CHARGE-TRANSFER SPECTRA OFLANTHANIDE IONS IN OXIDES

Henk Hoefdraad

FAKGK-'i KAKSFKK SPECTRA

• ••¥

LA>:Tli AN ;;!:•'. IONS IN OXIL^.S

typewerk : Mej. t.J.M. Tieiand

cekeningen : Mej. H.E. ELberse

fotografie : Hr. J.P. Hogeweg

offsetdruk : Hr. L.W. van de Voorde

• Hr. H.M. van Zoest

PRCMOTUK: r K O F . I )H. G . Ki.ASSH

CHARGE-TRANSFER SPECTRA

OF

LANTHANIDE IONS

IN OXIDES

Pioef sclirif t

ter verkrijging van de graad van doctor In

de Wiskunde en Na*-uurwetenschappen aan de

Fijksuniversiceit te Utrecht, op gezag van

de Rector Magnifvcus prof.dr. Sj . Groennian,

volgens besliiit van het Cc 1 1 ego van Dekanen

in het openbaar te verdedigen op voensdag

12 maart 1975 des namiddags te 2.45 uur

door

HENK EDWIN HOEFDKAAD

op 28 december 19A7 Le Willemstad,

Curasa

VOW £M

Graag ,;ou ik iedrreen wiilen beaanken die heeft

ineegewerkt aan het tot stand komen van dit proef-

schrift, daadwerkelijk of door bij te diragen aan

mijn wetenschaopelijke vorming en aan de prettige

sfeer waarin ik neb kunnen werken.

CONTENTS

CHAPTER I. INTRODUCTION , . 1

CHAPTER II. CHARGE-TRANSFER SPECTRA OF TETRAVALENT LANTHANIDE

IONS IN OXIDES j

CHAPTER III. THE CHARGE-TRANFER ABSORPTION BAND OF Eu"+ IN OXIDES.. 25

CHAPTER IV. EVIDENCE FOR THE INFLUENCE OF AN EFFECTIVE CHARGE ON

THE POSITION OF THE CHARGE-TRANSFER BAND OF EuJ+ IN

SOLIDS 3 3

CHAPTER V. GREEN EMITTING PRASEODYMIUM IN CALCIUM ZIRCONATF, 39

3AMENVATTING

SUMMARY

CURRUCULUM VITAE 50

Vhe invesc iga t ions described in t h i s thes i s w i l l be published in the

following j o u r n a l s :

Chapter I I . "Charge- t ransfer spect ra of t e t r a v a l e n t lanthanide ions

i;i oxides"

accepted by J. Inorg. Nucl. ("hem.

'•.•ir'f lii. "Tlit- charge-Lransfer absorption band cf Eu in oxides '

.-.ul"T. 11 l fd t- J . Solid State ulinm.

v.i.ijiiL'r :V. i-.K, Hoeldraad, l-'.M.A. Stogers and G. Blasse,

"Fvidence tor the influence of an effective charge on

Liu. position 01 the charge-transfer band of Eu in

so 1 ids"

submitted to Phys . ('hem. Let.

i in-,: Lei" V. :l,t. Koefdraad and G. Biassc,

"C.i'i'pii vm i tt i n . pr jscudyrr, iui;: in calcium zirconate".

>-.r.hir.i Tt..'d to Phys. Status Solidi 5.

CHAPTER I

INTRODUCTION

The study and interpretation of charge-transfer (c.t.) spectra

of chromophores has been much advanced in the last two decades.

Especially the ligand-to-metal c.t. spectra of complexes with a

central transition-metal ion and with halogen ligands have been

extensively studied by J(6rgensen (1). He noted a regular increase in

the position of the c.t. absorption bands of complexes of transition

mptal ions in the sequence of halogen ligands I, Br, Cl , F. The

quantitative differences are almost invariantly : 21 kK from F to

Cl, 6 kK fro.ti Cl to Br and 10 kK from Br to 1. Therefore he was able

to define an optical electronegativi ty (\ ) from the wavenumber {_-"•)

relating to the maximum of the c.t. absorption band.

[ 3° kK'in which X -M) is the uncorrected optical electronegativetv of

uncorr K 6

tV. - central metal ion M and X refers to a ligand.

It will be obvious that each value of v (x ) must beopt Auncorr

characteristic of the relevant ion iniependent of the surroundings or

the stereo-chemistry of the chromophore. Otherwise a set of optical

electronegativitias would not be of practical use. In other words,

if we consider a ligand-to-metal c.t. transition, the level from

which the electron is excited should be completely localized on

the ligands and the level to which it is excited should be localized

on the lisetiil ion.

Actually this situation does not exist. Most relevant molecular

oruitals (.ii.o.'s) of a complex will have iigand as well as metal

character to a certain extent. But even if both m.o.'s part ic.ipat iiv

in 3 transition were strictly localized, the value of X t (XI

would be influenced by the s tereochemi.- * ry of the complex. The

energetic separation between the several sets of T ligand m.o.'s

of different symmetry designation is caused by effects of ligand-

ligand repulsion and thus the energetic position of a particular

set of ligand rn.o.'s is also dependent on the X-X distances in a

complex.

So the spectral position of the c.t. absorption band of a

chromophore is not solely determined by the oxidizing character of the

central metal ion and by the reduc'ng character of the ligand.

but is also influenced by effects arising from the spatial arrangement

of the complex and m this way from interactions between metal ion

and ligand orbitals and between iigand orbitais mutu.i,.Ly.

Most data on c.t. absorption bands have been obtained from the

study of complexes in solution. Tnis has the advantage of easily

obtainable absorption spectra. On the other hand it is not easy

to vary coordination, stereochemistry and bond length in a controlled

way.

Working with solid samples presents technical difficulties, viz.

either the preparation of suitable single crystals, or the measurement

of diffuse reflection spectra on powders (in stead of absorption spectra).

Tiie advantage, however, is that it is relatively easy to incorporate

a given metal ion in several host lattices with varying coordination

number, different nearest neighbours and next-nearest neighbours .

A case of special interest is the incoi"porati3n of a given metal ion

into a series of isonicphous host lattices in which one of the host

3+ 3+ 3+ 3+ 3+lattic ions is varied (e.q. Eu in La , Gu , Y , La compounds).

Investigations of this type have been performed. Blasse and

Bril (2,3,4) have studied the charge-transfer spectra of the

5f r\— '}+ '—chromophures Nb [0 ] . and Eu [ CT ] in several host lattices

b n

as a function of their crystallographic surroundings. Day and

coworkers (5) studied the c.t. spectra of post-transition-metal

ions in some solids. All. these authors were not able, however, to

explain clearly the way in which the position of the c.t. absorption

bands depends on the crystallograp^ic surroundings.

The present work describes a study of the c.c. spectra of some

lanthanide ions in mixed metal oxides. This was started for the

following reasons:

(1) In our laboratory we are interested in the factors determining

the quenching temperature of characteristic luminescence. Blasse and

'Jril (2-4) have shown that the spectral position of the c.t. band

is one of the important factors that determine the value of this

quenching temperature.

(2) The fact that the first empty orbital in lanthanide ions is

an f-orbital provides a possibility to eliminate for a large part

the effect of destabiliaation of the m.o.'s mainly localized on the

metal ion. This enables us to estimate the influence of ligand-1igand

repulsion and of other effects that may be of importance.

(3) Preliminary work on the c.t. spectra of tetravalent lanthanide

ions in oxides (6) indicates a strong dependence of the spectral

position of the c.t. bands on the choise of the host lattice.

References

1. C.K. J«irgensen, Progr. Inorg. Chem. \2_, JOi (1970)

2. G. Blasse and A. Brii, Z. Physik. Chem. 57_, 187 (1968)

3. G. Blasse, J. Chem. Phys. 48., 3108 (1968)

4. G. Blasse, J. Solid State Chem. U_, 52 (1972)

5. P. Day, P.J. Diggle and G.A. Griffiths, J. Chem. S o c , Dalton

Transactions, 1974 page 1446

6. N. van Vugt, T. Wigmans and G. Blasse, J. Inorg. Nucl. Chem.

35, 2601 (1973)

CHAPTER II

CHARGE-TRANSFER SPECTRA Ci TETRAVALENT LANTHANIDE IONS IN OXIPF,?.

