!
!
Ene gy'domain-synch o on- adia ion-
Mössbaue -sou ce- o -physics-unde -
ex eme-condi ions-
!
!
DISSERTATION!
!
!
zu !E langung!des!akademischen!G ades!eines!
Dok o s!de !Na u wissenscha en!(D .! e .!na .)!
an!de !Bay eu he !G aduie enschule! ü !Ma hema ik!und!
Na u wissenscha en!(BayNAT)!de !Uni e si ä !Bay eu h!
!
!
o geleg ! on!
!
!
Vasily-Po apkin!
aus!Eka e inbu g!(Russland)!
!
!
!
!
!
Bay eu h,!2012!
!
!
Die o liegende A bei wu de in de Zei on Mä z 2009 bis Juli 2012 am
Baye isches Geoins i u de Uni e si ä Bay eu h un e de Lei ung on He n P o .
D . Leonid Dub o insky ange e ig .
Volls ändige Abd uck de on de Fakul ä ü Chemie/Biologie/Geowissenscha en
de Uni e si ä Bay eu h genehmig en Disse a ion zu E langung des G ades eines
Dok o s de Na u wissenscha en (D . e . na .).
Tag de Ein eichung: 20 June 2012
Tag de wissenscha lichen Kolloquiums: 22 Oc obe 2012
Am ie ende Dekanin: P o . D . Bea e Lohne
P ü ungsausschuß:
P o . D . Tomoo Ka su a, Uni e si ä Bay eu h (Vo si zende )
P o . D . Leonid Dub o insky, Uni e si ä Bay eu h (E s gu ach e )
P o . D . Alexei Bosak, ESRF (Zwei gu ach e )
P o . D . S. Kümmel, Uni e si ä Bay eu h
P o . D . J. Senke , Uni e si ä Bay eu h
Table!o !Con en s!
!
!
3!
Table&o &Con en s&
ZUSAMMENFASSUNG- 5-
SUMMARY- 9-
1.-INTRODUCTION-12-
1.1!ORIGIN!AND!THE!STRUCTURE!OF!THE!EARTH!!12!
1.2!MINERALOGICAL!MODEL!OF!THE!MANTLE.!(MG,FE)(SI,AL)O3!PEROVSKITE!AS!DOMINANT!
COMPONENT!OF!THE!EARTH!LOWER!MANTLE!!16!
-
2.-MOTIVATION-22-
2.1!SPIN!STATE!OF!IRON!IN!THE!LOWER!MANTLE!SILICATE!PEROVSKITE!22!
2.2!NEED!FOR!ENERGYYDOMAIN!SYNCHROTRON!MÖSSBAUER!SPECTROSCOPY!23!
!
3.-METHODS-AND-INSTRUMENTATION-25-
3.1!CREATION!OF!HIGH!PRESSURE!25!
3.2!MÖSSBAUER!SPECTROSCOPY!27!
!
4.-SYNOPSIS-(SCOPE-OF-THE-THESIS)-32-
4.1!THEORETICAL!CALCULATION!AND!EXPERIMENTAL!STUDY!OF!ANGULAR!DEPENDENCES!AND!
ENERGY,!TEMPORAL!DISTRIBUTIONS!OF!SYNCHROTRONYBASED!57FE!MÖSSBAUER!RADIATION!32!
4.2!THE!57FE!SYNCHROTRON!MÖSSBAUER!SOURCE!37!
4.3!SPIN!STATE!OF!FE3+!IN!THE!LOWER!MANTLE!39!
4.4!DOUBLE!SIDED!LASER!SYSTEM!FOR!SMS!AND!XYRAY!DIFFRACTION!42!
4.5!LIST!OF!MANUSCRIPTS!AND!STATEMENT!OF!AUTHOR’S!CONTRIBUTION!43!
MANUSCRIPTS-46-
5.1-MULTISPACE-QUANTUM-INTERFERENCE-IN-57FE-SYNCHROTRON-MÖSSBAUER-SOURCE--46-
!
5.1.1!ABSTRACT!!46!
5.1.2!INTRODUCTION!!47!
5.1.3!MULTISPACE!INTERFERENCE!OF!γ!RAY!PHOTON!IN!THE!UNIT!CELL!OF!57FEBO3!CRYSTAL.!!50!
5.1.4!NUCLEAR!EXCITON!POLARITON!UNDER!DIFFRACTION!CONDITIONS!!60!
5.1.5!ANGULAR,!ENERGY!AND!TIME!DISTRIBUTIONS!OF!FE!SM!RADIATION!!65!
5.1.6!SUMMARY!AND!CONCLUSION!!73!
!
5.2-ANGULAR-DEPENDENCIES-OF-SPECTRAL-AND-TEMPORAL-DISTRIBUTIONS-OF-NUCLEAR-
RESONANCE-RADIATION-OF-SYNCHROTRON'BASED-57FE-MÖSSBAUER-SOURCE--76-
-
5.2.1!ABSTRACT!!76!
5.2.2!INTRODUCTION!!77!
5.2.3.!EXPERIMENTAL!METHODS!!79!
5.2.4.!RESULTS!AND!DISCUSSION!!81!
5.2.5!SUMMARY!!85!
Table!o !Con en s!
!
!
4!
!
5.3-THE-57FE-SYNCHROTRON-MÖSSBAUER-SOURCE-AT-THE-ESRF--87!
!
5.3.1!ABSTRACT-!87!
5.3.2!INTRODUCTION!!88!
5.3.3!SYNCHROTRON!MÖSSBAUER!SOURCE-!91!
5.3.4!OPTICAL!SCHEME-!102!
5.3.5!PROPERTIES!OF!THE!SMS!RADIATION-!106!
5.3.6!APPLICATIONS-!108!
5.3.7!CONCLUSIONS!!111!
5.4-NO-SPIN-TRANSITION-OF-FERRIC-IRON-IN-THE-LOWER-MANTLE-112-
!
5.4.1!ABSTRACT-!112!
5.4.2!INTRODUCTION!!112!
5.4.3.!EXPERIMENTAL!METHODS-!113!
5.4.4.!RESULTS!AND!DISCUSSION!!114!
5.4.5.!SUPPLEMENTARY!INFORMATION!!119!
!
5.5-PORTABLE-DOUBLE'SIDED-LASER'HEATING-SYSTEM-FOR-ENERGY'DOMAIN-MÖSSBAUER-
SPECTROSCOPY-AT-SYNCHROTRON-AND-SINGLE-CRYSTAL-DIFFRACTION-EXPERIMENTS-WITH-
DIAMOND-ANVIL-CELLS--128-
!
5.5.1!ABSTRACT-!128!
5.5.2!INTRODUCTION!!128!
5.5.3.!DESIGN!OF!THE!LASERYHEATING!SYSTEM!!130!
5.5.4.!EXAMPLES!OF!APPLICATION!OF!THE!PORTABLE!LASERYHEATING!SYSTEM!!135!
5.5.5.!CONCLUSIONS!!!139!
!
5.6-FULL-LIST-OF-PUBLICATIONS-(WITH-PAPERS-NOT-INCLUDED-INTO-THE-THESIS-)--141-
!
BIBLIOGRPHY-142!
!
ACKNOWELEGMENTS-156-
ERKLÄRUNG-157
Zusammen assung!
!
!
5!
Zusammen assung
Eisen is eines de häu igs en Elemen e de E de und eine wich ige Komponen e
in Mine alien. Elek ische und magne ische Eigenscha en eisenhal ige Ma e ialien
beein lussen signi ikan P ozesse in de Tie e de E de. In den Ma e ialien, die den
Un e en Man el ausmachen, kann Eisen sowohl in e schiedenen Valenz- und
Spinzus änden als auch in e schiedenen k is allog aphischen Umgebungen
o kommen. Die meis en expe imen ellen Techniken ges a en es nich , die
En wicklung de Eigenscha en e schiedene Eisenplä ze indi iduell zu e olgen
ode sie sind nich geeigne ü Un e suchungen un e gleichzei igen Hochd uck- und
Hoch empe a u bedingungen. Un e diesen Bedingungen we den Un e suchungen zu
elek onischen S uk u on Eisen eine g oße He aus o de ung.
Die o liegende Dok o a bei is in zwei Haup abschni e gegliede . De e s e
Teil beschä ig sich mi de En wicklung eine Synch o on Mössbaue Quelle
(Synch o on Mössbaue Sou ce, SMS). Sie e laub eine ene gieabhängige
Mössbaue spek oskopie on P oben un e D ücken jensei s on 100 GPa in
lase geheiz en Diaman hochd uckzellen. De zwei e Teil de A bei is dem Ve hal en
on Eisen in eisen-/aluminiumhal igen Silika -Pe owski en un e Bedingungen des
Un e en Man els gewidme .
1. Synch o on Mössbaue Quelle
Es exis ie en meh e e Techniken, um elek ische und magne ische
Eigenscha en on Ma e ialien un e ex emen Bedingungen zu un e suchen:
Rön genemissionsspek oskopie (XES), Nahkan en-Rön genabso p ionsspek oskopie
(XANES), Ke n esonazs euung, usw. Die ausge ei es e, emp indlichs e und
geeigne es e Technik ü solche S udien is jedoch die (ene giedi e enzielle)
Mössbaue spek oskopie.
Die nied ige B illanz de dazu benö ig en adioak i en Quellen und die nied ige
An eiche ung des Eisens in den Mine alien des Un e en Man els e o de jedoch im
Falle de ene gieau gelös en Mössbaue spek oskopie lange Messzei en und limi ie
die Anwendung au mode a e D ücke. Diese Ums and wü de du ch die Kombina ion
de ause gewöhnlichen Eigenscha en de Synch o ons ahlung (hohe B illanz,
Möglichkei en zu ex emen Fokussie ung) mi denen de klassischen
Mössbaue spek oskopie gelös we den. In ku z: Es wi d eine Synch o onquelle mi
Zusammen assung!
!
!
6!
Mössbaue s ahlung benö ig . Die Kons uk ion eine solchen Quelle wa die
Haup au gabe meine P omo ionsa bei . Smi no e al. (1997) ha die Möglichkei
eine solchen Quelle an de Nuclea Resonance Beamline (ID18) de Eu opean
Synch o on Facili y (ESRF) demons ie . Die Quelle basie au einen Ke n e lexen
im an i e omagne ischen K is all FeBO3.
Die echnischen Haup au gaben meine P omo ionsa bei wa en (a) die
Kons uk ion eine pe manen e ügba en Quelle mi (b) de bes möglichen
Ene gieau lösung bei gleichzei ige höchs mögliche In ensi ä . Um diese Ziele zu
e eichen, wu den zunächs heo e ische Rechnungen du chge üh , um die bes e
S a egie zu Op imie ung de Quelle zu inden. Meh e e Expe imen e olg en, um die
heo e ischen E gebnisse zu e i izie en. Zudem wu den meh e e op ische
Ano dnungen de SMS ge es e , um die op imale Ano dnung zu bes immen.
Das E gebnis des wissenscha lichen En wicklungsp og amms is eine
wi kungs olle Synch o on Mössbaue Quelle ü Hochd uckanwendungen an de
Nuclea Resonance Beamline (ID18) de ESRF. Theo e ische und expe imen elle
E gebnisse zu Winkelabhängigkei de Ene gie- und Zei e eilung de einen
Ke n e lexe des Eisenbo a k is alles e laub en es, die SMS bezüglich höchs mögliche
In ensi ä und bes mögliche Ene gieau lösung zu op imie en. Die Ene giebandb ei e
de S ahlung de SMS is e wa 15 neV (3 Γ0), die In ensi ä e wa 2,5·10 4 Pho onen/s
und de ypische du chs immba e Geschwindigkei sbe eich e wa ± 12 mm/s (± 0,6
µeV). Im Gegensa z zu klassischen adioak i en Quellen kann die S ahlung de SMS
in den Be eich on zehn Mik ome e n okussie we den. Die SMS is ein 'in-line'
Monoch oma o , de pe manen in de op ischen Hü e de Beamline ins allie is und
nach Einb ingung in den Synch o ons ahl be iebsbe ei is . Sie kann in Ve bindung
mi allen e ügba en P obenumgebungen de Beamline be ieben we den.
Die Ve ügba kei de SMS e ö ne Möglichkei en, Sys eme mi komplexen
Hype eins uk u en un e ex emen Bedingungen zu un e suchen, wie z.B. un e
Ul ahoch akuum und Hochd uck. Wei e hin e laub es die Quelle, Spek en in seh
ku ze Zei , d.h. in wenigen Minu en, zu messen; ku z genug, um Da en auch in
Ve bindung mi Lase heizung zu nehmen. Meh e e Hochd uckmessungen und
kombinie e Hochd uck-, Hoch empe a u messungen wu den im Rahmen de
Dok o a bei du chge üh . Diese we den im zwei en Teil de Dok o a bei
besch ieben. Die as 100-p ozen ige ücks oß eie S ahlung de Quelle und ih e hohe
Zusammen assung!
!
!
7!
B illanz e laub iele SMS Anwendungen. Die SMS kann in allen Be iebsmoden
eines Synch o ons be ieben we den.
