Of

SL

a

ARRAA

KTRT

1

spittnRncatS[wiodlli

0h

Applied Surface Science 265 (2013) 180– 186

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

jou rn al h om epa g e: www.elsev ier .com/ locate /apsusc

ptical and structural characterization of pulsed laser deposited ruby thin filmsor temperature sensing application

atchi Kumari, Alika Khare ∗

aser and Photonics Lab, Department of Physics, Indian Institute of Technology Guwahati, Guwahati 781039, India

r t i c l e i n f o

rticle history:eceived 22 March 2012eceived in revised form 18 October 2012ccepted 26 October 2012vailable online 6 November 2012

eywords:

a b s t r a c t

The ruby thin films were deposited by pulsed laser deposition (PLD) technique in an atmosphere of oxy-gen using ruby pellet, indigenously prepared by mixing Al2O3 and Cr2O3 in appropriate proportion. Thecharacteristics R1 and R2 lines at 694.2 nm and 692.7 nm in the photoluminescence spectra of target pelletas well as that of PLD thin films, confirmed the ruby phase in both. The XRD and Raman spectra confirmeddeposition of c-axis oriented crystalline ruby thin film on sapphire substrate. Effect of deposition time,substrate and deposition temperature on PLD grown thin films of ruby are reported. The intensity of R1

◦

hin filmuby (Al2O3:Cr3+)emperature sensor

and R2 lines of PLD ruby thin films increased enormously after annealing the film at 1000 C for 2 h. Thefilm deposited on sapphire substrate for 2 h was 260 nm thick and the corresponding deposition rate was2.16 nm/min. This film was subjected to temperature dependent photoluminescence studies. The peakpositions of R1 and R2 lines and corresponding line width of PLD ruby thin film were observed to be blueshifted with decrease in temperature. R1 line position sensitivity, d�/dT , cm−1/K in the range 138–368 Kwas very well fitted to linear fit and hence can be used as temperature sensor in this range.

. Introduction

Ruby is a well known lasing material for high power Q-switchedolid-state laser. It consists of sapphire (Al2O3) in which a smallercentage of Al3+ ions has been replaced by Cr3+ ions. The Cr3+

on has three d electron in its unfilled shell. Transition amonghe levels of Cr3+ ion gives photoluminescence in visible region ofhe electromagnetic spectrum. The characteristic photoluminesce-ce occurs at 694.2 nm and 692.8 nm respectively called R1 and2 lines [1]. Ruby possesses favorable combination of relativelyarrow line width, a long fluorescent lifetime, high quantum effi-iency, and broad and well-located pump absorption bands. In ruby,valanches of phonons take place by stimulated emission withinhe Zeeman-split 2E levels [2,3]. Hence it is shown to behave as aASER (sound amplification by stimulated emission of radiation)4]. Nonlinear optical phenomena such as non degenerate two-ave mixing, spectral hole and slow and fast light are reported

n ruby crystal [5–8]. Single crystal of ruby is shown to act as fiberptic thermometer [9,10]. It has also been used as an ion-irradiationamage sensor [11]. Ruby thin films have been used to probe the

ocal density of states in complicated photonic systems [12]. The Rines of ruby are accompanied by nearby red shifted weak bandsn both emission and absorption spectra. These bands are referred

∗ Corresponding author. Tel.: +91 361 2582705; fax: +91 361 2582749.E-mail address: alika@iitg.ernet.in (A. Khare).

169-4332/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2012.10.163

© 2012 Elsevier B.V. All rights reserved.

as vibronic side bands. These vibronics side bands are predomi-nantly one-phonon transition. This fact makes ruby well suited forphonon-spectroscopy by optical means, which can be used as adetector or tunable generator for high frequency phonons [13–15].The dependence of R1 and R2 line intensity, wavelength and thecorresponding fluorescence lifetime as a function of the tempera-ture and pressure makes ruby as the basis for a variety of sensors[16–20]. In order to exploit these properties of ruby in the form ofminiaturized sensor for photonics and electro-optic applications,it is required to be grown in the form of thin film. Very high qual-ity epitaxial ruby thin films, free from strain and stress, are highlydesirable for these applications.