Abstract

The charge-transfer spectra of Ce , Pr and Tb in a number

of oxides are reported. It is noted that the position of the first

charge-transfer band is fixed for the metal ion in an oxygen

coordination of VI, but varies in V I I coordination as

a function of the host lattice. It is argued that this variation

is inherent to the VIII coordination itself.

Introduction

In the past decade quite some work has been dene on charge

transfer (c.t.) transitions in molecular or ionic groups in

which an electron is excited from a delocalized ligand molecular

orbital (m.o.) to an empty or Incompletely filled orbital

mainly localized on the central ion of the chromophore.

The position of the first c.t. band of a complex is determined

by the oxidizing character of the central ion and by the properties of

the set of ligands. Thus not only the reducing character of a

particular ligand is of importance, but also effects due to

ligand-ligand repulsion and to destabilization of the metal

ion orbitals due to interaction with ligand m.o.'s (1,2,3,A).

Although most d a t a r e f e r t j c . t . s p e c t r a of t r a n s i t i o n m e t a l

i o n s , c . t . t r a n s i t i o n s have a l s o been r e p o r t e d for the t r i v a l e n t

1 a n t h a n i d e s F.u , Sm , Tin and Yb i r. complexes w i t h h a l o g e n l i g a n d s

I.D) and ir. o x i d e s ( 6 ) , in which the [ i r s t band in the a b s o r p t i o n

3+ U

s p e c t r a i s a s c r i b e d to such a t r a n s i t i o n , i n c o n t r a s t w i t h Ce , Pr" aii.i

:':•• where the l i r s t band i s a s s i g n e d to a 4f -*• sd t r a n s i t i o n , a

-•arkeuiy r.i t : o r en i f e a t u r e be ing the b a n d w i d t h , which i s a lmost

: v i c e as lar ia 1 lo r a c . t . band compared w i t h an f ->• d band ( 7 ) .

S ince t o r .i g iven e lement the c . t . bands s h i f t to lower e n e r g i e s

with i n c r e a s i n g o x i d a t i o n s t a t e , and t r a n s i t i o n s between d i f f e r e n t

cat i o n i c subshcl iK (Rydberg t r a n s i t i o n s ) to h i g h e r e n e r g i e s , on?

might expec t to f ind the c . t . bands for complexes of the t e t r a v a l e n t

lan thani r ie (Ln) ions Ce~" , Pr ' and Tb"1 i n an e a s i l y a c c e s s i b l e

r eg ion o i the spec t rum. In f a c t some y e a r s ago Ryan and Jj5rgen&t:R (5)2- 2-

iuvo reported the absorption spectra of CeCi. and CeBr, andb b

J^rgor.son and Rittershaus (8) described the diffuse reflection spectra

oi Pr and Tb in TbO,. and Y.,i) in which the broad bands in the

visible region or the spectr'.iir, were ascribed to c.t. transitions.

Recently the spectra of these ions in inonoclinic ZrO? have been

reported (9). It was noted that in ZrO, and Y..0 the bands of

Pr and Tb are situated at markedly higher energies than in

ThO^, a fact which cannot readily be understood.

Most complexes have been studied in solution, but in this4+ 4+

case, since it is not possible to obtain Pr and Tb complexes

in solution, it was necessary to work in solid samples. Solutions

have the advantage that absorption spectra can be obtained, whereas solid

samples can only be studied by reflection spectroscopy, because it

is usually difficult to obtain single crystals. On the other hand

solid samples enable a systematic study of c.t. bands in several

crystal lattices, with the central metal ion in a wide range of

coordinations, and with different neighbouring ions of the complex.

In order to study the tetravalent ianthanide ions it is necessary

to have at one's disposal host lattices which contain tetravalent

ir.etal ions with ionic radii about equal to those of the Ln ' ions

and with their own c.t. bands far into the u.v. region. Furthermore

these tetravalent ions should occur in different lattices. Thus

4+ 4+ 4 + A •*•oxidic compounds of Zr , Ce 5 Hf and Th were chosen with

either the perovskite and K_,NiF, structure in which the tetravale.nl

ion is coordinated octahedrally bv six oxygen ions, or with the

fluorite, the zircon or the scheeiite structure with a tetravaient

ion coordination of VIII.

One of the main difficulties with reflection spectra is u-

determine the band position, especially in the case of broad overlapping

bands. It seemed useful, therefore, to try to resolve the broad

unstructured bands, extending from the visible region up to the

c.t. bands of the host lattices in the u.v. ,by the method of

dilution in the reference white standard (10).

Experimental

Samples were prepared by firing intimate mixtures of high

purity compounds in an oxygen atmosphere (1 atm). Starting materials

were BaCO SrCO-,, ZrO9, HfO,, ThO , GeO, and SiO^.x H.O. ZrO,,

was prepared from ZrOCl0. 8H 0 by recrysta]lization and subsequent

firing as described in ref.(ll). Rare earth oxides with nominal

: ' o n n u i , i I A ' O , , I ' r O , a n d T b - , 0 o f 9 9 , 9 1 ' p u r i t y o r b e t t e r w e r e u s e d .Oil L J

The rare-earth concentration in our samples varied up to 1 atomic

per rent. Samples were checked by X-ray analysis using a Philips

di 1 t"raetometer with Cu--Ku radiation. Diffuse reflection spectra

wore recorded on a Perkin-Elmer EPS /3T double-beam spectrophotometer

against a HgO white standard. Diluted spectra were measured by

diluting the sample in high-purity MgO until R varied between 90

i:id 100":', against the same high purity MgO as a inference.

An attempt was nude to study the behaviour of the absorption

bands at lower temperatures. To that extent samples were cooled

to about 130 K in a specially constructed sample holder by a flow

of liquid nitrogen.

Resuit:

The presence of the tetr.-ival.ent Ln ions m the host lattices

used, results in the occurence oi broad absorption bands in the spectra.

Judging WOT, the reflection spectra L':I-J intensity of the absorption

4 + 4 +bands it high. For Pr and Tb all samples with 1% Ln impurity

are intensively coloured. The colours range from brightly yellow

(BaZrO.j, ZrSiO,) through orange (ZrGeO,) to purple and violet

(ThO^, T'nGeO,) indicating allowed transitions. A summary is

given in tablesl and II. All data were derived from spectra of

samples diluted in MgO.

The compounds Ba ZrO. and SrnZr0, have the KnNiF. structure

<i 4 I 4 L 4

with Zr in an octahedral oxygen coordination. The compounds AMO

(M = Zr, Ce, Hf, Th) have the perovskite structure with the same

coordination for M. Most spectra of Pr and Tb in these host

4+Table I. Spectral data (in kK) of Ln ions in VI coordination

in some oxides.

VI coordination

o obs. (in kK)

Cc Pr Tb

IRa,ZrO

SrZrO.,

BaZrO.

BaCeO,

BaHfO,

BaThO,

32

33

-

32

-

-

32

30

2 5 ( s h ) ;

2 5 ( ? ) ;

2 5 ( ? ) ;

2 5 : 29 .

2 5 ( ? ) ;

2 4 . 8 ; 2 9

2 9 . b

JO

30

h

2 9 . 8

.6

25 ( sh)

22 ( sh )

2 5 ( s h )

22 ( s h ;

25

2 5 ; 2 9

2 4 : 2 9

:29

; - 5 ; 2 8 . 6

• n u

; ^ 5 ; 2 9 . 2

(sh) = shoulder (?) = uncertain

lattices show a band with two distinct maximn at more or less iixod

positions (table I, fig.l). No changes were detected at lower

4+ , 4+

temperatures. Notable exceptions are BaZrO :1b and Ba^ZrO^'.Tb

where an extra shoulder is found at 22 kK which does show some

temperature dependence (fig.2). In the case of Ce the band is found

at higher energies. A second band could not be observed, mainly

due to the appearance of the c.t. bands of the host lattices.

