HAL Id: jpa-00247534
https://hal.archives-ouvertes.fr/jpa-00247534
Submitted on 1 Jan 1991
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.
Calculation of Kr photoabsorption spectrum fine structure within the KN23 ionization threshold region
V. Sukhorukov, A. Hopersky, I. Petrov
To cite this version:
V. Sukhorukov, A. Hopersky, I. Petrov. Calculation of Kr photoabsorption spectrum fine structure
within the KN23 ionization threshold region. Journal de Physique II, EDP Sciences, 1991, 1 (5),
pp.501-509. �10.1051/jp2:1991184�. �jpa-00247534�
Classification
Physics
Abstracts 32 80FCalculation of Kr photoabsorption spectrum fine structure within the KNz~ ionization.threshold region
V L
Sukhorukov,
A. NHopersky
and I D. PetrovRostov Railway Engineers Institute, Chair of
Physics,
344017 Rostov-on-Don, USSR(Received5July
1990, revised 29January 1991,accepted
31January1991)
Absbact. The near edge
KN~~-fine
structure of Kr atom is calculated via theconfiguration
interaction method
including
a monopole rearrangement of electron shells The theoreticalspectn~m form is in good agreement with the expenmental one The calculated absolute values of the oscillator
strengths
and thephotoabsorption
cross section arepredictions
1. Introduction.
Inner shell
photoabsorption spectra
of atoms, molecules and solids near ionization threshold havecomplex
fine structures XANES(X-ray Absorption
NearEdge Structure).
XANES may be caused bothby
theshape
of thepotential
of environment[1, 2]
andby many-body
effects[3-6]. Therefore,
studies of the nature ofmany-body
effects and their effect onphotoabsorp-
tion spectra are of
special importance
It is fruitful toperform
these studies on atoms where effects of environment on XANES are excludedExpenmental
spectra of Ne and Arphotoabsorption
within the KX-threshold regions(X
is the additionalvacancy)
have been obtained m reference[3] ~Ne.
X=
L)
and reference[4]
(Ar
X=
fil~.
Asemi-empirical
assignment of spectra has been made in these papers. Amore accurate
assignment
ofKX-photoabsorption
spectra has beenpresented
m reference[5],
where it has been shown that
photo-double excitation/ionization
processesplay
asignificant
role in
forrnJng
the fine structure features. Cross sections of these processes may be calculated with agood degree
of accuracy within the approximation ofmonopole
rearrangement ofelectron shells
(MRES)
which accounts for the deformation of outer electron shells because of appearance of inner vacancy.The
expenmental study
of multi-electron transitions near the K-threshold of Kr has shown the existence of acomplex
fine structure which is causedby photo-double excitation/ionization
processes
[6].
The last conclusion of reference[6]
is based onsemi-empincal analysis
ofoptical
data and on a comparison ofKX-photoabsorption
spectra ofNe, Ar,
Kr.The main aim of this work is to give a more accurate assignment and to
predict
the absolute values of KrKN23-Photoabsorption spectrum
features on thegrounds
of anon-empirical
calculation. One addluonal aim is to
give
acomparative analysis
ofKX-photoabsorption spectra
m a seriesNe, Ar,
KrJOURNAL DE PHYSIQUE II T V MA> 1991 25
502 JOURNAL DE
PHYSIQUE
II M 52. Method and the results of calculation.
A method of calculation of
photo-double excitation/ionization
processes has beendeveloped
and described in detail m reference[5], therefore,
we present hereonly
the basic formulae with bnef conunents.Basic atomic orbitals
(AO)
It has been shown in reference[5]
that whencalculating
KX-ionizationlexcitation
cross sections onehas,
in the firstplace,
to take MRES into account, and to achieve tbJs it suffices to usenon-orthogonal
AO obtainedseparately
for the1nltial and the final states Thus AO radial parts of the imual state were obtainedby solving
Hartree-Fock(RF) equations
of Krground
state The final state AO were obtainedby solving
HFequations averaged
over the excitedconfiguration.
To obtain AO of discrete excited states, tile HF equations for theconfiguration
ls~4p~ nin'i'
were solved. To obtain AO of continuum(ep
a frozen coreapproxbnation
of theconfigurations
ls~4p~ 5p(6p)
for ls4p-5p(6p)
epprocess and of the
configuration ls~~ 4p~~
forls4p-epe'p
process had beenexploited
Within the
calculation,
the coreAO,
discrete and continuous excited stateenergies
and theenergies
ofls4p-nin'i'
excitations were obtained as the differences of totalenergies
ofcorresponding configurations.