2. Un e suchungen zum Spinzus and de Fe3+ Ionen in Pe owski en
Silika -Pe owski (Mg,Fe)(Si,Al)O3 is die am häu igs en o kommende Phase
im Un e en Man el de E de. Die Kenn nis ih e Eigenscha en is unabdingba ü das
Ve s ändnis des Ve hal ens des Un e en Man els. Dynamische, he modynamische
und T anspo eigenscha en de Silika -Pe owski e können emp indlich du ch den
Valenz- und Spinzus and des Eisens beein luss we den. Silika -Pe owski e mi 5-10
mol% Eisen (Fe3+ / ΣFe ~ 50-75%; McCammon e al., 1997) und Al is die dominan e
Phase im Un e en Man el de E de (~ 75 ol%)(Zhang e al., 2006; S ackhouse e al.,
2007). Dennoch bleiben die Ve ände ungen de elek onischen Eigenscha en des
Eisens un e ähnlichen Bedingungen wie des Un e en Man els wei e hin ums i en.
De zwei e Teil meine P omo ionsa bei is den Un e suchungen des Spinzus andes
in Fe3+ eichen Silika -Pe owski en un e Hochd uck gewidme . Vie e schiedene
Silika -Pe owski p oben mi un e schiedliche S öchiome ie wu den mi Hil e de
Synch o on Mössbaue Quelle un e such . Die SMS Spek en wu den bei
Raum empe a u und un e D ücken bis 122 GPa in eine Diaman -S empelzelle
gemessen. Es wu den P oben wie he ges ell bzw. auch he misch ausgeheil e (mi els
Lase heizung) benu z .
Die aus den Messungen gewonnenen Hype einwechselwi kungspa ame e ,
d.h., die Isome ie e schiebung und die Quad upolau spal ung, sind bei gleichem
D uck ü alle P oben diese A bei inne halb des expe imen ellen Fehle s gleich.
Zudem ände sich das Fe3+/ΣFe Ve häl nis nich ü die indi iduellen P oben übe
den gesam en D uckbe eich des Expe imen s. Die
Hype einwechselwi kungspa ame e des Fe3+-Duble s en sp echen einem
Hochspinzus and (Gü lich e al., 2011) und ih e lache Abhängigkei om D uck
deu e an, dass Fe3+ keinen Spinübe gang im gesam en D uckbe eich mach . Alle
beobach e en Ände ungen in den Spek en sind den Ände ungen des elek onischen
Zus andes des Fe2+ zuzusch eiben. Die Hype einwechselwi kungspa ame e des Fe2+-
Duble s mi de kleinen Quad upolau spal ung en sp echen einem Hochspinzus and
(McCammon e al., 2008), wäh end das Duble mi de g ossen
Quad upolau spal ung, dessen An eil mi zunehmendem D uck au Kos en des
Zusammen assung!
!
!
8!
Hochspinzus andes zunimm , en wede einem Zwischenspinzus and (IS) des Fe2+
en sp ich (McCammon e al., 2008) ode mi eine Ve ze ung de Umgebung des
Pla zes des Hochspin-Fe2+ e klä we den kann (Hsu e al., 2010). Ungeach e de
In e p e a ion des Fe2+ Spinzus andes bleib die Schluss olge ung bezüglich des nich
o handenen Spinübe ganges in Fe3+ dieselbe.
Diese E gebnisse zeigen, dass de kü zlich be ich e e Spinübe gang in Fe3+
nich s a inde , wenn Fe3+ Ionen die A-Plä ze bese zen. In diesem Fall bleiben die
Fe3+ Ionen in Silika -Pe owski en sowohl mi als auch ohne Aluminium bis mides ens
122 GPa im Hochspinzus and. Diese D uck en sp ich as den Bedingungen, die an
de G enze zwischen dem Un e en Man el und dem Äusse en Ke n he schen. Die
E gebnisse zeigen auch, dass Fe3+ Ionen un e Hoch empe a u ausheilung und
Hochd uck nich on den A-Plä zen in die B-Plä ze di undie en. Demzu olge gib es
auch kein Anzeichen ü einen Hochspin- zu Niede spinübe gang au G und eines
Pla zwechsels de Fe3+ Ionen. Dies s eh im Gegensa z zu den Fe2+ Ionen, die on
einem Hochspin- in einen Zwischenspinzus and übe gehen, ohne dabei einen
Niede spinzus and bei Raum empe a u in dem un e such en D uckbe eich zu
e eichen. Diese E gebnisse lassen e mu en, dass die seismischen
Geschwindigkei sanomalien im Un e en Man el nich einem Spinübe gang in Fe3+
zuzusch eiben sind.
Summa y!
!
!
9!
Summa y
I on is one o he mos abundan elemen s on Ea h, and i is an impo an
componen in mine als. Elec onic and magne ic p ope ies o i on-bea ing ma e ials
signi ican ly a ec p ocesses occu ing in he deep in e io o he Ea h. In he
ma e ials ha make up he Ea h’s lowe man le i on may exis in di e en alence,
spin s a es and c ys allog aphic en i onmen s. Mos o he exis ing expe imen al
echniques ei he do no allow o sepa a ely ollow e olu ion o di e en i on si es o
a e no sui able o measu emen s unde high-p essu e/high- empe a u e condi ions.
This makes s udies o i on elec onic s uc u e unde such condi ions ex emely
challenging.
The cu en Ph.D. hesis is di ided in o wo majo pa s. The i s pa is
dedica ed o he de elopmen o a Synch o on Mössbaue Sou ce (SMS). This de ice
allows ene gy domain Mossbaue spec oscopy o be pe o med on a sample unde
p essu es abo e 100 GPa in lase hea ed diamond an il cells. The second pa is
dedica ed o s udying he beha io o i on in i on/alumina-bea ing silica e pe o ski e
unde condi ions o he Ea h’s lowe man le.
1. Synch o on Mössbaue Sou ce
The e a e se e al echniques ha allow magne ic and elec onic p ope ies o
ma e ials unde ex eme condi ions o be p obed: X- ay Emission Spec oscopy
(XES), X- ay abso p ion nea edge s uc u e (XANES), Nuclea Resonance
Spec oscopes, e c. Fo elemen s in which obse a ion o Mössbaue e ec is possible
he mos ma u e, sensi i e, and sui able echnique o s udies o magne ic and
elec onic p ope ies is ene gy-domain Mössbaue spec oscopy.
Howe e , due o low b illiance o u ilized adioac i e sou ces and low na u al
abundance o i on in lowe man le mine als measu emen s using con en ional ene gy-
esol ed Mössbaue spec oscopy equi e e y long ime and usually a e limi ed o
mode a e p essu es. The p oblem can be sol ed by combining he ou s anding
p ope ies o synch o on adia ion (high b illiance, possibili y o ex eme ocusing)
wi h he ene gy- esol ed app oach. In b ie , wha is needed is a synch o on sou ce o
Mössbaue adia ion. Cons uc ion o such sou ce was he p ima y ask o my PhD
wo k. The possibili y o de elop such a sou ce was demons a ed a he Nuclea
Resonance beamline ID18 a he Eu opean Synch o on Radia ion Facili y (ESRF) by
Smi no e al. (1997). The sou ce is based on pu e nuclea e lec ions exis ing in
an i e omagne ic 57FeBO3 c ys als.
1.!In oduc ion!!
!
!
!
!
16!
cons ained by bo h i s bulk chemical composi ion (Table 1.1) and seismic p o iles,
p o iding he mos de ailed in o ma ion ega ding he s uc u e o he Ea h’s in e io .
Analyses o seismic wa es p o iles e ealed he p esence o he wo majo
discon inui ies wi hin he man le: (I) he so-called ansi ion zone a a dep h om 410
o 660 km, which de ines he bounda y be ween he uppe and lowe man le; (II) and
he D’’ laye ha lies jus abo e he co e-man le bounda y (Fig. 1.1). The o igin o
he man le discon inui ies will be discussed in he ollowing sec ion.
1.2. Mine alogical model o he Man le. (Mg,Fe)(Si,Al)O3 pe o ski e
as dominan componen o Ea h lowe man le.
E en h ough elemen composi ion o man le is qui e homogeneous, he
s uc u e o mine als om which is composed g adually change wi h p essu e. A he
small dep hs (<100 km) Ea h composed o 4 majo phases: oli ine, o hopy oxene,
clinopy oxene and ga ne . In he amewo k o he popula “py oli e model”
(Ringwood, 1975), i is belie ed ha he (Mg,Fe)2SiO4 oli ine composes abou 60%
o he Ea h’s uppe man le(Fig. 2) and ha he na u e o man le he e ogenei ies a e
ela ed o he dep h- a ying beha io o oli ine.
The i s ansi ion happens a he uppe bounda y o he ansi ion zone ( he
laye in he lowe mos uppe man le, cha ac e ized by anomalous beha io o seismic
wa e eloci ies) whe e (Mg,Fe)2SiO4 oli ine ans o ms o i s high p essu e
polymo ph – β-phase o wadsleyi e wi h a modi ied spinel s uc u e, which occu s a
a ound 14 GPa (410-km dep h) (Fig. 1.2). Then a abou 17 GPa (500-550 km dep h)
wadsleyi e ans o ms o a mo e densely packed spinel-s uc u ed phase –
(Mg,Fe)2SiO4 ingwoodi e. In u n, ingwoodi e disp opo iona es o (Mg,Fe)SiO3
pe o ski e and (Mg,Fe)O e ope iclase. This disp opo ion de ines he lowe man le
composi ion and he bo om o ansi ion zone, which is he 660-km discon inui y.
1.!In oduc ion!!
!
!
!
!
17!
Figu e 1.2
Mine al olume ac ion o he py oli e model man le (Ringwood 1989, 1991; S ix ude and Li hgow-
Be elloni 2005). Opx and cpx a e ac onyms o o hopy oxene and clinopy oxene espec i ely.
Modi ied a e F os , (2008).
The emaining non-oli ine componen s ( oge he hey make ~40% o uppe
man le olume): o hopy oxene, clinopy oxene and ga ne , unde go g adual
ansi ions a he dep h om 350 o 450 km: py oxene dissol es in ga ne ia MgSi →
Al2 subs i u ion, esul ing in he o ma ion o Al-deple ed ga ne (majo i e) a
p essu es abo e 16 GPa (I i une and Ringwood, 1987), which e en ually also
ans o ms o silica e pe o ski e (a 23 – 26 GPa o 660 - 750 km dep h) wi h
exsolu ion o Ca-bea ing silica e pe o ski e (Fig. 1.2). Since he dissolu ion o
py oxenes in ga ne is “smea ed ou ” o e mo e han 100 km, i leads o changes in
slope o he cu es o seismic eloci y e sus dep h a he han causing disc e e
discon inui ies.
1.!In oduc ion!!
!
!
!
!
18!
Figu e 1.3
Polyhed al models o pe o ski e (Ross and Hazen, 1989). Big solid g ey sphe es co espond o
Mg2+/Fe2+/3+ in e adecahed al si e; Si4+/Al3+ oc ahed a in he pe o ski e s uc u e a e shown in g een.
Summa izing all abo e, in he lowe man le we ha e h ee dominan phases:
(Fe,Mg)O e ope iclase, MgSiO3 silica e pe o ski e and CaSiO3 pe o ski e. A e
majo i ic ga ne ( he main ese oi o Al in uppe man le), becomes uns able, he
pe o ski e is he main candida e o a phase in lowe man le, which can inco po a e
Al. I on in he lowe man le is di ided be ween e ope iclase and silica e pe o ski e.
Finally, a p essu es and empe a u es close o hose o he co e-man le
bounda y magnesium silica e pe o ski e (space g oup Pnma) unde go a s uc u al
ansi ion in o he pos -pe o ski e phase (space g oup Cmcm) (Mu akami e al.,
2004). This ansi ion is belie ed o be “ esponsible” o he D’’ seismic discon inui y
a 2600-km dep h (Fig. 1.1 and Fig. 1.2).
The i on con aining silica e pe o ski e is a dominan phase in he Ea h lowe
man le, making ~80% o i s olume (almos ~ 48% o en i e olume o he Ea h).
The e o e, his pe o ski e gene ally de ines lowe man le seismic and densi y p o iles,
elec ical and he mal conduc i i y, edox s a e, e c. Thus unde s anding o he
silica e pe o ski e p ope ies is c i ical o cons uc ion model o he Ea h.
1.!In oduc ion!!
!
!
!
!
19!
The ideal cubic ABX3 pe o ski e c ys al s uc u e wi h space g oup Pm3m
consis s o a h ee-dimensional ne wo k o he co ne connec ed oc ahed a (B-si e)
and he ca i ies be ween he oc ahed a o m he e adecahed al si e (A-si e). The
lowe man le silica e pe o ski e has he GdFeO3 s uc u e ype, which is dis o ed
om he ideal cubic symme y by il ing o he oc ahed a and has he space g oup
Pnma (Ho iuchi e al., 1987). I has wo si es: he la ge dis o ed 8- o 12- old
coo dina ed A-si e p edominan ly occupied by a Mg2+ (big solid g ey sphe es in Fig.
1.3) (Ho iuchi e al., 1987, McCammon e al., 1992); and he smalle nea ly
symme ical oc ahed al B-si e p edominan ly occupied by a Si4+ (small colo ed in
g een in Fig. 1.3).