ThO.-, has the fluorite structure, ZrSiO4 and ThGeO, have the

zircon structure and ZrGeO, has the scheelite structure. In these

compounds Zr and Th are surrounded by eight oxygen ions. For

Pri+ and Tb 4 + two absorption bands can be distinguished of which

the one at lower energy clearly has two maxima (fig.3,4). The

1O

4+Fig. 1 Reflection spectrum of BaThO :Pr (298 K)

1 1

100

90

Ba2ZrO4:Tb4^

298K130K

20 30KK

40

Fig. 2 Reflection spectrum of Ba ZrO^:Tb4+

1S

100ThGeO4:Pr4*

Fig. 3 Reflection spectrum of ThGeO :Pr + (zircon structure) (298 K)

1OOr

o•*->

uQJ

<ba:

o

90L

20 30kK

40

A+Fig. L Reflection spectrum of ZrGeO^: Tb (298 K)

4+ . . . .Table II. Spectral da-a (in kK) of Ln ions m VIII coordination m

some oxides.

VIII coordination

! l S i' 4

ZrGeO;

IThCeO.; -4

Th°2

Ce

3!

31

31

-

% 5

,5

a obs

22.

21.

18.

19;

. in

Pr

i.

8;

22

kK

'3 S

23;

-4;

;20

29

30

29

o

Tb

T A -

21.

20;

8;23;

23;30

29

.A

lower maximum appears to shift as a function ot the host lattice.

For Ce"4 the position of the first band does not change (table II).

Under our experimental conditions it is not always possible

to incorporate the Ln-ions in the tetravalent state in the

desired host lattices. It turned out that their valency was

three in parL ot the lattices used. Even in those cases in which

incorporation was achieved, it sometimes was hard to perform. This

implies that part of the lanthanide dopant may still be in the lower

oxidation state, and necessitates to consider the possibility of

transitions between the oxidation states Ln (III,IV). J^rgensen and

Rittershaus (8);however, have given arguments which make an assignment

of the bands to this type of transition improbable. Furthermore

it is often possible to identify the f -*• f transitions of

trivhlent Pr in a reflection spectrum even at very low concentrations

(0,1 at . % ) . In cases where incorporation of the tetravalent ions

15

proved to be easy no trace of these bands could be found. In some

cases, however, the f -* f transitions of the trivalent ion were

superimposed upon the broad absorption bands due to the tetravalcnt

ion.

Cerium poses another problem because trivalent cerium

is known to have its f -* d absorption bands in the region

around 30 kK. This is also the region where the absorption

4+ 4+bands of the Ce complex may occur. In the case of Ce ,

therefore, only data are reported for compounds where in-

4+ 4+corporation of Pr and Tb presented no problems.

In not a single instance luminescence from the tetravalent

Ln ions could be detected for excitation with ultraviolet radiation,

not even at 5K.

Discussion

Since the possibility that the bands can be attributed

to mixed oxidation states Ln (III,IV) has already been excluded,

two alternative assignments to the absorption band(s) of the

4+ . .

Ln ions remain, viz. 4f -*• 5d transitions or c.t. transitions

from one of the delocalized ligand m.o.'s to a m.o. mainly

localized on the metal ion. In Ce" the 4f -*• 5d transitions3+ 3+

occur at about 30 kK (5,12) and in Pr andTb at even higher

wavenumbers, so that for the tetravalent ions these transitions

are expected at much higher energies. This leaves only the

latter assignment, i.e. c.t. transitions as proposed already

in ref. (8).

4+The spectra of the Ln ions in VI and VIII coordination

IB

will be considered separately. Afterwards some general

conclusions will be drawn,

a. The VI coordination

The Ln ions are expected to be incorporated into

the host lattices at the sites of the tetravalent metal ions

on the basis of the ionic radii. The existence of BaCeO_,

BaPrO and BaTbO indicates that such an incorporation should

present no problems.

The symmetry at the site of the tetravalent metal ion

is 0, in the perovskite structure and D in the K NiF, structure,h 4h 2 4

4+The similarity of the spectra of the Ln ion in both structures

indicates,however, that 0n is a good approximation in the latterh

case.

If we follow SchmidLke's topological treatment of an octahedron

of six ligands (13), the sequence of ligand m.o.'s is,

a, < 1 t, < t . < t , < t , = e < 2 t ,lg lu 2g 2u lg g lu

Thtse liganu m.o.'s are not yet perturbed by the metal ion orbitals.

This implies that the ligand e -orbitals are expected to be stabilized

considerably on interaction with the d-levels of the central ion

with e symmetry in 0, . Thus, undoubtedly, the first allowed

c,t. band is the transition t -»• f, because the transition

2 t lu -»• f is parity forbidden. In fact we ascribe the first

absorption band in the case of Pr and Tb + and the band

4+of Ce to this transition. The designation f, refers to the

lowest excited c.t. state oil the central ion. The assignment of the

second peak of the absorption band of Pr and Tb T is more

difficult. These central ions have several excited states within

their f configuration a few kK above the lowest and, in addition,

an overlap with a next band arising from the transition t -> f

can be expected.

Within the experimental error there is no influence of the

host lattice on the absorption maxima, neither in going from

BaZrO^ to SrZrO nor in going from BaZrO. to Ba ZrO, nor in the

sequence BaMO (M = Zr, Ce, Kf, Th). This excludes the possibility

of a considerable destabilizatien of che metal ion f orbitals, as

well as a consider-ble increase of ligand-ligand repulsion if

the Ln-0 bond becomes shorter, a matter to which we will refer

later on.

4+The nature of the shoulder at 22 V:K in BaZrO -Tb and

4+Ba^ZrO.-Tb is not completely understood, especially since this

shoulder seems to behave completely different from the remainder

of the band upon cooling. We have the impression that the intensity

of the shoulder decreases upon cooling. This could indicate the

possibility of a parity-forbidden transition, which has become

partially allowed through vibronic coupling. Further investigation

will, however, be necessary to clarify this problem.

h. The VIII coordination.

4+The site symmetry of the impurity Ln ion in the fluorite

structure is 0 , in the zircon structure D,, , and in the scheelite

4+ .structure S,. The similarity of the spectra of the Ln ions in

ThCK and ThGeO, is so strong that it seems reasonable to treat all

sites as having 0, symmetry and to consider the non-cubic crystal

field components as small perturbations.

It is evident that the shift of the first absorption band as

a function of the host lattice cannot be explained by a destabilization

of the orbitals mainly localized on the metal ion by mixing with

a ligand m.o. with the same symmetry designation. The f-orbitals are

usually considered to be mainly non-bonding and there is no

reason to believe this assumption to be incorrect. If such an

effect would occur at all it is expected to be much smaller than

in the d metal complexes. An indication that this is correct

follows from the spectra of the Ln ions in octahedral coordination,

where the absence of an appreciable shift excludes this possibility

as well as the possibility of strong ligand-ligand repulsion

(see above,i .

4+The positions of the c.t. bands of a Ln ion complex in

silicates and germanates may bt? different from those in "pure"

oxides, because the Si-O(Ge-O) bond is strongly covalent.

The first bands in ThO and ThGeO , however, are located at roughly

the same position. This does imply that such an effect cannot be

the major factor in explaining the shift.

It is noteworthy that in those compounds where the largest

Ln-0 distance can be expected (ThO?, ThGeO,) the first maximum

is located at the lowest energy. This tendency is directly

contradictory to an explanation based on strong ligand-ligand

repuls ion.

