Laser induced fluorescence in Nd3+ : LaCl3 II. — Energy transfer phenomena

HAL Id: jpa-00208907

https://hal.archives-ouvertes.fr/jpa-00208907

Submitted on 1 Jan 1980

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Laser induced fluorescence in Nd3+ : LaCl3 II. - Energy transfer phenomena

N. Pelletier-Allard, R. Pelletier

To cite this version:

N. Pelletier-Allard, R. Pelletier. Laser induced fluorescence in Nd3+ : LaCl3 II. - Energy transfer phenomena. Journal de Physique, 1980, 41 (8), pp.861-867. �10.1051/jphys:01980004108086100�.

�jpa-00208907�

Laser induced fluorescence in Nd3+ : LaCl3

II.

Energy ^transfer phenomena

N. Pelletier-Allard and R. Pelletier

Laboratoire Aimé-Cotton, C.N.R.S. II, ^Bâtiment 505, ⁹¹⁴⁰⁵ Orsay, ^France

(Reçu le 18 décembre 1979, accepté ^{le 13}

1980)

Résumé. 2014 Des techniques ^{utilisant un} laser monomode continu et accordable ont été appliquées ^à l’étude d’ions Nd3+ isotopiquement enrichis dans une matrice de LaCl3.

Des problèmes liés à l’existence de transfert d’énergie ^ont été étudiés : raies satellites, up-conversion, ^diffusion spectrale ^entre ions dans des sites différents.

Abstract.

Techniques involving ^a single ^{mode CW} tunable laser have been applied ^to isotopically ^enriched Nd3+ ions in a LaCl3 ^host.

Some problems ^{related to} energy migration ^{have been} investigated : ^satellite structure, energy up-conversion,

and spectral ^diffusion among ions in dissimilar sites.

Classification

Physics ^Abstracts

71.70G - 78.50

1. Introduction.

In the experiments ^{carried out}

on the hyperfine ^{structure in Nd3 + :} LaCl3, ^some problems ^arose, including ^a particular difficulty ⁱⁿ obtaining ^the FLN Zeeman spectrum ^{of the 5 848} A line, and, ^as pointed ^{out in the} previous paper [1]

(referred ^{to as} I), a puzzling hyperfine 2H11/2 --+ 41 9/2

spectrum ^{which was} sometimes observed when exci- tation was obtained through the cascade

process.

These phenomena ^{seemed to be connected} mostly

to the existence of satellite lines and to energy transfer effects. For a better understanding, ^{we carried out a} study on an even isotope, which has the advantage

of suppressing ^{the line} structures due to nuclear interactions. In order to keep ^{the same} experimental conditions, ^we performed ^all experiments using ^the

same single ^{mode CW} tunable laser. We thus avoid the introduction of additional parameters ^{such as} lifetimes and _power effects. The excited levels were

2Hll/2 ^and 4G5/2.

In this _paper we first show the existence of _energy

up-conversion. ^{Then we describe a} high ^resolution study ^{of the satellite structure of the 5 848} A line

(relative ^{to the} 419/2 +-+ 4G 5/2 ^{17 095 cm-1 1 transi-}

tion). Finally ^we investigate ^the behaviour of the 6 284.8 A fluorescence line (relative ^{to the}

transition) ^excited through ^{a cascade} process, and we

give ^an interpretation ^{of its} hyperfine ^structure

observed on odd isotopes.

2. Expérimental ^methods.

^{The source, the} cryo-

genic system ^{and the} preparation ^{of the} samples

have been described in I, section 2. The even isotope

used was 146Nd, 96% enriched. Observations were

made on fluorescence spectra ^{and on} excitation spectra 2 .1 FLUORESCENCE SPECTRA.

These are obtained when the laser is tuned on an absorption ^{line and}

the spectrometer ^{is scanned over} all the fluorescence

profile. ^The resolution is then limited by ^the spectro-

meter. In our experiments ^the fluorescence is analysed by ^{means of either a} grating spectrometer ^{for low} resolution spectra, ^{or a} pressure scanned Fabry-

Perot interferometer with effective resolution between 0.3 and 1.5 x 106. Calibration is performed by ^use

of a Michelson interferometer.