Ruby thin films were grown via solid phase epitaxy [21], elec-tron beam evaporation [22] and chemical vapor deposition [23].The deposition of polycrystalline ruby thin film on silicon substrateis reported via pulsed laser deposition (PLD) technique [24]. In thepresent paper, the effect of deposition time, substrate (quartz andsapphire), substrate temperature and post-annealing on the qual-ity of PLD deposited ruby thin films have been studied. Furtherits application as temperature sensor in the range 138–473 K isdemonstrated through photoluminescence studies.

2. Experimental details

The ruby pellets were prepared by mixing 0.05 wt%, 0.5 wt% and1 wt% of Cr2O3 by weight in � phase Al2O3 powder. The mixedpowder was annealed in the furnace for 6 h at 1000 ◦C to make

S. Kumari, A. Khare / Applied Surface Science 265 (2013) 180– 186 181

Table 1List of PLD deposited ruby samples under various deposition conditions.

Sample name Deposition parameters Thickness (nm) Deposition rate(nm/min)

Substrate Deposition time Substratetemperature

Post-annealing

Sample 1 Quartz 2 h RT – 190 1.5Sample 2 Quartz 4 h RT – 400 1.5Sample 3 (annealed sample2) Quartz 4 h RT 1000 ◦C for 2 h 400 1.5Sample 4 Quartz 1 h 650 ◦C – 100 1.6Sample 5 Sapphire 30 min 650 ◦C – 55 1.8Sample 6 Sapphire 1 h 650 ◦C – 120 2.0Sample 7 Sapphire 2 h 650 ◦C – 260 2.16

650 ◦C ◦

650 ◦C650 ◦C

ittfiprpc5l(sfiN∼ttwaAgwpdstsa

Sample 8 (annealed sample 5) Sapphire 30 minSample 9 (annealed sample 6) Sapphire 1 h

Sample 10 (annealed sample 7) Sapphire 2 h

t moisture free. Pellets of diameter 13 mm were prepared fromhe annealed powder and were sintered for 24 h at 1400 ◦C andhen at 1700 ◦C for 2 h. Long sintering time results in better dif-usion of Cr ions in alumina sites replacing the Al3+ ions by Cr3+

ons. Small quantity of the pellet (0.5 wt% doped) in the form ofowder was subjected to HRTEM studies for confirmation of theuby phase. The emission and excitation spectra of all the threeellets were recorded by Edinburg instrument, FS-920P (commer-ial fluorimeter with double monochromator) by exciting with32 nm light from a Xenon lamp. Confocal image of target pel-

et (0.5 wt%), was recorded by confocal laser scanning microscopeCarl Zeiss LSM 510 Meta) using 543 nm laser as the excitationource. The experimental set-up used for fabrication of PLD thinlms of ruby is shown in Fig. 1. A second harmonic of Q-switchedd:YAG laser (Quanta system model no. HYL-01), pulse duration of8 ns, was focused on to the ruby target mounted, inside the abla-

ion chamber. The focusing of high fluence ∼23 J/cm2 resulted inhe formation and expansion of plasma of the target material, andas deposited in the form of thin film on to the substrate (quartz

nd sapphire) placed at a distance of 5 cm away from the target.ll the samples of PLD ruby thin films were deposited in a back-round pressure of around 2 × 10−3 mbar of oxygen gas. Thin filmsere deposited for 2 h and 4 h on quartz substrate at room tem-erature and also for 1 h at 650 ◦C substrate temperature. The 4 heposited sample was annealed for 2 h at 1000 ◦C. On sapphire sub-trate, ruby thin films were grown for three different deposition

ime; 30 min, 1 h and 2 h at 650 ◦C substrate temperature. Theseamples were annealed at 1000 ◦C for 2 h. Table 1 lists the details ofll the PLD deposited thin films of ruby. The maximum deposition

Fig. 1. Schematic of pulsed laser deposition setup.

