Laser diode induced white light emission of γ-Al2O3 nano-powders

Gokhan Bilir a,b,n, Joseph Liguori b

a Department of Physics, Istanbul Technical University, 34469 Istanbul, Turkeyb Department of Physics, Boston College, Chestnut Hill, MA 02467, USA

a r t i c l e i n f o

Article history:Received 9 January 2014Received in revised form21 February 2014Accepted 26 March 2014Available online 1 April 2014

Keywords:Al2O3

White light emissionLaser diode excitation

a b s t r a c t

A broadband white light emission was observed ranging from 450 to 900 nm from nominally un-dopedγ-Al2O3 nano-powders under 803.5 nm and 975 nm laser diode excitations. The white light emission hasa strong dependence on both the environment pressure and the pumping laser power. We spectro-scopically characterized this white light emission in a systematic way by collecting spectra, measuringdecay- and rise patterns and changing parameters such as pumping power, pumping wavelength,environment temperature and pressure.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

The production of white light (WL) has challenged physicistsand engineers and has been widely studied due to its numerousapplications in lighting, displays and so on. Considerable attentionhas gone to binary and ternary lanthanides doped materials [1–4],mixing the emissions of LEDs that emit three primary colors (red,green, blue) [5] and using some phosphor based LEDs [6,7] toobtain WL. Some recent studies have also shown that certain ultra-small semiconductor nano-crystals can emit WL [8–10]. WLemission from metal oxide nano-powders has been achieved onlyby the addition of lanthanides doubly or triply and, to the best ofour knowledge, there is no information about WL emission fromun-doped metal oxides in the literature.

2. Experimental

γ-Al2O3 nano-powders with the crystallite size of o50 nm andthe surface area of 440 m2/g were purchased from Sigma-Aldrich.

The continuous emission spectra of the samples were obtainedby pumping the samples with the output of a Laser Drive Inc.Model LDI-820 diode laser operating at 975 nm or 803.5 nm. Thesignal was directed toward the entrance slit of a 1 m McPhersonModel 2051 monochromator and chopped at a frequency of250 Hz before entering the slit. The monochromator provided a

resolution of 0.8 Å with the slits set at 50 μm and a wavelengthreproducibility of 0.1 Å.

The optical signal was detected by a Hamamatsu R1387 photo-multiplier tube with an S20 response, sent to a EG&G Model 5210lock-in amplifier and recorded in a computer. The experimentaldata on the responses of the samples to sudden changes inexcitation were obtained by employing a shutter to interrupt thediode laser beam and a Tektronix Model TDS3052B oscilloscope.

In order to obtain data on the temperature dependence of this WLin the range of 34–300 K, the samples were mounted on the coldfinger of a closed cycle Helium refrigerator. The powdered sample wasput in the indentation of a copper sample holder attached to the coldfinger of a dewar (see Fig. 1). A copper sample holder with a small holein the middle into which the sample was filled was used to attach tothe cold finger of the dewar. This system uses a Janis Research ModelRD dewar connected to a Leybold Model RW2 compressor. Thetemperature of the sample holder was controlled using a Lake ShoreCryotronics 805 Model controller.

For comparison with the original sample, an agate mortar and apestle were used to grind down the grains into a finer powder asγ-Al2O3 came in a granular form similar to table salt. A littlepressure was applied so that the grains simply broke apart. Allmeasurements were performed before and after grinding.

An Avantes AvaLight-Hal-Cal calibration light source was usedto correct the spectrum for the sensitivity of the system.

An Allied Scientific Pro ASP-MK350 Model Illuminance Meterwas also used to allow measurements of the International Com-mission on Illumination (CIE) coordinates, the Correlated ColorTemperature (CCT), the Color Rendering Index (CRI) and theilluminance values of WL to be obtained.

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jlumin

Journal of Luminescence

http://dx.doi.org/10.1016/j.jlumin.2014.03.0650022-2313/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author: Permanent Address: Department of Physcis, İstanbulTechnical University, 34469, İstanbul, Turkey. Tel.: +90(212)2857235.

E-mail addresses: bilirg@itu.edu.tr, bilir@bc.edu (G. Bilir).