Thus Is4p-spot excitationlionlzation
energy at the threshold(e =0)
is 14344.7eV,
that of Is4p-6pei
is 143482eV and the threshold energy Is4p-eie'i
is 14 351.9 eV. A relativistic correction to the energy of Is level of reference
[7]
was taken into account, it is 224.0 eV.
To calculate the
KN~~ photoabsorption
spectrum of Kr thefollowing
excitation had been considered~~~~ ~5p~~ ~4d~~ ~~~~~~ ~5s~~
~~~The choice of the excitations
(I)
is based on the expenencegained
in the calculation of KX-spectra
ofNe,
Ar[5]
and on using the smallness parametercharacterinng
themixing
of the wavefunctionQVF)
ofcorresponding configuration
with the WF of ls~4p~ 5p~ configura- tion,
because the transitions from theground
state into the states of tinsconfiguration
have thegreatest
oscillatorstrengths.
The smallness paranJeter was determined with thehelp
of theexpression
(5p~ V[ nin'i')
~~~~'~'
AE(5p~ nin' I'
'where
(5p~ V( nin'i')
is the mean interaction between theconfigurations ls~~ 4p~~ 5p~
and ls~
4p~ nin' I'
whichwas estimated as
(f~i f~,1, )~ (f~i being
the mean radius of thecorrespondJng AO),
andAE(5p~ nin'f')
is the difference of energies ofcorresponding
excitations. Thus the calculation of
K~2/K~ii
hadgiven
I
=
) ~~~~~ ~~
=
~~ ~'~~
~'~'~ ~'~
~~= l10
(2)
K4F
f~ AE(5p
4d(4.65
a-u0,37
~VExpression (2)
allows one to assume that the effect of is4p-4P
exaltation on the spectrum istwo orders less than that of
ls4p-4d2
excitation and not include it in calculation. Ananalogous
selection had been made for all the excitations included in the calculation.Together
with the excitations(I)
thefollowing
channels of continuous spectrum were takeninto account in the calculation: Is
4p-(spep,6pep, epe'p).
The channels ofdiffenng
symmetnes
were not included because the transitionamplitudes
for them are small ascompared
with theamplitudes
of the transiuons into the above mentioned channels.Energies
andwavefunctions of
the excited states To construct the full WF of the atom in the final state, basic vectors ofLS-coupling
were used(spin-orbital
interacuon for4p, ni
andn'i'
electronsbeing weak)
a
LS)
= ls~4p~ (Lo So) nfn' f' (L£) LS)
Here the
ls-4p
Coulomb interaction had beenneglected
because thesplitting
of ~P and~P terms of the
configuration
ls~4p~
is 0.90 eV which is less than the natural width of Kr Is level amounting to 2.69eV[8].
Aftercalculating
via the method[9]
the matrix elements(a
' LS V aLS)
of theoperator
V=
H~
V~~(K),
where H~ is a Coulomb operator and
V~~(K)
is the average over theconfiguration
KTable I
Spectral features of
KrKN~~-XANES
m the regionof
double Is 4 p-n jii n~f~
photoexcitation (see Fig. I).
x x io4, Mb «iP) (2)
2 41 0 11 0 30 035p21'D) + 0 14d2('D) +
+ 5s 4d[0 9('D) + 0 2(3D)]
2 0 16 0 52 38 5p2[0 2('D) 0 5(iS) 0 5(3P)] + 4d2[061'S) 0 21'D)]
+ 5s 5d[0 3('D) 0 1(3D)]
3 0 00 2 09 5 51 5p2[0 1('S) 0 3('D) 08(3P)] 0 25s 5d('D) +
+ 4d~10 5('D) o I('s)j
4 0 78 0 63 66 5p2[0 2('S) + 0 7('D)] 054d2(3P)
0.14d 5d(~D) 0 35s
4d('D)
5 81 0 46 22 5p2[0 2('D) 0 2('S) + 0 1(3P)] + 4d2[0 3('D) + 04(3P)]
+ 5p 6p[0 1('S) + 0 2('P) 02(35) + 03(3P) + 0 5(3D)]
4d 5d[0 1('D) + 02(3P) + 0 3(3D)]
6 2.31 0 18 0 46 0 25p2('S) 4d 5d[0 4('S) + 0 4(D) 06(3P) 0 3(3D)]
+ o24d2j( is) + (3P)j 5p 6pjo 2('P) + o i('D) o4(3P)j
7 2 46 0 78 2 05 0.25p2~'S) 5p 6p[0 1(lP) 0 7('D) + 02(35) 04~3P) + 0 4(3D)]
