.

This work was financially supported by the U.S.Office of Naval Research (ONR) through the Multi-disciplinary University Research Initiative (MURI)program (N00014-05-1-0432).

References1. W. A. Yarbrough and R.Messier, “Current issues and problems

in the chemical vapor deposition of diamond,” Science 247,688–696 (1990).

Fig. 9. CH rotational temperatures at different CO2 laser powersfor C2H4=C2H2=O2 combustion flames with (a) 10.591 and(b) 10:532 μm CO2 laser excitations.


2. J. C. Angus and C. C. Hayman, “Low-pressure, metastablegrowth of diamond and ‘diamondlike’ phases,” Science 241,913–921 (1988).

3. D. Das and R. N. Singh, “A review of nucleation, growth andlow temperature synthesis of diamond thin films,” Int. Mater.Rev. 52(1), 29–64 (2007).

4. Y. Hirose, 1st International Conference on the New DiamondScience and Technology, R. Roy, R. Messier, J. E. Butler, and J.T. Glass, eds. (Materials Research Society, 1988), p 38.

5. Y. Hirose, S. Amanuma, and K. Komaki, “The synthesis ofhigh-quality diamond in combustion flames,” J. Appl. Phys.68, 6401–6405 (1990).

6. K. Tanabe, Y. Nishibayashi, T. Imai., A. Ikegaya, and N.Fujimori, “Deposition of diamond film from the O2—CH4

system,” in Science and Technology of New Diamond, S. Saito,O. Fukunaga, and M. Yoshikawa, eds. (KTK Scientific Pub-lishers/Terra Scientific Publishing Company, 1990), pp. 71–77.

7. J. S. Kim and M. A. Cappelli, “Temperature measurements inlow-pressure, diamond-forming, premixed flames,” J. Appl.Phys. 84, 4595–4602 (1998).

8. M. D. Welter and K. L. Menningen, “Radical density measure-ments in an oxyacetylene torch diamond growth flame,” J.Appl. Phys. 82, 1900–1904 (1997).

9. S. J. Firchow and K. L. Menningen, “Radical density measure-ments in an atmospheric pressure oxyacetylene torch,” J.Phys. D 32, 937–941 (1999).

10. S. Roy, J. DuBois, R. P. Lucht, and N. G. Glumac, “Hydroxylradical concentration measurements near the deposition sub-strate in low-pressure diamond-forming flames,” Combust.Flame 138, 285–294 (2004).

11. A. G. Gaydon, The Spectroscopy of Flames (Chapman &Hall, 1957).

12. S. Pellerin, J. Koulidiati, O. Motret, K. Musiol, M. de Graaf, B.Pokrzwka, and J. Chapelle, “Temperature determination

using molecular spectra simulation,” High Temp. Mater. Pro-cesses 1, 493–509 (1997).

13. C. de Izarra, “UV OH spectrum used as a molecular pyrom-eter,” J. Phys. D 33, 1697–1704 (2000).

14. S. Pellerin, K. Musiol, O. Motret, B. Pokrzywka, andJ. Chapelle, “Application of ð0; 0Þ Swan band spectrum of C2

for temperature measurement,” J. Phys. D 29, 2850–2865(1996).

15. J. S. Kim and M. A. Cappelli, “Diamond film growth in lowpressure premixed ethylene-oxygen flames,” Appl. Phys. Lett.65, 2786 (1994).

16. J. Asmussen and D. Reinhard, Diamond Film Handbook(Marcel Dekker, Inc., New York, 2002), pp. 303–304.

17. S. Y. Moon and W. Choe, “A comparative study of rotationaltemperatures using diatomic OH, O2, andNþ

2 molecular spec-tra emitted from atmospheric plasmas,” Spectrochim. ActaPart B 58, 249–257 (2003).

18. G. Herzberg, Molecular Spectra and Molecular Structure(Litton Educational, 1950), pp. 465–472.

19. I. Kovacs, Rotational Structure in the Spectra of DiatomicMolecules (American Elsevier, 1969), pp. 115–135.

20. J. H. Walker, R. D. Saunders, and A. T. Hattenburg, “Spectralradiance calibrations,” NBS Special Publication 250-1(National Bureau of Standards, 1987).

21. F. Tuinstra and J. L. Koenig, “Raman spectrum of graphite,” J.Chem. Phys. 53, 1126–1130 (1970).

22. R. J. Nemanich, J. T. Glass, G. Lucovsky, and R. E. Shroder,“Raman scattering characterization of carbon bonding in dia-mond and diamondlike thin films,” J. Vac. Sci. Technol. A 6,1783 (1988).

23. R. E. Shroder, R. J. Nemanich, and J. T. Glass, “Analysis of thecomposite structures in diamond thin films by Raman spec-troscopy,” Phys. Rev. B 41, 3738–3745 (1990).


spectroscopic determination of rotational temperature in c_2h_4/c_2h_2/o_2 flames for diamond growth...

Documents