1000 C for 2 h 55 1.8 1000 ◦C for 2 h 120 2.0 1000 ◦C for 2 h 260 2.16

rate observed was 2.16 nm/min on sapphire substrate. The thick-ness of the deposited films were measured by Veeco Dektak 150profilometer and listed in Table 1. The emission and excitation spec-tra of the PLD thin films were recorded by using Edinburg, FS-920P fluorimeter. The samples were excited using 532 nm light froma Xenon lamp. The fluorescence lifetime measurement was per-formed by using a micro second flash lamp of fluorimeter. Ramanspectra were recorded by Horiba Jobin Yvon, LabRam HR800 micro-Raman spectrometer using 488 nm of argon ion laser. Temperaturedependent PL spectra measurement was carried out, in the range138–473 K, by low temperature Linkam stage, THMS 600 integratedwith LabRam having a resolution of 0.2 cm−1, using 632.8 nm He:Nelaser as the excitation source.

3. Result and discussion

3.1. Characterization of pellets

The intense red confocal image of one of the pellet (0.5 wt%of Cr2O3), on excitation with 543 green He–Ne lasers, is shownin Fig. 2(a) confirming the ruby phase. HRTEM image from thepellet is shown in Fig. 2(b), the measured d-spacing of 0.261 nmfurther confirmed the ruby phase [25]. Raman spectra of aluminaand chromium oxide powder, used to prepare the pellet and the sin-tered pellet are shown in Fig. 3. Absence of any well-defined bandin the Raman spectrum of alumina indicated the �-Al2O3, whichis Raman inactive [26]. Fig. 3(b) shows the Raman bands of com-mercial chromium oxide powder (Cr2O3) used for the pellet. Theobserved bands at 306.2, 342.7, 390.6, 541.2, 597.8 and 670.6 cm−1

are in agreement with the Raman bands of Cr2O3 [27]. The Ramanbands of sintered pellet (0.5 wt% of Cr2O3) are shown in Fig. 3(c). Theobserved Raman bands are located at 341.2, 352.7, 381.4, 417 and645.8 cm−1. The Raman bands at 381.4, 417.5 and 645.8 cm−1 corre-sponds to the corundum phase [27,28]. The LO band at 417.5 cm−1

in the pellet is signature of ruby phase [28]. Fig. 4(a) and (b) showsthe excitation and emission spectra of ruby pellets for three dif-ferent chromium concentration 0.05 wt%, 0.5 wt%, and 1.0 wt% ofCr2O3 in Al2O3. Two broad absorption bands centered at 402 and554 nm in Fig. 4(a) correspond to the blue and green absorptionbands of ruby. These bands at 402 and 554 nm are associated withthe spin-allowed transitions from the 4A2 ground state to the 4F2excited state (U-band) and to the 4F1 excited state (Y-band) respec-tively as shown in Fig. 5. The weak, sharp B lines in absorption states,Fig. 4(a) are associated with the spin-forbidden transition to thedoublet levels [1]. These are between states of the same crystal-field

orbital configuration and take place through single-electron spinflips. Fig. 4(b) shows the photoluminescence spectra of ruby pelletconsisting of well resolved R1 and R2 lines at 694.2 and 692.7 nmrespectively. In ruby, this bright red light emission results from

182 S. Kumari, A. Khare / Applied Surface Science 265 (2013) 180– 186

Fig. 2. (a) Confocal image of target pellet and (b) HRTEM of pellet.

Fig. 3. Raman spectra of (a) � phase alumina powder and (b) Cr2O3 powder. (c)Sintered pellet (0.5 wt%) showing ruby phase.

Fig. 4. Excitation and PL spectra of ruby pellet used as the target for PLD. (a) Excitation spectra and (b) photoluminescence spectra. (c) Decay time of R1 line, solid lineexperimental curve, broken line (blue) exponential fit. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of thearticle.)