2.2 EXCITATION SPECTRA.

These are obtained when the spectrometer ^is fixed and the monomode laser frequency is scanned over all the absorption profile. The observation can be made on the total

fluorescence, or on one of the fluorescence lines of the spectrum, whether the excitation of the emitting

level is direct or through ^{a cascade or an} up-conversion

process.

For this technique of monochromatic excitation of the fluorescence, ^the apparatus function is limited

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphys:01980004108086100

862

only by the laser effective linewidth, ^{and the} absorp-

tion spectra ^{thus obtained show the} inhomogeneous profile. ^{In our} experiments ^the scanning ^{is obtained} by passing ^{from one mode of the} cavity ^{to the other.}

An electronic signal ^{is recorded at each} change ^of mode, providing ^a calibration of the spectrum. ^The fluorescence is observed with the spectrometer ^centred

on individual lines. A simultaneous recording ^{of the}

total fluorescence is used as a spectral reference of the laser position ^{at each moment.}

3. Energy up-conversion ^effects.

^{When the} 2 Hll/2 (15907 cm-1) ^and 4GS/2 (17 095 cm -1) ^{levels are}

laser excited, fluorescence transitions of higher energy have been observed, demonstrating the existence of energy up-conversion processes.

Much interest has recently ^{centred on such} pheno-

mena, particularly ⁱⁿ systems involving ^{Pr3+ and} Nd3+ doped ⁱⁿ lanthanum halide crystals. ^In

the production ^of up-converted fluorescence from the 3p 0 ^and 3p 1 levels of Pr3 + has revealed several

up-conversion processes : ^a sequential two-photon

excitation process in which the two sequential ^exci-

tations are coupled by ^internal relaxation, ^{and an}

energy transfer mechanism in which the excitation of two nearby ions annihilate each other resulting ⁱⁿ

one ion in an excited state and the other in a lower state.

In our experiments the laser frequency ^{was fixed}

and spectra ^{were recorded in} the shorter wavelength region up ^to 3 000 A. Fluorescence transitions have been observed from the 4D3/2, 2G9/2, 4 G7/2 ^and 4GS/2

levels. They ^{have been} schematically represented ^on figure ^{1. A} quantitative measurement of their relative intensities has not been carried out, ^as the spectral

response of the various analysers ^{was not known.}

As a qualitative indication of the intensities, ^{it can}

be noted that the fluorescence in the entire visible range is ^seen by ^{the naked} eye when the _{pump power} is about 15 to 30 mW over a few MHz when focussed in the sample by ^{a 0.15 m} focal length ^lens.

Fig. ^1.

Up-converted fluorescence transitions observed when laser absorption transition is 15 907 cm-1

17 095 cm-1. The level diagram is from Dieke [3]. ^For clarity only the lowest Stark levels of the multiplets ^{have been} represented. ^The pendant ^half

circles indicate fluorescing ^levels.

Processes linked to :a, simultaneous excitation of two ions are not considered since, owing ^to nearly

resonant accidental coincidences, ^a sequential ^two- photon process ^can explain the result : the initially

excited ion decays ^{to a} lower metastable level ; ^from

this level, continuously populated by relaxation, ^the absorption ^{of a second} photon places ^{the ion in a} higher ^{excited state. Such} possible processes ^are summarized in table I. Given the energy mismatches of a few cm -1, this mechanism could be linked with satellite structures related to pairs ^of ions, ^{and this}

idea is consistent with the results presented ^{in sec-}

Table 1.

Possible two-photon excitation processes responsible for the observed _energy up-conversion.

