Journal of Luminescence 113 (2005) 100–108 Crystal growth, upconversion, and infrared emission properties of Er 3+ -doped KPb 2 Br 5 U. Ho¨mmerich a, , Ei Ei Nyein a , S.B. Trivedi b a Department of Physics, Hampton University, Hampton, VA 23668, USA b Brimrose Corporation of America, 5020 Campbell Blvd., Baltimore, MD 21236, USA Received 12 July 2004 Available online 27 October 2004 Abstract We report on the material preparation and optical properties of Er 3+ -doped KPb 2 Br 5 (KPB). KPB has a maximum phonon energy of only 138 cm 1 and is non-hygroscopic, which makes KPB an attractive candidate for solid-state laser applications. The preparation of Er doped KPB was based on a careful purification of starting materials followed by self-seeded Bridgman crystal growth. Under 975 nm diode laser pumping, Er:KPB revealed intense blue upconversion emission. For comparison, Er-doped KPb 2 Cl 5 (KPC), which has a maximum phonon energy of 203 cm 1 , exhibited a dominant green Er 3+ upconversion emission. The blue upconversion from Er:KPB can be attributed to emission from the 4 F 7/2 excited state of Er 3+ , which is quenched in most solid hosts due to strong multiphonon non-radiative decay. Due to the small phonon energy of KPB, the 4 F 7/2 level becomes highly radiative with a room-temperature lifetime of 85 ms and an estimated quantum efficiency of 94%. For comparison, the 4 F 7/2 decay time in Er:KPC was only 11 ms at room temperature and the radiative quantum efficiency was estimated to be 9%. Infrared (IR) emission bands were observed at 1.5 mm( 4 I 13/2 - 4 I 15/2 ), 1.7 mm( 4 I 9/2 - 4 I 13/2 ), 2.0 mm( 4 F 9/2 - 4 I 13/2 ), 2.7 mm( 4 I 11/2 - 4 I 13/2 ), 3.6 mm( 4 F 9/2 - 4 I 9/2 ), and 4.5 mm( 4 I 9/2 - 4 I 11/2 ), indicating the potential of Er:KPB for IR laser applications. The absorption and IR emission properties of Er:KPB were investigated in terms of transition linestrengths, branching ratios, radiative decay rates, and emission cross-sections using the Judd–Ofelt method. r 2004 Elsevier B.V. All rights reserved. PACS: 42.70.Hj; 42.55.Rz; 78.55.m Keywords: Laser spectroscopy; Infrared luminescence; Upconversion 1. Introduction The development of rare-earth (RE)-doped low- energy phonon host materials is of significant current interest for solid-state laser applications including direct ultraviolet and visible lasers, upconversion ARTICLE IN PRESS www.elsevier.com/locate/jlumin 0022-2313/$ - see front matter r 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jlumin.2004.09.111 Corresponding author. Tel.: 757 727 5829; fax: 757 728 6910. E-mail address: [email protected](U. Ho¨mmerich).
Transcript
ARTICLE IN PRESS
0022-2313/$ - se
doi:10.1016/j.jlu
�CorrespondiE-mail addre
(U. Hommerich
Journal of Luminescence 113 (2005) 100–108
www.elsevier.com/locate/jlumin
Crystal growth, upconversion, and infrared emissionproperties of Er3+-doped KPb2Br5
U. Hommericha,�, Ei Ei Nyeina, S.B. Trivedib
aDepartment of Physics, Hampton University, Hampton, VA 23668, USAbBrimrose Corporation of America, 5020 Campbell Blvd., Baltimore, MD 21236, USA
Received 12 July 2004
Available online 27 October 2004
Abstract
We report on the material preparation and optical properties of Er3+-doped KPb2Br5 (KPB). KPB has a maximum
phonon energy of only 138 cm�1 and is non-hygroscopic, which makes KPB an attractive candidate for solid-state laser
applications. The preparation of Er doped KPB was based on a careful purification of starting materials followed by
self-seeded Bridgman crystal growth. Under 975 nm diode laser pumping, Er:KPB revealed intense blue upconversion
emission. For comparison, Er-doped KPb2Cl5 (KPC), which has a maximum phonon energy of 203 cm�1, exhibited a
dominant green Er3+ upconversion emission. The blue upconversion from Er:KPB can be attributed to emission from
the 4F7/2 excited state of Er3+, which is quenched in most solid hosts due to strong multiphonon non-radiative decay.
