Journal ~l A Ilo vs and Compoumls, 192 (1993) 17 - 21 17

J A L C O M 2002

Luminescence of rare earth ions at the end of the century

G. Blasse
DebA,e Research Institute, Universi O' ~)1 UtrechL P.O. Box. 80.000, 3508 TA Utrecht (Netherhmds)

Abstract

Recent applications of rare earth luminescence are reviewed in connection with their physics background. These
relate to luminescent lighting, X-ray photography and other medical applications. Some problems of the future are
outlined: cheaper phosphors for lamps, higher-elticiencies, time characteristics in the field of X-ray phosphors, and
synthesis procedures.

I. Introduction lamps possible, for example those with a 10-mm diame-

ter tube. The choice of phosphors has been made, and
As an undergraduate student in the fifties I studied a the physics is in principle understood. Let us summarize
textbook on inorganic chemistry of some 1000 pages here some aspects.
[1]. One page was devoted to rare earths, but we were The blue and the red phosphor (Eu 2+ and Eu 3+
allowed to skip it. H o w drastically has the situation respectively) are of the simple type in which excitation
changed in the second half of this century. When and emission occur in one and the same luminescent
rare-earth oxides of high purity first became available, centre. High absorption of the 254 nm Hg-discharge
the first application of rare-earth luminescence was a radiation is guaranteed by allowed optical transitions
real breakthrough, i.e. the use of Eu3+-activated phos- ( 4 f - 5 d and charge transfer respectively). High quantum
phors as the red-emitting component in colour televi- efficiencies are achieved by using stiff host lattices, so
sion tubes [2]. that the excited state cannot expand too much. Other-
Perhaps even more impressive is the replacement wise this would result in unacceptably large nonradia-
of the classic lamp phosphor Cas(PO4)3(F, CI): tire losses [4, 6]. Recently it has become clear that in
Sb 3", Mn 2+ by a mixture of three phosphors all acti- some host lattices the Eu 2 ion shows photoionization
vated with rare earth ions [3]. A present-day lumin- followed by impurity-bound exciton emission which
escent lamp produced in my country will contain five shows a very large Stokes shift [7]. This type of lu-
rare-earth elements: blue-emitting BaMgAl~oO~y:Eu 2+, minescence is not suitable for lamp applications.
green-emitting GdMgBsO~o:Ce 3+, Tb 3+ and red-emit- The green phosphors have a physically more compli-
ting Y203:Eu 3+. Drastic improvements in efficiency, cated nature because energy transfer plays an important
colour rendering and lifetime have been obtained. Later role. In (Ce, Tb)MgAlliOl9 and (La, Ce, Tb)PO4
followed the introduction of rare-earth luminescent mainly one-step transfer from Ce 3. to Tb 3- occurs, but
materials in fields with medical applications, for exam- in (Gd, Ce, Tb)MgBsOlo the Gd ~- ions play an inter-
ple X-ray photography and luminescence immunoassay mediary role. Excitation occurs in the Ce ) ion, which
[4, 5]. transfers its energy to the Gd 3- sublattice. Energy
The aim of this keynote address is to review the migration over the Gd 3+ ions transports the excitation
present status of rare-earth luminescence from the point energy to the activator Tb 3+ [4--6]. The transfer rates
of view of application and physical insight, and to deal are now known. The Gd 3~ --Gd 3~ transfer rate is 10 6 -
with some problems which will be under study in the 107s l which is much higher than the Gd 3+ radiative
near future. rate (102-103s ~). A special problem in this type of
phosphor is to have the Ce 3+ emitting level in reso-
nance with the Gd 3- levels, whereas the Ce 3~ absorbing
2. State-of-the-art level should allow strong absorption of 254 nm rad-
iation. Lattices like GdF3, GdBO3, GdB306 and
2. I. L u m i n e s c e n t lighting GdMgBsOjo satisfy this requirement, but many do not
The three-colour lamp based on rare earth lumines- ([4 6] and references cited therein). It remains surpris-
cence has become a big success, making all types of ing that phosphors with such a series of complicated