4d2[0 2('S) 0 1('D)] + 0 24d 5d[('S) ('D)]
8 2 72 0 66 75 045p2('S) + 0 15s2(lS) + 4d2[04(lS) + 02(3P)] +
+ 5p 6p[04('S) + 0 3(lD) + 0 1(3P) 02(3D)] +
+ 4d 5d[- 0 2('S) + 0 2('P) + 0 5(lD) 02(3P)]
N is the number of component in figure I
(I) Zero of the photoexcitation energy corresponds to the pos1tlon of the line N
= 3, w
=
14 342 3 eV (2) Final state of photoexcltation (only the components with (a~ i ~~t~j~s~ m 01 are listed)
~~~~~
~j
~l ii n2t2(LS) ~~ ~~ ~~~~~~~)fll ~l~2~2(£S),
~P)~l 1n2 2 LS
504 JOURNAL DE
PHYSIQUE
II M 5potential,
andsolving
secular equation, the energies(li~
and WF of the final states wereobtained
(ELS)
=
£
a~~aLS) (3)
The
energies
and wavefunctions of the final states for the transitionshaving
thegreatest
oscillatorstrengths
are listed m table1.Cross sections
of excitation/ionization
processes The oscillatorstrength
of the transition from theground
state into the state of(3)
has been calculated via the formula :fi~~~(w )
= ~ w£ a~~(aLS[[D
[0)
,
(4)
3
~
"
~
where the
photon
energy w=
E
Eo
is in a-u-,Eo
is the HF energy of theground
state, andthe
amplitudes
A=
(aLS[[ Q
[0)
are determinedaccording
to[5] by
the expressions.~4~~))~~)$i~
"(~l
~~~~ ~ ~2 ~~~~~(LS)
n'12 ~~2(l~' ~), ~l'
(~/
(°, ~~)
""
(~
i)~
Nimax
~~ ~ ~~ l') ~~
~ ~ ~~ ~~~~) ~'~~) (D12
+
(~
i)~'~~ C12) (5)
In
(5) 0, ~5j
is theground
state WF,
p =
ii
+i~
+S,
N is aproduct
ofoverlap integrals
of the AO not involved in the transition,i~~~
= max(ij, i~)
;1~12 "
(n~2~4~
~l~l) (n'~2~~2~2),
ci~
=(n~ i~l1411
niii) (n~ i~Ini~),
the first order expression for the matnx element of the operator is
(F
is a Fermllevel) [10]
(n~i~[[d[[
niill
"
(n2 f2(~(
~iii) i ~~~~~~~~( )i~~~
~~ ~~~ ~~~nj<n<F ~~ 2 ~ ~~
(ni in' I')
=
)~ P~i (r) P~,
I,(r)
dr,
(nf jr
n'I' ) oo
=
P~i (r) P~,
I,(r)
r dro
If one of the
electrons,
i-e np orn'p,
is in a continuum, one mayneglect
the electrostaticsplitting
of the core and sum upsquared amplitudes (5)
In tins case the expression of theexcitation/ionization
cross section is~npep(~
4
~ 2
~
~2
~j~ npep
(~)
IS4p
j
0 0 is4p ,l~~())~~
ii ~~ ~max
(~
~2 +1) (D12
+C12) ~1~12 f~12)
,
(~)
where
ao15
a fine structure constant, ao is Bohr's radius If both electrons go to continuunl, double ionJzation cross sections are calculated with the expressionw -JPj~~~
"ls4p(W )
""~i~~(W ) dE, (9)
0
Table II.
ls-ep
and Is4p- (~) pe'p photoabsorption
cross sections in Kr m the regionof
e
KN~~-threshold (see Fig. I).
«(f((
x 10~, Mbw
('),
eV«[f
x 10~(2),
Mbn = 5 n
=
6 n
=
e'
14 338.0 154 12 0 50 0 12 0.00
14 342.3 145 44 2 88 0 25 0.00
14 347.3 137 24 16 50 50 0.00
14 354.0 129 55 19 25 4 38 75
14 358.0 125.05 18 75 3 75 4.25
(1)
Kr K-level relativistic correction Am~~j =
224.0 eV
[7]
is taken into account.(2)
Results of references[11, 12].
ax ~o~/tl6
~
sis4j ntej
6
,----,
,'
--,-,_
/" ~'sly-sjtj
/ , i i
<4.a5 k~liev i i ,
I i I
/~I
Il~
I
j
j
,'~
,'
,'
/
,~,' ,,
"
'~_~ ,,is4j -iii)
fi.55
~ J) )jL(~
i 3 ~ 5
Fig
I K-XANES of Kr atom(-)
this work's theory,
(-)
doublephotoexcitation
spectrum (- -) excitation/ionization and doublephotoionization
channels Insert expenmental curve of reference ,[6] o~absorbed
photon
energy Lorentzianamplitudes
are calculated via(10)
withrj~
= 2 69 eV [8]
The notation of the final states labelled here with numbers is given in table1.