These arguments lead to the conclusion that the shift is

inherent to the VIII coordination itself, and thus we again

consider the sequence of ligand m.o.'s using the topological

treatment described by Schraidtfce. Considering only nearest

neighbour interaction, we found for the orbital sequence in a cubic

chromophore with 0 symmetry,

2g < 2 t l u < t l g

(2)

Again this is the sequence of 1igam! m.c.'s unperturbed by metal

ion orbitals. We note that there are two ligand m.o.'s of t

symmetry. A comparable situation exists in the case of a d-metal

ion in an octahedral surroundings with two ligand m.o.'s of t

symmetry. In this case the first c.t. band in the absorption

spectrum must be attributed '.o the transition 2t, -* d J not tolu

t« -*• d (4). This can partially be explained by a mixing of the

two ligand m.o.'s resulting in a dostabi1ization ot the highestt ligand m.o., (see 1) (13). But in the case of ,i cube, stilli u

not considering interactions ot the lieand rr..o.'s with the

central ion orbitals, the two sets of ligand t? orbitals do not

mix. This is caused by the fact that we cannot distinguish

between the two different se-'s of ligand atomic orbitals as in

octahedral coordination, where one set is used in constructing

the ligand a m.o.'s and the other in constructing the ligand r,

m.o.'s (15).

We now consider the pertubation caused by the central ion

orbitals. The important feature is the mixing of the i?etal ion t ^

orbitals from the 5d-level with both ligand m.o.'s. This mixing

destroys their mutual orthogonality, resulting in a rearrangement.

As a consequence the position cf the highest ligand t.5 m.o. s

relative to the t. ligand m.o.'s depends strongly on the energetic

position and the extent f delocalization of the d-orbitals

ot the l.n"* ion. This in fact should be the most important

tactor i .• explaining the shift of the first c.t. band of the

l.v.~* iii'is in VTII coordination.

O:i the basrs or .arguments given above the first maximum in

4+ 4+VI!! .. .-u-rdi nation for i'r and i'h should be ascribed to the

;ransition 2 t. » : and the maximum at about 22.5 kK to t, -* f,

although in ZrSiO, the order could very well be reversed. The

band at about 2 *5 kK poses the same problem as the second maximum

for VI coordination.

4+For Cc ihi! situation is markedly different. The band

lioes not shift and is situated ..t more or less the same position

as in VI coordination. Tho iact that <> has an £ confif ition,

-:;;Kes it ditii.uit to predict tin- m-iiav i our ol the 5d orbitals.

rurtiuT analysis wi,i be necess.-ry u> i.in>i('i"s t and tl-.L;' behaviour

•'triple t ely.

c. Optical electronpgativity.

It we ascribe the first maximum at about 24.5 kK in the

4+ 4+absorption spectra of Pr and Tb in VI coordination to the

transition t -• 1, and if we accept for the optical electronegativity

•) _

of oxygen y (0" ) = 3.2 (91, wo arrive at a value of \ = 2 . 4opt • Auncorr4+ 4+

ior both Pr and Tb , We have used Jtfrgonsens -lef ini tion (15)

i't the uncorrecU'd optical e 1 ect runegat i vi ty .

• • = [ > . .(>; ) " X (M)] .30 kK,opt uncorr

whore - is the position cf the c.t. band, x (X) the optical

electronegativity of the anion and y (M) the uncorrected

optical electronegativity. This is slightly lower than values

calculated from II1-IV oxidation potentials in chromophores with

4+halogen ligands (16). For Ce we get x = 2.1 if we take

33 kK for the position of the first band, in excellent agreement with

the values reported in reference (8).

It will be evident that a calculation cf v of Pr and'uncorr

4+Tb from the position of the first band for VIII coordination would

be meaningless, since this optical electronegativity would not be

a property charateristic of a particular ion because it would aiso

characterise a specific surroundings.

If we calculate \ from the transition t, -*• i we get\inccrr lg 6

values which are consistent with the values obtained for VI ccu.--rdin.-it i on;

for Pr"4* and Tb + >- =2.45 and for Ce**+ • = 2,15.•uncorr uncorr

This procedure limits of course the practical value ot the optic;) i

electronegativity in predicting the approximate position of the first

c.t. band of an ion in VIII coordination. This, however, is a minor

limitation compared with a situation where the value would be

different in each compound.

d. The Eu ion.

It should be noted that an unexpected variation of the

3+position of the first c.t. band has also been noted for Eu

(6). Closer analysis reveals that here again the position

of the band is more or less fixed in VI coordination and that

the strong variation occurs for VIII coordination, so that the

same line of reasoning probably applies here too. Elsewhere we

will disruss the Eu ion further (17).

Luminescence

It is well known that highly--,.-nursed cations like Nb ,

A'~ and U often display lumi nescence (18,19,20), which can bo

atributed to transitions between, rietal ion orbitals and ligand

m.n. 's. The Eu ion or. the other hand displays a red luminescence

which can he acributed to transitions within the f configuration.

We note that tho tetravaient Ln L. ns display no luminescence at

all. It is hardly conceivable that the energy could be lost

by radiationless processes without another intermediary excited

state. It is unlikely that this intermediary is one of the higher

excited f-states of the central ion, so that one is led to

conclude, '-hat either an excited d-stuto or a state which arises

from 3 hole in another ligand m.o. must play a part in the process.

Acknow t edgmenLs

The author is much indebted to Prof. Dr. G. Blasse -for

encouraging this work and for many stimulating discussions.

Mr. F.".A. Stagers is gratefully acknowledged for the

preparation of and spectral measurements on the samples.

References

(!) C.K. JjSrgensen and W. Preetz, Z. Naturfor.sch. 22a, ^45 U967)

(2) C.K. J^rgensen, Modern Aspects of Ligand Field Theory,

North Holland Publishing Comp., Amsterdam 197 1

(3) H.H. Schmidtke, Coord. Chcm. Rev. 2_, 3 (1967)

(A) C.K. JsSrgensen, Progr. Inorg. C'm-n:, _i_ , 101 (]y70)

(5) J.L. Ryan and C.K. Jflrgensen , J. I'hys. Cher.1.. 2£s — 45 (196b,)

(h) G. Blasse, J. Solid Si;Ui' Cliem. - , 32 (.1972; and references

citec there in.

(7) J.L. Nugent, R.D. Baybar::, J.L. Burner and J.L. Ry.;:i,

-I. Piiys. Chern. 7_7_, 1 iJB U ° 7 J )

(8) C.K. J«!>rgensen and E. Rittersh^"S, Mat. Fys. Medd. Dan.

Vid. Selskab 35, no 15 (I'jhl)

(9) N. van Vugt, T. Wigmans and G. Klassi', J. iR^rL,. .'o.. 1 . Chen.

25_, 2602 (!973)

(10) C. Kor'iirr., Kel K:c tanco S t. c t roscopy ; ^prin^er-V: rl a;; Berlir. 1 9 < -•

iiij j.K. :-,i;vcr, J. El ec I r.-.tn:-... M--. 113, !2- U96*O

(12) C.K. Jrfrgensen, Mol. Phys. _5> 2 7 1 C1962)

(13) H.H. Schmidtke, J. Chcm. Phys. £5, 3920 (1966)

(14) G.N, Henning, A.J. McCafiory, P.N. Schatz and P.J. Stephens

J. Chem. Phys. _48, 565b (1968)

(15) C.K. J«5rgensen "Absorption Spectra and Chemical Bonding"

PL I'gainon Press , New York 1962

(16) L.J. Nugent, R.D. Bayharz, J.L. Barnett and J.L. Ryan,

J. Inorg. Nucl. Chem. 32^ -503 (1971)

(17) Chapter ill of this thesis

(18) C. Blasse and A. Bril, Z. phys. Chem. N.F. 57_, 187 (1968)

(19) F .A. K r o g e r : "Some A s p e c t s of Luminescence of S o l i d s "

E l s e v i e r Amsterdam 1948

(JO) J .Th .W. de Ha i r and G. B l a s s e , J . Luminescence 8, 97 (1973)

CHAPTER I I I

TKF CHARGE-TRANSFER ABSORPTION BAX1> OF E u " + IN OXIDES.

Ab s t r a c t

3 +The position of the charge-transfer banci ol Eu in I he abs^rpti

spectrum of a number of oxides is discussed. It is shown that this

position is more or less fixed in octahedral VI' eoordination and

that it varies in VIII and XII coordinations as a function 01 the

effective ionic radius of the relevant host lattice ion.