0925-8388/93/$6.00 ~" 1993 ElsevierSequoia. All rights reserved

18 G. Blasse / Luminescence of rare earth ions

luminescence processes are able to show such high tivity due to the use of a photomultiplier tube instead of
quantum efficiencies (about 95%). Below we will show film, and the processing of the digitized images by a
that the present problem of the lamp industry is not in computer. The disadvantage up till now is the lower
the first place to find new phosphors with higher quan- resolution.
tum efficiencies. The photostimulable phosphor used at the moment is
BaFBr:Eu 2+. A physical mechanism for the PSL in
2.2. Cathode-ray tubes Eu2+-activated barium fluorohalides has been proposed
For conventional colour television tubes the definite by Takahashi et al. [12]. Optical, electron paramagnetic
choice of phosphors was made years ago: ZnS:Ag + resonance and photoconductivity studies indicate that
(blue), ZnS:Cu, A1 (green) and Y202S:Eu 3+ (red) [8]. upon X-ray irradiation some of the holes are trapped
New cathode-ray phosphors have been developed for by Eu 2+ ions, giving Eu 3+, and some of the electrons
use in projection television tubes. These phosphors are trapped in halide vacancies, giving F centres. Illumi-
operate under much higher excitation density than in nation in the F-centre absorption band stimulates re-
conventional tubes. Under such high loading the combination of the trapped electron with the hole
efficiency of many phosphors decreases with increasing trapped on Eu 2+, resulting in Eu 2+ in the excited 4f65d
excitation density as a result of saturation effects. Sev- state. The Eu 2+ ion returns to the ground state radia-
eral explanations have been proposed for this effect tively and the characteristic Eu 2+ emission around
[8, 91. 390 nm is observed.
The blue-emitting ZnS:Ag + is also used in projection Recent developments concern two topics: a refinement
television tubes, although it suffers from strong sat- of the theory and a search for new and better storage
uration. However, there seems to be nothing better. phosphors. These developments are reviewed here briefly.
Red-emitting Y203:Eu 3+ satisfies the requirements rea- It was soon realized that the simple model sketched above
sonably. As green-emitting components several Tb 3+- is an oversimplification. Studies of the recombination
activated oxides have been proposed (Gd202S, mechanism suggest that electron-hole recombination in
Y3(Ga, A1)50~2, InBO3) [8]. BaFBr:Eu 2+ does not take place via an electron entering
Clearly this is an application where the requirements the conduction band, but via tunnelling. The evidence for
are so severe that it is not possible to satisfy these this originates from different sources:
completely. (i) De Leeuw et al. [13] have measured the decay of
the photostimulated luminescence under continuous
2.3. X - r a y (storage) phosphors stimulation in the temperature region 100-300 K and
In the field of X-ray phosphors applied as an intensi- found it to be temperature independent.
fying screen in combination wtih film the development (ii) Von Seggern et al. [14] have measured the in-
of new materials seems to slow down [10]. From the crease in the photostimulated luminescence intensity
classic C a W O 4 to the rare-earth activated Gd202S: with increasing X-ray dose. This increase appears to be
Tb 3+ (green) and LaOBr:Tm 3+ (blue) was a large step. linear, which points to a tunnelling process.
Strangely enough, YTaO4:Nb 5+ (blue) has latterly also (iii) Meijerink and Blasse [15] have measured the
been successful [10]. This is peculiar because the physics thermally stimulated luminescence. Analysis of these
of this material are essentially the same as those of results points also to tunnelling.
CaWO 4. Other efficient rare-earth phosphors which An important contribution to the theoretical model
have been proposed are GdTaO4:Tb 3+, Gd2SiO5 :Tb 3+ formation originates from the use of vacuum-ultraviolet
and Gd3GasOlz:Tb 3+ [9]. radiation from a synchrotron source (Riiter et al. [16]).
X-ray storage phosphors, however, are in rapid de- It was shown that photostimulable centres in BaFBr:
velopment after their introduction in the early eighties Eu 2+ can be created by irradiation into the vacuum-
by Sonoda et al. [ 11]. The technique is based on the use ultraviolet region ( > 6 . 7 eV), i.e. in the excitonic and
of a photostimulable storage phosphor screen. In such a interband region of the host lattice.
phosphor the X-ray created free charge carriers are The model originating from this work invokes as a
trapped in electron and hole traps. At 300 K the pro- first step the relaxation of a free exciton in the neigh-
bability of thermal escape should be small and a bourhood of a lattice distortion induced by Eu 2+. This
metastable situation is created. leads to an e + Vk centre (where e is an electron and Vk
The stored energy can be released optically (photo- is the Vk hole centre consisting of a Br£ molecule on
stimulated luminescence (PSL)). In practice this is done two Br sites). Subsequently an off-centre self-trapped
with a focussed H e - N e laser. Advantages over conven- exciton is formed, i.e. a nearest-neighbour F - H pair
tional X-ray photography are the linearity of the re- (where F is the well-known F centre, in which an
sponse over at least four decades of the X-ray dose electron is trapped at an anion vacancy, and H is the H
(preventing over- and underexposure), the higher sensi- centre, which can be regarded as an X2 molecule
 G. Blasse / Luminescence qf rare earth ions 19