506 JOURNAL DE
PHYSIQUE
II M 5where the integration is over the surface
e + e'
= w IP
i~~~, e and e' are the
energies
of the electrons in continuum andIPj~~~
is ls4p
ionizationpotential
of Kr.Results
of
the calculation and comparison with the experiment Oscillatorstrengths
of the transitions from theground
state into the states of(3)
were calculated using(4)
and arepresented
in table I. We also present theamplitudes
of Lorentzian curves calculated via theexpression
1(W )
=
4 7r«o
aifis4p(W )/ris, (lo)
where
ri~
is the Kr ls level width. This was done to make the theoretical spectrum convenientfor companng with
expenment.
Each component of thespectrum
ispresented
with theLorentzian curve, and the curves are summed up. The cross sections of the transitions Is
4p-
npep and Is
4p-epe'p
were calculated via the formulae(5-9), they
are given in table II. The theoreticalsingle
ionization cross sections for Kr[I1, 12]
are alsopresented
in the sanJe table to make itpossible
to compare with theexpenment
the sum ofsingle
and doubleionizationlexcitation
cross sections.In
figure
I the theoreticalKrKt§~
spectrum obtained in the way descnbed above is shown.The measured spectrum of reference
[6]
is alsogiven.
One may see agood qualitative
agreement of the spectra. One may,therefore,
state that the assignment of the spectrum which is obvious(see
Tab. I andFig I)
is reliableenough Generally,
ourassignment
is thesame as the
qualitative
one of reference[6],
but one should note on thebnghtest peak
A(the peak
at w=
14 343 eV in
Fig. I)
that its ongin is connected notonly
with the excitation of electrons into discrete states but also withphotoiomzation
of atom.3.
Comparison
of calculated KXphotoabsorpfion
spectra ofKr,
Ar and Ne.The calculation of the
photoabsorption spectra
in theregions
of KX-ionization for Ne(X= L),
Ar(x= M)
and Kr(X= N)
had made itpossible
to understand the main mechanismsforming spectral
structuresSpectral
intensities The main mechanismdetermining integral
intensities of spectra is amonopole
rearrangement of the electron shells at the appearance of ls- and X-vacancies. To charactenze the relative value of excitation and ionJzation processes we list in table III several values :f(A)
=zfisn~p(W (i1)
Table III
Changes
m relative valueof
excitation and ionization processes m the rowA
=Ne, Ar,
Kr.A Ne Ar Kr
f(A
x 10~(1)
2 37 0 44 0.06« j
(A
(2) 2 75 0 53 0 07x(A )
x10~~
(3) 0.29 0.30 0 31(ii, (2), (3)
See the formulae(11)-(13)
of this work.Ii
"I(A)
"
z "I$§ d~, (12)
n
o x
(A )
=
«i(A )/(2 7r~
«oaif(A)) (13)
where the oscillator
strengths
and ionization cross sections are determinedby (4)
and(7)
;A =
Ne, Ar, Kr,
nj=
2, 3, 4, Integration
in(12)
is made in a-u One may see from table IIIthat, although f
and «j values decrease with thegrowth
of the atomicnumber,
their ratiocharactenzing
the relative value of the exaltation and ionization cross sectionchanges weakly Among
thepecuhanties
of the calculation one should note theimportance
of the inclusion of additional terms in(6)
and the necessity to demand the WF ofhigher lying (in energy)
configurations
to beorthogonal
to the lowerlying
WF of the same symmetryThus, calculating
ls4p-5p2
exaltation oscillatorstrengths,
the inclusion of additional terms led toincrease the value of
f by
479b,
and inclusion oforthogonahty
condition for the WF ofconfigurations
ls~4p~ 5p2
and ls~5p
led to a 15 ifi decrease of the valueoff-
Inall, taking
additional terms in
(6)
andorthogonahty
conditions into account increased the cross sectionsby
30 9b This value is somewhat less than those forf/f~~(Ne )-72
ifi andf(f~~(Ar )-43
ifi. This allows us to conclude that the role of additional terms andorthogonahty
conditions m the calculation of excitation and ionizationprobabilities
decreases withgrowth
of atomicnumber,
and this conclusion may be useful in calculations ofphoto-double excitation/ionization
processes in multi-atomic systems where inclusion of the effects under discussion is difficult
Table IV.