Introduction

The red luminescence ci the Eu ior. has i ound technical

application. As a consequence a considerable amount of optical

data on this ion in a iarge variety oi compounds has become

available in the last decade. It has been suggested that the

efficiency of this luminescence is determined by the spectral

position of the charge-transfer (c.U) band in the u.v. region. (1,2)

Up till now the variation of the position of this c.t. land as

a function of the lattice was not understood. It has been noted

that the band position varies considerably from ~*48 kK in

ScPO,:Eu~+('3) to 30,5 kK in SrLaLiWO,:Eu + (4). This experimental

fact is not in accordance with Ji6rgens<?n' s relation between the

position of a c.t. band (a) and the optical electronegativities (v)

of the central ion (M) and the ligand (X) (5):

ae

[ , fXi - \ (M)l . 30 kK.-opt ' uncorr

in reference (6; we have proposed a model which explains the

variation of the c.t. band position o! tetravaLent lanthanide (Ln)

ions in VIi1 coordination in oxidic lattices. It seemed interesting

to I t"v to extend this model to the Lu' c.t. band. Furthermore

we vi Li trv to give some data -.one erning tlie position of this band

. n , .'"ipounds in vhic.n the Ku ior. lias the same coordination number,

nut . :::: tennt s tc: reocheiiii M t ry .

Kesuiti

All data were taken from trie literature (3,4, 7-15) and

are suirar.arized in tables I, 11 , ar.ti FIX. We use only data

relating to the Ku ion on the site of trivalent lanthanide

..3+ . . 3+ions or jn the si>:e ci L ana Sc iii oxidic lattices.

TaSji..' i shows t!-.at Liie posiuun ol the L U ' e.t. band does

not s hi In appreciably Ln the case of .) (nore or less regular)

octnliedral coordination of the Hu ion by oxygen ions in lanthanide

compounds (including Y and Sc compounds). The ionic radii of the

relevant h.'st lattice ions vary from Sc (rVT = 0.745 A) to

^d (r = 0.938 A ) ; the ionic radius of Eu in VI coordination

is rv j = !.).')47 A (14). This is consistent with observations

lor Letravaient Ln ions(see chapter II). For a regular octahedron

this band can be ascribed to the transition fit -* f (see II-1J .

Table I. Position of the c.t. band of Eu in octahedral VI

coordination in oxides.

compound

ScBO3 :

NaLuGeO,4

LiLuSiO,-4

KaLuSiO,

LiLuO., :

LiLuGeO,4

NaYGeO.4

LiYSiO,

YBO : E

En 3 +

: Eu3 +

: Eu 3 +

: Eu 3 +

Eu3 +

3+: Eu

: Eu 3 +

: Eu 3 +

u 3"

crystal

structure

calcite

olivine

olivine

olivine

a-LiFeO.,

olivine

olivine

oiivine

calcite

reference

7

8

8

8

Q

8

8

8

10

a(kK)

43

4 3.4

43.1

43. 1

4 3.0

41 .6

4 ;i. i

4 Z . 7

4 j . 7

3+LiYO? : Eu

Y2°3 : Eu3+

CaYBO4 : Eu 3 +

LiYGeO^ : Eu

: Eu 3 +

3+

NaGdGeO4 : Eu

3+

3+

LiGd0o : F.u

: Eu3+

NaGdO., : Eu3+

.i-LiFeO.,

bixbyite

Y.,BeO.2 4

olivi ne

vaterite

olivine

a-HA109

bixbyite

] 1 41.7

4 1 . h

41.3

41.1

41. 1

Accepting an average band position of 42 kK in octahedral VI

coordination and using JjSrgensen's definition (5) we arrive at

a value of . = 1.8 for the Eu ion with x ,. (°" ) = 3.2.'uncorr opt

This is only slightly lower than the value reported for Eu3+

complexes with halogen ligands in solutions, viz. 1,89 for Eu

in EuClf" (15).

Turning to VIII coordination it is seen in table 11 that the

position of the first exitation band shifts from very high energies

for Eu )+ on the. site of S c ^ vr = 0.67 A) to very low energies

for tv/ on the site of I,a (r = i.18 A ) .

The same line of reasoning as in the case of The tetravalent

Ln ions applies l-ere. Again the position of the 2t.n ligand m.o.

Table ii. Position of the c. r. . band of Eu" ' in VIII coordination in

,xlde1 .

".•ompouniic-rys t a 1

structure

reference o(kK)

SrPO : En"

YPO, : Eu J H

Y.Ga.O : EuJ D 1 1

YTaO.. : Eu J +

Gd.,Zr,O.. : E u 3 +

LaPO, : Eu

LaTaO,: Eu

3-t

3+

zircon 3

zircon 1 I

garnet 12

fergusonite !3

disordered fluorite 12

CePO,

own type

I 1

13

~ 48

- 45

42.5

40.8

38.1

37

36

Table III. Position of the c.t. band of Eu in XII coordination

in oxides.

compound

GdGaO :

GdAl\)3 :

LaA10 3 ;

SrLaLiWO

Eu 3 +

Eu 3 +

Eu 3 +

crystal

perovski

perovski

perovski

ordered

te

tc

te

perovski te

reference

12

12

i 2

U

a(kk)

AO. 5

38

32.3

30.5

(see 11-2) relative to the t ligand m.o. is strongly dependent

on the energetic position nnd the extent of delocalization of the

metal ion d-orbitals and tho magnitude of the interaction terms

of both sets of ligand t. m.o.'s with the metal t d-orbitals.

Thus the Eu-0 distance becomes a factor which strongly determines

the ct. band position in the case of cubic VIII coordination.

For an oxygen XII coordination with (more or less regular)

cubic symmetry only data for Eu on the site of a Gd or a La

ion have been reported. The data point to the existence of an

effect similar to that observed for Eu "*" in VIII coordination

(see table III) . This is not unexpected since in this coordination

there are also two sets of ligand t» m.o.'s as well as two sets

of ligand e m.o.'s; furthermore it should be noted that the

crystal-field splitting in both coordinations is similar (16)-

3 D

We thought it interesting to compare data on the r.t. land

of Eu ions with the same coordination number but with different

3+stereochemistry. Only a few data are known for the Eu c.t. band

in a trigonal prismatic VI coordination, viz. 40.5 kK in

YA1,B4O ~ (11), 40.1 kK in GdAl B ^ , , (12) and 41.3 kK in

NaGdSiO (8). Here the position of the c.t. band appears to be4

more or less fixed as in octahedral VI coordination. Accepting

an average band position of 40.5 kK from the available data we

3+arrivc at v = !.85 for the Eu ion in trigonal prismatic

Auncorr °

VI coordination. This is somewhat higher than the value for

octahedral VI coordination and well within the limits of experimental

error.

One clear discrepancy should be mentioned. For Eu on

a La site in octahedral VI coordination only one datum is

reported, viz. NaLaOn:Eu with the « - LiFeO. structure whichi. L

is an ordered rocksa]t structure. The c.t. band is reported to

be situated at 36 kK (9) which is much toe low for an octahedral

VI coordination (r (La ) = 1.045 A ) . At the moment we cannot

explain this discrepancy. It seems worthwhile to reinvestigate

NaLaO9:Eu3+.

In reference (6) we found for the tetravalent Ln ions in VIII

coordination a second maximum which we attributed to a transition

ligand t. -* f. For Eu in oxides nn reliable data are available,

one of the difficulties being the absorption of the host lattice,

which is often expected in the same region of the spectrum.

It has been suggested that the variation of the c,t.

31

3+ .absorption band of Eu in oxides can be rv,-1 -j by the

i _

variation of the potential at the 0^ anion (i2). This, however,

cannot be a major factor, since the consequence of this would

be that the magnitude of the variation should be roughly the same

for all coordinations.

In conclusion we can state the following concerning the

position of the Eu c.t. band in oxidic lattices:

(1) The coordination of the Eu ion is a determining factor.

In octahedral VI coordination the band position is more or less tixeo,

in cubic VIII and XII coordinations the band position varies as a function

of the host lattice.

