Characterization of fluoride nanocrystals for optical refrigeration

Characterization of fluoride nanocrystals for optical refrigeration

Elton Soares de Lima Filho1,*, Marta Quintanilla2, Fiorenzo Vetrone2, Galina Nemova1, Kummara Venkata Krishnaiah1, and Raman Kashyap1,3

1Department of Engineering Physics, ÉcolePolytechnique de Montréal, P.O. Box 6079, Station Centre-ville, Montréal, QC H3C 3A7, Canada

2Nanobiotechnology, Université du Québec, Institut National de la Recherche Scientifique – Énergie, Matériaux et Télécommunications, 1650 Boul. Lionel-Boulet, Varennes, QC J3X 1S2,

Canada 3Department of Electrical Engineering, ÉcolePolytechnique de Montréal, P.O. Box 6079, Station

Centre-ville, Montréal, QC H3C 3A7, Canada *[email protected]

ABSTRACT

This paper reports on the characterization of nanocrystalline powders of ytterbium doped YLiF4 for applications in optical refrigeration. Here we used powders with nanocrystals of Yb3+ concentrations of (10, 15, 20) mol % and lengths (70, 66, 96) nm. Our preliminary spectroscopic measurements did not show an enhancement in the absorption at the long-wavelength tail of the spectra of the nanocrystalline powder when compared with bulk Yb:YLiF4, indicating that the increase of the phonon-assisted excitation is not large enough to play a significant role in cooling in the present conditions. One advantage of nanocrystalline powders over bulk crystals is the possibility of enhancing the absorption by the realization of cavity-less pump recycling through photon localization [1]. While photon localization also increases the reabsorption of the fluorescence depending on the quantum efficiency of the material and can mitigate cooling, it allows the use of crystals of low enough concentrations to avoid deleterious effects such as ion-ion energy transfer followed by quenching. The pump intensity enhancement favors upconversion luminescence to visible wavelengths, which can be used for optical refrigeration and extends the scope of the application for the material. We observed both green and blue emission from the samples and investigate the processes which lead to it. We present the experimental investigation of the nanocrystals’ absorption and emission spectra and the first excited state lifetime measurements, which are used to estimate the nanocrystal’s photoluminescence quantum efficiency.

Keywords: optical refrigeration, ytterbium yttrium lithium fluoride, nanocrystalline powder, absorption enhancement

1. INTRODUCTION Optical refrigeration with nanocrystalline powders was proposed in [1] and it was experimentally achieved through infrared-to-visble upconversion, in an neodymium-doped KPb2Cl5 nanocrystalline powder [2, 3]. The optical cooling of nanocrystalline powders can increase considerably the range of possible optically-cooled devices, for example by mixing the powder in a fluid or embedding it in a polymer or glass matrix for microfluidics, localized cooling or making of devices of arbitrary shape. Embedding the nanocrystals in glass is an alternative scheme for producing high-quality glass-ceramics for optical cooling. The use of nanocrystals also allows one to explore new phenomena that were not yet obtained in bulk materials for cooling. For instance, the random scattering of photons in the powder can lead to photon localization, allowing one to achieve high intensities without the use of a high power pump source or a cavity. The small size of the crystals and the high surface-to-volume ratio can lead to higher absorption in the long-wavelength tail of the spectra, when compared with the bulk counterpart. That would allow the use of a pump at a longer wavelength, thus providing a higher cooling efficiency than the one observed in the bulk. In this work we investigate the spectroscopic properties of ytterbium-doped nanocrystalline powders for optical refrigeration. The preliminary results indicate that in the present conditions, no significant distortion or enhancement of the absorption spectra occurs. More studies are necessary to obtain nanocrystals with enough quantum efficiency to show optical cooling. It is yet to be determined if the source of the small quantum efficiency is due to metallic and rare-earth impurities, trapping or water or other adsorption of contaminants on the powder’s large surface.