S. Kumari, A. Khare / Applied Surface Science 265 (2013) 180– 186 183

eabsiialqnvbstttiw6sbieccotiFfldfii0ho

3

v

3

st

substrate temperature of 650 ◦C (sample 4) shows Raman bands at379 cm−1, 416.9 cm−1, 511.2 cm−1, 623.4 cm−1 and 746.4 cm−1 asshown in inset of Fig. 7. The LO band is observed at 416.9 cm−1,downshift 2.9 cm−1, in the Raman mode is attributed to the stress

Fig. 5. Energy levels diagram of ruby.

lectronic transitions that take place exclusively at the Cr3+ ionscross 2E and 4A2 states. The 2E states are excited via absorptionands (pump bands). The lifetime of the pump bands, are extremelyhort, with the ions transferred to a metastable 2A and E statesmmediately. The two transitions, E → 4A2 and 2A → 4A2, emittingn red end of the visible spectrum at 694.2 and 692.8 nm are referreds the R1 and R2 lines respectively. The separations of these twoines are about 29 cm−1. The R lines fluorescence is strong and theuantum efficiency is high, these lines are Zero Phonon lines witho stokes shift. The sharp R lines fluorescence from the pellet pro-ides the direct evidence that a substantial proportion of Cr3+ ionsehave as substitutional solute on the Al3+ sites in the corundumtructure. It is evident from the spectra, Fig. 4(b) that by increasinghe concentration of chromium from 0.05 to 0.5 wt%, results in mul-ifold increase in emission and excitation spectra. But on increasinghe concentration further (1.0 wt%) there is slight decrease in PLntensity due to Cr3+ concentration quenching in ruby [10]. Along

ith the sharp Zero Phonon lines, broad vibronic side bands at99, 701, 704 nm and few more weak bands with very low inten-ity were observed in the emission spectrum of ruby. These broadands are associated with multiphonon vibronic transitions orig-

nating from the 2E level from the electrostatic interaction of thelectronic states of the ions with vibrations of the host lattice andontains information not only on the phonon normal modes of therystal lattice but also on the strength and symmetry propertiesf the electron–phonon coupling. It can be seen from Fig. 4(b) thathe vibronic side bands are becoming more pronounced by increas-ng the chromium concentration due to increase in Cr3+ pairs [14].ig. 4(c)–(e) shows the decay of R1 after exciting with microsecondash lamp. The decay time is observed to be 3.38 ms for 0.05 wt%oped pellet, which reduces to 3.20 ms for 0.5 wt% doping and thennally it drops down to 2.51 ms for 1.0 wt% doped pellet and is

n excellent agreement with literature [10]. Since the pellet with.5 wt% is showing maximum PL emission as evident from Fig. 4(b)ence it was used as target for pulsed laser deposition of thin filmsf ruby on quartz and sapphire substrate.

.2. Characterization of PLD deposited thin films

Table 1 list the details of the all the thin films samples preparedia PLD in the present manuscript.

.2.1. XRD of PLD ruby thin filmsThe XRD spectra of ruby thin films grown on sapphire and quartz

ubstrate, along with the pellet are shown in Fig. 6. The XRD spec-ra of pellet are showing peaks at 25.5◦ (0 1 2), 35.14◦ (1 0 4), 37.7◦

Fig. 6. XRD of pellet and ruby films grown on quartz and sapphire.

(1 1 0), 43.3◦ (1 1 3), 52.5◦ (0 2 4), 57.4◦ (1 1 6), 61.2◦ (2 1 1) and 68.2◦

(2 1 4) [ICSD Collection code: 089665] confirming its polycrystallinenature. The thin films grown on quartz substrate (samples 1–4)are of amorphous nature as these do not show any diffractionpeak in the XRD spectra. The as deposited PLD thin film films onsapphire substrate (samples 5–7) shows diffraction peak at 43.2◦

corresponding to the c-axis growth. The XRD spectra of sample 4and sample 7 are shown in Fig. 6. On post annealing the sample 7at 1000 ◦C for 2 h (referred as sample 10) shows enhanced signal at43.3◦ as shown in Fig. 6.

3.2.2. Raman spectra of PLD ruby thin filmsThe as deposited thin films of ruby on quartz at RT do not show

any prominent peak in Raman spectra (sample 1 and sample 2).Even post annealed films on quartz (sample 3) do not show anyRaman peak. The film grown on quartz substrate at an elevated

Fig. 7. Raman spectra of bare sapphire substrate and PLD ruby thin films corre-sponding to sample 8, sample 9 and sample 10. Inset shows the Raman spectra offilm grown on quartz substrate; sample 4.