Le ’s look in he mo e de ails; so he base composi ion o he lowe man le
pe o ski e is he MgSiO3 (MgP ), whe e Mg2+ and Si4+ ca ions a e balanced by O2-
anions. In he lowe man le, i is assumed ha he MgP inco po a e Fe (in wo s a es
Fe3+ and Fe2+) and non-negligible amoun s o aluminum. Le ’s look in he mo e
de ails how his ca ions can en e pe o ski e.
The aluminum ions Al3+ due o hei small ionic adius (0.53 pm) can en e
only he B-si e o he silica e pe o ski e h ough he Si4+ subs i u ion. Howe e , such
subs i u ion c ea es he imbalance in he cha ge o he sys em. So i should be coupled
wi h c ea ing o an oxygen acancy, o he wise a 3+ ion should subs i u e he Mg2+ in
he A-si e. In case o Fe3+ he si ua ion is mo e complica ed. I s ionic adius o 0.63
pm is close o ha o silicon han ionic adius o e ic ions. Thus i was p oposed
ha e ous i on could en e B-si e as well as A-si e o MgP and subs i u e bo h Mg2+
and Si4+. The Fe3+ ion, which en e s he A-si e, allows he Al3+ o en e he B-si e
wi hou c ea ing oxygen de ec s in he MgP . The Fe2+ ions can en e he A-si e o
MgP subs i u e he Mg2+ one o one. On he o he hand, e ous i on can’ en e B-
si e due o ac ha Fe2+ ionic adius o 0.77 pm is bigge han 0.4 pm in case o Si4+
and because o needs o balance cha ge o he o sys em. Se e al mechanisms o such
subs i u ion we e p oposed, namely, Mg2+A + Si4+B o Fe3+A + Fe3+B, Mg2+A + Si4+B o
Fe3+A + Si4+B+□2- (oxygen acancy) and Mg2+A + Si4+B o Fe3+A + Al3+B; whe e A
and B subsc ip s indica e he co esponding MgP si e. The las mechanism a ac s
inc easing a en ion as; silica e pe o ski e wi h 5-10 mol% o Fe and Al is dominan
1.!In oduc ion!!
!
!
!
!
20!
phase in Ea h lowe man le (~75 ol%) (Zhang e al., 2006; S ackhouse e al., 2007)
and i allows o bo h ca ions o en e MgP wi hou c ea ing o he oxygen acancies
which a no ene ge ically a o able unde high-p essu e.
I was ound, ha he silica e pe o ski e in i s s abili y ield has conside able
amoun o he Fe3+, up o Fe3+/ΣFe ~50–75% (McCammon e al., 1997). En iching o
pe o ski e in Fe3+ is belie ed o be due o he c ys al chemis y; in p esence o he
aluminum i on p edominan ly is in Fe3+ s a e (McCammon e al., 1997) and no due
o oxida ion s a e o he lowe man le (F os e al., 2004; McCammon e al., 2005).
The e o e, he i on/aluminum con aining silica e pe o ski e (FeAlP ) is he mos
app op ia e sys em o s udy a ole o i on in he lowe man le p ope ies.
Beha io o he i on is di e en o i s a ious alence s a e and depends on
which c ys allog aphic si e i occupies in MgP . The e is no ag eemen in he
p essu es o he spin ansi ion o he Fe3+ epo ed by p e ious s udies, b oadly a
high-spin (HS) ( i e unpai ed d elec ons) o low-spin (LS) (one unpai ed d elec on)
ansi ion is p edic ed when Fe3+ occupies he B-si e, while Fe3+A is p edic ed o
emain in he high-spin s a e a all p essu es h oughou he lowe man le (Zhang e
al., 2006; S ackhouse e al., 2007; Ca alli e al., 2010 and 2011; Hsu e al., 2011).
E en hough Fe3+ is p edic ed o occupy exclusi ely he A-si e in lowe man le
composi ions o FeAlP (Vanpe eghem e al., 2006), a HS-LS ansi ion o Fe3+ was
epo ed o occu in FeAlP a high p essu e, po en ially due o exchange o Fe3+ om
he A- o he B-si e (Ca alli e al., 2011; Hsu e al., 2011).
I on is playing an essen ial ole in he modeling o he lowe man le
p ope ies. This ansi ion elemen in luences chemical pa i ioning and eac ions
among man le mine als, as well as he mal, elec ical, and mechanical p ope ies a
dep h (Fy e, 1960; Bu ns, 1970; Ga ney and Ande son, 1973; She man, 1988; Lin e
al., 2005, 2006; Goncha o e al., 2006; Kepple e al., 2007). Fo example, i was
ound ha p esence o he Fe3+ could con ol he adia i e conduc i i y o he
pe o ski e in he lowe man le (Kepple e al., 2007). Al eady mo e han 50 yea s
ago, Fy e (1960) p edic ed ha a spin-pai ing ansi ion o i on akes place wi hin he
lowe man le. Indeed he ansi ion was expe imen ally obse ed in he e ope iclase
(Bad o e al., 2003; Lin e al., 2005, 2006; Speziale e al., 2005; Kan o e al., 2006a)
1.!In oduc ion!!
!
!
!
!
21!
and he silica e pe o ski e (Bad o e al., 2004; Jackson e al., 2005; Li e al., 2006). I
was ound ou ha hese ansi ions s ongly a ec an elas ic p ope ies o he
e ope iclase; o example change o he shea eloci ies eaches ~4% (Mu akami e
al., 2012). This indica es ha exis ence o he spin ansi ion o he i on in he MgP
may lead o he change in i sha e eloci ies. Thus, exis ence o a so-called “Spin
ansi ion zone” in he lowe man le could explain phenomena like, o example, a
small seismic he e ogenei ies (Hedlin e al., 1997; Kaneshima e al., 1999). The
essence o hese he e ogenei ies can also be ela ed o a small luc ua ions in a Mg/Fe
o a (Mg,Fe)/Si a ios (Bina, 2003). Such a luc ua ion can be, in u n, caused by he
spin ansi ions in he lowe man le assemblage (Lin e al., 2007).
The Ea h mine alogy model is cons ained h ough compa ison o sound
eloci ies ob ained in labo a o y s udies o he seismological da a. Recen ly Mu akami
e al., (2012) had measu ed Vs o i on- ee MgP con aining Al and compa ed esul s
o P elimina y Re e ence Ea h Model (PREM). They came o conclusion ha he
lowe man le on a leas 93% consis s o he MgP . So, in he lowe man le a Mg/Si
a io is a ound 1. Such a high p opo ion is mo e consis en wi h he chond i ic man le
model han wi h con en ionally used pe ido i ic man le model. Such, chemical
s a i ica ion implies laye ed man le con ec ion wi h limi ed mass anspo .
Howe e , in ha s udy i on ee MgP was s udied, such composi ion does no
co espond o he eal li e composi ion o he Ea h lowe man le. As was s a ed
be o e he spin ansi ion o he i on in he MgP could lead o change in he Vs
alues o i . Thus, de e mina ion o he i on s a e is c i ical o de e mina ion o he
co ec man le model, which, in u n would lead o cons uc ion o he co ec model
o he con ec ion in he Ea h man le.
2.!Mo i a ion!!
!
!
!
!
22!
2. Mo i a ion
2.1 Spin s a e o i on in he lowe man le silica e pe o ski e
Lowe man le make up o 60% o Ea h olume. This makes lowe man le he
la ges (by olume) en i y o he Ea h. Thus, i is impossible o cons uc a eliable
model o he Ea h wi hou p ope unde s anding o lowe man le p ope ies. I is
belie ed ha Ea h lowe man le inco po a es in a e age ~10 mol.% o i on in o m o
Fe2+ and Fe3+ (S u hahn e al. 2005). I on is inco po a ed in wo majo phases o
lowe man le ( >90% o i olume, Fig. 1.2): e ope iclase and Al-bea ing
magnesium silica e pe o ski e. The mal and elec ical conduc i i y, and elas ic
p ope ies o he lowe man le may s ongly depend on i on oxida ion and elec onic
s a e (Xu e al. 1998, Lin e al. 2005, Keple e al. 2008). Thus, i on is p obably he
mos impo an ansi ion elemen in he physics o he Ea h. Cons uc ion o eliable
model o he Ea h is impossible wi hou cons aining he e olu ion o i on p ope ies
in e ope iclase and silica e pe o ski e unde lowe man le condi ions.
The beha io o i on in e ope iclase is well cons ained (Kan o e al, 2006,
Mu akami e al., 2012). On he o he hand, epo s on beha io o i on in silica e
pe o ski e emain s ongly con o e sial. The e a e h ee main easons o ha . Fi s ,
i on exis s in pe o ski e in wo alence s a es simul aneously (2+ and 3+ alence
s a es). Second, he s uc u e o magnesium silica e pe o ski e con ains wo si es,
la ge 8+4 coo dina ion si e (“A”) occupied by Mg2+ and a smalle oc ahed al si e
(“B”) occupied by Si4+. E en h ough i on expec ed o eside in he A-si e, he e a e
epo s, which sugges ha Fe3+ ions can also en e he B-si e (Ca alli e al. 2010).
Thi d, all i on elec onic s a es and c ys allog aphic en i onmen may depend on
p essu e and/o empe a u e. All oge he make ha i ons bea ing magnesium silica e
pe o ski e ex emely challenging sys em o s udy spin s a e o i on in i . E ec o Al
on Fe s a e should be men ioned he e as well.
The e a e se e al di e en scena ios desc ibing he spin ansi ion o i on in
silica e pe o ski e:
1) Bo h Fe3+ and Fe2+ ions unde go a high-spin o low-spin ansi ion (Bad o e al.
2004).
2.!Mo i a ion!!
!
!
!
!
23!
2) Only Fe3+ ions loca ed in A-si e unde go high spin o low-spin ansi ion (Jackson
e al. 2005, Zhang e al. 2006, S ackhouse e al. 2007).
3) Fe2+ ions unde go high-spin o in e media e spin ansi ion, while Fe3+ s ay in
high-spin s a e (McCammon e al. 2008, Na ygina e al. 2010, Lin e al. 2008).
4) Only Fe3+ ions loca ed in B-si e unde go high spin o low spin ansi ion (Ca alli
e al. 2010).
5) Fe2+ ions s ay in high spin s a e in lowe man le (Zhang e al. 2006, Cohen e al.
1997, Li e al. 2005), while obse ed changes in elec onic s uc u e o Fe2+ ions
a e explained by c ys allog aphic dis o ions o he A-si e (Lin e al. 2012).
6) Tha a Fe3+ ions a high-p essu e and high- empe a u e “di use” o B-si e a he
same momen as unde go high spin o low spin ansi ion (Ca alli e al., 2011).
The e is no ag eemen on which o hese models desc ibes he beha io o i on in
he lowe man le.
We ha e pe o med he in si u measu emen s o i on hype ine pa ame e s in
i on/alumina bea ing silica e pe o ski e samples wi h di e en s oichiome y in a
lase -hea ed diamond an il cells (DACs) in o de o in es iga e he spin s a e o Fe2+
and Fe3+ a condi ions o Ea h lowe man le, namely, unde p essu e up o ~ 130
GPa and empe a u e up o ~ 3500 Co. In o de o pe o m such s udy and achie e his
goal we de eloped new expe imen al ool ha was se up a ID18 a ESRF –
Synch o on Mössbaue Sou ce, as usually used me hods a e no e y sui able o his
kind o expe imen s.
2.2 Need o ene gy-domain synch o on Mössbaue spec oscopy
Beha io o he i on in silica e pe o ski es unde high-p essu e can be
in es iga ed by many di e en echniques: X- ay Emission Spec oscopy (XES),
Nuclea Fo wa d Sca e ing (NFS), X- ay abso p ion nea edge s uc u e (XANES)
and ela ed X- ay Magne ic Ci cula Dich oism (XMCD), con en ional Mössbaue
spec oscopy and X- ay powde and single c ys al di ac ion (XRD). Un o una ely
applica ion o any o hese powe ul echniques does no allow unambiguous
2.!Mo i a ion!!
!
!
!
!
24!
de e mina ion o he spin s a e o i on in lowe man le pe o ski e. Fo example,
al hough XANES is sensi i e o ions alence s a e and XMCD gi es an es ima e o
<Lz> and <Sz>, he e a e no well sui ed o he cases when he same elemen can be
ound in di e en alence s a e and c ys allog aphic si es (like i on in silica e
pe o ski e). XRD is no a di ec me hod o es ing o spin s a e. XES p o ides
in o ma ion on he bulk on densi y o elec onic s a es. This makes sepa a ion o he
indi idual con ibu ions om ions wi h di e en oxida ion s a e and/o
c ys allog aphic si es ex emely challenging. On he o he hand ene gy domain 57Fe
Mössbaue spec oscopy in many cases enables an unambiguous esolu ion o all
hype ine pa ame e s o ions in di e en oxida ion and elec onic s a es and di e en
c ys allog aphic en i onmen . High-p essu e measu emen s using con en ional
adioac i e poin sou ces encoun e se e al di icul ies such as long coun ing ime
(gene ally mo e han one week pe spec um a ~50 GPa p essu e ange), di icul ies
o ocus he beam, and s ong backg ound. All hese di icul ies g ea ly educe he
quali y o he spec a and hus esolu ion and sensi i i y o he echnique. The hi d
gene a ion synch o on acili y o e s a solu ion in he o m o ime-domain
coun e pa o con en ional Mössbaue spec oscopy (i.e., nuclea o wa d
spec oscopy, NFS). Howe e NFS is no well sui ed o high-complexi y sys ems,
which con ain i on nuclei in di e en spin, alence s a es and c ys allog aphic si es.