Due to the small phonon energy of KPB, the 4F7/2 level becomes highly radiative with a room-temperature lifetime of
�85ms and an estimated quantum efficiency of �94%. For comparison, the 4F7/2 decay time in Er:KPC was only
�11ms at room temperature and the radiative quantum efficiency was estimated to be �9%. Infrared (IR) emission
bands were observed at 1.5 mm (4I13/2 -4I15/2), 1.7mm (4I9/2-4I13/2), 2.0mm (4F9/2-
4I13/2), 2.7 mm (4I11/2-4I13/2),
3.6mm (4F9/2-4I9/2), and 4.5mm (4I9/2-
4I11/2), indicating the potential of Er:KPB for IR laser applications. The
absorption and IR emission properties of Er:KPB were investigated in terms of transition linestrengths, branching
ratios, radiative decay rates, and emission cross-sections using the Judd–Ofelt method.
The development of rare-earth (RE)-doped low-energy phonon host materials is of significant currentinterest for solid-state laser applications includingdirect ultraviolet and visible lasers, upconversion
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108 101
lasers, and mid-infrared (MIR) lasers [1–5]. Com-pared to traditional laser hosts based on oxides andfluoride crystals, RE activated low-energy phononmaterials exhibit significantly reduced non-radiativedecay rates between closely spaced RE energy levels[1–5]. The reduced multiphonon relaxation rates leadto higher emission efficiencies for known REtransitions as well as the possibility of RE emissionbands at new wavelengths. Laser materials withintrinsic low maximum phonon energies includehalides, sulfides, and chalcogenides.
Recently, potassium lead chloride (KPb2Cl5,KPC) has emerged as a new low-energy phononlaser host for RE ions [4–17]. KPC is non-hygroscopic and readily incorporates trivalentRE ions. The maximum phonon energy of KPCis 203 cm�1 [4,9], which results in efficient MIRemission due to small multiphonon relaxationrates. Laser action has been demonstrated fromDy:KPC at 2.43 mm [7], Nd:KPC at 1.06 mm [8],and Er:KPC at 1.7 and 4.5 mm [15]. The overallefficiencies of the reported RE:KPC lasers, how-ever, was relatively low (o10%) mainly due topoor crystal quality and not optimized REconcentrations and pumping schemes.
We are currently investigating KPb2Br5 (KPB)as a potential new solid-state laser host material.Initial results of Nd-, Er-, and Tb-doped KPBwere recently presented at conferences by severalresearch groups [18–20]. KPB has a maximumphonon energy of only 138 cm�1 [18,19], whichpromises small non-radiative decay rates and highemission efficiencies for RE transitions with smallenergy gaps. In the following, we present results onthe material preparation and optical spectroscopyof Er doped KPB including transmission andabsorption, upconversion emission, infrared emis-sion, and lifetime measurements. Using Judd–O-felt (JO) theory, transition linestrengths, radiativelifetimes, branching ratios, and cross-sections werecalculated and will be discussed.