Laser Refrigeration of Solids VIII, edited by Richard I. Epstein, Denis V. Seletskiy, Mansoor Sheik-Bahae, Proc. of SPIE Vol. 9380, 93800Q · © 2015 SPIE

CCC code: 0277-786X/15/$18 · doi: 10.1117/12.2079856

Proc. of SPIE Vol. 9380 93800Q-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/11/2015 Terms of Use: http://spiedl.org/terms

2. THEORY The laser cooling process in the two level Yb3+-doped crystallline system with high concentration of ytterbium ions, where cooperative effects such as energy migration and cooperative luminescence are taken into account, can be described by the rate equation

( ) ( ) ( )( ) ,222

12 XNNex

pepaph

pIpa

phpI

dtdN

−⎥⎥⎦

⎤

⎢⎢⎣

⎡++−=τ

νσνσν

νσν

(1)

where N2 is the population of the excited 2F5⁄2 level normalized by the Yb3+ ion concentration, NYb, in the sample. νp is the pump frequency, Ip is the pump intensity. σa(νp) and σe(νp) are the absorption and emission cross sections at the pump frequency, respectively. τex is the lifetime of the excited level, 2F5⁄2. X is the rate of cooperative emission. The term 2

2XN indicates the loss of population of the excited level by the cooperative radiation of Yb3+ ion pairs. The high ion concentration influences the radiative lifetime, τex of the excited 2F5⁄2 level of Yb3+ ions [4].

( ) ,

291

12

0

0

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛+

+=

NN

lN

Yb

Ybacoop

π

σττ

(2)

where l is the effective length of the cavity l = ls(1-R+γ), where ls is the sample’s length (bulk case), R is the mirror reflectivity and γ is the sum of the contributions from all sources to the cavity loss. N0 is a critical concentration for self-quenching, τ0 is the measured lifetime at very weak concentration. Here we considered the general case of a cavity, to account for the possibility of trapping. The increase in the ion concentration causes an increase in the re-absorption process of the emitted anti-Stokes radiation. This process is described by the term (1 + σalNYb) in relation (2). The process of migration of the excited electron from the ion to ion also increases with the increase in ion concentration. It is described by the term

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛+

2

0291

NNT

π in the denominator of relation (2). For Yb3+:YAG the critical concentration N0 = 2.3

× 1021 ions-cm-3, i.e. 17 at.%, and 55 at.% in Yb3+:YLF. Rare-earth ions in nano-crystals smaller than the wavelength and surrounded by a medium with refractive index ns have a lifetime of [5]:

.coop

sex n

n ττ = (3)

Contrary to bulk samples, the phonon DOS of nanocrystals is quantized into discrete levels and low energy acoustic phonons are cut-off. Nanocrystals can have different shapes but for simplicity let us consider nanocrystals spheres. The spheroidal modes (a vibration with dilation) and torsional modes (a vibration without dilation) in this system can be described using equations 4 and 5 [6, 7]:

[ ] ,0)12)(1()()(

)2)(1(22)(

)()1(

)()(

)2)(1(2 2124

112 =−−++−−+−⎭⎬⎫

⎩⎨⎧

⎥⎦

⎤⎢⎣

⎡+−+−+ +++ η

ηη

ηηηξξ

ξηη

ηη llj

jlll

jj

lj

jll

l

l

l

l

l

l (4)

for ,0≥l and

,0)(1)(1 =

−−+ η

ηη ll jlj (5)