occupying an X anion site; the anion concerned here is ties make the use of a multiwire proportional chamber
Br ). This pair is assumed to be stabilized by the possible. The number of measured accidental coinci-
presence of a substitutional Eu 2+ ion which is much dences is minimized by the short decay time (positron
smaller than Ba 2+. The photostimulable centre is then emission tomography). Unfortunately BaF~ also shows
an Eu 2÷- F H complex. Photostimulation is thought to a slower, ultraviolet (300 nm) emission band. The latter
occur as follows. Upon excitation the F centre induces emission is due to self-trapped exciton decay, the for-
a certain relaxation, which destabilizes the H centre, mer to a crossover transition [24].
and an excited e + Vk centre is formed. The excitation Such a transition corresponds to recombination be-
energy is transferred to the nearby Eu 2+ ion and emis- tween a hole in the lower 5p (Ba 2+ ) core band and an
sion follows. electron in the 2p ( F ) valence band. It is a localized,
Many other X-ray storage phosphors have been intrinsic transition.
proposed. The most promising among these are The rare earths offer an alternative for short wave-
BasSiO4Br6:Eu 2+ and BasGeO4Br6:Eu 2+ [15, 17]. Oth- length emission with short decay time. The N d ~ ion in
ers are Ba2BsOgBr:Eu 2- [15], Ba4OBr6:Eu 2+ [18], fluorides, for example, shows 175 nm emission with a
Y2SiOs:CeS~ [19] and RbBr:TI- [20]. decay time of a few nanoseconds [25]. This is an
Interestingly enough, Koschnick et al. [21] have re- allowed 4f:5d ~ 4f 3 transition. In addition, Ce ~' in
cently presented experimental evidence from cross-re- fluorides may show possibilities in this direction. This
laxation spectroscopy which shows that the PSL centre field is still under development.
in BaFBr:Eu 2~ consists of a spatially correlated but
undistorted F centre, and Or-" centre and a Eu~,, ~ 2.6. Luminescence immunoassav
centre. These results illustrate how difficult it is to Certain Eu 3+ cryptates can be used in luminescence
unravel the storage mechanism in a simple compound immunoassay, a technique different from the field of
like BaFBr. Since these studies were performed on medical diagnostics. The cryptate is bonded to an anti-
crystals with low Eu concentrations, it cannot be ex- body which couples in a specific way with a particular
cluded that in the powder particles in a screen the biomolecule. If the labelled antibodies encounter the
mechanism is different. In the other proposed storage appropriate biomolecules they bind to these molecules
phosphors the mechanism may also be different. There- which, alter suitable treatment of the sample, are de-
fore we conclude that neither the physical basis nor the tected by the red luminescence of the Eu 3- cryptate.
ultimate choice of the material is yet known. The technique is very sensitive and can compete with
radioactive labelling. Requirements are that the absorp-
2.4. Computed tomography tion strength for ultraviolet radiation is high and that
Computed t o m o g r a p h y (CT) is another example re- the quantum efficiency of the Eu)--cryptate lumines-
lating to medical diagnostics. In this technique the cence is high (up to now, values of 10 20% have been
transmission of X-ray radiation through the body is obtained [26]).
measured. Using a large number of transverse transmis- A cryptand is an organic molecle with a hole in it; it
sion profiles, a three-dimensional density distribution can thus act as a cage li)r fitting species. In certain
can be determined. The method is fast and accurate, cryptates this species can be a rare earth ion. The use of
but requires a relatively high radiation dose per investi- an organic cage around a luminescent rare earth ion
gation. For detection an X-ray phosphor is used [22]. A presents some interesting phenomena, from a funda-
severe restriction is the requirement that the afterglow mental as well as from an applied point of view.
should be fast (after 2 ms less than 0.5% of the lumines- (i) The cage restricts expansion upon excitation. In
cence intensity after the X-ray dose should be present). this way nonradiative transitions can be suppressed. If
The afterglows of NaI:T1 and CsI:T1 are too long and the rare-earth-metal ion fits the cage, the Stokes shift
their stabilities too low. The efficiencies of C d W O 4 , and, therefore, the nonradiative rate are small. This has
ZnWO4 and Bi4Ge30,x are too low. Rare-earth acti- been shown experimentally for the case [Ce c 2.21] 3+
vated materials i.e. (Y, Gd)203:Eu 3+ and Gd202S: [271.
Pr, Ce, F [22] are now considered. (ii) The cage shields the luminescent ion t¥om its
further surroundings. This is of considerable impor-
2.5. Scintillators tance in aqueous solutions. The rare-earth-metal ion
Scintillators for detection of g a m m a radiation are, emission is easily quenched if high-tYequency vibrations
for example, NaI:T1, CsI:T1, CsI:Na, Bi4Ge3Ol2 and reside in the direct neighbourhood of the ion [26].
BaF2 [9]. The latler shows a special type of lumines- Water molecules present very high vibrational frequen-
cence. The emission is situated in the vacuum ultravio- cies and quench the greater part of the rare-earth-metal
let with bands peaking at 220, 195 and 179 nm and an ion emission. This nonradiative loss can be suppressed
extremely short decay time of 600 ps [23]. These proper- by caging the ion.
20 G. Blasse / Luminescence of rare earth ions