In o de o s udy spin ansi ion in pe o ski e unde lowe man le condi ions
we ha e de eloped an ene gy-domain synch o on Mössbaue sou ce. New
me hodology allows pe o ming o ene gy domain Mössbaue measu emen s using
synch o on as a sou ce o gamma adia ion. SMS inhe i s all o he ad an ages o he
con en ional Mössbaue spec oscopy and on op o ha o e s a numbe o new
bene i s: high lux, a beam diame e o a ew mic ons and ze o backg ound. SMS
allows o apid measu emen o ene gy-domain Mössbaue spec a unde ex eme
condi ions wi h a quali y gene ally su icien o unambiguously decon olu e e en
highly complex spec a. These ea u es we e success ully demons a ed in ou
in es iga ion o i on beha io in Al-bea ing silica e pe o ski e unde high-p essu e.
3.!Me hods!and!Ins umen a ion!!
!
!
!
!
25!
3. Me hods and Ins umen a ion
3.1. C ea ion o high p essu e
The i s expe imen s wi h diamond an il cells we e pe o med in la e 50s o
he 20 h cen u y (Wei e al. 1959). The me hod has s a ed o gain popula i y only
a e a disco e y o a eliable p essu e s anda d in he beginning o 70s (Fo man e al.
1972). Since he echnique appea s i has been cons an ly e ol ing, and he numbe o
possible applica ions inc eases. Now he lis o DAC applica ions includes
measu emen s o magne ic p ope ies, esis i i y, nuclea magne ic esonance, x- ay
di ac ion, x- ay abso p ion, op ical spec oscopy, con en ional Mössbaue
spec oscopy, Nuclea Fo wa d Sca e ing e c. (Che in e al. 1995, Mi o e al. 2001,
Ga g e al. 2004, Haase e al. 2009).
Ope a ion o diamond an il cell is e y simple and can be desc ibed as below.
The sample chambe is p essu ized by applica ion o o ce on wo uni-axially aligned
diamonds, his leads o comp ession o a sample, a p essu e ansmi ing medium and
a p essu e s anda d. The o ce can be applied in a ious ways. One o he mos
popula ways is igh ening o he sc ews (Me ill and Basse 1974) ano he one is
ising p essu e in a memb ane o he memb ane cell (Che in e al., 1995). Basic
diamond an il cell (modi ied a e Me ill-Basse design) is shown in Figu e 3.1.
In ou expe imen s, wo ypes o DACs we e used; hey a e shown in Figu e
3.2. I is he ou pin modi ied Me ill-Basse cell and he pis on-cylinde cell,
de eloped in Baye isches Geoins i u (BGI). The ypical diame e o hese cells is
Ø50mm and he a e age heigh is be ween 40 and 45 mm.
The uppe limi o p essu e gene a ed by diamonds depends on wo ac o s,
i s -on he size o he diamonds cule s and second-on he design o diamonds. In
p ac ice he simple ule is used: he smalle is diamonds cule size – he highe
maximum p essu e can be eached. In he wo ks desc ibed below, usually diamonds
wi h a cule diame e Ø250-300 µm we e employed.
To gene a e high-p essu e wi hou b eaking o he diamond an ils se e al
hings a e impo an : diamond suppo ing pla es (sea s), co ec ma e ial o he gaske
and good mechanical s abili y o he diamond an il cell. Design o diamond backing
pla es depends on expe imen al echnique and i manu ac u e . In BGI DACs used o
expe imen ha dened ungs en ca bide pla es we e employed, wi hou any speci ic
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
32!
4. Synopsis (Scope o he Thesis)
This chap e p o ides a summa y o he esul s p esen ed in i e pape s
comp ising he Chap e 5. This chap e is o ganized in ollowing way: Subsec ion 4.1
is dedica ed o heo e ical and expe imen al in es iga ion o angula , ene gy and
empo al p ope ies o pu e nuclea e lec ion o i on bo a e c ys als nea i s Néel
empe a u e. In subsec ion 4.2 we p esen a desc ip ion o SMS de eloping and
examples o i s applica ion beyond high-p essu e physics. Subsec ion 4.3 con ains
epo on spin s a e o Fe3+ ions in pe o ski e unde lowe man le condi ions. Finally,
in sec ion 4.4 combina ion o SMS wi h po able double-sided lase sys em is
desc ibed. This combina ion allows us o pe o m ene gy-domain Mössbaue
measu emen s unde high-p essu e and high- empe a u e ex eme condi ions
simul aneously.
I w o e h ee ou o i e pape s p esen ed in Chap e 5 as a i s au ho . Fo
hem I had p epa ed and conduc ed all expe imen s, analyzed he da a, and pe o med
calcula ions. I had de eloped he Synch o on Mössbaue Sou ce in close
collabo a ion wi h o he coau ho s. I had also w i en he ex o he publica ions
(alone o wi h ad ices o he co-au ho s) In pe cen s my con ibu ion o Chap e s 5.2
and 5.3 is 80% and o 5.4 is 75%. Two manusc ip s (Sec ions 5.1, 5.5) we e no
w i en by me as a i s au ho . Fo he pape desc ibed in Sec ion 5.1 I had pa icipa e
in calcula ions, da a analysis, and ex p epa a ions. My o al con ibu ion o he pape
is a ound 45%. Fo i h pape desc ibed in Chap e 5.5 I ha e p epa ed and
conduc ed lase -hea ing expe imen s wi h SMS, acqui ed and analyzed measu ed
spec a, and con ibu ed o he pape p epa a ion, which make my con ibu ion on he
le el o 33%.
4.1 Theo e ical calcula ion and expe imen al s udy o angula
dependence, ene gy and empo al dis ibu ions o synch o on-based
57Fe Mössbaue adia ion
We had pe o med a heo e ical calcula ion o he angula dependences and
ene gy, empo al dis ibu ions o pu e nuclea e lec ions in c ys al o i on bo a e
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
33!
(57FeBO3) a a ious alues o in e nal magne ic ield. These calcula ions we e
pe o med in o de o look a he beha io o hese dis ibu ions while in e nal
magne ic ield app oaching i collapse, which happens a Néel poin . I was ound ha
ollowing he collapse o he in e nal magne ic ield angula dependence o he
e lec ed adia ion ( ocking cu e) becomes signi ican ly b oade and change i s shape
om Gaussian-like o complex s uc u e wi h wo peaks (Fig. 4.1.2-a). On he o he
hand, ene gy and ime dis ibu ions demons a e s ong angula dependence. Only in
angula posi ion o he minimum be ween wo peaks (which co esponds o exac
alue o he B agg angle) ene gy dis ibu ion consis s om jus one line (he ea e
main line). Ou side o his na ow angula egion, he main line is accompanied by
sa elli es (Fig. 4.1.1-b). In he egion o he lowe angle sa elli e has a lowe ene gy
han main line, in he egion o highe angles sa elli e ene gy is sligh ly highe . An
analysis o ime dis ibu ions gi es (Fig. 4.1.3) an addi ional insigh a he in ol ed
p ocesses. The non-exponen ial decay shape o ime dis ibu ions in angula posi ion
o he minimum e eals a ue na u e o he main line, which was ob ained nea he
Néel poin . In ac his is pseudosingle line o med as he esul o he collapse o he
hype ine magne ic s uc u e o nuclea le els. In his pa icula case, he ime
dis ibu ion ollows such dependence:
, (4.1.1)
whe e τ – li e ime o he le el in he exci ed s a e, – ime, I( ) – in ensi y. The Eq.
(4.1.1) ac ually is a module o he Fou ie ans o m aken om ollowing equi a ion
(in limi o → 0):
, (4.1.2)
whe e – ampli ude o he sca e ed wa e, – di e ence in ene gy be ween wo
lines, – wid h o he line (
€
τ
×Γ≈
). The Eq. (4.1.2) desc ibes in e e ence be ween
wo Lo en zian lines wi h opposi e sign. In i s app oxima ion, he module aken
om o Eq. (4.1.2) a he limi o → 0 desc ibes he main line (which is ou
pseudosingle line) and in he i s app oxima ion ollowed squa ed Lo en zian
dis ibu ion:
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
34!
, (4.1.3)
whe e – ene gy, – a alue o esonance ene gy, L(E) – in ensi y.
The esul s o he expe imen s pe o med in almos pe ec condi ions (high-
quali y i on bo a e c ys al, e y small di e gence o he inciden beam e c.) a e in
e y good ag eemen wi h heo e ical p edic ions. Indeed in icini y o Néel
empe a u e, he wid h o he c ys al ocking cu e becomes signi ican ly b oade and
ocking cu e changes i s shape om Gaussian-like o complica ed s uc u e wi h wo
peaks (Figu e 4.1.2). Fu he mo e, he angula and ime dis ibu ions acqui ed a
s ong angula dependence (Figu e 4.1.3). As p edic ed, in he angula egion in he
minimum be ween wo peaks o ocking cu e he ene gy dis ibu ion consis o only
one main line ee o sa elli es. The ull wid h a hal maximum (FWHM) o his line
is o ~10 neV (2 ). Analysis o ime dis ibu ion showed ha he dis ibu ion
ob ained in he minimum angula posi ion can be i ed using Eq. 4.1.1 (Figu e 4.1.3-
2b). Thus, in i s app oxima ion he expe imen ally ob ained main line obeys squa ed
Lo en zian dis ibu ion.
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
35!
Figu e 4.1.1
Calcula ed angula , ene gy and empo al dis ibu ions o (333) pu e nuclea e lec ions o i on bo a e
c ys al. (a) – Rocking cu es calcula ed wi h di e en alue o in e nal magne ic ield. Ene gy and
empo al dis ibu ion was calcula ed o alue o he ex e nal magne ic ield equal o 2kOe. The dip
zone ( aken om Fig.4.1.1a) is shown on he uppe panel on Fig. 4.1.1b. The ene gy and ime
dis ibu ions on he lowe panels e e o he le and he igh sides o he dip zone, L and R
espec i ely, as shown on he uppe panel. The ene gy and ime dis ibu ions a he exac dip posi ion
a e displayed on bo h L and R panels (bold line cu es). Each nex dis ibu ion co esponds o an
angula shi o he c ys al om he dip posi ion by 5 μ ad.!Fo be e isualiza ion he ene gy and ime
dis ibu ions a e equally spaced along he e ical axis. (b) – Ene gy dis ibu ions a di e en angula
posi ion. Only in he angula posi ion, which co esponds o he exac B agg angle, main line is
p ac ically ee om sa elli es. (c) – Tempo al dis ibu ions a angula posi ions a which ene gy
dis ibu ions was calcula ed.
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
36!
Figu e 4.1.2
Angula dependence o he e lec ed adia ion o i on bo a e (333) pu e nuclea e lec ion o IB c ys al
(a) a RT (b) 348.9 K. Ci cles show expe imen al poin s, solid line shows heo e ical i . On panel b
le e s ( om a o d) ma k angula posi ion o he c ys al ela i e o he inciden beam in which ene gy
and ime dis ibu ion was measu ed.
Figu e 4.1.3
Measu ed Ene gy and ime dis ibu ions o nuclea B agg di ac ion o i on bo a e (333) e lec ion.
[1] - Ene gy dis ibu ions o nuclea B agg di ac ion o i on bo a e (333) e lec ion measu ed in
di e en angula posi ion o he c ys al ela i e o he inciden beam. Oli e ci cles - expe imen al
poin s, ed line - heo e ical i . Ene gy dis ibu ion was measu ed using s anda d Mössbaue
K2Mg57Fe(CN)6 single line abso be . [2] - Time dis ibu ions o nuclea B agg di ac ion o i on
bo a e (333) e lec ion measu ed in di e en angula posi ion o he c ys al ela i e o he inciden
beam. Oli e ci cles shows expe imen al poin s, ed line - heo e ical i , blue dash line shows i o
da a using Eq. 4.1.1. Measu emen s pe o med a ou angula posi ions (4.1.2b): small peak (a),
minimum (b), big peak (c) and he igh slope o he big peak (d).
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
37!
4.2 The 57Fe Synch o on Mössbaue Sou ce
Using esul s summa ized in Sec ion 4.1 (Chap e s 5.1 and 5.2) we design a
57Fe Synch o on Mössbaue Sou ce. I allows o pe o m ene gy-domain Mössbaue
measu emen s using synch o on adia ion. The SMS is based on pu e nuclea
e lec ion o an i on bo a e c ys al (57FeBO3). Figu e 4.2.1 shows he op ical scheme
o high-p essu e expe imen wi h a DAC using he SMS based on (333) pu e nuclea
e lec ion. The SMS is he in-line monoch oma o , i.e., he beam emi ed by he
sou ce is di ec ed almos exac ly along he synch o on adia ion beam om high-
hea load monoch oma o . We de eloped design o wo in-line op ical schemes based
on (111) and (333) pu e nuclea e lec ions o he 57FeBO3.