2. Material preparation and basic material
properties
The preparation of KPb2Br5 was based oncommercial starting materials of PbBr2 and KBr
with 99.999% purity. The PbBr2 material wascarefully purified through a combination of direc-tional freezing and horizontal zone-refinement[21,22]. Subsequently, KPB was synthesized usingstoichiometric amounts of purified PbBr2 andKBr. The synthesized KPB material was furtherpurified through zone-refinement. 1wt% of ErBr3(99.9% purity) was then mixed with the purifiedKPB. The resulting material was loaded into aquartz ampoule and sealed under vacuum. TheEr:KPB crystal was grown using Bridgman tech-nique at a translation speed of �1–2mm/h. Afterthe growth process, the resulting crystal was slowlycooled over a time period of 2 days. The Erconcentration was measured to be �0.14wt% atGalbraith Laboratories using inductively coupledplasma optical emission spectroscopy (ICP–OES).For comparative spectroscopic studies, Er-dopedKPC with 1wt% ErCl3 was prepared using asimilar procedure as described before.There are only a few reports in the literature on
the crystal structure of KPB [23–25]. KPB belongsto a family of congruently melting ternary alkalilead bromides including KPb2Br5, RbPb2Br5, andCsPbBr3. According to Beck et al. [24], KPB ismonoclinic (space group P21/c) with an angle bvery close to 901. The unit cell parameters area=0.9264 nm, b=0.8380 nm, c=1.3063 nm,b=90.061, and Z=4 [24]. The melting point ofKPB is �382 1C and a phase transition occurs at242 1C [23]. The density of KPB is 5.53 g/cm3 [25].Pb2+ ions occupy two non-equivalent lattice sitesof low symmetry, one site is a distorted octahedronand the second site a distorted trigonal prism [24].Similar to RE-doped KPC, RE ions most likelyoccupy Pb2+ lattice sites in KPB with K+
vacancies providing the necessary charge compen-sation [7–13].
3. Spectroscopic studies
3.1. Experimental details
The transmission measurements were performedusing a Cary 5 spectrophotometer. Spectral emis-sion studies were carried out using diode lasersoperating at 660, 810, and 975 nm. The visible
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108102
emission was dispersed in a 0.5m monochromatorand detected with a photomulitiplier tube (PMT).The infrared emission measurements employed a0.25m monochromator in conjunction with a InSbdetector. The emission spectra were recorded usingstandard lockin-technique. For temperature-de-pendent emission studies the sample was mountedon the cold-finger of a two-stage closed-cyclehelium-refrigerator. Luminescence decay measure-ments were performed using the output (ns pulses)of an optical parametric oscillator (OPO) system.The PL decay signals were detected by a PMT orInSb detector and averaged using a digitizingoscilloscope.
3.2. Transmission, absorption, and JO analysis
The transmission spectra of undoped KPB andEr-doped KPB are shown in Fig. 1(a) and (b). Ascan be seen from Fig. 1(a), the fundamentalabsorption edge of KPB is located at �400 nm.At longer wavelength the transmission rangedbetween �75–77% without any significant absorp-tion features. It is well known that binary andternary lead halides are prone to contaminationcaused by oxy-halides and other impurities, whichcan lead to infrared absorption features [21,22,26].Singh et al. reported distinct absorption features at3.1, 5.7, and 7.2 mm in PbBr2 due to residualimpurities of water, oxygen and carbon [22].
0.0
0.2
0.4
0.6
0.8
1.0
2000 4000 6000 8000 100000.0
0.2
0.4
0.6
0.8
1.0
(a)
Wavelength (nm)
Tran
smis
sio
n
undoped KPB
(b)
Er: KPB
Fig. 1. Room-temperature transmission spectra of undoped
KPB (a) and Er-doped KPB (b).