(iii) By the use of cryptands with very high absorp- value of qcv close to 100% [28]. However, other Eu 3+
tion strength, which transfer this energy to the rare- ions have a much lower qca-. Although it has become
earth-metal ion, it is possible to obtain a high clear that cheaper lamp phosphors are not easy to find,
luminescence output from the molecular unity. This has this is not impossible.
been realized by using cryptands containing 2,2'-
bipyridine [26]. Excitation occurs into the bipy group, 3.2. Luminescence immunoassay
which relaxes to the lowest triplet state. From here the Materials for this purpose suffer from too low an
energy is transferred to e.g. the Eu 3+ ion, which subse- efficiency. However, the means to overcome this are in
quently shows red emission. These types of cryptates principle clear. The charge-transfer state of Eu 3+ should
are promising for application in medical diagnostics, be high in energy, and shielding from the direct sur-
since their absorption strengths are very high as a result roundings (with high vibrational frequencies) is neces-
of the presence of the bipy groups. The highest quan- sary. The organic chemist's preparative abilities may
tum efficiency obtained up to now (20% at room tem- bring considerable improvements. This is the more so if
perature) has been obtained for the cryptate with two not Eu 3+ but Tb 3+ is used as the central luminescent
2,2'-bipyridine units and one 3,3'-biisoquinoline-2,2'- ion, since it suffers less from radiationless processes
dioxide unit. [26]. It is interesting to note that the considerations on
radiationless processes in rare-earth-doped solids can
also be of use in these molecular complexes [29].
3. Problems for the remaining part of the century
3.3. Storage phosphors
In this section we discuss some of the problems which Despite some very fine research on the X-ray storage
are or soon will be in the centre of interest. In fact the phosphor BaFBr:Eu 2+ (see above), the exact storage
contents of this section are strongly coloured by the mechanism in commercial materials remains unclear.
author's personal interest and experience. The results known up to now do not show a way of
finding better phosphors of this type. Therefore, the
3.1. Lamp phosphors development of new and/or improved storage phos-
The luminescent materials in the present three-colour phors remains for the large part a problem of trial and
lamps are rather expensive, and this is reflected in the error. A fundamental approach of considerable breadth
price of the product. Therefore research should be seems necessary to improve the situation.
started in order to develop cheaper materials. Another
approach is to increase the absorption of the exciting 3.4. Computed tomography
radiation, since at present this does not seem to be Much said on the storage phosphors is also true for
optimal. Quantum efficiencies are now close to 100%, the scintillator materials used in CT. It seems to us that
so that no further progress can be expected from studies in this field no systematic approach to find the optimal
directed to higher efficiencies. material has been performed, nor that a fundamental
In our laboratory we have performed a study on the study has been undertaken. A reduction of afterglow
luminescence of Eu 3+ in calcium and zirconium com- remains a formidable problem since such a phe-
pounds in order to find a cheaper replacement for Y203 nomenon depends critically on the presence of extrinsic
[28]. In these hosts the Eu 3+ ion carries an effective and intrinsic defects, the concentration of which may be
charge: Eu~a or Euzr. The positive effective charge is low.
detrimental for the quantum efficiency of the Eu 3+ In fact, the last two examples illustrate one of the
emission upon charge-transfer excitation (qc'r); the neg- present problems in materials science: that there is only
ative effective charge is favourable. A qualitative model a restricted amount of time and money available to
based on the configurational coordinate diagram has solve fundamental problems.
been proposed in order to explain these results. If the
Eu 3+ ion has a positive effective charge, the ground 3.5. Synthesis
state shrinks more than the excited charge-transfer Although phosphor preparation has always been del-
state, so that the value of qcT decreases relative to that icate, see for example Ref. 30, the importance of new
of Y203:Eu 3+. The only way to overcome this effect is and more easily controllable preparation procedures
to create a very stiff environment around the Eu 3+ ion. has been increasingly realized during the last decade.
In this way qcT values higher than 60% have been We expect this to increase in the coming years.
obtained (MgO:Eu 3+ [28]). In rare-earth activated phosphors the impurity level
For the negative effective charge more promising should be extremely low in certain cases, even in photo-
effects are to be expected and these have in fact been luminescent materials which are excited in the activator.
observed. Some Eu 3+ ions in BaZrO3:Eu 3+ have a For example, the iron level should be 1 ppm or less in
 G. Blasse / Luminescence 0,1"rare earth ions 21