184 S. Kumari, A. Khare / Applied Surface

Fs

p(tgwtssP5cTt(idtscofi

3

dpiow(btionFEta2ocvia

ig. 8. Photoluminescence spectra of PLD thin film of ruby samples 1–4 and 6. Insethows the vibronic side bands.

resent in the film due to lattice mismatch between the substratequartz) and the deposited film. Broadening in the peak is attributedo the amorphous nature of the grown film on quartz. The ruby filmsrown on sapphire substrate (samples 5–7) show Raman bandshich get enhanced on annealing. Fig. 7 shows the Raman spec-

ra of post annealed PLD thin films of ruby deposited on sapphireubstrate, corresponding to samples 8–10 (Table 1), along with bareapphire substrate. The spectra of bare substrate as well as that ofLD ruby thin films shows 6 bands located at 379, 418, 431, 450,77 and 750 cm−1. The Raman bands at 379 cm−1 and 431 cm−1

orrespond to the transverse optical (TO) modes of Eg vibrations.he other bands at 450 cm−1, 577 cm−1 and 750 cm−1 correspondso longitudinal optical (LO) modes of Eg vibrations. The 418 cm−1

LO) A1g vibration in which the motion is primarily along the c-axiss the most intense Raman active band. This suggests that the PLDeposited films on sapphire substrate are c-axis oriented [30]. Allhe observed Raman modes of substrate are in agreement with thepectrum of c-plane of (�-Al2O3) sapphire [28,29]. No appreciablehange in the peak position and line width in the Raman spectraf film w.r.t. sapphire substrate confirm that these PLD ruby thinlms are of highly crystalline in nature and stress free.

.2.3. Photoluminescence (PL) spectra of PLD ruby thin filmsFig. 8 shows the photoluminescence spectra of ruby films

eposited on quartz (sample nos. 1–4) and sapphire substrate, sam-le 6 (Table 1). The sharp and distinct peaks at 694.2 and 692.8 nm

n corresponding to R1 and R2 lines along with vibronic side bandsf ruby were observed in all the samples. The PL intensity of R-linesas found to increases on increasing the deposition time from 2 h

sample 1) to 4 h (sample 2) as observed from Fig. 8. This coulde due to thick and uniform film resulting from longer deposi-ion time. The PL intensity of post annealed sample 3 was found tomprove drastically. Annealing of the film improves the crystallinityf the film which results in enhancement of PL intensity. There iso appreciable change in relative intensities of R1 and R2 lines andWHM for these two samples, sample 2 and post annealed sample 3.mission spectra of film grown on quartz for 2 h at 650 ◦C substrateemperature (sample 4) shows drastic enhancement in PL emissions compared to that of 4 h grown film at room temperature (sample) and post annealed at 1000 ◦C (sample 3). The relative intensitiesf R1 and R2 lines increased from 1.75 to 1.79 but no appreciable

hange in FWHM of the peaks were observed. Deposition at ele-ated temperature helps in c-axis growth. Thus substrate heatings more effective than growing the films at RT for longer durationnd then post-annealing at higher temperature. PL spectra of all

Science 265 (2013) 180– 186

the PLD ruby films were fitted to Voigt line shape and the Gaussianand Lorentzian width were compared. The ratio of Gaussian width(Gw) to Lorentzian width (LW), listed in inset of Fig. 8, was foundto decrease from 1.02 to 0.08 from sample 1 to sample 6, accord-ingly the PL intensity was also observed to increase as shown inFig. 8. Since the Gaussian component accounts for the inhomoge-neous broadening of the spectral line profile, it can be concludedthat the stress in the film is reduced from sample 1 to sample 6and the quality of film has improved, which is also supported bythe increase in PL intensity. The vibronic side bands of sample 4are shown in the inset of Fig. 8. As compared to the target pelletas shown in Fig. 4(b), these bands are more pronounced in the thinfilm. PL signal from the film grown on sapphire substrate, sample6, is much higher than that of on quartz, due to the lattice match-ing with sapphire substrate, enhancing the crystalline nature of thefilm.