Figu e 4.2.1
The op ical scheme o a high-p essu e expe imen wi h DAC using he Synch o on Mössbaue Sou ce
based on he (333) pu e nuclea e lec ion. U – undula o ; HHLM – high-hea -load monoch oma o ;
CRL – compound e ac i e lens; SMS – he Synch o on Mössbaue Sou ce: HRM – high esolu ion
monoch oma o , De – Si (311) de lec o , IB – he i on bo a e c ys al inside he u nace wi h he ou
magne s and moun ed on he Mössbaue ansduce ; KBM - Ki kpa ick-Baez mi o s; DAC – diamond
an il cell; D – a alanche pho o diode de ec o .
The SMS is op imized o highes in ensi y and bes ene gy esolu ion, which is
achie ed by collima ion o he inciden synch o on adia ion beam and hus
illumina ion o he high-quali y i on bo a e c ys al wi hin a na ow angula ange
a ound an op imal posi ion o he ocking cu e. The SMS is pe manen ly loca ed in
an op ics hu ch and is ope a ional immedia ely a e mo ing i in o he inciden beam.
The bandwid h o adia ion p o ided by he SMS is ~3 Γ
0 o ~15 neV (in eal
expe imen s we wo k wi h ene gy esolu ion lowe han possible o ha e mo e
in ensi y), he in ensi y is ~2×104 pho ons/s and he ypical scanning eloci y ange is
abou ±12 mm/s (±0.6 µeV). In con as o con en ional adioac i e sou ces, he SMS
gi es he possibili y o ocus he beam o ens o µm ( he bes achie ed by now is 9x6
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
38!
µm2). As in-line monoch oma o SMS can be used wi h all exis ing sample
en i onmen s in he expe imen al hu ches downs eam o he beamline.
Figu e 4.2.2
Spec a o (Mg0.857Fe0.2)O e ope iclase a a p essu e o 29 GPa a oom empe a u e (g een ci cles)
and a 1580 K (blue iangles).
I implemen ed SMS in o de o s udy elec onic p ope ies o ma e ials unde high-
p essu e ex eme condi ions. One o he bigges SMS ad an ages is ha i allows
collec ing spec a apidly, on he scale o se e al minu es. Fo example, a high-quali y
Mössbaue spec um o he silica e pe o ski e sample in he DAC a 93 GPa was
acqui ed o abou 10 min, compa ed o mo e han one week using a adioac i e poin
sou ce o he same sample. This, in combina ion wi h possibili y o ocus beam
p oduced by SMS, enables o pe o m high- empe a u e s udies using lase hea ing
(Figu e 4.2.2). The almos 100% ecoilless esonan adia ion deli e ed by he sou ce
and i s high b igh ness allows a b oad ield o SMS applica ions. Due o a e y s ong
supp ession o elec onic sca e ing o pu e nuclea e lec ions (~10-9) SMS ope a ion
does no equi ed any ga ing o he p omp elec onic sca e ing. Thus, SMS can be
u ilized in any mode o s o age ing ope a ion.
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
39!
In his wo k SMS was used o s udy o he spin ansi ion in Fe3+ ions in i on
con aining silica e pe o ski e, which is he dominan phase o he lowe man le o he
Ea h (by mos ecen epo s ~93% (Mu akami e al.)).
4.3 Spin s a e o Fe3+ ions in he lowe man le
Using de eloped ene gy domain synch o on Mossbaue sou ce we conduc ed
a s udy o se e al i on con aining silica e pe o ski e samples (bo h Al con aining and
Al- ee). We ha e s udied ou di e en silica e pe o ski e samples:
Mg0.63Fe0.37Si0.63Al0.37O3 (~80% Fe3+/ΣFe); Mg0.63Fe0.37Si0.63Al0.37O3 (~70%
Fe3+/ΣFe); Mg0.78Fe0.2Al0.05Si0.97O3+δ (~50% Fe3+/ΣFe); Mg0.94Fe0.06SiO3+δ (~20%
Fe3+/ΣFe). In ou expe imen s SMS spec a a oom empe a u e we e collec ed up o
122 GPa using diamond an il cells, wi h o wi hou lase annealing o he samples.
De ails o sample syn hesis and expe imen me hodology a e gi en Sec ion 5.4.
Fi s , ene gy-domain Mössbaue s udy o Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e
was pe o med. I con ains i on dominan ly as Fe3+, measu ed spec a a e ex emely
well esol ed (Fig. 4.3.1). We i he da a using model ha consis o h ee quad upole
double s, one assigned o Fe3+ and wo assigned o Fe2+ (high and low quad upole
spli ing, QS). Unlike in case o NFS s udies, SMS allows unambiguously sepa a e
con ibu ions o Fe3+ and Fe2+ based on double s cen e shi s (CS). As i is clea
e en om isual inspec ion o Mössbaue spec a (Fig. 4.3.1) he e a e no changes in
he shape and ela i e a ea o he double co esponding o he Fe3+ o e he en i e
p essu e ange.
The hype ine pa ame e s o he low QS Fe2+ double (blue double in Fig.
4.3.1) co esponds o he high-spin s a e (McCammon e al., 2008), while he double
wi h high quad upole spli ing (black double in Fig. 4.3.1) co esponds ei he o
in e media e-spin (IS) Fe2+(McCammon e al., 2008) o a dis o ion o he si e
occupied by high-spin Fe2+ (Hsu e al., 2010). I espec i e o he in e p e a ion o he
Fe2+ spin s a e, ou conclusions ega ding he absence o a spin ansi ion in Fe3+
emain he same.
The i ing model ob ained o sample #1 spec a was implemen ed o all
o he samples. The hype ine pa ame e s o Fe2+ and Fe3+ o all samples s udied a e
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
40!
he same wi hin expe imen al e o . Mo eo e he e is no change in Fe3+/ΣFe o
indi idual samples o e he en i e p essu e ange o he expe imen (Fig. 4.3.2). The
hype ine pa ame e s o he Fe3+ double a e consis en wi h he high-spin s a e
(G eenwood and Gibbs, 1971), and hei smoo h a ia ion wi h p essu e indica es ha
Fe3+ does no unde go any spin ansi ions wi hin he en i e p essu e ange. No ably
he QS alue epo ed o low-spin Fe3+ om bo h expe imen al (Ca alli e al., 2010;
Ca alli e al., 2011) and heo e ical (Hsu e al., 2011) s udies ( ed dashed line, Fig.
4.3.2) a e mo e han wice ou obse ed alues.
!
Fig. 4.3.1
SMS spec a o Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #1 a oom empe a u e showing hei
e olu ion wi h p essu e (a) 2.5 GPa; (b) 38 GPa; (c) 67 GPa; (d) 93 GPa. The da a we e i o one Fe3+
double ( ed) and wo Fe2+ double s (blue and black), and he i esidual is shown below each
4.!Synopsis!(Scope!o ! he!Thesis)!
!
!
!
!
41!
spec um. A ea asymme y is due o p e e ed o ien a ion o he sample, and he eloci y scale is gi en
ela i e o α-i on.
Figu e 4.3.2
P essu e a ia ion o FeAlP hype ine pa ame e s (cen e shi [CS] and quad upole spli ing [QS]) (a)
high-spin Fe3+; (b) high QS Fe2+. Sample da a a e indica ed as ollows: Mg0.63Fe0.37Si0.63Al0.37O3
pe o ski e sample #1 ( ed and blue ci cles); Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #2 (o ange and
blue squa es); Mg0.78Fe0.2Al0.05Si0.97O3+δ pe o ski e (b own and da k blue ci cles); Mg0.94Fe0.06SiO3+δ
pe o ski e (pu ple and g een ci cles). The expec ed quad upole spli ing o low-spin Fe3+ is shown as a
ho izon al ed dashed line. (c) P essu e a ia ion o Fe3+/ΣFe as de e mined om he ela i e a eas:
pe o ski e sample #1 ( ed ci cles); Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #2 (o ange squa es);
Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e (b own ci cles); Mg
0.94Fe0.06SiO3+δ pe o ski e (pu ple ci cles).
Ho izon al lines a e guides o he eye. Values ob ained a e lase hea ing o Mg0.63Fe0.37Si0.63Al0.37O3
pe o ski e sample #2 and Mg0.78Fe0.2Al0.05Si0.97O3+δ pe o ski e a e indica ed by g ey s a s. In all cases,
pa ame e s emained unchanged om hose be o e hea ing wi hin expe imen al e o , demons a ing
ha he e is no high-spin o low-spin ansi ion in Fe3+ a he p essu e- empe a u e condi ions o he
lowe man le
Se e al s udies (Ca alli e al., 2010; Ca alli e al., 2011; Fujino e al., 2012)
ha e sugges ed ha he educed olume o LS Fe3+B could lead o a edis ibu ion o
Fe3+ om he A- o he B-si e in he pe o ski e s uc u e wi hin he lowe man le. To
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
48!
c ys al, he nuclea pola i on gene a es a - ay beam. The dis ibu ion o he emi ed
quan a in ime and in space is dic a ed by he empo al and spa ial de elopmen o
he nuclea pola i on, which can exis inside he a ge in o ms o ei he unning o
s anding wa e. The unning wa e mode is ealized in he p ocesses o he o wa d
sca e ing, while he s anding wa e pa e n is o med o nuclea B agg e lec ions.
The in ensi y o adia ion emi ed a e each SR pulse is decaying in ime. In space,
he adia ion o ms usually he low-di e gen beams bo h in he o wa d sca e ing
and in B agg e lec ion di ec ions (Kagan, e al., 1979).
The delayed esonan adia ion is, howe e , p eceded by a high-in ense x-
ay pulse con ibu ing a huge non- esonan backg ound in di ec ion o he p ima y
beam. E en a e ex eme monoch oma iza ion o he inciden adia ion (e.g., down o
eV) he signal- o-noise a io is s ill 10 - 10 ( wi h he na u al wid h o
nuclea esonance 5 neV). As o B agg e lec ion, an x- ay pho on can be
e lec ed by a c ys al h ough ei he nuclea esonan o Rayleigh elec onic sca e ing
p ocesses. I is he elec onic sca e ing ha b ings he la ge backg ound in he B agg
di ec ion on op o he nuclea esonan signal.
Fo una ely he na u e o nuclea B agg e lec ions o e s a way ou o his
disappoin ing si ua ion. The solu ion o he p oblem can be ound using elec onically
o bidden bu nuclea allowed B agg e lec ions, which exis owing o an unusual
pola iza ion dependence o nuclea esonance sca e ing in he p esence o hype ine
in e ac ion. Such pu e nuclea e lec ions we e p edic ed o an i e omagne ic
c ys als (T amell, 1961; Belyako and Ay azian, 1968) as well as o he c ys als
wi h elec ic ield o de ing (Belyako and Ay azian, 1969), and obse ed o he i s
ime in e s. (Smi no e al., 1960; Mi zababae e al., 1971). I was a pu e nuclea
e lec ion ha was applied in he i s expe imen whe e nuclea esonance exci a ion
by synch o on adia ion was unequi ocally obse ed (Ge dau e al., 1985).
Since he pu e nuclea e lec i i y is in ima ely ela ed o magne ic o elec ic
hype ine in e ac ion, he nuclea a ay in a c ys al beha es as a mul iline adia o due
o hype ine spli ing o nuclea le els. This is, o cou se, an incon enien
cha ac e is ic o a Mössbaue sou ce, which can no be accep ed. Luckily, a pa icula
case o hype ine in e ac ion, which is well ma ched o he idea o he gene a ion o a
single-line Mössbaue adia ion, does exis . Pu e nuclea e lec i i y wi hin an ene gy
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
49!
band o abou na u al linewid h o he Fe nuclea exci ed le el has been ob ained
when an i on bo a e FeBO single c ys al was hea ed close o Néel empe a u e
(Smi no e al., 1986). Tempo al p ope ies o pu e nuclea di ac ion in he icini y
o we e s udied in Re . Chumako e al., (1990) using he pulsed s uc u e o SR.
App oaching he ansi ion o he pa amagne ic s a e a Néel empe a u e ( o
FeBO T is 348 K), he hype ine s uc u e o he Mössbaue di ac ion
spec um o FeBO collapses and he e lec ion no mally disappea s abo e T
due o des uc i e spa ial in e e ence. Howe e , an applica ion o an ex e nal weak
magne ic ield o he c ys al esul s in a d as ic ans o ma ion o he in e e ence
condi ions. Acco ding o he quan um mechanical p inciple o supe posi ion o s a es,
a e abso p ion o a SR x- ay pho on e e y nucleus o he nuclea ensemble can be
ound on one o he ou allowed sublele s o he exci ed s a e. Fu he mo e, due o
mixing he nuclea spin s a es, each o hese ou suble els is cha ac e ized by wo
allowed spin p ojec ions. Thus, he emission o a delayed gamma- ay pho on is
desc ibed by he mul i-space in e e ence o he ele an p obabili y wa es, whe e
geome ical, ene gy and spin domains a e in insically in ol ed. The combined
mul ipa h in e e ence in space, ene gy, and spin domains esul s in a pseudo single-
line esonance s uc u e, which p o ides a basis o he c ea ion o a single-line
synch o on Mössbaue (SM) sou ce. An analysis o he hype ine s uc u e unde
hese condi ions is gi en in Re . Smi no e al. (2000).