For ,1≥l respectively. Here ,tvRΩ=η and ,lR vξ = Ω where Ω is the phonon frequency, lv and tv are the sound velocities of the longitudinal and transverse modes, respectively. jl is the lth order spherical Bessel function. Using (4)



and (5) we have estimated that the phonon cut-off frequency for the crystals of dimensions exceeding ~20 nm is less than 15 cm-1. Such phonons in the sample can bridge the Stark sub-levels providing thermalization in the samples. For example in a similar system of Yb3+:YAG nanocrystals, the energy levels are similar to the energy levels of bulk Yb3+:YAG, as one can see in [8, 9]. Stark sub-levels of the 2F7⁄2 level of Yb3+:YAG nanocrystal are 0 (0), 563 (581), 606 (619), 786 (786) cm-1. Stark sub-levels of the 2F5⁄2 level of Yb3+:YAG nanocrystal are 10298 (10327), 10638 (10634), 10905 (10927) cm-1 [8]. The cooling power density in the system of RE-doped host can be estimated as a difference between the absorbed pump power density and the power density removed from the system with anti-Stokes spontaneous emission, stimulated emission, and cooperative emission

( )( ) ( )22

2 2 2 ,cool p a p Yb e pF ex coop Yb

Nhc hcp I N N N N XN

σ λ σ λλ τ λ

⎡ ⎤= − − − + +⎣ ⎦ (6)

where σe(λp) is the emission cross section at the pump wavelength λp, λF is the mean fluorescence wavelength, and λcoop is the wavelength of the cooperative emission. Numerically analyzing (6) one can conclude that the power density of the cooperative emission is very small in comparison with the power density of anti-Stokes spontaneous emission and stimulated emission and can be neglected in the system. Assuming that the surface of the sample is in direct contact with air, there is natural or free convection in the system. The equilibrium temperature of the sample can be calculated as

( ) ,cool r sp V hS T T= − (7) where Tr is the room temperature, Ts is the equilibrium temperature of the system, h is convective heat transfer coefficient, V is the volume of the sample, and S is the total surface area of the sample. It is important to note that the cooling power density, pcool, in equation (7) is a function of the temperature [10]. Using the transcendental equation (7) in conjunction with (6), one can estimate the equilibrium temperature of the sample. Alternatively one can choose the optimum sample radius or dopant concentration before the experiment. If one expects an equilibrium temperature Ts to be achieved, it can be shown that the nanocrystal radius must exceed Rmin,

( ) ( )( )

( ) .13

minYbpF

expa

pesr

Nhh

TThRR

νν

τνσνσ

−

⎟⎟⎠

⎞⎜⎜⎝

⎛+−

=> (8)

Following computer simulations, a Yb3+ ion concentration of ~30 % can be recommended for small (nano-) samples for optimal cooling. Nanocrystals can be arranged into nanopowders with a filing factor ranging from 0 to 1. The pump power can be enhanced in the nano-powder as a result of the trapping process. However, anti-Stokes fluorescence will undergo trapping and degrade the cooling process.

3. EXPERIMENTS 3.1 Preparation

For the first measurements, the samples were dispersed in water inside a 1-mm-long cuvette by vigorous shaking and immersion in a sonicator for 5 min, then immediately transferred to a Perkin-Elmer 1050 spectrometer in order to measure the transmission spectra. When compared to the transmission spectra of pure deonized water, no significant difference was observed, due to the small concentration of nanocrystals. Also, we found the crystals to precipitate relatively fast, thus compromising the accuracy of the measurement. Therefore it is better to work with the dry form of the powders so that one can ensure reproducibility.



The Yb:YLFtemperature corder to incre 3.2 Absorpt

Although repmeasurementto measure ththe relative afrom [11].

Figure 1 Acoefficienarbitrary

In the Figurethe resulting wavelength tpeaks, due towhere the brcoefficients mtail, as can bphotons is nohowever be m 3.3 Photolu

The photolumwe used the nkept the samare shown.