Energies ^and labelling of the levels are taken from reference [4].

tion 4.3. However other models are possible. ^The experiments ^{that we have} recently ^{carried out on}

either Pr or Nd or both in various hosts have shown similar phenomena, ^even in the absence of quasi

coincidences. This suggests ^the possibility ^{of a}

second photon ^absorbed by phonon sidebands. In any ^case it is obvious that in the absence of _any

systematic investigation ^{of flux} dependences ^and decay times, ^{it is not} possible ^to attribute the energy up-conversion process ^{to a} definite mechanism.

4. Satellite structure of the 5 848 A line.

In the close vicinity ^{of an} electronic crystal field transition there _may be a great ^{number of other} lines, ^called satellites. These satellite lines have been frequently

observed in rare earths salts and attributed to impu-

rities or defects or to the existence of ion pairs.

The sharp satellites of the electronic transitions of the Nd3 + : LaCl3 absorption spectrum ^{have been}

extensively ^studied [5]. They ^{are found to be asso-}

ciated with the sharpest ^and strongest transitions, and are separated ^{from the} parent ^lines by energies

of 0.1 to 20 cm-1. Unambiguous conclusions have not been drawn. However the behaviour and the structure of the satellite lines are found to be indi- cative of near-neighbour interactions, ^{and are taken}

to be further evidence for both clustering ^of impu- rity ^ions and the distortion of neighbouring ^rare

earth sites.

4.1 FLUORESCENCE SPECTRUM RELATIVE TO THE

4GSJ2 ^{(17 095 cm-’) ,} 4IgJ2 ^{(0 CIn 1) TRANSITION}

(5 ^{848 A} LINE).

^The homogeneous width of the non-lowest Stark levels of the multiplets usually ^are

of the same order of magnitude ^{as the} inhomogeneous

width. As a consequence, pumping ^{into an} absorp-

tion line which corresponds ^{to a} transition to such

a level excites ions belonging ^{to a} large ^{number of}

different sites. Due ^to fast relaxation to a lower metastable level, fluorescence from this level exhibits its inhomogeneous profile. Transitions involved in such a process ^are schematized in figure ^2a.

Using ^this technique ^the 4G 5/2 ^{(17 165 cm -1 ) level}

has been excited (exciting wavelength ^{5 824} A). ^Obser-

vation of the satellite structure of the 5 848 A line has been achieved by recording ^the inhomogeneous

fluorescence profile ^{of the}

transition. The results obtained _agree with the results published by Prinz and Cohen from the

absorption spectrum.

4.2 EXCITATION SPECTRUM RELATIVE TO THE

419/2 -+ 4G5/2 TRANSITION.

The absorption spec- trum of the 5 848 A line has been obtained using ^the technique of monochromatic excitation of the fluo-

rescence. The transitions involved in the _process are

schematized on figure ^2b.

Fig. ^2.

Levels and transitions involved in the experimentally studied spectra.

Observation has been made with the spectrometer centred on various fluorescence lines. The recorded patterns agree with the fluorescence spectrum, ^the intensity of the main line being ^{about two order of} magnitude greater ,than ^the intensity of the satellite lines. One of them, corresponding ^{to the}

Fig. ^3.

Monochromatically ^excited fluorescence of the 4GSj2

(17 095 cm -1 ) level observed

the direct 5 888 A (16 980 cm -1 )

fluorescence line (dashed lines) ^and

^the up-converted ^{5 263} A

(18 993.6 cm-1) fluorescence line (full line).

864

transition (5 ⁸⁸⁸ A line), ^{is shown on} figure ³ (dashed lines) ; ^{the two curves} correspond ^{to the same fluo-}

rescence line recorded with different sensitivities.

4.3 EXCITATION SPECTRUM RELATIVE TO THE

4I9/2

4G5/2 TRANSITION OBSERVED AFTER AN UP- CONVERSION PROCESS.

Using ^{the same} technique

of monochromatic excitation of the fluorescence,

observations have been made on up-converted ^fluo-

rescence lines. Figure 2c schematizes the involved transitions.