The i s sou ce o SR-based Mössbaue adia ion has been de eloped
(Smi no e al., 1997) in 1997 a he Eu opean Synch o on Radia ion Facili y. The
sou ce emi s a di ec ed beam o ully ecoilless and pola ized adia ion. These
p ope ies ha e been e ec i ely used in he i s applica ion o he SM sou ce o
s udying magne ic ield dis ibu ion in i on alloys (Pankhu s e al., 2001). A la ge
p og ess in he u he de elopmen o he SM sou ce has ecen ly been achie ed a
SP ing-8 (Mi sui e al., 2009).
The physics o SMS is a ascina ing example o mul ispace quan um
in e e ence, i.e. he in e e ence o a gamma-quan um in geome ical, ene gy, and
spin spaces. This pape is de o ed o he de ailed analysis o his cohe en p ocesses,
which de e mine he p ope ies o he SM adia ion. Namely, he angula , ene gy, and
ime dis ibu ions o he SM adia ion a e analyzed in o de o de ine necessa y
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
50!
condi ions o an op imal pe o mance o he SM sou ce. The emainde o he pape
is o ganized as ollows. Sec ion 1 gi es a de ailed analysis o he p ocess o he
mul ispace in e e ence o nuclea esonan sca e ing o SR by he uni cell o
57FeBO3. In sec ion 2, p ope ies o he nuclea pola i on o he di ac ion p ocess o
adia ion by he nuclea a ay a e conside ed. In Sec ion 3, he angula , ene gy, and
ime dis ibu ions o he Fe SM adia ion a e calcula ed. A summa y and
conclusions a e gi en in Sec ion 4.
5.1.3 Mul ispace in e e ence o γ- ay pho on in he uni cell o
57FeBO3 c ys al
Table 5.1.1
Nuclea ansi ions be ween he g ound and exci ed s a es wi h magne ic quan um numbe s
o a ious ansi ion ene gies and a ious nuclei in he uni cell o he IB c ys al.
.
We conside he emission o a - ay pho on by he wo Fe nuclei belonging
o he uni cell o FeBO - I on Bo a e (IB) - c ys al. Le he nuclei be exci ed by a
sho pulse o synch o on adia ion much sho e han he li e ime o he exci ed
nucleus. Then he p ocesses o nuclea exci a ion and de-exci a ion a e wo sequen ial,
empo ally well decoupled, e en s. A e he p omp abso p ion o a p ima y pho on
and some dwelling in he in e media e exci ed s a e, he ansi ion back o he g ound
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
51!
s a e occu s wi h emission o a seconda y pho on. In acco dance wi h he gene al
p inciples o exci a ion, bo h nuclei a e exci ed in he in e media e s a e wi h equal
p obabili y ampli udes, he phases o which in he case o a plane synch o on
adia ion wa e a e de e mined by he scala p oduc , whe e is he wa e
ec o o he inciden adia ion and is he ec o indica ing he posi ion o a
nucleus in he uni cell ( =1,2). In o de o calcula e he emission in he di ec ion ,
one should add up he ampli udes o he wo - ay wa ele s wi h he accoun o hei
ela i e phase, which is de e mined by he di e ence in hei co esponding op ical
pa hs. The phase di e ence is hen gi en by he exp ession . Thus, in
o de o e alua e he esul o he in e e ence in geome ical space, one has o
conside wo pa hs. The phase di e ence o wa es sca e ed along he wo pa hs plays
a c ucial ole o he in ensi y o emission.
Each o he wo spa ial pa hs is b anched u he in o sepa a e pa hs wi hin he
ene gy domain. Due o magne ic dipole in e ac ion o he nuclea spin wi h he
in e nal ield, he nuclea le els in he IB c ys al a e spli in o suble els. The le el o
he g ound s a e ha ing nuclea spin spli s in o wo suble els; and he le el o
he exci ed s a e ha ing spin spli s in o ou suble els. The selec ion ules o
magne ic dipole ansi ion (M1), ele an o ou case, allow one o obse e six nuclea
ansi ions be ween he g ound and exci ed s a es wi h he change o he magne ic
quan um numbe . He e a e he magne ic quan um
numbe s o he exci ed and g ound nuclea s a es, espec i ely.
The IB c ys al is a can ed an i e omagne . The magne ic and c ys alline uni
cells o he c ys al a e he same (like o Fe2O3 hema i e c ys al, see Fig. 16 in Re .
Shull e al. (1951). The magne ic ields ac ing upon nuclei in he uni cell a e equal in
magni ude bu almos opposi e in di ec ion. Fo his eason, he suble els o he
exci ed and he g ound s a es ha e he same ene gies o bo h nuclei bu he s a es
app op ia e o he equi alen suble els a e cha ac e ized by he opposi e signs o he
magne ic quan um numbe s. The e o e, he ansi ions a he s and nd nuclei wi h
he same esonan ansi ion ene gies ( ) ha e he opposi e signs in he change
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
52!
o he spin magne ic quan um numbe s , see Table 5.1.1. He e he spin quan iza ion
axis is chosen o be di ec ed along he magne ic ield a he s nucleus.
Fo each a bi a y ene gy o he adia ion spec al componen all six
ansi ions will gi e i s speci ic con ibu ion o he in e e ence ield depending on he
ene gy dis ance be ween and a esonan ene gy , whe e . He e one
de ini ely aces a b igh case o in e - esonance nuclea in e e ence. Fo he
Lo en zian shape o each esonance, he summa y ampli ude o sca e ing is hen
p opo ional o , whe e s ands o he ene gy
de ia ion om h esonance in he uni s o he esonance na u al hal -wid h ( o he
sake o simplici y we ha e assumed so a ha he ampli udes o di e en ansi ions
a e equal o each o he ). Thus, o he wo nuclei in he uni cell he combined p ocess
o - ay in e e ence in geome ical and ene gy spaces in ol es wel e con ibu ions.
The a angemen o a oms in he IB c ys al does no only esul in he
o ma ion o he in e nal magne ic ield a he i on si es bu also in he c ea ion o a
non-uni o m c ys alline elec ic ield. The elec ic ield g adien (EFG) in he c ys al
exhibi s axial symme y. The p incipal axis o he EFG enso is o hogonal o he
in e nal magne ic ield a bo h i on si es. The Fe nucleus in he i s exci ed s a e
possesses bo h a magne ic and a quad upole momen , while in he g ound s a e i has
only a magne ic momen . So, he nucleus expe iences a pu e magne ic dipole
in e ac ion in he g ound s a e and a combined magne ic dipole and axially symme ic
elec ic quad upole in e ac ion in he exci ed s a e.
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
53!
Figu e 5.1.1
Ampli udes o con ibu ions o he pu e spin s a es , in o he mixed spin s a es o
di e en ene gy suble els o he exci ed nuclea s a e o di e en nuclei in he uni cell o he IB
c ys al, see Table II. Fo pa ame e see explana ion o Eq. (5.1.1).
The spli ing o nuclea s a es in an IB c ys al unde condi ions o he
combined magne ic dipole and elec ic quad upole hype ine in e ac ion has been
analyzed in e . Smi no (2000). The ou dimensionless ene gies o he hype ine
in e ac ion in he exci ed s a e a e gi en by
(5.1.1)
whe e , is he ene gy o in e ac ion o he magne ic momen
o he nucleus in he exci ed s a e wi h he in e nal magne ic ield ( is he
nuclea - ac o in he exci ed s a e, is he nuclea magne on) and
is he ene gy o in e ac ion o he nuclea elec ic
quad upole momen wi h he elec ic ield g adien ( is he elec ic ield
po en ial a he nucleus, is he cha ge o p o on); a e gi en in he uni s o he
quad upole in e ac ion ene gy . Since in he exci ed s a e o Fe he - ac o is
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
54!
nega i e, he sub-le el ene gies a e o de ed in he ollowing way: .
Conce ning he alue o he in e nal magne ic ield, i is a unc ion no only o
empe a u e bu also o he applied magne ic ield. The applica ion o he ex e nal
ield plays a signi ican ole in he nea icini y o Néel empe a u e whe e
an i e omagne ic o de ing o he a omic spins can be induced in he c ys al (Bo o ik-
Romano and Ozhogin, e al., 1960).
Fo pu e magne ic in e ac ion pu e spin s a es (i.e., he s a es wi h a de ined
spin p ojec ion ) can exis . Bu unde condi ions o combined magne ic and elec ic
hype ine in e ac ion he exci ed nuclea s a es may be mixed o e he spin
p ojec ions. In he pa icula case o he axially symme ic EFG wi h he main axis
o hogonal o he magne ic ield and z-axis di ec ed along he magne ic ield, each
nuclea exci ed s a e is mixed o e wo spin p ojec ions. In his case he Hamil onian
eigen alues gi en by Eq. (5.1.1) a e app op ia e o he eigen unc ions o he exci ed
s a es ep esen ed in Table 5.1.2.
Table 5.1.2:
The mixed-spin eigen unc ions o a ious exci ed nuclea s a es wi h he eigen ene gies and o
a ious nuclei in he uni cell o he IB c ys al.
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
55!
The exci ed nuclea s a es wi h ene gies a e mixed o e spin p ojec ions
, whe eas he s a es wi h ene gies a e mixed o e spin p ojec ions
Table 5.1.2 shows ha only ou ampli udes a e equi ed o
desc ibe he eigen unc ions o all exci ed s a es o bo h nuclei. Following he
analysis p esen ed in Re . (Smi no , 2000), hese ampli udes can be ob ained in an
explici o m
(5.1.2)
They a e de ined by a single pa ame e . The magne ic hype ine ield in
he IB c ys al is a unc ion o empe a u e. App oaching he ansi ion o he
pa amagne ic s a e a Néel empe a u e , he ield magni ude apidly dec eases.
A oom empe a u e, he magne ic in e ac ion domina es o e he elec ic one,
pa ame e , and he exci ed nuclea s a es a e nea ly pu e spin s a es wi h
and . On he con a y, in he ange o he exci ed
s a es a e al eady signi ican ly mixed o e spin p ojec ions. The e olu ion o he
mixed-spin s a es is illus a ed in Fig. 5.1.1. The dependences o he pu e spin s a e
con ibu ions in o he mixed s a es a e shown wi hin he ange o whe e
he magne ic dipole and elec ic quad upole hype ine in e ac ions a e ge ing
compa able. Figu e 5.1.1 clea ly shows ha in app oaching Néel empe a u e
( ), he exci ed s a es a e ge ing s ongly mixed o e spin p ojec ions. Because
he wo spin p ojec ions a e mixed in each exci ed subs a e he esonance sca e ing
ia a sepa a e nuclea ansi ion in e e y nucleus b anches ou u he in o he ou
pa hs.
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
56!
Figu e 5.1.2
Sca e ing pa hs ia he pu e spin s a es o he exci ed nuclea sub-s a e ha ing ene gy in he 1s
nucleus o he IB c ys al uni cell. On he uppe panel he ansi ions, up and down, be ween he g ound
and he exci ed s a es a e displayed. On he lowe panel ou possible pa hs a e shown, whe e changes
o spin p ojec ions in di e en pa hs a e indica ed. a e he changes o magne ic
quan um numbe in sepa a e ansi ions; a e he magne ic quan um numbe s o he exci ed
and g ound nuclea s a es espec i ely.
Fo elas ic sca e ing, he p ocess conside ed he e, he ini ial and inal s a es
a e he same. The e o e, i measu emen s do no pe mi o es ablish, which
in e media e spin s a e was in ol ed in he sca e ing, he sca e ing pa hs ela ed o
all in e media e s a es should in e e e. He e one mee s an in e es ing case o in a-
esonance nuclea in e e ence in he spin domain. Fig. 5.1.2 shows an example o
he sca e ing pa hs ia he nuclea ansi ion wi h he lowes ansi ion ene gy o
he -s nucleus. As i is seen om he igu e, in o de o build up he in e e ence
ield o a gi en ansi ion in a single nucleus, ou wa ele s o he same equency
bu di e en in pola iza ion s a es and sca e ing ampli udes should be added. The
ampli ude o a wa ele o - h nuclea ansi ion in - h nucleus con ains he
complex G - ac o gi en by he ollowing gene al exp ession
(5.1.3)
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
57!
whe e designa es a Clebsch--Go dan coe icien , which is a
p obabili y ac o o he ansi ion be ween suble els o he g ound and exci ed s a es
aking in o accoun he angula momen um ans e o he (nucleus+pho on) sys em;
i also depends on he spins o he g ound and he exci ed s a es and on hei z-
p ojec ions. In his manusc ip we a e conside ing di ac ion wi h (M1) mul ipola i y
(magne ic dipole ansi ion) and ha e chosen he magne ic basis o he pola iza ion
ec o : is he scala p oduc o he magne ic pola iza ion ec o o he
wa e p opaga ing in he di ec ion ( s ands o he di ec ion along he
inciden wa e, - o he di ec ion along he exi wa e), wi h pola iza ion
( ec o s lie in he sca e ing plane, while ec o s a e pe pendicula o i ) and
he sphe ical uni ec o in he coo dina e sys em ela ed o he hype ine ields
a h nucleus in he uni cell: , whe e a e
mu ually o hogonal uni ec o s ela ed o he hype ine ields in he uni cell, is
di ec ed along ec o (see Fig. 5.1.3); inally, a e he ampli udes o spin s a es
in he h sub-le el o he exci ed nuclea s a e. Below he - ampli udes a e
ela ed o he de ined abo e ampli udes.