F nanopowderchamber with ease the emittin

tion spectra

producible, it t is only relativhe transmissionabsorption spec

Absorption specnt values are novalue so that it m

e 1(a) it was asgraph one ca

tail of the speco the effective road absorptionmatch at a wavbe seen in Figuot favored in thmade once an a

minescence

minescence spenanopowders m

me when using

s were dried dry air. The reng centers in th

is relatively ve. The microsn. After correcctra shown in

ctra of the 15 moormalized to the matches the bulk

sumed that then see that the

ctra when compbroadening ca

n in the nanopvelength of 96ure 1(b). In thehe nanocrystalsabsolute value

ectra of the sammixed in waterdifferent samp

by drop-castinesulting dispershe optical prob

hard to measscope slides wicting for scatteFigure 1. Also

ol % Yb:YLF naarea under the

k absorption at a

e broadband abnanopowder d

pared to the buaused by the rapowder is much0 nm, we still e light of the as relative to thefor the absorpt

mples were mer, so that the g

ples. In Figure

ng in the top sed dry powderbe volume.

sure the probeith dry powderr and Fresnel ro shown in the

anopowder, and curve. (b) The wavelength of 9

bsorption is thedoes not preseulk form. Alsoandom orientath larger than tsee a relative

above results we bulk form; thtion coefficient

asured by pumgeometry of the

2 the emission

of microscoper was accumul

ed thickness ors were placed reflection frome picture is the

a 5 mol % YLFabsorption of th

960 nm.

e same for bothent an enhanceo a decrease intion of the nanthat observed ireduction in th

we can concludhe opposite hapt in the nanocr

mping with a 98e probing region spectra of the

e slides and llated in a singl

of the powderin a Perkin-Elm

m the microscope absorption sp

F bulk from [11]he nanopowders

h forms, bulk aed absorption cn the absorptionnocrystals. In a in the bulk forhe absorption ide that the absppens. A definirystalline powd

80 nm wavelenon and the dene samples in a

left to dry in le ~ 3mm-wide

r, thus the abmer 1050 specpe slide, we ca

pectra of bulk

]. (a) The absorp is multiplied b

and nanocrystacoefficient in n can be notice

a less realistic srm, and the abin the long wasorption of lowitive conclusioder is obtained.

ngth laser. For tnsity of the emi

1 wt. % conce

a room-e spot, in

bsorption ctrometer alculated Yb:YLF

ption y an

als. From the long ed at the scenario, bsorption avelength w energy n should

this part, itters are entration



45

40 -

35 -

30 -

25 -

20 -

15-

10-

5-

2Fa2

0 , 1 T900 920

ÑEw

/

/

Ii_i_m_rT180 1000 102

D Wavelengtf

- LiYF4:Yb3t (11

- LiYF4:Yb3+ (11

- LiYF4:Yb3t (21

:0 1040 1060

(nm)

D mol %)

5 mol %)

D mol %)

1080 1100

Figure

As can be sethis, two effethis case a sularger are thethis case, i.e.certain extensynonymous favored at higin a nonradiaat higher conintensity decconcluded thNeverthelessFor the quanvisible spectroptical spectrlaser operatinvisible and in

e 2 Emission spe

en, the sampleects must be courface-related e nanocrystals. 15 mol. % and

nt, a larger amoto shorter Yb

gher concentraative decay, anncentrations thcrease that is ohat for the sys, this value countification of thra using a Thorum analyser (ng at a wavelenfrared spectra

ectra of the 2F3/2

e that shows thonsidered. On effect. TEM im Since smallerd 20 mol. % ofount of Yb3+ i

b3+-to-Yb3+ distations. Energy nd this effect cohere are more observed when ynthesis paramuld be differenhe visible emisorlabs SPLICCOOSA). In orde

ength of 960 nma are shown in F

→ 2F7/2 transitio

he highest emisthe one hand,

mages showedr nanoparticles f Yb3+. Additioions will accoutances, which transfer can prould slow dowions that can e[Yb3+] is incr

meters appliednt for different psion, we used O CCD spectrr to relate the mm and a whiteFigure 3.