Using these optimized growth conditions, epitaxial ruby thinfilms were deposited on sapphire substrate at 650 ◦C substratetemperature for three different time duration 30 min, 1 h and 2 h,samples 5–7. Fig. 9(a) shows the PL spectra of these films. With theincrease in deposition time, the intensity of R1 and R2 lines as well asrespective relative intensity has increased from sample 5 to sample7 as shown in Fig. 9(a). The PL spectrum of post annealed film sam-ple 8 (annealed sample 5) shows multifold increase in the intensityof R1 and R2 lines as compared to corresponding as deposited sam-ple (sample 5), because post-annealing improves the crystallinityof the film and makes it free from stress. Fig. 9(b) shows the PL spec-tra of post annealed samples 8–10, the Lorentzian fitting for sample10 is shown in Fig. 9(b). It was found to fit very well with doubleLorentzian line shape with r2 ∼ 0.999, confirming highly crystalline,stress free nature of the PLD deposited film.

3.2.4. Thickness measurement of PLD ruby filmThe thickness and the corresponding deposition rates for all

the PLD deposited thin films of ruby on quartz and sapphire sub-strate are listed in Table 1. Sample 1 and sample 2, depositedon quartz substrate at RT, were found to have a thickness of190 nm and 400 nm respectively. No observable changes in thick-ness were observed after annealing. The deposition rate was foundto be ∼1.5 nm/min. Sample 4, deposited at elevated temperatureof 650 ◦C was found to have a thickness of 100 nm and the cor-responding. The deposition rate was found to increase marginallyincreased from 1.5 nm/min to 1.6 nm/min. The thin films depositedon sapphire substrate; sample 5, sample 6 and sample 7 werefound to have a thickness of 55 nm, 120 nm and 260 nm respec-tively. The deposition rate was found to increase from 1.8 nm/minto 2.16 nm/min, on increasing the deposition time from 30 min to2 h, from sample 5 to sample 7 respectively. The effect of substrateon deposition rate can be observed from sample 4 (quartz) andsample 6 (sapphire), which are deposited under similar deposi-tion conditions. The deposition rate was found to increase from1.6 nm/min in case of quartz to 2 nm/min, in case of sapphire. Thiscould be due to lattice matching on sapphire substrate and thusefficient growth as compared to quartz substrate.

3.2.5. Temperature dependent PL studies of epitaxial ruby film viaPLD

The sample 10 was subjected to temperature dependent PL stud-ies. Fig. 10 shows the temperature dependent photoluminescencein the temperature range of 138–473 K for sample 10. It can be seenfrom Fig. 10 that the R-lines are becoming sharp and peak center isblue shifted at lower temperature as reported in case of bulk ruby

[31]. The observed profile is in accordance with two-phonon Ramanscattering process [32]. All the spectra were very well fitted to dou-ble Lorentzian line shape with the coefficients of determination forthe fits r2 ≥ 0.999. This reveals that the sample is stress free and

S. Kumari, A. Khare / Applied Surface Science 265 (2013) 180– 186 185

Fig. 9. PL spectra of film deposited on sapphire subst

Fig. 10. PL spectra of sample 10 at seven different temperatures.

Fig. 11. Variation of R lines (a) width with temperature, (b) position with temper

rate for (a) samples 5–8 and (b) samples 8–10.

the line broadening is only due to the temperature variation. Thecalculated line width at 298 K is ∼11.4 cm−1, which is comparableto the bulk ruby (11 cm−1) at 300 K. The ruby film grown via solidphase epitaxy was observed to have a line width of 15.8 cm−1 [20],higher than that of PLD deposited thin film of present experiment.Thus PLD grown film is of high optical quality. The variation of linewidth, peak position, PL intensity and splitting of R1 and R2 lines asa function of temperature are shown in Fig. 11. The line width andpeak position is calculated from the Lorentzian fitted data. It canbe observed from Fig. 11(a) that the width of both lines increaseswith increase in temperature. The width of R2 line is observed to besmaller than that of R1 as reported in literature in case of bulk [17].The R1 and R2 line position as a function of temperature is fittedwell to cubic equations (1) and (2) as shown in Fig. 11(b).