Fig. 1(a) demonstrates that the zone-refined KPBprepared in this study was free of residualimpurities and of high optical quality. Assumingthat the maximum transmission was only limitedby fresnel reflection losses, the refractive index ofKPB was estimated to be n�2.1. This value isbetween the refractive indices reported for KPC(�2.016) [9] and PbBr2 (�2.3) [27], respectively.Fig. 1(b) depicts the transmission of Er-doped
KPB. The overall transmission of the crystaldecreased to roughly 65–70% in the infraredregion, most likely due to slight cracks in thematerial. Several intra-4f Er3+ absorption bandsare noticeable in the region from �400–1600 nm,indicating the successful incorporation of trivalentEr ions. In addition, impurity absorption bands�3, �3.4, and �6.3 mm can be noticed. Similarabsorption bands around 3 mm were recentlyreported for zone-refined Er-doped KPC crystals[16] and are most likely related to OH impurities[28]. Despite the existence of residual impurities,the prepared Er:KPB sample was very stable underambient conditions without any noticeable degra-dation of the surface quality or optical propertiesfor several months.The room-temperature absorption spectrum of
Er:KPB with Er3+ intra-4f transitions originatingfrom the 4I15/2 ground state are shown in Fig. 2.The spectrum was corrected for background lossesdue to fresnel reflections. Prominent Er3+ absorp-tion features are readily assigned based on theexisting literature on Er-doped crystals and glasses[1,29]. Since the 4I9/2 absorption bands wasrelatively weak, its energetic position and band-shape was clarified through near-infrared excita-tion studies shown in the inset of Fig. 2. SevenEr3+ absorption bands were chosen to determinethe JO intensity parameters [1,30,31]. The 2H11/2
absorption band was not considered in the JOanalysis because it is hypersensitive [1,12]. Themagnetic dipole contribution in the 4I15/2-
4I13/2transition was subtracted from the experimentallinestrength value [1,32]. Table 1 shows the Er3+
transition wavelengths, integrated absorptioncoefficients, experimental linestrengths, andthe calculated linestrenths for Er-doped KPB.The three JO parameters obtained from thisanalysis were O2 ¼ 0:15� 10�20 cm2; O4 ¼ 3:64�
ARTICLE IN PRESS
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108 103
10�20 cm2; and O6 ¼ 2:08� 10�20 cm2: The rootmean square (rms) error for the electric dipoletransitions was calculated to be �0.38� 10�20 cm2,which is comparable to rms-values reported forother Er-doped solids [33]. Table 1 shows areasonably good agreement between observedand calculated linestrengths for Er-doped KPB.
3.3. Upconversion emission
The first noteworthy observation in the emissionstudies of Er doped KPB was the upconversion
400 600 800 1000 1200 1400 1600
0
1
2
3
4
5
6
7
800 850 900 950 1000
4 I11/2
4 I9/2
Wavelength(nm)
2G9/2
4F5/24F3/2
Er: KPB
4I9/2
Abs
orp
tion
Co
effic
ien
t(cm
-1)
Wavelength (nm)
4F7/2
4F9/2
2H11/2
4S3/2
4I11/2
4I13/2
Fig. 2. Absorption spectrum of Er:KPB showing the main
intra-4f transition of Er3+ ions originating from the 4I15/2ground state. The spectrum was corrected for fresnel losses. The
inset shows an excitation spectrum covering the 4I9/2 and4I11/2
absorption bands.
Table 1
Er3+ energy levels, average absorption wavelengths, integrated abso
*The magnetic dipole contribution in the 4I15/2-4I13/2 transitio
linestrength value.
behavior as depicted in Fig. 3. Under 975 nm diodepumping into the 4I11/2 state of Er
3+, a bright blueupconversion emission was observed fromEr:KPB. Under the same pumping conditions,Er:KPC exhibited the well-known green upconver-sion as it has been reported for many other Er-doped solids [13,34,35]. A detailed study of theupconversion emission from Er:KPC under 4I9/2(�800 nm) excitation was recently reported byBalda et al. [17]. The remarkable difference in theupconversion spectra of Er:KPB and Er:KPC
rption coefficients, measured electric-dipole line strengths, and
cm) Smeas (� 10�20 cm2) Scalc (� 10�20 cm2)
3.25* 3.41
0.84 0.83
0.77 0.65
2.86 2.91
0.74 0.46
2.13 1.84
1.22 0.54
n (�0.73� 10�20 cm2) was subtracted from the experimental
400 500 600 700
4F 7/
2->4
I 13/2
(721
nm)
Er: KPC
Er: KPB
4S
3/2->
4 I 15/2
(551
nm)
2H
11/2-
>4 I 15/2 (
531n
m)
Wavelength (nm)
300K
4F
7/2-
>4I 15
/2 (49
4nm
)
I int
ensi
ty (
a.u.
)
Fig. 3. Upconversion emission spectra of Er:KPB and Er:KPC
under 975 nm excitation. Er-doped KPb2Br5 (KPB) exhibited a
dominant ‘‘blue’’ upconversion emission due to the 4F7/2-4I15/
2 transition of Er3+. On the contrary, the upconversion
spectrum from Er-doped KPb2Cl5 (KPC) exhibited mainly
green emission lines from the transitions 2H11/2/4S3/2-
4I15/2.