on after 980 nm l

ssion intensity smaller partic

d that in the prhave larger su

onally the effecunt for a highemeans that the

rovoke the migwn the intensity

emit. In our careased from 15d here, [Yb3+

particles sizes.the dried sampometer (CCS),measured pow

e light source in

laser excitation f

is the one withcles are generaresent conditiourface-to-volumct of Yb3+ concer intensity. Nee energy transgration of lighty increase in thase, this last e5 mol. % to 20+] = 10 mol. ples inside an , and the infrar

wer by these twon conjunction

for the different

h [Yb3+] = 10 mally better quenons, the higherme ratio, the qucentration shouevertheless, a fer processes b

t from ion to iohe emission asseffect is probab0 mol. %. Up

% is the op

integrating sphred spectra usino instruments, with a large a

samples in wate

mol. %. To unnched. Quenchr the concentrauenching is struld be considerlarger concentbetween Yb3+

on, eventually fsociated to the bly responsibleto this point, i

ptimum conce

here, and measng an Ando Awe used a Ti:S

area photodetec

er.

nderstand hing is in ation, the ronger in red. To a tration is ions are

finishing fact that

e for the it can be entration.

sured the AQ6317B

Sapphire ctor. The



1

V

500 F

wal550 600velength (nm

650 7 8Q

inte

nsity

(lo

g. a

rb. u

nits

)8

óó

0

l..io

900 9E

t1

30 1000wavelengtl-

-L-P

10501 (nm)

Figure 3 sphere.

The 480 emi2F5/2-2F7/2 empresence of number of nothat Thuliumto bridge the0.1% of the idifficult to es 3.4 Lifetime

For lifetime frequency of The choppedwhich avoidsto a silicon pamplifier (Thradiation is fthe telescopeFor the 10 mdispersed (Fi

Visible (a) and

ission of Yb3+:mission from th

Er3+. One canon-radiative pa

m contaminatione power measuinfrared emissistimate an exac

e

measurementsf 14.00 Hz. In td beam is collis both transmitphotodiode of horlabs PDA2filtered by place and the photo

mol % sample, tigure 4 b and c

d near infrared (

YLF is due to e ytterbium ion

n also see the athways, and can does not affe

urement by theion. Since the ct value for the

, the beam frothe present conimated, attenuatted light and s13 mm2 area (00C) and disp

cing a 975 nm diode, rejectinthe powder wa) to reduce pos

(b) spectra of th

thulium [12], n. The 520 nm535 nm and 5an lead to a decect the coolinge OSA and thevisible spectru

e emission.

om a Ti:Sapphinfiguration, theated to ~ 20 mspecular reflect(Thorlabs SM0played on an owavelength lo

ng 99.85 % of tas concentratedssible optical c

he Yb:YLF nano

and it takes thm, 541 nm, 542537 nm associcrease in the ex

g efficiency sige CCS, we estium is above a

ire laser is foce signal drops f

mW of power ation, a 4-f teles05PD1B). Theoscilloscope (Tongpass interferthe on-axis 940d in a hand-madcavity effects.

ocrystals pumpe

hree subsequen nm, 550 and 6iated to the prxternal quantum

gnificantly [13]mate the visibstrong backgro

cused onto a 2-from 90 % to 1and loosely refscope collects signal is amp

Tektronix TDSrence edge filt

0 nm wavelengde glass microv

ed at 960 nm, in

nt excitations u670 nm emissiresence of Ho3

m efficiency. I]. Using the brble upconversioound from the

-slot chopper w10 % of its mafocused into thand directs the

plified by a benS7104). The strter (Edmund Ogth light. vial (data of Fi

nside an integra

using photons ons are assigne3+ [13]. That it was previousoadband photoon power to beTi:Sapphire la

wheel with a caximum value ihe sample. At e light from thenchtop transimrongly scattere

Optics 86-065)

igure 4 a), and

ating

from the ed to the increases sly found odetector e around aser, it is

chopping in 12 μs. an angle e sample

mpedance ed pump between

d then re-



Figure 4 Dlaser opea(a). In (b)