Journal of Luminescence 153 (2014) 350–355

3. Experimental results

The WL spectrum consists of a wide band that varies between450 and 900 nm as shown in Fig. 2(a) and (b). Fig. 2(a) gives thespectrum directly obtained from the monochromator. This spec-trum needed to be corrected for the response of the system using acalibration light source, and this corrected spectrum is given in

Fig. 2(b). The broad spectra appear to be structureless, except forthe fact that there is a spike in the spectrum due to the excitinglaser diode. Since the monochromator in the system operates inthe second order a spike at half of the excitation wavelengthappeared in the spectrum. The WL spectrum of γ-Al2O3 aftergrinding is also given in Fig. 2(a) and (b). A shift in the peakpositions and a slight change in shape can be seen in both Fig. 2(a) and (b). The value of the blue-shift after grinding is 6.7 nm. Thisdifference between the peak positions before and after grindingcorresponds to a temperature difference of 40 K according toWien's Displacement Law.

TheWL spectrumwas measured also using the illuminance meter.These results are given in Fig. 2(c). The corrected spectrum and thespectrum measured using the illuminance meter look similar.

WL output from the sample under 803.5 nm excitation wasobtained, but its persistence was very short and no meaningfulmeasurements could be carried out with the apparatus. However,the WL spectrum was obtained using the illuminance meter,shown in Fig. 2(d).

The CIE coordinates were found to be x¼0.392 and y¼0.334under 975 nm excitation and x¼0.365 and y¼0.324 under803.5 nm excitation. These values lie in the yellowish region andFig. 1. Scheme of the sample environment.

Fig. 2. (a) Spectral profile of the WL emission from γ-Al2O3 nano-powders before and after grinding, (b) corrected spectra for the system response, (c) measured spectrumunder 975 nm excitation by using illuminance meter, and (d) measured spectrum under 803 nm excitation by using the illuminance meter.

G. Bilir, J. Liguori / Journal of Luminescence 153 (2014) 350–355 351

correspond to CCT values of approximately 3285 K and 4012 K,each with a CRI of 86. The total incident luminous flux values werefound to be 53 and 392, correspondingly. The results are summar-ized in Fig. 3.

A strong dependence on the pumping laser power was alsoobserved. The threshold pumping power value to obtain WLemission was 3.5 W. When the power was increased threefold,the WL intensity increased by 75 times of its original value. Theseresults are given in Fig. 4. The power dependence of the intensitycan be approximated by the relation I¼ APn, where I is the peakintensity of the spectrum, A is a constant, P is the diode laserpower and the exponent n is found to be 4.75.

A strong dependence on the laser pumping power can also beseen from Fig. 5, which shows the pumping power dependenceof the WL spectrum. A blue-shift in peak positions as pumpingpower increases can also be seen in Fig. 5 with a value of21.6 nm. This difference between the peak positions at lowand high power values corresponds to a temperature differenceof 129 K according to Wien's Displacement Law.

The WL spectrum from γ-Al2O3 was collected as a function oftemperature of the sample holder, ranging from 34 K to 300 K.No dependence of the WL spectrum on this temperature wasobserved.

A strong dependence on the atmospheric pressure surroundingthe sample was observed. It was not possible to obtain WLemission at room pressure. This pressure dependence of WLemission is given in Fig. 6. As seen from the figure, when thepressure is reduced 400 times, the WL intensity increases25 times.

In order to provide further characterization the decay patternsand rise patterns of this emission were also measured by changingthe parameters that this emission is sensitive to. The results aresummarized below:

1. It was observed that the decay patterns did not change as afunction of pumping power (see Fig. 7), and the rise patternsseemed to increase in length with decreasing pumping poweras seen from Fig. 8.

Fig. 3. (a) CIE coordinates of WL under 975 nm excitation, (b) CIE coordinates of WL under 803 nm excitation, (c) CCT, CRI and illuminance values of WL under 975 nmexcitation, and (d) CCT, CRI and illuminance values of WL under 803 nm excitation.

Fig. 4. Pumping laser power dependence of the WL emission intensity.

G. Bilir, J. Liguori / Journal of Luminescence 153 (2014) 350–355352

2. The decay patterns of WL at different wavelengths were foundto differ as seen in Fig. 9 but there was no change in the risepatterns as seen in Fig. 10.

3. The sample holder's temperature dependence of the decay andrise patterns was also measured. There was no relevant change

in both the decay and rise patterns with this temperature asseen from Figs. 11 and 12.

4. A correlation was observed between grinding the grains andthe corresponding decay and rise times of WL. It was found thatgrinding caused the decay and rise patterns to be shorter than

Fig. 5. Pumping laser power dependence of the WL spectrum.

Fig. 6. Pressure dependence of the WL emission intensity.

Fig. 7. Pumping laser power dependence of the decay pattern of the WL emission.

Fig. 8. Pumping laser power dependence of the rise pattern of the WL emission.