(5.1.4)
In he example, depic ed in Fig. 5.1.2, he in e e ence in spin space leads o he
ollowing exp ession o he - ac o
(5.1.5)
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
64!
esonance in e e ence s a s playing he essen ial ole e en a oom empe a u e. On
he o e lapping slopes o he 1s and he 3 d lines, as well as o he 4 h and he 6 h
lines he deep and sha p alleys (i.e., minima o he sca e ed in ensi y) a e obse ed
which a e due o he esonance abso p ion o adia ion a he ene gies . Fo
he ansi ions M = 0 he nuclea di ac ion is cancelled due o des uc i e
in e e ence o he cons i u ing wa es in geome ic space.
When due o he c ys al hea ing he in e nal ield dec eases down o kOe,
he ole o he in e - esonance in e e ence ises signi ican ly, Fig. 5.1.4b. The e ec
o he in e e ence is seen in a much b oade angula ange. The in e e ence causes a
s ong asymme y in bo h ene gy and angula dis ibu ion o he emi ed adia ion.
Beyond he B agg angle ange he ene gy spec um con ains only h ee peaks a he
lowe angles, - he 3 d line is supp essed o e he e, while a he highe angles he
ou lines a e p esen bu wi h he s ongly weakened inne ones. Wi hin he B agg
angle ange a complica ed in e e ence pa e n is obse ed whe e one can s ill
dis inguish he ou lines o abou equal s eng h. On he ene gy scale, o he
esonance ange a ex ended wings a ac a en ion. They a e also p esen on he
o me igu e. The wings ep esen an axial symme ic con igu a ion. Such a
symme y is due o a ying condi ions o he in e e ence o he wa es sca e ed by
nuclei and is ela ed o he ac ha he eal pa o he nuclea ampli ude is une en
unc ion o he ene gy de ia ion om esonance.
Finally, we come nea o he Néel poin wi h he magne ic ield o only 2 kOe
(Fig. 5.1.4c). The mul i-line spec um o he emi ed adia ion has signi ican ly
collapsed he e. In he icini y o he B agg angle a pseudo single line esonance
s uc u e is obse ed, which is c ea ed by he des uc i e in e -nuclea in e e ence (in
geome ical space) as well as by a s ong in e - esonance (in ene gy domain) and
in a- esonance (in spin-domain) in e e ence o adia ion componen s discussed in
he p e ious sec ion.
As seen in Fig. 5.1.4 emission occu s o e a ange o se ings o he c ys al
a ound he B agg angle. To e eal he de ails o he in e e ence pa e n in he
ollowing sec ion we ha e analyzed he angula unc ion o he cohe en emission and
de e mined he angula b ea h o s ong emission o di e en magni udes o he
in e nal magne ic ield. In addi ion he spec al composi ion and he ime dis ibu ion
o - ay pho ons adia ed by he nuclea a ay in SM sou ce we e examined.
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
65!
5.1.5 Angula , ene gy and ime dis ibu ions o Fe SM adia ion
Ini ially we discuss he geome ical p ope ies o he in e e ence pa e n. Fi s
o all we wish o know how he in ensi y o he adia ion emi ed by he c ys al
depends on i s angula se ing in he icini y o he B agg angle. To es ima e he
heo e ical angula wid h o he emission unc ion o a pe ec c ys al we can neglec
he angula wid h o he inciden beam, assuming a plane wa e is inciden on he
c ys al. To ind in his way he emission angula unc ion we ha e o in eg a e he wo
dimensional emission unc ion ( o i s de ini ion see ex a e Eq. (5.1.14))
o e equency, assuming SR in ensi y o be equency independen alue
o e he whole ange o he nuclea esonance. The emission angula
unc ion is hen gi en by he ollowing in eg al
(5.1.15)
whe e in he in ensi y unc ion he angula and equency a iables a e eplaced
adequa ely: and . One should unde line ha he exp ession
o Eq. (5.1.15) gi es he emission angula dependence in eg a ed o e he whole
esonance ange. The emission angula unc ions o di e en magni udes o he
in e nal magne ic ield a e displayed in Fig. 5.1.5. The cu e on he bo om panel o
Fig. 5.1.5 ha ing a symme ic shape ep esen s he angula dependence o emission a
oom empe a u e. The cen e o angula ange o p onounced emission is shi ed wi h
espec o he angle gi en by he o iginal B agg e lec ion law by 15.5 ad. This
shi is due o e ac ion o he inciden beam a he en ance in o he c ys al. The
wid h o he e lec i i y cu e is e y small, 5 ad.
When hea ing he c ys al owa ds Néel empe a u e he b ead h and shape o
he emission angula unc ion is d as ically changed. Th ee main ea u es can be
ecognized (Smi no , 2009):
• signi ican dec ease o he emi ed in ensi y,
• s ong b oadening o he angula ange o emission,
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
66!
• c ea ion o he double hump s uc u e o he angula cu es wi h he dip
exac ly a he B agg angle co ec ed o e ac ion.
Figu e 5.1.5
The emission angula dependence o adia ion om he IB c ys al in eg a ed o e he whole
esonance ange. The angula se ing o he IB c ys al is changed in he icini y o B agg angle o pu e
nuclea e lec ion (333). The angula unc ions a e displayed o di e en alues o he in e nal
magne ic ield dec easing om he bo om o he op in app oaching Néel empe a u e.
A he lowes ield (i.e., he highes empe a u e o he c ys al) he emission
in ensi y is only o he o de o o he in ensi y a oom empe a u e. Such
diminu ion o he cohe en emission nea Néel empe a u e was expec ed because o
he des uc i e na u e o in e e ence a his empe a u e.
Along wi h he dec ease o he cohe en emission he angula ange o
emission is being signi ican ly enla ged: om ad a he ield o kOe ( he
bo om panel in Fig. 5.1.5) up o ad a he ield o kOe ( he uppe panel in
Fig. 5.1.5). This happens because he s ong o - esonance wings o nuclea sca e ing
(loca ed mainly in he na ow angula in e al a ound he exac B agg posi ion, i.e.,
a ound he peak o he ocking cu e, see Figs. 5.1.4a,b) become hea ily supp essed
by des uc i e in e e ence when empe a u e app oaches (Fig. 5.1.4c). On he
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
67!
con a y, he wide angula dis ibu ion o he on- esonance sca e ing is less a ec ed
by des uc i e in e e ence (Fig. 5.1.4c). Thus, i p o ides he wide ocking cu e
a ound . I is o in e es o no e ha a u he dec ease o he ield and o he ela ed
emission in ensi y is no accompanied any longe by a change o he shape and he
wid h o he angula cu e. A kind o sa u a ion is eached in he b ead h o he
angula ange o emission o ays o e he en i e esonance egion nea he Néel
poin .
Fig. 5.1.4c shows ha he ou e wings o he landscape alls ab up ly in he
na ow angula in e al nea he B agg angle. Compu e analysis o he emission
unc ion shows ha he wings a e shi ed ou o he immedia e B agg ange mo e and
mo e as he collapse o he hype ine s uc u e occu s. They s a he e o e o gi e
hei con ibu ions o he emi ed adia ion only when he angle o incidence is shi ed
om he cen al zone o he sides. This is he eason o he o ma ion o he double
hump s uc u e and he appea ance o he dip in he emission angula unc ion.
The angula dis ibu ion o he emi ed adia ion is de e mined bo h by he
emission angula unc ion and by he ins umen al unc ion, i.e., by he angula
dis ibu ion o synch o on adia ion in he inciden beam. When he angula
di e gence o he exci ing adia ion is la ge in compa ison wi h he angula wid h o
he ange o emission, he emission occu s h oughou he whole allowed ange. I he
SR in ensi y is a cons an alue in his ange, he angula dis ibu ions o he emi ed
adia ion jus epea hose depic ed in Fig. 5.1.5. In he opposi e case o e y small
di e gence o he exci ing adia ion in compa ison wi h he angula wid h o emission,
he angula dis ibu ion o adia ion ep oduces he ins umen a ion unc ion o m.
To ge he angula dis ibu ion o he emi ed adia ion in an in e media e case one
mus ind he p oduc o he emission angula unc ion and o he ins umen al
unc ion a an a bi a y angle o incidence o he exci ing synch o on adia ion.
An example o he angula dis ibu ion o he emi ed adia ion is shown in Fig. 5.1.6,
assuming he ins umen al unc ion o be o Gaussian o m
. The nuclei a e exci ed by SR exac ly a he dip o
he emission angula unc ion, ad and ad. The angula
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
68!
dis ibu ion o adia ion emi ed in he icini y o Néel empe a u e o he in e nal
magne ic ield o 2 kOe is compa ed wi h he ins umen al angula unc ion. As seen
in he igu e, unde he conside ed eal condi ions, he angula b ead h o he emi ed
adia ion exceeds sligh ly he angula wid h o he ins umen al unc ion. The op o
he dis ibu ion is sligh ly modula ed by he emission angula unc ion.
We u n now o he ene gy dis ibu ion o he emi ed adia ion o a ixed
angle o incidence o he exci ing SR beam. A he beginning we conside a simple
model o he des uc i e in e - esonance in e e ence. As shown in e . Smi no
(2000), due o he combined mul ispace in e e ence only wo esonance lines su i e
close o Néel empe a u e. They belong o he 3 d and 6 h esonances in he o iginal
hype ine s uc u e in he FeBO . The ene gies o he ele an ansi ions nea ly
coincide and he s eng hs o he lines a e almos equal. Thus he spec um in ou
model can be p esen ed by he sum o wo Lo en zians, each ende s a single line
adia ion emi ed a a sepa a e esonance ansi ion,
(5.1.16)
As , wha co esponds o a collapse o he hype ine s uc u e, he emission
in ensi y (as he in eg and in Eq. (5.1.15)) anishes. Unde hese condi ions he
double peak s uc u e is being ans o med in o a pseudo-single line, which shape is
gi en app oxima ely by he ollowing exp ession
Figu e 5.1.6
Angula dis ibu ion o he emi ed adia ion - solid line; angula dis ibu ion o he exci ing
synch o on adia ion -do s (see ex ). The i s dis ibu ion is scaled in in ensi y o compa ison wi h
he second one. The c ys al is se a he posi ion o angula dip ( op panel in Fig. 5.).
(6.1.17)
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
69!
whe e . Thus he ene gy dis ibu ion is p esen ed in ou model by a
Lo en zian-squa ed unc ion. The line wid h in his case is less han , namely
. One should no e ha he in e e ence is cons uc i e a he ene gies lying in
be ween he esonance ones and , while, in con as , i is des uc i e on bo h le
and igh ou e sides o esonance egion ( he ampli udes in Eq. (5.1.16) end o cancel
each o he o e he e). Whence, - ay emission comple ely disappea s when
(abo e Néel empe a u e). Howe e , in he eal case he applica ion o a small
ex e nal magne ic ield es o es he an i e omagne s uc u e o he c ys al and - ay
cohe en emission esumes (Chumako e al., 1986). Due o cohe en e ec s he eal
wid h o he ene gy dis ibu ion can be la ge han .
To ob ain he ene gy dis ibu ion o he emi ed adia ion a he angle o
incidence o one should calcula e he ollowing in eg al
(5.1.18)
All he alues en e ing in o Eq. (5.1.18) a e de ined abo e. Se e al ene gy
dis ibu ions o adia ion emi ed nea Néel empe a u e a di e en angles o
incidence o he SR a e displayed in Fig. 5.1.7. The cen al pa o he emission
angula unc ion a ound ad is shown on he uppe panel in Fig. 5.1.7,
whe e he conside ed angula in e al is limi ed by he L,R ba s. I is mos ly he
egion in be ween he wo humps o he angula cu e. The di e gence o he exci ing
SR beam is aken as 5 ad.
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
70!
Figu e 5.1.7
Ene gy dis ibu ions o he emi ed adia ion a di e en angles o incidence o he exci ing SR in
he icini y o he dip zone o he emission angula unc ion o he (333) e lec ion and o he in e nal
magne ic ield o 2 kOe. The dip zone ( aken om Fig. 5.1.5) is shown on he uppe panel. The ene gy
dis ibu ions on he lowe panels e e o he le and he igh sides o he dip zone, L and R
espec i ely, as shown on he uppe panel. The ene gy dis ibu ion a he exac dip posi ion is displayed
on bo h L,R panels - bold line cu es. Each nex dis ibu ion co esponds o an angula shi o he
c ys al om he dip posi ion by 5 ad. Fo be e isualiza ion he ene gy spec a a e equally spaced
along he e ical axis.