The other 15These showesummarized changes in thdifferent volu

Figure 5 Dpumped w

The logarithmprocesses takfrom the curv

Decay curves ofarting at a wavel) the beam is foc

mol. % and 20ed a much shoin Table 1. T

he measured lifumes, and as fo

Decay curves ofwith a chopped l

mic curves obsking place betwvature differen

f the near-1μm fllength of 940 nmcused and in (c) t

0 mol. % samporter lifetime t

The samples wfetime. Since tor the 10 % sam

f the near-1μm wlaser wavelength

served in Figurween the Yb3+

nces between th

luorescence intenm. The curves (bthe beam is unfo

ples, were keptthan the 10 m

were pumped inthe powder wample, the volum

wavelength fluorh of 940 nm.

re 5 do not fit + and impurity he data for the

nsity of a 10 mo) and (c) were m

ocused.

t on top of a mimol. % sample,

n different regas accumulatedme does not pl

escence intensity

to straight lineions. Howeve 15 mol. % an

ol % Yb:YLF nanmeasured using s

icroscope slide, as can be segions in (20 %d by hand, we ay a significan

y of 15 mol % an

es, which couler, the exact prnd 20 mol % sa

nopowder pumpsignificantly less

e, as in the preven in Figure 5

%, a) and (20 assume that di

nt role in the m

nd 20 mol % Yb

ld be an indicarocess is not clamples. Also, t

ped with a chopps powder than fo

vious cases (b)5. The lifetime%, b), but shoifferent regions

measured lifetim

b:YLF nanopow

ation of energylear, as one cathe same was o

ped r

) and (c). e data is owed no s present

me.

ders

y transfer an notice observed



for small-amsignals of thegood fit, and

We can nowconcentrationtogether with

Figure 6 Mfit to the from refe

It is clear thamol. %. Thesimilar numbinconclusive.mol. %. The [18], where tof fast diffusthe ion-concmeasurementnot as high a

mounts of the 1ese measuremethe difference

Table 1. F

Samp10 m10 m10 m15 m20 m20 m

w observe hown and compareh the best fit of

Measured lifetimliterature data (

erences [15, 16],

at the fluoresce data is strongbers to other r. The critical cexact lifetime

the energy is trsion and slow dcentration of t should be pers in the bulk, a

0 mol. % sampents. Fitting th between the re

Fluorescence life

ple, measuremmol % Yb, a mol % Yb, b mol % Yb, c mol % Yb mol % Yb, a mol % Yb, b

w the measurede it with resulf Eq. (2) to our

me (blue crosses)(circle) in dark gwith no informa

cence lifetimesgly dependent reported valueoncentration, f dependence oransferred to indiffusion (Eq. the nanocrystrformed. We cand no enhance

ple (Figure 4, he data to a sinesults from the

etime, single-exp

ment

d lifetime of tlts of bulk samdata as well as

) of the Yb:YLFgray, and the fit ation provided fo

s of the nanocron the Yb3+ c

es for low confrom the fit foron the concentrntrinsic nonrad2) processes. W

tals. To confican conclude hement in the ab

b and c), whicngle exponentiae different arran

ponential fit to d

Lifetime

the excited stamples from ths to the literatu

F nanocrystals, aof our data (cro

or dopant concen

rystalline samponcentration, Encentrations. Fr the nanocrystaration should adiative centers.We believe thairm this, a hihowever that thbsorption was o

ch could be an al to the data fngements is ne

decay curve of Y

e and σfitting (μs2405.4(132417.6(292383.1(28814.4(13507.0(46494.4(43

ate of Yb3+ ine literature. T

ure data.

and results from oss) in yellow. Hntrations.

ples is quenchEquation (2) c

For nanocrystaals data is 0.25also take into c

However, the at the problemigh resolution

he quantum effobserved. Thus

indication of for the 10% Ybegligible, as on

Yb:YLF nanocrys

s) 3) 9) 8) 3) 6) 3)

n YLF nanopowhese results ar

literature [14-17Horizontal lines

ed for concentcan be fitted toals, the fit div5 mol. %, whileconsideration apresent data c

m arises from a n transmissionficiency of the s we do not exp

errors due to tb samples give

ne can see in Ta

stals.