Fig. 9. Decay times measured at different wavelength positions in the WL spectrum.

Fig. 10. Risetimes measured at different wavelength positions in the WL spectrum.

G. Bilir, J. Liguori / Journal of Luminescence 153 (2014) 350–355 353

they were before grinding the grains down. The lengths of thedecay patterns before and after grinding were found to be490 ms and 332 ms, respectively.

5. All measurements were repeated to see if grinding the sample hadany effect on these results. There were no differences except for thepeak position shift in the spectrum and the decreasing length ofthe decay and rise patterns as discussed above.

4. Discussion of the results

The production of WL has already been achieved and reportedin the literature. In some recent works, Strek et al. [11–13] showedthe possibility of obtaining WL emission from lanthanide ionsdoped nano-materials. Wang and Tanner [14] reported WL emis-sion from lanthanide oxides. However, to the best of our knowl-edge, there is no information in the literature about WL emissionof un-doped metal oxides.

The production of WL emission by laser diode excitation of theun-doped γ-Al2O3 nano-powders was obtained under low sur-rounding pressure. When discussing the mechanism for theproduction of WL, absorption and emission processes need to beconsidered separately. As γ-Al2O3 is transparent in a wide range ofwavelengths [15], the first problem with obtaining WL under975 nm excitation is addressing the absorption mechanism of

the excitation wavelengths, namely the deposition of energy.There could be some type of impurities which are responsiblefor the deposition of energy, but we do not have any idea aboutthe impurities that exist in the sample at this point. Someresearchers have reported a relatively broad absorption band ofH2O molecules peaking at 760 nm and 800 nm [16]. This could, inprinciple, be the mechanism responsible for the deposition ofenergy in this case. However, since low pressure was required forobtaining WL emission and it is well known that water can easilyevaporate even at room temperature at low pressure values, thisexplanation seems unlikely. As mentioned above, γ-Al2O3 has awide bandgap of 7 eV; it is clear that the absorption process ismultiphotonic due to the sub-band energy of the exciting photon.This is confirmed by the fact that WL production requires anattentive focusing of the diode laser beam on the powder. Theprocess of absorption could be resonant with the band gap ornonresonant; in the latter case it would bring electrons in theconduction band.

A result that confirms the multiphoton aspect of absorption isthe dependence of the intensity of WL on the power (4.75) of theexcitation wavelength (Fig. 4): the required 177 nm (7 eV) energynecessary for the excitation of alumina [15] requires five photonsof wavelength 975 nm (1.27 eV).

The 21.6 nm blue-shift of the peak position with increasingpumping power (see Fig. 5), which corresponds to a temperaturechange of 129 K, is due to the change of the surface temperature ofthe sample.

As seen from Fig. 6, the emission of WL increased withdecreasing pressure. This behavior cannot be explained on thebasis of optical emission due to the wide transparency region of γ-Al2O3, as well as the unknown impurities. However, a thermallyoriginating mechanism could be related with this observation.Radiative thermal emission from lanthanide ions doped materials[11] and from silicon [17] nano-particles has been observed in theliterature but it has never been observed for un-doped metaloxides. A similar intensity–pressure dependence has beenreported previously in the literature for SiC nano-particles andmechanically milled Si which was then attributed to thermallyoriginating emission and related to the low thermal conductivitiesof the nano-materials [17–19].

It is important to note that laser induced partial melting of thesamples was observed during the experiments. This observationindicates that the sample had reached temperature close to orabove its melting point (this value was provided from Sigma-Aldrich to be 2040 1C). Researchers have reported experimentaland theoretical studies on the relationship between laser inducedheating and melting of some metal nano-particles and the opticalproperties of the materials [20].

There is no dependence of the decay patterns of the emissionon the applied pumping power, but the rise patterns seem to beslightly dependent on this applied pumping power. They becomelonger with decreasing pumping power, seemingly because, as thepumping power decreases, the sample surface takes longer to heatup. This behavior can also be considered as a feature of thermalradiation.

Another important observation is the variation of the decaypatterns measured at different wavelengths. Decay patternsbecome longer as the wavelength values on the spectrum increase.A possible explanation of this is that the temperature of thesample decreases during the measurement of the decay pattern,thus shifting the spectrum toward longer wavelengths and causingthe disappearance of the short wavelength emission.

The measurements of the lengths of the decay patterns beforeand after grinding were found to be 490 ms and 332 ms, respec-tively. Also we observed a 6.7 nm of blue-shift of the peakpositions after grinding. This may be due to a possible change of

Fig. 11. Environment temperature dependence of the decay pattern of WL.