o r d e r to o b t a i n for Y 2 0 3 : E u 3~ the o p t i m u m light out- 9 G. Blasse, in A. H. Kitai (ed.), ~Tsible Luminescence, (,'hap-
put [31]. man and Hall, London, 1992, in press.
10 L. H. Brixner, Mater. Chem. Phys., 16 ( 19873 253.
C o p r e c i p i t a t i o n techniques are b e c o m i n g increasingly
11 M. Sonoda, M. Takano, J. Miyahara and H. Kato, RadiohJgy,
i m p o r t a n t in o r d e r to distribute the a c t i v a t o r in the host 148 (1983) 833.
on an a t o m i c scale. A g a i n Y 2 0 3 : E u 3. is a striking, but 12 K. Takahashi, J. Miyahara and Y. Shibahara, J. Electrochem.
n o t the only, e x a m p l e [32]. O x a l a t e s are n o w p o p u l a r , Soc., 132 (1985) 1492.
a n d mixed oxides have b e c o m e available c o m m e r c i a l l y . 13 D. M. de Leeuw, T. Kovats and S. P. Herko, J. Electrochem.
H o w e v e r , the particle size a n d particle size distribu- Soe., 134 ( 19873, 491.
14 H. von Seggern, T. Voigt, W. Kntipper and G. Lange, J. Appl.
tion are also o f great i m p o r t a n c e , especially for screen Phys., 66(1988) 1405.
technology. T o these should be a d d e d m o r p h o l o g y a n d 15 A. Meijerink and G. Blasse, J. Phys. D, 24 (1991) 626.
m i c r o s t r u c t u r e o f the luminescent materials. 16 H. H. ROter, H. von Seggern, R. Reiniger and V. Saile, Phys.
In certain a p p l i c a t i o n s , for e x a m p l e in CT, the lu- Rev. Lett., 65(1990) 2438.
minescent m a t e r i a l is a p p l i e d as a ceramic plate which H. H. ROter, Thesis, Hamburg, 1991.
17 A. Meijerink, G. Blasse and L. Struye. Mater. Chem. Phys., 21
should p r e f e r a b l y be t r a n s p a r e n t [22]. In scintillator (1989) 261.
a p p l i c a t i o n s the use o f large single crystals is often 18 W. J. Schipper, Z. A. E. P. Vroon, G. Blasse, Th. Schleid and
imperative. In this w a y every a p p l i c a t i o n a n d every G. Meyer, Mater. Chem. Phys., 30 (1991) 43.
p h o s p h o r requires its own p r e p a r a t i o n m e t h o d which is 19 A. Meijerink, W. J. Schipper and G. Blasse, J. Phys. D, 24
usually far f r o m easy. F o r t u n a t e l y , the n u m b e r o f tech- ( 1991 ) 997.
20 H. yon Seggern, A. Meijerink, T. Voigt and A. Winnacker, J.
niques available is large [33]. Nevertheless, m u c h chem- Appl, Phys., 66 ( 19893 4418.