R1(T) = 14, 420 + 6.34 × 10−2 T − 5.11 × 10−4 T2

+ 4.087 × 10−7 T3 (1)

R2 (T) = 14, 450 − 1.86 × 10−2 T − 1.96 × 10−4 T2

+ 5.648 × 10−7 T3 cm−1 (2)

ature, (c) PL intensity with temperature and (d) splitting with temperature.

186 S. Kumari, A. Khare / Applied Surface

Fp

(cfiTclawtt2adw(ttivhu

4

pPdattrts

[

[[

[

[[

[[[

[

[[

[[[

[

[[

[[[

ig. 12. Sensitivity of R1 and R2 line position with temperature. Inset shows R1 lineosition fitted well to the linear fit in the range 138–368 K.

The coefficient of determination for the fits, given by Eqs. (1) and2), r2 ≥ 0.999 indicating that these fits can be used with very highonfidence. The PL intensity of R-lines as observed from Fig. 11(c)rst increases on increasing the temperature and then decreases.he R-line splitting with temperature is shown in Fig. 11(d). Thehanges in separation with temperature indicate that the R1 and R2ines do not shift exactly in the same manner on increasing temper-ture. This can be due to anisotropic thermal expansion of ruby, alsoith increase in temperature, the Cr3+ ion shifts slightly in relation

o the surrounding oxygen. The splitting increases from 29.2 cm−1

o 29.6 cm−1 slowly on increasing the temperature from 138 K to50 K, and then rapidly reaches a maximum value of 31.5 cm−1

t 400 K. On further increasing the temperature the splitting fallsown as shown in Fig. 11(d). The observed splitting is in agreementith the bulk ruby crystal [17]. The sensitivity; line shift per Kelvin

cm−1/K), d�/dT , of R-line position is shown in Fig. 12. The sensi-ivity of R1 line position in the range 138–368 K is very well fittedo linear fit with coefficient of determination r2 ∼ 0.978 as shownn inset of Fig. 12. So in low temperature range, R1 line position canery well act as a temperature sensor with linear sensitivity. In theigh temperature range of 370–475 K the R1, R2 line width can besed for the temperature sensing application, Fig. 11(a).

. Conclusion

Ruby thin films are deposited via PLD onto the quartz and sap-hire substrate. The observation of prominent R1 and R2 lines inL spectra confirms the ruby phase. Substrate heating during theeposition improves the quality of the film. Sapphire substratellows the epitaxial growth of the film due to lattice matching withhe ruby. The Raman spectrum of deposited film shows the vibra-

ional modes of ruby. The temperature dependent studies in theange 138–473 K shows that the R lines are becoming sharp as theemperature is lowered and the peak position is blue shifted. Thepectra fitted to double Lorentzian line shape, confirms absence of

[

[

Science 265 (2013) 180– 186

stress in the film. The narrow line width of 11.4 cm−1 of R-line atroom temperature confirms the very high optical quality of grownPLD thin films of ruby. The sensitivity of R1 line position in the range138–368 K shows linear behavior and hence the deposited film canbe used as a linear temperature sensor.

Acknowledgements

This work was supported in part by Department of Scienceand Technology Govt. of India Grant no. SR/S2/HEP-0019/2008.Physics department IIT Guwahati, is acknowledged for providingprofilometer facility funded by DST, New Delhi under FIST pro-gram (Ref no.: SR/FST/PSII-020/2009). Central instrument facility,IIT Guwahati, is acknowledged for micro Raman facility.