The absorption spectra ^{of the 5 848} A line, per- turbed by ^the energy up-conversion process, often do not exhibit identical patterns. Wave numbers of the satellite lines remain unchanged, but their relative intensities with regard ^to the main line intensity vary

according ^to the transition from which fluorescence is observed. In figure ^{3 an} example ^{of such a} spectrum is shown (full line). ^It corresponds ^{to the}

transition (5 ²⁶³ A line). ^{For such a} spectrum, ⁱⁿ which there is a relative quenching ^{of the main line}

with regard ^to the satellite lines, ^some lines close to the main line are then visible. Their energies ^{can be}

added to the table Ic of the _paper by ^{Prinz and Cohen}

.

(table II).

Table II.

Additional satellite lines (in cm-1) of

the 419/2 ^{(0 cm -1 )}

4G 5/2 ^{(17 095 cm -1 ) electronic}

transition.

To date no extensive study ^{has been} systematically

undertaken. Nevertheless the few results obtained suggest ^a connection between the energy up-conver- sion mechanism and the neighbouring ion interac- tions. It is interesting ^{to notice} that, ^{with a} quite

different approach, similar conclusions have been

recently drawn in the case of Pr3 + : LaF3 [6] ^con- cerning the existence of a relationship ^between

energy transfer and satellite lines.

5. Fluorescence spectrum ^{of the 6} 284.8 À line excited through ^a cascade process.

The 6 284.8 A

line is associated with the

transition. Its fluorescence profile ^has been studied when excitation was on the metastable level of a

higher multiplet, ^more precisely ^the 4G5/2 ^level

(17 095 cm -1) (exciting wavelength ^{5 848} À). ^This

particular ^{case is} schematized on figure ^2d.

In the odd isotope experiments, ^{an unusual beha-}

viour of the spectra ^{recorded in this} way ^was observed.

The spectrum obtained when the exciting frequency

was on the low _energy side of the inhomogeneous profile ^of the 5 848 A absorption line has been _pre- sented in 1 section 5.2. It has an unexpectedly ^well

resolved structure. However the phenomenon ^was

not symmetric ^with respect ^{to the centre of the} 5 848 A _line; the various components increasingly overlapped when excitation was displaced ^{to the} high energies. This is the reason why ^we have undertaken

a systematic study

^{on an even} isotope

^{of the}

variation in frequencies ^and linewidths of the 6 284.8 A

fluorescence line as a function of the exciting ^fre-

quency inside the 5 848 A absorption ^{line. The}

results are used to help explain ^odd isotopes spectra.

5. 1 METHODS OF INVESTIGATION. - The fluo-

rescence spectra ^were analysed ^{with a} Fabry-Perot spectrometer. ^{The free} spectral range ^was 0.287 cm -1 and the instrumental width of 0.007 cm-1 was ade- quate ^for the linewidth measurements.

For measuring ^the shifts of the excitation and fluorescence wavelengths, ^a relative calibration was

used. The unknown line and a reference line (2°Ne

is convenient because of its small linewidth) ^were simultaneously ^{incident on a} Fabry-Perot cavity. By recording ^all spectra ^{in the same} order of the Fabry-

Perot a relative displacement of the line under study

is easily measured with a 0.001 cm -’ ¹ _accuracy.

With this method the distance between two modes of the laser cavity ^{was estimated as 0.013 cm-l.}

Measurement of the fluorescence wavelength ^shifts

was made with a slightly ^{different method. The}

fluorescence line was used as a reference, ^{and the} Fabry-Perot ^was continuously pressure scanned over

about ten free spectral ranges, with the exciting wavelength changed after each recording ^{of a fluo-}

rescence spectrum, ^the energy separation ^between

two successive excitation lines being ^{0.013 cm -1.}

Under these conditions, ^{if the} frequency ^{of the}

fluorescence was constant, ^the distance between- two consecutive spectra ^{would be} equal ^{to the free} spec- tral _range D6 of the Fabry-Perot. ^{Thus, the} frequency

shift of the fluorescence is obtained from the diffe-

rence between Au and the real distance of two conse-

cutive spectra.