(2) The variation in VIII and XII coordinations is proportional

with the Eu-0 distance. With increasing bond length the band

shifts to lower energies.

These rules may be used as strong indication for the exclusion

3+of certain surroundings in unknown structures. Eu in Y SO , for

example, has the c.t. land at 37 kK (11). This seems to exclude

VI coordination for Y in this compound.

Finally these rules relate the value of the quantum efficiency

of Eu luminescence for excitation into the c.t. band to the

structural details of the host lattices applied.

Acknowledgement

The author is much indebted to Prof. Dr. G. Blasse for

suggesting this work and for many stimulating discussions.

3 S

References

1. G. Blasse and A. Bril, J. Electrochem. Soc. 115, 1067 (1968)

G, Blasse, A. Bril and J.A. de Poorter, J. Chem. Phys. 53,

4430 (1970)

j. G. Blasse and A. Jri 1 , J. Chem. Phys. _50_, 2974 (1969)

4. D. Krol and G. Blasse, unplubiishcd result.

5. C.K. J(6rji,£'nsen, "Orbitals in Atoms and Molecules" Academic Press

New York 1962

6. H.E. Hoefdraad, J. Inorg. Nucl. Chem., in press.

7. A. Bril aid W.L. Wanmaker, J. Klectrochem. Soc. Ill, 1363 (1964)

8. G. Blasse and A. Bril, J. Inorg. .Nm:l. Chem. 29.. 2 2 3 1 ('967)

9. G. Blasse and A. Bril, J. Chem. Phys. 4_5, 3327 (1969)

10. G. Blasse, J. Chem. Phys. 5J_, 5529 (1969)

11. G. Blasse, J. Solid State Chem. 4, 52 (1972)

12. C. Elasse, J. Chem. Phys. J O , 235b (1966)

13. G. Blasse and A. Bril, J. Luminescence 3_, 109 (1970)

14. R.D. Shannon and C.T. Prewitt, Acta Cryst. B2_5_, 925 (1969)

15. C.K. JjSrgensen in "Halogen Chemistry" Vol. _1_, V .Gutmann , Ed.,

Academic Press, London 1967

If). J.S. Griffith, "The theory of transition-metal ions" Cambridge

University Press, 1961, page 204.

3 3

CHAPTER IV

EVIDENCE FOR THE INFLUENCE OF AN EFFECTIVE CHARGE OK THE POSITION

OF THE CHARGE-TRANSFER BAND OF Eu + IN SOLIDS

Summary

Evidence is given for the influence- of the effect, ivir char

on the position of the charge-transfer band of the Eu ion

occupying the crystallogrqphicsite of cations with a different

formal charge (viz. Ca , Zr ) in oxides.

The spectral position of the first charge-transfer (c.t) band

J! J molecular or ionic complex is determined by the oxidizing

character ol the central metal ions and by the properties of the

set o! iiu.iads (1,2.). In solids the presence of an

eifc-ctivo charge may also >.' or importance. The central metal

ion bears a formal effectivj ciiaree if it replaces a

netal ion with different charge in the host lattice concerned l. :> i .

The Eu' ion on the bite of a Ca ion in a calcium compound, tor

example, is effectively positive but on the site of a Zr ion in

a zirconium compound it is effectively negative. It is to be expected

that 1igand-to-metal c.t. bands are situated at relatively low energy

if the metal ion bears an d f ec t .ve lv positive charge and at relatively

high energy if the metal von bear;; an effectively negative charge.

Almost no attention has been paid to this effect, which may be

due to the fact that much o;' the work on c. i.. spectra ha;; been

performed in liquids when.-- it is -.nposs ible to define an effective

charge. An example where such an effect has been claimed is the system

CaSO,-V,Tb whefn the fact that the Tb emission is not quenched by the

Tb -V c.t. state has been explained by noting that Tb on a

Ca* site has an effectively positive charge and V"^ on a S site

has an effectively negative charge (A,). Thus, it was argued, the

c.t. band should be expected at higher energies than in YPO,-V,'lb

where the Tb luminescence is quenched rapidly if P is replaced

by V (5,6).

3 5

We now present some direct evidence for the existence of such

an effect. We have recently proposed a model to explain the strong

variation of the position of the first c.t. band of Eu (7) and oi li,

tetravalent lanthanides (2) in oxides, It was noted that in octdheur.il

VI coordination the position of the first c.i. band is more uv less

fixed (42 *_ 1 kK) thus obeying Jargons ens rules concerning the

optical electronegativities (8). A suitable system to study tin-

effect of effective charge on the spectral position of c.t. hands i.- ,

therefore, the Eu ion incorporated in three different compounds

on octahedral sites which are occupied in the pure compound by

di-, tri- and tetravalent metal ions, respectively. Tr limit t':.<.-

2 +effect of different ionic radii we considered Ca as t K divalent

4+

and Zr as the tetravalent ion.

The results are the following: m ("aO-K.i the c.t. hand is

located at - 36 kK (9), in Y ,0 -Ku'^ at 4 ', . " kV (10, .;ud in /.rP 0,-Lu '*

at 45 kK or higher. The spec:r:il shift due :••• i!;e -pr-.-s-.'nce ,.-t ,;n

effective charge (A-5 kK} is much larger than the small variation

in the position of the c.t. band of octahedral ly coordinated Eu"

in a large number of compounds where it docs not bear an effective

charge (1-2) kK. These date indicate, therefore, directly the

existence of an influence of the effective charge en the oxidizing

character oi antral metal ion.

Unfortunately it is extremely diilicult tn obtain more exauu-

ules. In VI coordination the c.t. bands .a the host :.!ti.ice niton

interfere, because they appear in trie same region oi the spectrum.

36

Furthermore it cannot be excluded that in other cases different

effects will obscure the effective charge influence. We consider

Ku ' * in VIIT-coordination to illustrate this remark. For this

o.v u'd in.it ion the position of the c.t. band is also determined by

ihe Ku-n distinct- C7) , which in turn is influenced by the formal

riu-.'fU1 or the !-"u" ion.

3+ 3 +

As an i '. '.v.stration ve L-o'.npo.ro YPO -tvi and CaSO,-Eu ,P. In both

• ".••rv.iour. ••::•> Ki: .s in VIII coord: n.v. i on. In the former it is effectively

neutral ,in the latter effectively positive. The position of the

.. t. b.-ind has been reported for VPO,-Ku' at •-( 5 kK (10) and determined

nv \;s i.ir CaSO -Eu ,? at US kK. This large shift may partly be due

i;'- the i nil nonce oi the effective charge, but on the other hand the

Ku-0 distanie in the calci-e.r: I'oinpound cannot easily be compared with

Liuit in rhe yttrium compound on >_!:..• basis o\ the ior,ic radii of both.

hi st tons. Therefore, the nagn; lude ol t'rie shitt of the r.t.banti•J-i . .

-.liii- to an eile " i ve charge or. *. ho Ku i ,'m w. \ ill ,: oordi n.it l on

.-.innriL easily be estimated.

We I'ur.i lude that the ini lueiice t>f ,-'.n effpctive i:harge on the

position o] c.t. bands is not easy to d e m o n s t r a t e , but that the most

suitable example lur such a study, viz. oetahedrally coordinated

3 *Ki: m oxides, presents strong eviderue for such an effect which mustbe quile genera i.

r i menl d )

ZrP_,O_-Eu was prepared by firing recrys tallized ZrOCl.,. 8 H^O,

(NH4).,HPCV and Eu^O^ in oxygen ol" I1OO°C. CaS0,-EuJ+r was prepared

by firing CaCn^, ( N H ^ S O , , Eu 0 and ( N H ^ H P O in oxygen at

JOOOUC. Samples were checked by X-ray analysis using CuK \ radiation.