ical inventiveness will still be needed to i m p r o v e present 21 F. K. Koschnick, J. M. Spaeth, R. S. Eachus, W. G. McDugle
d a y p h o s p h o r synthesis. and R. H. D. Nuttall, Phys. Rev. Lett., 67( 19913 3751.
22 W. Rossner and B. C. Grabmaier, J. Luminesc., 48/49 (1991)
29.
23 M. Laval, M. Moszynski. R. Allemand, E. Cormoreche, P.
4. Conclusion Guinet, P. Odru and J. Vacher, Nucl. hlstrum. Methods, 206
(1983) 169.
In o r d e r to m a k e available new a n d / o r i m p r o v e d P. Schotanus, C. W. E. van Eijk and R. W. Hollander, Nucl.
r a r e - e a r t h - a c t i v a t e d luminescent materials, a c o m b i n a - hTstrum. Methods, A269 (1988) 377.
24 Ya. A. Valbis, Z. A. Rachko and Ya. Yansons, JETP Lctt., 42
tion o f f u n d a m e n t a l physical research a n d original chem- (19853 172.
ical synthesis is necessary. In this way, a n d with a little 25 P. Dorenbos, C. W. E. van Eijk, R. W. Hollander and R.
bit o f luck, m a n y new d e v e l o p m e n t s are still possible. Schotanus, IEEE Trans. Nucl. Sci., 37(1990) 119.
26 N. Sabbatini, Coord. Chem. Rev. (1992) in press.
N. Sabbatini, Mater. Chem. Phys., 31 (19923 13.
27 G. Blasse, G. J. Dirksen, N. Sabbatini and S. Perathoner,
References lnorg. Chim. Acta, 133 (1987) 167.
28 D. van der Voort and G. Blasse, Chem. Mater., 3(1991) 1041.
l A. F. Holleman and E. H. Biichner, Leerboek der Anorgan- J. Alarcon, D. van der Voort and G. Blasse, Mater. Res, Bull.,
ische Chemie, 15th edn., J. B, Wolters, Groningen, 1952. 27 (1992) 467.
2 A. K. Levine and l=. C. Palilla, Appl. Phys. Lett., 5 (1964) 118. 29 G. Blasse, Photochem. Photobiol., 52(19903 417.
3 J. M. P. J. Verstegen, D. Radielovic and L. E. Vrenken, J. 30 J. L. Ouweltjes, Modern Material,s', Vol. 5, 1965, p. 161,
Electrochem. Soe., 121 (1974) 1627. Academic Press, New York.
4 G. Blasse (ed.), Mater. Chem. Phys., 16 (3-4) ( 1987); an issue 31 W. van Schaik and G. Blasse, Chem. Mater., 4 (19923
on New Luminescent Materials. 410.
5 G. Blasse, Chem. Mater., I (1989) 294. 32 P. Maestro, D. Huguenin, A. Seigneurin, F. Deneuve,
6 G. Blasse, Prog. Solid State Chem., 18 (1988) 79. P. LeLann and J. F. Berar, J. Electrochem. Soe., 139(19923
7 B. Moine, C. Pedrini and B. Courtois, J. Luminesc., 50(19913 1479.
31. 33 See, for example, A. R. West, Solid State Chemistr), and Its
8 T. Welker, J. Luminesc., 48/49 (1991)49. Applications, Chap. 2, Wiley, New York, 1984.