5.2 RESULTS OBTAINED WITH AN EVEN ISOTOPE.

Using the method just described, ^{a set of fluores-}

cence lines relative to excitation lines regularly spaced ⁱⁿ frequency has been recorded. A 0.009 cm -1 shift of the fluorescence (6 ^284.8 A) ^{line was measured}

for each 0.013 cm-’ displacement ^{of the} exciting

radiation inside the inhomogeneous ^{5 848} A absorp-

tion line, showing ^a proportional dependence ^between

the _energy variation of the 4G 5/2 ^{(17 095} cm-1) ^and 2H, 1/2 (15 907 cm - 1) ^levels according ^to the sites.

As the number of the recorded spectra ^{is too} large

to be presented, ^{we show on} figure ^{4a some charac-}

teristic examples of the fluorescence profiles : ^{0 cor-,} responds ^{to the 6 284.8 A} fluorescence observed when the laser is tuned ^on the centre of the 5 848 À line. On the left and the right ^{we show the fluores-}

cence spectra obtained ^{when the laser} frequency ^is

shifted respectively ^to the red and ^to the blue in a

spectral range of 0.2 cm-’ ¹ on both sides of the line centre.

Fig. 4. - 6 284.8 A (15 ⁹⁰⁷ cm -1 ) fluorescence spectra ^recorded for various positions of the laser _energy inside the inhomogeneous absorption profile ^{of the 5} 848 A ₍₁₇ 095 cm-’) line. From left to

right increasing energies ^{of the} exciting radiation ; ⁰ corresponds

to the laser centred in the 5 848 À profile. (a) Spectra ^obtained

with the

146Nd isotope. (b) Spectra obtained with the odd 145Nd isotope. ^The exciting radiations corresponding ^{to the} spectra have been chosen here

that characteristic profiles

shown; therefore the exciting energies ^intervals

^not regular.

5 . 2 .1 Linewidths.- The spectra ^marked A 1 ^to A3 ^and Ai A2 ^{exhibit a} gradual increase in line- width for excitation going ^{from the} centre to the

wings ^{of the 5 848 A} line. More precisely, if the laser is centred, ^the fluorescence linewidth has the same

order of magnitude ^{as the} instrumental linewidth,

whereas it is of the order of the inhomogeneous

linewidth (- ^0.025 cm -1) if the laser frequency ^{is in}

the wings. ^This gradual destruction of the homo- geneous character of the fluorescence in the cascade process 4 G.5/2 --+ 2Hll/2 ^{has no obvious} interpreta-

tion. It can be accounted for as a gradual ^{increase of}

the spectral diffusion, corresponding ^{to a} gradual

lost of site sélection, when the excitation _goes from the centre to the wings. ^Another point of view is to consider that the random components ^{of the} crystal

field at different sites become more and more uncor-

related when excitation départs ^{from the centre of}

the line. This also accounts for the observed broa-

dening ^when excitation is in the wings. But, ^at this stage ^{of the} study ^{it is not} possible ^{to draw} any definite conclusion.

5.2.2 Satellite lines.

An isolated 6 284.8 A fluo-

rescence line is expected ^{when the} excited ions are

even Nd3+ isotopes. ^However spectra A3 A4 AS

exhibit structure. They ^{are obtained when excitation}

is on the high energy side of the electronic transition, and figure ^{3 shows that, in this} region the satellite structure of the 5 848 A line is closely spaced ^{and the} profiles partially overlap. ^{It is then easily understood}

that the laser line can simultaneously ^{reach the} wing

of the main line and one or more satellite lines. Iso- lated ions and pairs ^of neighbouring ^{ions are then} selectively excited. Under these conditions the _pre-

sence of two or three fluorescence lines account for the existence of levels of different energies ^{for iso-}

lated ions and for pairs, ⁱⁿ other words of a satellite structure of the electronic 6 284.8 A _line, although

it has not been detected by ^{means of a} direct excita- tion.

5 . 3 ^RESULTS OBTAINED WITH ODD ISOTOPES.

5 . 3.1 Experimental spectra.