Changes
m thedkgree of effect ~((l'~ of
the excitationY=np~-md~,
np~-mj
sm~ d on K-XANES m the row A=
Ne, Ar,
KrA Ne Ar Kr
n, m j 3 4 5
m,m~ 3 3 4
E~(md~)
(~),
eV 6.50 1.17 0 37E~(mi
sm~d),
eV 0 75 2.20 2 20R[(md~),
eV 2 51 2.44 2 16R((md~),
eV 46 1.57 37R/(m
sm~ d),
eV 2 89 2.32 1.23~/(md~) (2)
0.39 2.09 5.84~
((m
sm~ d)
3.85 1.05 0.56(~)
E~( Y)
"
E~~(np~j E~~( Y).
(2) See the formula
(14)
of this work508 JOURNAL DE
PHYSIQUE
II M 5Structure
ofspectra.
The structures of KXphotoabsorption
spectra are dominatedmainly by multiplet splitting
of theconfigurations
ls" njp~ npn'p
andby
theirpolanzation by
npn'
p-mdm~
d andnpn'
p-mj sm~ d excitations. Coulomb
ls-nj
p interaction issignificant only
mNe,
the role ofnp-n'p
Coulomb interaction isapproximately equal (the splitting
ofl-2
eV),
and the role ofnpn'
p-mj
dm~
d andnpn'
p-mj sm~ d is shown m table IV. Listed m this table are Slater
integrals R((Y) charactenzing
the value ofmixing
for the excitations Y=
np~-md~, np~-mj
sm~ d(n
=
3, 4,
5 forNe, Ar, Kr),
and the energies of these excitationsE~(
Y~(that
is the differences between centres of gravities ofcorresponding configurations).
The values
~i(Y)
=
iRj( Y~/En(Y) (14)
give the
degree
of Y excitation effect on spectrum. It can be seen from table IV that the effecton spectra of
np~-md~
excitations grows because ofdecreasing
energy denonfInator and that ofnp~-mj sm~d sharply
decrease because of energy denominator's increase anddecreasing integrals R((npnp,
mi sm~
d).
4. Conclusion.
This
study
allowed us to give anassignment
of Kr KXphotoabsorption spectrum
based on anon-empirical
calculation. Ourassignment
isbasically
the same as that of reference[6]
It was shown that the main fine structure features are determinedby mu1tlplet splitting
of thediscrete excitations and
by
opening of channels ofsingle
and double ionization. Theseprocesses' probabilities
may be described within the MRESapproximation and,
to calculate the fine structure, one has to considermultiplet splitting
of theconfiguration
ls~4p~ npn'
pand its
polanzation by npn'p-m
j
dm~d
excitationsA companson of the results of KX spectra calculations m
Ne, Ar,
Kr has shown that MRESis the main mechanism
allowing
one to descnbephoto-double excitation/ionization probabili-
ties near the KX
threshold,
but relativesignificance
of the main fine structureforming
mechanisms vary somewhat m the row
Ne, Ar,
Kr Thlsstudy
may be useful in theanalysis
of K-XANES m 10 « Z « 36 elements which arebeing widely enough
used ininvestigation
of matter's electronic structureAcknowledgments.
The authors are
grateful
to Dr. A G Kochur forhelpful
discussions on some of the abovereported
results.References
[II
SUKHORUKOV V L, DEMEKHINA L A, YAVNA V A, DEMEKHtN V F, Phys Tver Tela USSR21
(1979)
2976[2j VEDRINSKI R V, BUGAEV L A, GEGUSIN I. I, KRAISMAN V L., NOVAKOVICH A A,
PROSANDEEV S A., RUUS R E., MAISTE A A., Solid State Comm 44
(1982)
1401[3] ESTEVA J. M, GAUTHE B, DHEz P., KARNATAK R C, J
Phys
lr 16(1983)
L-263[4] DESLATTES R D, LA VILLA R E, COWAN P. L, HENINS A, Phys Rev. A 27 (1983) 923
(5j
SUKHORUKOV V L, HOPERSKY AN,
PETROV I. D, YAVNA V. A, DEMEKHIN V F., JPhys.
France 48
(1987)
1677[6] BERNIERI E, BURATTINI E.,
Phys
Rev A 35(1987)
3322.[7] KUCHAS S A., KAROSENE A V., KARAzIYA R I., Liihuanian Phys.
Papers
IX(1978)
593[8] KESKI-RAHKONEN O, KRAUSE M. O, Atomic Data Nucl Data Tabl 14
(1974)
139[9] JUCYS A P
,
SAVUKINAS A J
,
Mathematical Foundat1onS of the