ARTICLE IN PRESS
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108104
reflects on the smaller phonon energies in bro-mides compared to chlorides [36–40]. The blueupconversion line at �494 nm from Er:KPB can beattributed to emission from the 4F7/2 excited stateof Er3+ as illustrated in Fig. 4. The 4F7/2 state islocated only �1300 cm�1 above the 2H11/2 stateand is strongly quenched in oxide and fluoridehosts due to rapid multiphonon decay [34,35]. Inhalides, however, the 4F7/2 can become partiallyradiative [36–40]. A weak emission at �493 nmfrom the 4F7/2 was also observed from Er:KPC,but with significantly lower intensity then inEr:KPB. The intensity ratio between the 4F7/2
(blue) and 2H11/2/4S3/2 (green) lines was �1:10 in
Er:KPC and �15:1 in Er:KPB, respectively. Thislarge difference in the intensity ratio indicates alarge difference in the non-radiative decay rates ofthe 4F7/2 level in both crystals. Under the assump-tion that the radiative branching between the 4F7/2
and 2H11/2/4S3/2 is negligible (see Table 1),
observation of emission from the 2H11/2/4S3/2
states indicates the existence of non-radiativedecay from the 4F7/2 level. The green emissionfrom the 2H11/2/
4S3/2-4I15/2 transitions can be
considered to be mainly radiative due to the largeenergy gap [1,29]. It can further be assumed thatthe transition strengths between chloride andbromide materials vary only slightly [38]. Theintensity ratio between green and blue upconver-sion lines can then be used to estimate the radiative
4I11/2
2H11/2
4I15/2
4S3/2
4F7/2
4I11/2
2H11/2
4I15/2
4S3/2
4F7/2
975n
m97
5nm
Er: KPC Er: KPB
Fig. 4. Schematic energy level diagram for the upconversion
emission from Er:KPC and Er:KPB under 975 nm excitation.
quantum efficiency (Z) of the 4F7/2 state [Z�1/(R+1)] [9], which yielded a value of �94% forEr:KPB and �9% for Er:KPC. The obtainedquantum efficiency for the 4F7/2 level in Er:KPC isin good agreement with results reported byTkachuk et al. [13]. Further support for the largedifferences in the 4F7/2 quantum efficiencies wasobtained from decay time studies shown in Fig. 5.The room-temperature lifetimes of the 4F7/2 weremeasured to be �85 and �11 ms in Er:KPB andEr:KPC, respectively. The high quantum efficiencyof the 4F7/2 in Er:KPB is also in agreement withthe JO analysis shown in Table 2, which yielded aradiative lifetime of �56 ms for the 4F7/2 level. Thesmall difference in experimental and calculatedlifetimes for the 4F7/2 could possibly be due toradiation trapping in the investigated Er:KPBsample.The lifetime transients following pulsed excita-
tion provide further information on the upconver-sion mechanism in Er:KPB and Er:KPC [17,39].There are two possible upconversion processes,which are excited state absorption (ESA) andenergy transfer (ET) between excited Er3+ ions.Since ESA is a radiative process, its time evolutionis directly linked to the duration of the pumppulse. On the contrary, upconversion through ETis a radiationless process, which depends on the
0.0 0.1 0.2 0.3 0.4
0.01
0.1
1
1/e~0.011ms
1/e~0.085ms
4F7/2
->4I15/2
300K
Er: KPC
Er: KBP
Em
issi
on In
tens
ity (
a.u.
)
Time (ms)
Fig. 5. Room temperature decay transients of the 4F7/2-4I15/2
emission from Er:KPB and Er:KPC under pulsed (�10 ns) laser
excitation at �980nm. The 4F7/2 decay times were 0.085ms for
Fig. 6. Room-temperature infrared emission spectrum of
Er:KPB in the 1.5–5mm spectral region. The following emission
bands were observed: (1) 1.5mm (4I13/2-4I15/2), (2) 1.7mm (4I9/
2-4I13/2), (3) 2.0 mm (4F9/2-
4I13/2), (4) 2.7mm (4I11/2-4I13/2),
(5) 3.6mm (4F9/2-4I9/2), and (6) 4.5mm (4I9/2-
4I11/2).