Most of the results have been obtained for the l4sNd isotope ^{as it was} 90 % enriched. The spectra ^{of the} 143Nd isotope, 75 % enriched, ^{are in} agreement with these results.

Although ^less accurately determined, ^{the same} dependence is found between the _energy variation of the 4G S/2 ^and 2Hl 1/2 ^{levels according to the sites}

that has been observed for even isotopes.

Figure ^{4b shows some} characteristic examples ^of

the recorded fluorescence hyperfine patterns ^corres- ponding ^{to various} positions ^{of the} exciting ^radiation

inside the 5 848 A absorption profile. ^{The additional}

intense line on spectra 0, BI ^and B2 ^{is due to the}

residual even isotopes.

A superficial examination suggests ^a similar beha- viour for the 6 284.8 À _spectrum for odd and even

isotopes. Spectra Bl,2,3 ^exhibit progressively ^unre-

solved structure when the laser frequency is shifted to the blue, consistent with the even isotope ^results.

A careful study ^{of the} spectra ^{0 and} B1, however,

shows that a more complex explanation ^{has to be} found, ^{as :} i) ^{in the} 0-spectrum, ^{relative to an}

excitation centred in the inhomogeneous ^{5 848} A absorption profile, ^the linewidths of the hyperfine

components ^{are much} broader than the instrumental

linewidth; ii) ^{in the} B*1 spectrum, ^{relative to an} excitation shifted to the red, ^the linewidth of the

hyperfine components ^{are narrower} than in the

0-spectrum. This result disagrees ^{with the} symme- trical behaviour of the line broadening ^with respect

to the 0-spectrum ^{observed in the 146Nd fluores-}

cence spectra.

5. 3.2 Interpretation of ^the hyperfine- spectra.

The paradox ^{of a} fluorescence spectrum appearing

to correspond ^{to the} emission of ions belonging ^to only ^one class of sites when the ions of a large ^number

of sites are in fact excited, ^{has been} emphasized ⁱⁿ

paper I. The study of the linewidths presented ⁱⁿ

this second _paper allows us to ascertain that the

profiles of the emission spectra imply ^the participa-

tion of ions belonging ^to several classes of sites, using

866

the assumption ^{that the} fluorescence spectra ^relative

to each class are very slightly shifted with respect ^to each other.

Therefore we have studied the distribution of the selected excited sites, using ^{for this} the distribution of the nuclear states inside the 4I9/2 ⁽⁰ cm-1) ^and

4G _5/2 (17 ⁰⁹⁵ cm -1) electronic levels. They ^are repre- sented in figure ^{5. On this} diagram ^the ground ^level

structure is known from EPR and FLN experiments (see I); ^the excited level is assumed to be a pure

4G5/2 (11 = ^± 1/2) level and the hyperfine ^structure

is calculated using ^the a4f parameter derived from the ground ^level study.

Fig. ^5.

Hyperfine levels of the 4I9/2 ^{(0 CM-1) and} 4GSl2

(17095 cm-1). ^{On the} right ^side

^{the nuclear} wavefunctions.

The scale is relative to the 14’Nd isotope.

FLN experiments have shown that all the hyper-

fine levéls of the ions are not simultaneously ^excited

in all the involved crystalline sites. This _agrees with the inhomogeneous width of the 5 848 A line,

Ay 0.165 cm -1, which is the order of magnitude

of the hyperfine ^{structure of the 17 095 cm-’ tran-}

sition. Therefore we have considered the different conditions of excitation (exciting energy ^on the low

or on the high energy side of the inhomogeneous profile ^{of the 5 848 A} absorption line).

Figure ⁵ shows that both levels involved in the transition are composed ^{of two} groups of hyperfine

levels. They ^are distinguished with letters on the

diagram. An excitation on the low _energy side of the 5 848 A profile is relative to a Y -+ M transition.