References

[1] R.C. Powell, Physics of Solid State Laser Engineering, AIP Press (Springer), 1998.[2] L.G. Tilstra, A.F.M. Arts, H.W. de Wijn, Physical Review B 68 (2003) 144302-

1–144302-7.[3] H.W. de Wijn, Journal of Luminescence 125 (2007) 55.[4] L.G. Tilstra, A.F.M. Arts, H.W. de Wijn, Physical Review B 76 (2007) 024302-

1–024302-13.[5] I. McMichael, P. Yeh, P. Beckwith, Optics Letters 13 (1998) 500.[6] L.W. Hillman, R.W. Boyd, J. Krasinski, C.R. Stroud, Optics Communication 45

(1983) 416.[7] M.S. Bigelow, N.N. Lepeshkin, R.W. Boyd, Physical Review Letters 90 (2003)

1139031-1–1139031-4.[8] G. Feng, X.U. Jing Jun, Z. GuoQuan, Z. WenDing, Chinese Science Bulletin 55

(2010) 473.[9] K.T.V. Grattan, R.K. Selli, A.W. Palmer, Review of Scientific Instruments 58

(1987) 1231.10] H. Aizawa, N. Ohishi, S. Ogawa, T. Katsumata, S. Komuro, T. Morikawa, E. Toba,

Review of Scientific Instruments 73 (2002) 3656.11] Q. Wen, N. Yu, D.R. Clarke, Journal of Applied Physics 80 (1996) 3587.12] T.M. Hensen, M.J.A. de Dood, A. Polman, Journal of Applied Physics 88 (2000)

5142.13] U. Rothamal, J. Heber, W. Grill, Zeitschrift für Physik B: Condensed Matter 50

(1983) 297.14] G.F. Imbusch, Journal of Luminescence 53 (1992) 465.15] R.C. Powell, B. Dibartolo, B. Birang, C.S. Naiman, Physical Review 155 (1967)

296.16] Y.M. Gupta, X.A. Shen, Applied Physics Letters 58 (1991) 583.17] D.D. Ragan, R. Gustavsen, D. Schiferl, Journal of Applied Physics 72 (1992) 5539.18] Ma Dong-Ping, L. Yanyun, Ma Ning, C. Jurong, Journal of Physics and Chemistry

of Solids 61 (2000) 799.19] C. Winnewisser, J. Schneider, M. Borsch, H.W. Rotter, Journal of Applied Physics

89 (2001) 3091.20] U.J. Gibson, M. Chernuschenko, Optics Express 4 (1999) 443.21] N. Yu, Q. Wen, D.R. Clarke, P.C. Mclntyre, H. Kung, M. Nastasi, T.W. Simpson, I.V.

Mitchell, D. Li, Journal of Applied Physics 78 (1995) 5412.22] Q. Wen, D.R. Clarke, N. Yu, M. Nastasi, Applied Physics Letters 66 (1995) 293.23] C. Pflitsch, D. Viefhaus, B. Atakan, Chemical Vapor Deposition 13 (2007) 420.24] H. Aizawa, M. Shibasaki, S. Komuro, Y. Miyazaki, T. Katsumata, International

Conference on Electrical Engineering, 2009.25] A.R. Zanatta, C.T.M. Ribeiro, U. Jahn, S.B. Aldabergenova, H.P. Strunk, Journal of

Applied Physics 100 (2006) 113112-1–113112-7.26] A. Aminzadeh, H. Sarikhani-fard, Spectrochimica Acta Part A 55 (1999) 1421.27] S.-H. Shim, T.S. Duffy, R. Jeanloz, C.-S. Yoo, V. Iota, Physical Review B 69 (2004)

144107-1–144107-12.28] R.S. Krishnan, Proceedings of the Indian Academy of Science A XXVI (1947) XIX.29] M. Kadleikova, J. Breza, M. Vesely, Microelectronics Journal 32 (2001) 955.30] (a) S.P.S. Poroto, R.S. Krishnan, Journal of Chemical Physics 47 (1967) 1009;

(b) H. Yao, C.H. Yan, S.P. Denbaars, J.M. Zavada, Materials Research SocietySymposium Proceedings 512 (1998) 411.

31] T.-H. Huang, C.-C. Hsu, C.-T. Kuo, P. Lu, W.-S. Tse, D.P. Wang, T.C. Chou, A.Y.G.Fuh, Journal of Applied Physics 75 (1994) 3599.

32] D.E. McCumber, M.D. Sturge, Journal of Applied Physics 34 (1963) 1682.