wders dependre shown in F

7]. Also shown is

corresponds to

trations as smao the bulk dataerges and is te for the bulk ia fast diffusioncannot fit a sim

spatial dependn electron mic

current nanocrpect optical co

the small es a very able 1.

s on the Figure 6,

s the data

all as 15 a, giving therefore t is at 52

n process mple sum dence on croscopy rystals is

ooling on



these samples in the present form. One possible alternative to use nanoparticles for optical cooling is to embed, or thermally grow them inside a bulk glass host [19]. That would prevent surface-related problems present in bare nanocrystals, such as adsorption of contaminants.

4. CONCLUSION We have investigated the spectroscopic properties of ytterbium-doped nanocrystalline powders for optical refrigeration applications. We have synthetized Yb:YLF at three different dopant concentrations, and different nanocrystal sizes. From the absorption spectra, we do not see an improvement in the long-wavelength absorption tail, as it was theoretically suggested, which could enhance optical refrigeration. The samples show visible emission to a level between 0.1 % and 1 % of the infrared emission, and indicates the presence of other rare-earth impurities, leading to an increase in the possible non-radiative pathways, and which could futrther preventing cooling. From the infrared photoluminescence spectra of the samples, we could find a concentration which gives the most intense emission, which happens to be the least concentrated one. This show that quenching is present for much lower concentrations than expected. Lifetime measurements confirm that the emission of the nanocrystalline samples is strongly quenched at relatively low concentrations (~15 mol. %) for this material. The much shorter lifetimes of the nanocrystals compared to the bulk counterparts further indicates that the quantum efficiency of the samples is too low to allow optical cooling to occur in their present form. Further investigation on the dopant concentration, and spatial distribution in the crystals needs to be performed in order to find an agreement between the theoretical model and the experimental data. Further purification of the samples may lead to different results, but so far there is no clear indication that the properties of the nanocrystalline samples are more promising than those found in bulk hosts.

ACKNOWLEDGEMENTS

RK acknowledges support from the Natural Sciences and Engineering and Research Council (NSERC) of Canada’s Strategic grants program, NSERC’s Discovery Grants program, Canada Council for the Arts’ Killam Research Fellowships program, and the Government of Canada’s Canada Research Chairs program.

REFERENCES

[1] X. L. Ruan and M. Kaviany, "Enhanced laser cooling of rare-earth-ion-doped nanocrystalline powders," Physical Review B, vol. 73, p. 155422, 04/21/ 2006.

[2] A. J. Garcia-Adeva, R. Balda, M. Al Saleh, and J. Fernandez, "Local internal and bulk optical cooling in Nd-doped crystals and nanocrystalline powders," 2010, pp. 76140A-76140A-12.

[3] A. J. Garcia-Adeva, R. Balda, M. Al Saleh, S. Garcia-Revilla, and J. Fernandez, "Local internal and bulk optical cooling in Nd-doped crystals and nanocrystalline powders revisited," 2011, pp. 795106-795106-10.

[4] F. Auzel, F. Bonfigli, S. Gagliari, and G. Baldacchini, "The interplay of self-trapping and self-quenching for resonant transitions in solids; role of a cavity," Journal of Luminescence, vol. 94–95, pp. 293-297, 12// 2001.

[5] H. P. Christensen, D. R. Gabbe, and H. P. Jenssen, "Fluorescence lifetimes for neodymium-doped yttrium aluminum garret and yttrium oxide powders," Physical Review B, vol. 25, pp. 1467-1473, 02/01/ 1982.

[6] X. Y. Chen, H. Z. Zhuang, G. K. Liu, S. Li, and R. S. Niedbala, "Confinement on energy transfer between luminescent centers in nanocrystals," Journal of Applied Physics, vol. 94, pp. 5559-5565, 2003.