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108 105
decay time of the intermediate excited state ofEr3+ (i.e. the 4I11/2) and can therefore evolve afterthe pump pulse is switched off [39]. As can be seenin Fig. 5, the decay transients of Er:KPB andEr:KPC under pulsed excitation (�10 ns) did notexhibit any significant rise-time, which indicatesthat upconversion through ESA is the dominantprocess for both samples. More detailed time-resolved emission and excitation studies as afunction of temperature will be carried out in thefuture.
3.4. Infrared emission (IR)
IR emission properties of Er:KPB were investi-gated for the spectral range from �1.5–5 mm. Anoverview of the normalized room temperatureemission spectra and relevant IR emission transi-tions is given in Figs. 6 and 7. Diode lasersoperating at �660, 810, and 975 nm were em-ployed for resonant excitation into the 4F9/2,
4I9/2,and 4I11/2, respectively. The following emissionbands were observed: 1.5 mm (4I13/2-
4I15/2),1.7 mm (4I9/2-
4I13/2), 2.0 mm (4F9/2-4I13/2),
2.7 mm (4I11/2-4I13/2), 3.6 mm (4F9/2-
4I9/2), and4.5 mm (4I9/2-
4I11/2). The corresponding emissionlifetimes at 300K for the 4F9/2,
4I9/2,4I11/2, and
4I13/2 excited states were 0.1, 1.2, 2.1, and 4.6ms,respectively. The decay transients were slightlynon-exponential, which could possibly be due tooverlapping emission contributions from differentEr sites and/or due energy transfer processesrelated to impurities (e.g. OH) [28].
ARTICLE IN PRESS
4I9/2
4I15/2
4I13/2
4I11/2
4.5
1.7 2.7
1.5
4F9/2
3.6
2.0
Fig. 7. Schematic energy level diagram for infrared transitions
in the 1.5–5mm region for Er:KPB. The corresponding IR
emission wavelengths are given in units of mm.
1500 1550 1600 1650
4I13/2->4I15/2exc: 980nm
300K
130K
90K
50K
10K
Wavelength (nm)
KPB:Er
Fig. 8. Temperature-dependent emission spectra of the 1.5 mmemission from Er:KPB.
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108106
In order to gain more insight in the incorpora-tion and local environment of Er3+ ions in KPB,the 1.5 mm emission (4I13/2-
4I15/2) was furtherevaluated through temperature dependent, high-resolution emission studies as shown in Figs. 8and 9. The 1.5 mm emission is relatively broad atroom temperature with a full-width at half-maximum (FWHM) of �45 nm.With decreasingtemperature, the emission spectra exhibited fine-structure and several Stark-levels were resolved at10K. For direct comparison of Er:KBP andEr:KPC, high-resolution emission spectra wererecorded for both crystals as shown in Fig. 9. Itwas previously suggested for Er:KPC, that Er3+
preferentially occupies only one of the two Pb2+
lattice sites in KPC [12,13]. In point symmetrieslower than cubic, all inter-manifold degeneraciesare lifted and the electronic energy levels areKramer doublets [1,29]. The number of Starklevels for the 4I13/2 excited state and 4I15/2 groundstate are then seven and eight, respectively. At10K the emission arises from the lowest Starkcomponent of the 4I13/2 leading to eight emissionlines terminating in the ground state manifold.Fig. 9 reveals the existence of at least 16 emissionlines at 10K for Er:KPB, which is not consistentwith a single emitting Er3+ center. A similar
observation was made for our Er:KPC sample.Crystal-field splittings of the 4I15/2 level in Er-doped halides have been reported to be around200–260 cm�1 [36–40]. Therefore, it appears un-likely that the emission lines observed above1600 nm originate from isolated Er3+ centers.The existence of different Er3+ sites and possiblyEr3+-pair centers in KPB, as observed for RE-doped CsCdBr3 [41,42], will be further explored infuture investigations.Branching ratios, radiative lifetimes, and peak
emission cross-sections for the observed infraredtransitions for Er:KPB were calculated using JOtheory and are listed in Table 2. A reasonablygood agreement was obtained between measuredand calculated lifetimes for the 4I11/2,
4I9/2, and4F9/2 excited states. The experimental lifetime ofthe 4I13/2 was significantly longer compared to thecalculated radiative lifetime, which could be due to
ARTICLE IN PRESS
1500 1550 1600 1650
x10
x10
4I13/2
->4I15/2
10K
Er: KPC
Er: KPB
exc: 980 nm
PL
In
ten
sit
y (
a.u
.)