3+The spectra! position OL the .:. t . band ot the Eu ion wa:- dot enr.ine,

from the excitation spe. ira of the red "D - V ,, emissiono •

a s w e r e a i l l i t e r a t u r e v a l u e s o i l e d . T h e e x c i t a t i o n s p e c t r . ;

w . - r o m e a s u r e d o n .-; P e r k i n - E I t r e r M J ' F - J ' L s p e . ' t r o t ' u o r I m e t e r

J O ' V. / . n o i - j r : ' - . ' . ' L e d f o r L i : , l . i n i p v : i l e i i s i t y v ; s i n i ; s o d i u m t :.'; i c y "l a t i - .

D u e : . : t h o l o w i a t r . p i n t e n s i t y a b e v e ^ 0 k K w o c o u l d o n l y d e t e r m i n e

a l o w e r l i r i i t ; o r t i i e c . r . b a n d p o s i t i o n o t / ' r P , 0 _ - E u J .

R e i e r e n ' . ' 1 ' s

! . C . i . . J s i i . ' . c i i s e i i , M o d e r n A s p e c L r - ' - ' l

N \ r ; ;i n o '. i a n d ; ' ^ : i i . C e . A r v s ;. ,-ri.-. a r

_ . i . !• . r i c o j r a a d , J . I ' . i o r p . N o e l .

; ' . I r . . \ . K r o g e r , "Thi-- C h o a u s t r v o l I r e p e r l e o t C r y s t a l s " , N o r t h

• i o l l a n d T i i b l . C o . , A m s t e r d a m i 9 6 i>.

W . T . D r a a i a n d U . B l a s s e : P l i y s . S L a t . S o l . ( a ) 2]_, 5 0 9 ( 1 9 7 -

3 . R . G . He L v j s h , T . Y . T i e n , : - : . F . G i b b o n s , P . J . Z a c a m a n i d e s

unc H . L . S t a d l e r , j . C h e m . P ' n y s . _53_> 6 ° : l i Q 7 0 )

6 . r , , ;;. i a :•; s e , J . L u m i n e s c e n c e 1 ,2 7 6 6 ( 1 9 7 0 )

7 . H . I " . H o e i d r a a d ( C o b e p u b l i s h e d )

b . i i . K . J i f r f ! e n s e n , " A b s o r p t i o n S p e c t r a a n d C h e m i c a l b o ; i v i i n ^ "

P e r g a n i c n P r e s s , New Y o r k , 1 9 o d

9 . W. L e h m a n n , J . L u m i n e s c e n c e b_, 4 5 5 ( 1 9 7 3 )

1 0 . G . B l a s s e , J . S c l i . d S t a t e C h e m . 4 , 5 2 ( 1 ^ 7 2 )

( R A F T E R V

GREEN KM I T T INC PKASEOUYVIirM IN CALCHJM 7. i RCON.-YIT

on 1 v LTI. t r . ITII s s i

T i ; e 1 uir,i n e s c e n ^ ' e o r P r : n C i Z r i ) . i s r e p o r t e d ,

i o n i r . i ' a / ' . r O . , s h o w s i" . Li ie v i s i b l e r t - ^ i 1

f v '. i ; i n . i l i ; n ' . i r- . :"; r l i t- r' i i.". f . '. . • " i r - : r v i c L

w h i c h y r e e n and r e d l u r a i n e s e e n c e i s I'.mi Li .ed .

T i u P r

The Pr3+(4f2) ion is known to emit in the visible region

a red and a green luminescence originating from the D,, and

S levels, respectively (fig.l). Sometimes the green emissiono

is f-vt-n absent. In the course of our studies on the spectra of

u

>U)

LJJ

50r

40-

30rJ20

10

Ol

's,

'4IS

fig.

The energy level scheme of Pr"3+

3H:

tetrav.i lent lanthanide ions (I) we happened to find a compound

in whirh the Pr ion shows green 1 umini'sceu.-e only, viz.Ca/.rl).. ~l'r ,

with perovskite structure. Fig 2 shows the emission spectrum under

ultraviolet excitation. The maximum of the excitation band is

situated at 245 nm, and can be ascribed to a Af-*5d transition (2,3).

Host lattice absorption occurs at snorter wavelength only.

J1

\

1

1

1

400 500 600Mnm)

3+fig. 2 Emission spectrum of CaZrO -Pr at room temperature for

245 nm excitation. $ denotes the spectral radiant power

in arbitrary units.

We compare this result with the luminescence of calcium-

modified ZrO -Pr3+ with the fluorite structure. The exact composition

is 0.85 ZrO . 0.15 CaO. In this compound both the red and the

green emission occur. In striking contrast with CaZrO^-Pr , che

red emission has almost the same intensity as the green emision.

The ultraviolet excitation band peaks at 295 nm and is also

attributed to a 4f->-5d transition.

In CaZrO -Pr ^ the lowest level of the 4f->-5d configuration

is situated at relatively high energies (41 kK), even higher than

in Y Al 0 ~Pr , where the lowest 4f5d level is situated at about

35 kK '4). The compound Y A1.0 -Pr shows upon excitation into

the 4f5d state radiative emission from the lowest lfvel of the 4f5d

configurar- ir. to the ground state. This emission extends from

300-450 ran (4) . Ws could not detect any 5d->-4f luminescence in

CaZrO ~Pr , not even at 4 K. A similar situation exists in

CaF^-Pr and BaF.^Pr . where the lowest 4f5d level is situated

slightly below the S level (5). Here also 5d->4f luminescence

is absent.

We conclude that in CaZrO -Pr the lowest 4f5d level is de-

excited into the levels of the 4f*" configuration between P., and

P , but certainly not into the V)n level . We have drawn this

situation s.chematicly in fig. 3.

In Ca-modified ZrO^-Pr the 4f5d state is situated at much

lower energies (34 kK). It is obvious to assume that a non-radiative

transition from the lowest 4f 5d level to the D7 state has now a

higher probability. This has been indicated in a schematic way

in fig, 3 and explains the occurence of red emission in Ca-modified

Zr0o-Pr , which is absent in CaZrO^Pr +.

On the basis of this model one would expect the green emission

to dominate in both compounds upon selective excitation into the

2 3 3levels of the 4f" configuration between P and P , since

3 1obviously, the probability of a non-radiative transition P -»- D«

/CaZrO,

;Ca-ZrO3

fig. 3 Schematic configuration-coordinate diagram for Pr

2Drawn parabolas refer to the relevant levels of the Af

configuration, the upper broken parabola to the lowest level

of the 4f5d configuration of Pr in CaZrO^ and the lower

broken parabola to the lowest Af5d level of Pr ' in Ca-stabi1ized

ZrO_. Their equilibrium distances have been taken

arbitrari ly.

4 4

is low. Upon verification this assumption proved to be correct:

the red to green ratio for Ca-modified ZrO -Pr under this condition

(at room temperature) is about 0.2.

Our results suggest that the emission of Pr upon excitation into

the 4f-*5d absorption band depends strongly on the characteristics of

the lowest 4f5d level and in this way on the choice of the host

lattice.

Experimental

Ca-modified ZrO2~Pr + was prepared by firing CaCO3> ZrO2

(prepared from recrystallized ZrOCl«.8H_O) and Pr,0 (nominal

formula) at 1500 C and -ubsequent quenching into air. The

composition ratio was 85 mol. percent ZrO and 15 mol. percent CaO.

The compound CaZrO -PrJ+ was prepared from (NH.) C_0,.H 0,

Zr0Clo.8Ho0, CaCl. and PrCl- as described in ref. (6). The

praseodymium concentration was 1 at. percent. Samples were

checked by X-ray analysis using CuKa radiation. Emission and

excitation spectra were measured on a Perkin-Elmer MPF-3L spectro-

fluorimeter at 300 K. The emission spectra were corrected for

photomultj.plier sensitivity, the excitation spectra for lamp

intensity. The heliumcryostat was an Oxford C.F. 100.

Acknowledgement

Mr. G.P.M. van den Heuvel is gratefully acknowledged for the

preparation of the samples.