Fig. 12. Environment temperature dependence of the rise pattern of WL.

G. Bilir, J. Liguori / Journal of Luminescence 153 (2014) 350–355354

the surface shape or of the grain sizes as a result of this grinding. Ithas been shown in the literature that even grinding done using anagate mortar and a pestle may cause relevant structural changes ofthe materials [21–23].

The most important result that points to a thermally originatedWL is the similarities of the obtained WL spectrum with that of anincandescent lamp.

5. Conclusions

The production of WL under laser diode excitation has beenachieved using commercially obtained γ-Al2O3 nano-powderswith a particle size of o50 nm. Strong surrounding pressure andpumping power dependences of the WL emission have beenobserved. The decay and rise patterns of the emission have alsobeen measured with lengths of several hundreds milliseconds andseveral tens of second, respectively. In accordance with the resultsand observations, the WL emission from γ-Al2O3 has been attrib-uted to thermal radiation.

References

[1] H. Yang, G. Lakshminarayana, Y. Teng, S. Zhou, J. Qui, J. Mater. Res. 24 (2009)1730.

[2] R.M. Martinez, E. Alvarez, A. Speghini, C. Falcony, Thin Solid Films 518 (2010)5724.

[3] N.K. Giri, K. Mishra, S.B. Rai, J. Fluoresc. 21 (2011) 1951.[4] H. Hu, Y. Bai, J. Alloys Compd. 527 (2012) 25.[5] S. Muthu, F.J.P. Schurmans, M.D. Pashley, IEEE J. Sel. Topics Quantum Electron.

8 (2002) 333.[6] R. Mueller-Mach, G. Mueller, M.R. Krames, H.A. Hoppa, F. Stadler, W. Schnick,

T. Juestel, P. Schmidt, Phys. Status Solidi A 202 (2005) 1727.[7] W.M. Yen, S. Shionoya, H. Yamamoto, Phosphors Handbook, CRC Press

Taylor&Francis Group, Boca Raton (2006) 534–540.[8] S. Sapra, S. Mayilo, T.A. Klar, A.L. Rogach, J. Feldman, Adv. Mater. 19 (2007) 569.[9] A.D. Dukes, P.C. Samson, J.D. Keene, L.M. Davis, J.P. Wikswo, S.J. Rosenthal,

J. Phys. Chem. A 115 (2011) 4076.[10] Y.T. Chen, C.L. Cheng, Y.F. Chen, Nanotechnology 19 (2008) 445707.[11] L. Marciniak, W. Strek, A. Bednarkiewicz, A. Lukowiak, D. Hreniak, Opt. Mater.

33 (2011) 1492.[12] W. Strek, L. Marciniak, D. Hreniak, A. Lukowiak, J. Appl. Phys. 111 (2012)

024305.[13] W. Strek, L. Marciniak, A. Bednarkiewicz, A. Lukowiak, D. Hreniak, R. Wiglusz,

Opt. Mater. 33 (2011) 1097.[14] J. Wang, P.A. Tanner, J. Am. Chem. Soc. 132 (2010) 947.[15] R. Ahuja, J.M. Osorio-Guillen, J.S. de Almedia, B. Holm, W.Y. Ching,

B. Johansson, J. Phys.: Condens. Matter 16 (2004) 2891.[16] J.A. Curcio, C.C. Petty, J. Opt. Soc. Am. 41 (1951) 302.[17] P. Roura, J. Costa, Eur. J. Phys. 23 (2002) 191.[18] P. Roura, J. Costa, M. López-de Miguel, B. Garrido, J. Fort., J.R. Morante,

E. Bertran, J. Lumin. 80 (1999) 519.[19] P. Roura, J. Costa, G. Sardin, J.R. Morante, E. Bertran, Phys. Rev. B 50 (1994)

18124.[20] S. Bruzzone, M. Malvaldi, J. Phys. Chem. C 113 (2009) 15805.[21] T. Ban, K. Okada, T. Hayashi, N. Otsuka, J. Mater. Sci. 27 (1992) 465.[22] W.A. Yarbrough, R. Roy, Nature 322 (1986) 347.[23] P.J. Sanchez-Soto, A. Wiewiora, M.A. Aviles, A. Justo, L.A. Perez-Maqueda,

J.L. Perez-Rodriquez, P. Bylina, Appl. Clay Sci. 12 (1997) 297.

G. Bilir, J. Liguori / Journal of Luminescence 153 (2014) 350–355 355