[7] H. Lamb, "On the Vibrations of a Spherical Shell," Proceedings of the London Mathematical Society, vol. s1-14, pp. 50-56, November 1, 1882 1882.

[8] J. Amami, D. Hreniak, Y. Guyot, W. Zhao, and G. Boulon, "Size-effect on concentration quenching in Yb3+-doped Y3Al5O12 nano-crystals," Journal of Luminescence, vol. 130, pp. 603-610, 4// 2010.

[9] G. Boulon, L. Laversenne, C. Goutaudier, Y. Guyot, and M. T. Cohen-Adad, "Radiative and non-radiative energy transfers in Yb3+-doped sesquioxide and garnet laser crystals from a combinatorial approach based on gradient concentration fibers," Journal of Luminescence, vol. 102–103, pp. 417-425, 5// 2003.

[10] G. Nemova and R. Kashyap, "Laser cooling in Yb3+:YAG," Journal of the Optical Society of America B, vol. 31, pp. 340-348, 2014/02/01 2014.



[11] D. V. Seletskiy, S. D. Melgaard, R. I. Epstein, A. Di Lieto, M. Tonelli, and M. Sheik-Bahae, "Precise determination of minimum achievable temperature for solid-state optical refrigeration," Journal of Luminescence, vol. 133, pp. 5-9, 2013.

[12] L. C. Courrol, I. M. Ranieri, R. E. Samad, S. L. Baldochi, L. Gomes, A. Z. de Freitas, et al., "Enhancement of blue thulium emission on Nd:Yb:Tm-doped YLF crystals," 2006, pp. 610010-610010-9.

[13] A. Di Lieto, A. Sottile, A. Volpi, Z. Zhang, D. V. Seletskiy, and M. Tonelli, "Influence of other rare earth ions on the optical refrigeration efficiency in Yb:YLF crystals," Optics Express, vol. 22, pp. 28572-28583, 2014/11/17 2014.

[14] A. Bensalah, Y. Guyot, A. Brenier, H. Sato, T. Fukuda, and G. Boulon, "Spectroscopic properties of Yb3+: LuLiF4 crystal grown by the Czochralski method for laser applications and evaluation of quenching processes: a comparison with Yb3+: YLiF4," Journal of Alloys and Compounds, vol. 380, pp. 15-26, 10/20/ 2004.

[15] S. Chénais, F. Druon, S. Forget, F. Balembois, and P. Georges, "On thermal effects in solid-state lasers: The case of ytterbium-doped materials," Progress in Quantum Electronics, vol. 30, pp. 89-153, 2006.

[16] L. D. DeLoach, S. A. Payne, L. L. Chase, L. K. Smith, W. L. Kway, and W. F. Krupke, "Evaluation of absorption and emission properties of Yb3+ doped crystals for laser applications," Quantum Electronics, IEEE Journal of, vol. 29, pp. 1179-1191, 1993.

[17] N. Coluccelli, "Diode pumped passively mode-locked Yb:YLF laser," Opt. Express, vol. 16, pp. 2922-2927, 2008.

[18] M. Ito, G. Boulon, A. Bensalah, Y. Guyot, C. Goutaudier, and H. Sato, "Spectroscopic properties, concentration quenching, and prediction of infrared laser emission of Yb3+-doped KY3F10 cubic crystal," Journal of the Optical Society of America B, vol. 24, pp. 3023-3033, 2007/12/01 2007.

[19] E. Soares de Lima Filho, K. V. Krishnaiah, Y. Ledemi, Y.-J. Yu, Y. Messaddeq, G. Nemova, et al., "Ytterbium-doped glass-ceramics for optical refrigeration," Optics Express, vol. 23, pp. 4630-4640, 2015/02/23 2015.



Characterization of fluoride nanocrystals for optical refrigeration

Documents