References

1. H.E. Hoefdraad, Chapter II of this thesis

J. Inorg. Nucl. Chem. (in press)

2. C.K. Jiirgensen, Mol. Phys. 5, 271 (1962)

3. E. Loh, Phys. Rev. \h]_, 332 (1966)

4 . M . J . Weber, S o l i d S t a t e Conrn. J_2, 741 (1973)

5 . J . L . Sonunerdi jk , A . B r i l , and A.W. de J a g e r , J . Luminescence

9_, 288 (1974)

6 . J . Chanewaye and G. Boulon, C.R. Acad. S c i . P a r i s 2 7 1 , B486 (1970)

4B

SAMENVATTING

Dit proefschrift beschrijft de resultaten van een onderzoek naar

de ladingsoverdrachtspektra (charge-transfer = c.t. spektra) van

vierwaardige lanthanide ionen en van Eu in oxiden.

Hoofdstuk II is gewijd aan de c.t. spektra van de vierwaardige

44 4+ 4+lanthanide ionen Ce , Pr en Tb . Deze zijn hiertoe ingebouwd

in vaste metaaloxiden. Uit de reflektiespektra van poedervormige

monsters blijkt dat:

a. de positie van de eerste c.t. absorptieband praktisch

onafhankelijk is van de keuze van het inoederrooster indien net lanthanide

ion ingebouwd is op een oktaedrisch zesomringde positie, en dat

b. de positie van de eerste c.t. absorptieband afhankelijk is van

de afstand tussen het lanthanide ion en de omringende zuurstof

ionen indien het lanthanide ion ingebouwd is op een achtomringde

positie. Er is een model opgesteld dat deze afhankelijkheid van de

c.t. absorptiebanden van de aard van het moederrooster goed kan

verklaren.

In hoofdstuk III wordt het model dat is opgesteld om het

gedrag van de c.t. absorptiebanden van de complexen van vierwaardige

lanthanide ionen in oxiden te verklaren, toegepast op driewaardig

europium in oxidische roosters. Zoals verwacht blijkt ook hier de

positie van de oerste c.t. band in VI coordinatie vrijwel onafhankelijk

van de keuze van het moederrooster te zijn en in VIII en XII

4 7

coordinatie te schuiven als funktie van de Eu-0 afstanci. De verschuiving

is aanzienlijk groter dan in het geval van de vierwaardige lanthanide

ionen.

In hoofdstuk IV wordt een experimented aanwijzing besproken

voor de invloed van een effektieve lading ter plaatse van het metaal-

ion op de positie van c.t. absorptiebanden. Als voorbeeld wordt

gebruikt de reeks CaO-Eu +(36 kK), Y 0 -Eu + (4 2 kK) en

^ - E u 3 * (> 45 kK).

In de loop van het onderzoek naar de c.t. spektra van vier-

waardige lanthanide ionen werd een verbinding gevonden waarin

driewaardig praseodymium in het zichtbare gebied alleen een

groene luminescentie vertoonc onder ultraviolette 4£-»5d excitatie,

n.l. CaZrO -Pr . Meestal wordt naast deze groene ook rode emissie

waargenomen. In hoofdstuk V is een model opgesteld dat dit kan

verklaren; bovendien wordt beargumenteerd dat de overgangswaar-

schijnlijkheid voor een stralingsloze overgang P -»• D? relatief

klein is.

4 B

SUMMARY

Chapter 1 of this thesis is a general introduction on charge-

transfer (c.t.) spectra of chromophores, specifically ligand-to-

metal charge-traasfer.

4+ 4+ 4+In chapter II the c.t. spectra of Ce , Pr and Tb in a

number of oxides are reported. It is noted that the position of

the first c.t. band is fixed for the metal ion in an oxygen coordination

of VI, but varies in VIII coordination as a function of the host

lattice. It is argued that this variation is inherent to VIII

coordination itself.

3+In chapter III the position of the c.t. band of Eu in the

absorption spectrum of a number of oxides is discussed. It is

shown that also this position is more or less fixed in octahedral

VI coordination and that it varies in VIII and XIII coordinations

as a function of the Eu-0 distance.

In chapter IV evidence is given for the influence of an effective

charge at the metal ion on the position of c.t. absorption bands. The

series CaO-Eu3+ (36 kK), Y 0 -Eu3+(42 kK) and ZrP^-Eu 3* (> 45 kK)

is used as an example.

During our study on tetravalent lanthanide ions a compound

was found in which the trivalent praseodymium ion shows in the visible

region of the spectrum only green emission, originating from the

P level. Usually red emission from the D level is observed too.

This has been discussed using the properties of the lowest Af5d

level.

5O

CURRICULUM VITAE

De schrijver van dit proefschrift behaalde zijn

H.B.S.-B diploma in 1965 aan het Peter Stuyvesant

College op Curasao. Kort daarop vertrok hij naar

Nederland waar hij aan de Rijksuniversiteit Utrecht

chemie ging studeren. Na in 1970 het kandidaats-

eksamen (So) te hebben afgelegd, deed hij in 197 2

het doktoraal eksamen met specialisatie Theoretische

Chemie en bijvak Vaste Stof Chemie. Op 1 december

1972 begon hij op basis van een promotiebeurs aan

een onderzoek bij de vakgroep Vaste Stof. Op 1

februari 1974 kreeg hij een tijdelijke aanstelling

aan de Rijksuniversiteit Utrecht op een deelpost.

De resultaten van het onderzoek zijn in dit

proefschrift vastgelegd.

H.E. Hoefdraad RECEPTIE

Oude Gracht 209 na afloop van de promotie

Utrecht in het Universiteitsgebouw

Domplein 29, Utrecht

STELLINGEN

bij het proefschrift: Charge-tvansfer spectra of lanthanide ions in oxides

1. Lever's uitspraak dat de metaal-naar-1igand charge-transfer band

van een metaal complex naar lagere energie schuift als het

coordinatiegetal lager wordt is niet in overeenstemming met de

experimentele gegevens.

A.B.P. Le.ver, "Inorganic Electronic Spectroscopy" Hfdst 8

Elsevier Amsterdam 1968.

A.B.P. Lever, J. Cbem. Ed. 5J_, 612 (1974)

2. De vorm van de asymetrische interne valentietrilling van

geordende perovskieten is ten dele zeker toe te schrijven aan

een deformatie van het rooster.

M. Liegeois-Duyckaerts en P. Tarte, Spectrochim. Acta

30A, 1771 (1974)

A.F. Corsmit, H.E. Hoefdraad en G. Blasse, J. Inorg. Nucl.

Chem. 34_, 3401 (1972)

3. Het is een oversimplifikatie kristalgroei volgens de Verneuil

methode op te vatten als groei uit de vloeibare fase.

M,A. Verneuil, An. Chim. Phys. _3» 2 0 ('904)

F.A. Reiss, Applied Opties 5_» ™ 2 (1966)

4. Aan de energiebeschouwingen van Khetan met betrekking tot de

plaats van protonering in quinolon-derivaten moet niet al te

veel betekenis worden gehecht.

S.K. Khetan, Chem. & Ind. 182 (1973).

5. Het is onjuist de groene luminescentie van het uranyl ion

toe te kennen aan een overgang binnen de 5f schil.

R.N. Shchelokov, Yu.I. Krasilov en V.E. Karasev

Sov. Phys. - Solid State _1£, 1885 (1973)

6. De wijze waarop de doof temperatuur van de rode uranium

luminescentie in de reeks BaWO,-U, SrWO.-U, CaWO -Ua i) 4

varieert met het aardalkalion is tegengesteld aan wat verwacht

zou worden; de verklaring die Morozova en l'eofilov geven is

echter aanvechtbaar.

L.G. Morozova un P.P. footilov, Opt. Spektrosk. 35,

789 (1973;

Het verschil tussen de kristalstrukturen van SrCr„O, en SrGa 0,

is beter te verklaren met behulp van de voorkeur van Cv voor

oktaedrische coördinatie dan met de grootte van de ionstraal

in Cr

H. Pausch en rik. Mul ier-Ruschbaum, Z. Anorg, Allg. Chem.

405, i (1974;

K.E. Hoefdraad Utrecht, 12 maart 1975