Wavelength (nm)
Fig. 9. High-resolution emission spectra of the 1.5mm emission
from Er:KPB and Er:KPC at 10K. The emission was excited
with the 975 nm output of a diode laser.
U. Hommerich et al. / Journal of Luminescence 113 (2005) 100–108 107
radiation trapping in the first excited state of Er3+
[12,43,44]. The 4.5 mm transition (4I9/2-4I11/2) of
Er3+ has recently received significant interest forMIR laser applications [15,45] and lasing at roomtemperature has been demonstrated from Er:KPC[15]. The 4I9/2 is located roughly 2200 cm�1 abovethe 4I11/2 level, which leads to strong non-radiativedecay in Er doped oxides and fluoride hosts. Dueto the low-phonon energy of KPB, the 4I9/2 levelbecomes highly radiative in Er:KPB. Based on theJO analysis the emission efficiency can be esti-mated to be near unity. However, there are severalchallenges and drawbacks related to the 4I9/2-
4I11/2 transition including a low branchingratio (o1%), a relatively small peak emissioncross-section (�0.15� 10�20 cm2), and the self-terminating nature of this transition. In addition,the absorption cross-section at 800 nm is small(�0.2� 10�20 cm�2), which limits the pump effi-ciency at this wavelength.
4. Conclusions
Er-doped KPb2Br5 (KPB) was synthesized andgrown by Bridgman technique. Strong blue upcon-version emission was observed under 975nm diodelaser pumping, which is due to the 4F7/2-
4I15/2transition of Er3+. The 4F7/2 excited state is radiativedue to the low maximum phonon energy of138 cm�1 in KPB. The lifetime of the 4F7/2 wasmeasured to be �85ms with an estimated radiativequantum efficiency of �94%. In contrast, the 4F7/2
lifetime in Er:KPC was only �11ms indicatingstrong multiphonon decay. The IR emission proper-ties of Er:KPB were investigated and several intenseIR bands between �1.5 and 4.5mm were observed.Radiative rates, branching ratios, and peak emissioncross-sections were calculated for transitions arisingfrom the first four excited states of Er3+ (4I13/2,
4I11/2,
4I9/2, and 4F9/2) using JO theory. The 4.5mmemission originating from the 4I9/2-
4I11/2 transitionof Er3+ remains of great interest for MIR laserapplications. Due to the low-phonon energy of KPB,the observed 4.5 mm emission can be considered asmainly radiative with a quantum efficiency nearunity. However, this transition has some inherentdrawbacks including a relatively low emission cross-section, a small branching ratio, and a weak pumpefficiency at 800nm. Two-step excitation of the 4I9/2through an upconversion process [5] and sensitizingschemes through co-doping with other RE ions (e.g.Tm, or Ho) [45] have been suggested for other low-phonon hosts and will be further explored for Er-doped KPB.
Acknowledgement
The authors from H.U. acknowledge financialsupport by the National Science Foundationthrough Grant HRD-0400041 and the ArmyResearch Office through Grant W911NF-1-0302.
References
[1] A.A. Kaminskii, Crystalline Lasers: Physical Processes and
Operating Schemes, CRC Press, New York, 1996.
[2] H.U. Gudel, M. Pollnau, J. Alloys Comp. 303–304 (2000)