The distribution of the nuclear states in these two groups is nearly identical. As afl the AMj transitions

are allowed and the AMI

⁰ selection rule are

respected, the number of the sites involved in the excitation _process is high, ^{but it can be} easily ^con-

ceived that they ^are nearly équivalent, ^{that is to} say

they have very close crystal field parameters. ^An accurate calculation shows _{that any} excited level of the _{group M} characterized by ^a given ^{nuclear wave-}

function and relative to all these sites lies within a

spectral range close to 0.014 cm-1.

Under these conditions each recorded 6 284.8 A pattern ^is really ^the juxtaposition of spectra arising

from the various excited sites. The spectra ^are slightly

shifted relative to each other, ^{but are in other} ways identical. Since the relative energy variation according

to the sites, ^of 2Hll/2 ^{with respect to} 4G5/2 ^{is 9/13,}

the spectral dispersion ^{of the} simultaneously ^recorded fluorescence spectra is less than 0.010 cm-1. Given

an instrumental width of 0.007 cm-’, ^the experi-

mental width 0.017 cm -1 of the components ^{of the} B*1 spectrum perfectly agrees with this interpretation.

The same type ^of argument ^{can be} applied ^when

the exciting radiation is in the blue wing. ^{In this}

case the absorption ^{is relative to a X - N} transition.

The distribution of the nuclear states in these ^two groups is such that the excited sites differ _appre-

ciably. ^{It is} easily calculated that the excited levels lie in a spectral range of 0.03 to 0.04 cm -1. The

spectral dispersion ^{of the} fluorescence 6 284.8 A is then about 0.025 cm-1, ^{that is to} say of the order of magnitude ^{of the} hyperfine splittings. ^{A resolved}

structure is thus not expected, and this result is in agreement ^{with the observed} B2 ^and B3 spectra.

The 0 and B, spectra ^are the intermediate expected

cases.

The study ^of energy transfer _processes carried out

on an even isotope ^{has thus} provided ^an interpre-

tation of the hyperfine spectra ^{of odd} isotopes. ^The

very simple pattern obtained which seemed so

puzzling ^at first has been satisfactorily interpreted,

but it has been shown that this simplicity ^{was, not}

so much a general property ^of spectra ^excited through

a cascade _process, but rather the result of a favourable combination of particular circumstances.

6. Conclusion.

Hyperfine ^structure studies of Nd3 + ions in a LaCl3 host had shown evidence for the existence of some phenomena attributed to energy transfer processes. The study of these processes, carried out on an even isotope, ^and working ^{in a}

CW regime, ^is presented ^here.

The results of this study ^have provided ^{an inter-} pretation ^{of the} hyperfine ^{structure of odd} isotopes.

Moreover, interesting results have been obtained on

i) energy up-conversion ;

ii) the destruction of the homogeneous ^character

of the fluorescence in a cascade _process which is

more or less important according ^to the localization

of the excitation in the wings ^{or at the centre of the}

inhomogeneous profile ^{of the} absorption ^line.

These experiments have therefore opened up inte-

resting prospects, ^and Nd3 + : LaCl3 appears ^{as a} suitable material for _energy transfer investigations.

An extension of the study ^{is now} being considered, using high resolution time resolved techniques.

References

[1] PELLETIER-ALLARD, N., PELLETIER, R., J. Physique ⁴¹ (1980)

855.

[2] ZALUCHA, D. J., SELL, ^J. A., FONG, ^F. K., J. Chem. Phys. ⁶⁰ (1974) 1660.

[3] DIEKE, ^G. H., unpublished, quoted by CROSSWHITE, ^H. M.

and Moos, H. W., Optical properties of ^{ions in} crystals (Interscience Publishers) p. 3.

[4] CROSSWHITE, ^H. M., CROSSWHITE, H., KASETA, F. W., SARUP, R.,

J. Chem. Phys. ⁶⁴ (1976) ^1981.

[5] PRINZ, G. A., COHEN, E., Phys. ^{Rev. 165} (1968) ^335.

[6] VIAL, ^J. C., BUISSON, R., MADEORE, F., POIRIER, M., J. Physique

40 (1979) 913.