The s ep in change o he angle o incidence is o he same alue. The
ene gy dis ibu ions o he emi ed adia ion a e shown on he lowe panels. The
dis ibu ions co esponding o he shi s owa ds he lowe hump o he emission
angula unc ion ( o he le o he dip zone) a e deno ed by and hose owa ds he
highe hump ( o he igh o he dip zone) a e deno ed by . The ene gy dis ibu ion
o he emi ed ays in he case o exci a ion o he nuclea a ay a he exac dip
posi ion is ep esen ed by he bold-line cu e on bo h and g aphs.
The na u e o he double hump s uc u e in he emission angula unc ion
p omp s he idea o selec jus he dip zone o ob aining a single line SM adia ion
spec um. Indeed, as seen om he igu e, in his ange i is p edominan ly a single
line ha ing a wid h close o (ho izon al ba on he hal -heigh ). A s ongly
supp essed sa elli e is no iced on he le -hand side o he main peak. When mo ing
o he cen e o he angula cu e o i s le hump a b oade line is being o med in
he ange o he lowe ene gies o he adia ion spec um. I s ela i e con ibu ion is
ising as he angula shi is inc easing. When mo ing o he cen e o he angula
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
71!
cu e owa ds i s igh hump he changes o he adia ion spec um a e mos ly he
same bu now a b oade line appea s a he igh -hand side o he cen al maximum
and he small sa elli e is no smoo hing ou as in he L dis ibu ions. Appea ance o
he side b oad lines in bo h L and R spec a is due o exci a ion o he nuclea a ay a
he angles o incidence a which he wings he emission landscape in Fig. 5.1.4c a e
al eady ouched.
Figu e 5.1.8
Time dis ibu ions o he emi ed adia ion a di e en angles o incidence o he exci ing SR in he
icini y o he dip zone o he emission angula unc ion o he (333) e lec ion and o he in e nal
magne ic ield o 2 kOe. The dip zone ( aken om Fig. 5.1.5) is shown on he uppe panel. The ime
dis ibu ions on he lowe panels e e o he le and he igh sides o he dip zone, L and R
espec i ely, as shown on he uppe panel. The ime dis ibu ion a he exac dip posi ion is displayed
on bo h L,R panels - bold line cu es. Each nex dis ibu ion co esponds o an angula shi o he
c ys al om he dip posi ion by 5 ad. Fo be e isualiza ion he ime dis ibu ions a e equally
spaced along he e ical axis.
Finally we come o he angula esol ed ime dis ibu ions o he emi ed
adia ion. They we e calcula ed by pe o ming he ollowing in eg a ion
(5.1.19)
whe e is eal ime and is he na u al li e ime o he exci ed nuclea s a e.
The inne in eg al ep esen s he in e se Fou ie ans o m o he ene gy dis ibu ions
o he emi ed adia ion ampli udes. The ob ained ime dis ibu ions a e displayed in
Fig. 5.1.8. This igu e is delinea ed in he same way as he o me one. All ime
dependences o e he e co espond o he ene gy dis ibu ions shown in Fig. 5.1.7. The
ime dependence o he emi ed adia ion in he case o exci a ion o he nuclea
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
72!
a ay a he exac dip is ma ked by he bold lines on bo h and g aphs. The non-
exponen ial shape o his dependence e eals he ue na u e o a single line spec um
ob ained nea by he Néel poin . In ac i is a pseudo-single line o med in he p ocess
o he combined in e e ence o adia ion pa hs in he geome ic space, and in ene gy
and spin domains. In pa icula , he - ay wa es emi ed by nuclei in he uni cell
igh upon exci a ion cancel each o he , because hese wa es a e o equal ampli udes
and opposi e in sign due o he phase shi be ween hem o a he ze o ime. The
ela i e phase be ween he wa es is changing due o a di e ence in hei ca ie
equencies wi h ime owa ds hei cons uc i e in e e ence, he e o e a quan um
bea pa e n could be obse ed. Howe e , a he same ime he cohe en decay o he
exci ed s a e is p oceeding, so ha he quan um bea and decay p ocesses a e
o e lapped in ime. In ou model o in e e ence o he wo Lo en zian lines, see Eq.
(5.1.16), using he in e se Fou ie ans o m, we can ob ain he ollowing ime
dependence
(5.1.20)
whe e quan um bea ing and decay a e p esen ed by he p e-exponen ial ac o and he
exponen ial one, espec i ely. Quan um bea pe iod he e is . The
esul an pa e n depends s ongly on he ela ion be ween he cha ac e is ic imes o
he wo p ocesses, he bea pe iod and he na u al decay ime . In he case whe e
he bea pe iod and decay ime a e o he same o de he wo p ocesses compe e.
In he eal case one should conside he ela ion be ween he bea pe iod and
he cha ac e is ic ime o he cohe en emission . In he icini y o Néel
empe a u e he bea pe iod can be se e al imes la ge han , since in his
empe a u e ange he hype ine spli ing is al eady ac ional o b ead h o he
cohe en esonance line. The ansi ion om des uc i e o cons uc i e in e e ence
p oceeds he e slowe han he decay. In his app oxima ion he ollowing ime
dependence is alid: . The o m o he ime dependences in Fig.
5.1.8 is s ill de e mined by he compe ing p ocesses. Inc ease o he in ensi y due o
posi i e-going in e - esonance in e e ence is compensa ed by he speeded up
cohe en emission. This compe i ion esul s in he o ma ion o he emission
maximum in he ime dependences. When app oaching he imes o ully cons uc i e
5.1!Mul ispace quan um in e e ence in 57Fe synch o on Mössbaue sou ce!
!
!
!
73!
in e e ence mos o he s o ed ene gy is eleased al eady h ough he incohe en
channels (Chumako e al, 1990).
When mo ing om he cen al angula zone owa ds he humps he ime
dis ibu ions a e de e mined in addi ion o he discussed easons by he in e e ence o
he na ow cen al componen wi h he b oade lines a he sides in L and R g aphs o
Fig. 5.1.7.
5.1.6 Summa y and conclusion
Elec onically o bidden bu nuclea -allowed B agg e lec ions can be employed
o gene a e he ecoilless adia ion by exci ing nuclea a ay o i s isome le el wi h
he help o he synch o on adia ion. Pu e nuclea e lec ions exis owing o a s ong
spin dependence o nuclea sca e ing in he p esence o hype ine ields in c ys als.
This way a sou ce o Mössbaue adia ion al e na i e o he adioiso ope sou ces can
be ealized. The cohe en exci a ion o he nuclea a ay leads o he c ea ion o
nuclea exci on pola i on s a es which gene a e a he exi o he c ys al he cohe en
adia ion ee o a non- esonan con ibu ion.
A pa icula case o pu e nuclea e lec ion is conside ed whe e a single line
Fe Mössbaue adia ion is gene a ed. I is accomplished in an an i e omagne
57FeBO3 c ys al (I on Bo a e (IB) c ys al) hea ed up o he icini y o i s Néel
empe a u e.
When conside ing he in e e ence o adia ion while nuclei in he uni cell o
he IB c ys al de-exci e one can ind 48 con ibu ions in o he in e e ence ampli ude.
Two con ibu ions come om he de-exci a ion o he wo nuclei p esen in he uni
cell. Each o hem con ains six con ibu ions in he ene gy domain whe e nuclea
le els a e spli due o combined magne ic and elec ic hype ine in e ac ion in he IB
c ys al. Finally nea Néel poin whe e he magne ic dipole and elec ic quad upole
in e ac ions in he IB c ys al a e o compa able s eng h he nuclea suble els in he
exci ed s a e become s ongly mixed in spin p ojec ions. The wo p ojec ions o
nuclea spin a e allowed a each sub-le el. The e o e each o he 6 con ibu ions in he
ene gy domain in ol es in addi ion he 4 con ibu ions in he spin domain. Thus he
combined mul i-pa h in e e ence in geome ic space, in ene gy, and spin domains
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
80!
he c ys al se ing ela i e o he inciden beam. The ele an angula dependence o
he e lec ion in ensi y is e e ed as ocking cu e.
Figu e 5.2.1
Schema ic iew o he expe imen al se -up. SR - beam a e HRM; T, Mössbaue ansduce which
used o oscilla ion o he c ys al; F - u nace combined wi h magne s a angemen ,which con ains
57FeBO3 c ys al; H - magne ic ield o 110 eO; A - s anda d single line abso be K2Mg57Fe(CN)6 ; D -
APD de ec o .
The beam e lec ed om he c ys al was inciden on o a single line abso be
(A) and hen o de ec o (D) (Fig. 5.2.1). To moni o he ene gy dis ibu ion o
adia ion emi ed by he i on bo a e c ys al we used a s anda d Mössbaue
K2Mg57Fe(CN)6 single line abso be , which had an a ea densi y o esonan Fe
iso ope o 50 mg/cm and line wid h o 2.1 ( is he na u al wid h o he
esonance line). All measu emen s we e pe o med o (333) PNR o he i on bo a e
c ys al. Because he e is no di e ence in physics o nuclea sca e ing be ween
di e en PNR in FeBO , he ob ained esul s a e applicable o any o hem, wi h a
co ec ion on he di e ence in he ocking cu e wid h o each e lec ion.
The ex e nal magne ic ield o 110 Oe was applied o he c ys al in he
di ec ion no mal o he sca e ing plane in o de , i s , o maximize he e lec ion
coe icien , second, o b oaden he magne ic ansi ion egion. We used he u nace
and he magne s a angemen o a special cons uc ion (Po apkin e al., 2012) in o de
o a oid g adien o he empe a u e and o he magne ic ield. Thanks o he e y
s ong supp ession o elec onic sca e ing o PNR (10 ), i was possible o pe o m
all measu emen s wi hou any ga ing o he p omp elec onic sca e ing. Fo mo e
de ails conce ning he expe imen al se up o c ys al en i onmen see Re . (Po apkin e
al., 2012).
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
81!
5.2.4. Resul s and Disscusion
Fig.2 shows he ocking cu e measu ed a oom empe a u e (a) and a 348.9
K (b), i.e., sligh ly abo e 348.35 K. A oom empe a u e, he ull wid h a hal -
maximum (FWHM) o he angula depend ence (Fig. 5.2.2a) was 10.5(2) ad. The
heo y p edic s FWHM o 5.9(2) ad a oom empe a u e. When he FeBO
c ys al was hea ed app oxima ely o i s Néel empe a u e he ocking cu e wid h
g ew by a ac o o 3, om 10.5(2) o 32.1(2) ad. Fu he mo e, he ocking cu e
acqui ed double-peak s uc u e (Fig. 5.2.2b). The angula dependence nea
becomes asymme ic; he peak a he smalle angles is less han ha a he bigge
angles.
Figu e 5.2.2
Angula dependence o he e lec ed adia ion o i on bo a e (333) pu e nuclea e lec ion o IB c ys al
(a) a RT (b) 348.9 K. Ci cles shows expe imen al poin s, solid line shows heo e ical i . On igu e
2(b) le e s ( om a o d) ma k angula posi ion o he c ys al ela i e o he inciden beam in which
ene gy and ime dis ibu ion was measu ed.
To compa e he expe imen al esul s wi h he heo e ical p edic ions, he
expe imen al da a we e i wi h he use o code based on he heo e ical o malism
(Smi no e al., 2011). The wo a iable pa ame e s we e used in he i ing o
ocking cu es, namely, he in e nal magne ic ield H and he pa ame e which
includes di e gence o he inciden beam and possible bending o he c ys al. In he
expe imen di e gence o he inciden beam was ixed. Thus, a ia ion o could
simula e only bending o he eal c ys al wi h espec o he pe ec one.
The bes i o he ocking cu e a oom empe a u e was ob ained wi h
= 5.3(2) ad (Hin was ixed o 330 kOe, which is he oom empe a u e alue o in
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
82!
Hin he i on bo a e c ys als). The inc ease o he e ec i e di e gence om 2 ad o
5.3 ad e eals he sligh c ys al bending. In icini y o he Néel empe a u e he
bes i o he ocking cu e was eached wi h H =12 kOe and = 5.9(2) ad.
The ob ained alue o in e nal magne ic ield was used as ixed pa ame e o i ing
he ene gy and he ime dis ibu ions o he emi ed adia ion in he ange o he
B agg angle nea he Néel empe a u e.
Figu e 5.2.3
Ene gy dis ibu ions o nuclea B agg di ac ion o i on bo a e (333) e lec ion measu ed in di e en
angula posi ion o he c ys al ela i e o he inciden beam. Oli e ci cles - expe imen al poin s, ed line
- heo e ical i . Ene gy dis ibu ion was measu ed using s anda d Mössbaue K2Mg57Fe(CN)6 single
line abso be . Measu emen s pe o med a ou angula posi ions (Fig. 5.2.2b): small peak (a),
minimum (b), big peak (c) and he igh slope o he big peak (d).
The measu emen s o ene gy and ime dis ibu ions we e pe o med a a
se ies o angula posi ions along he ocking cu e. He e we p esen he
measu emen s aken a he ou angula posi ions indica ed in (Fig. 5.2.2b): small
peak (a), minimum (b), big peak (c) and he igh slope o he big peak (d). Mössbaue
spec a o he single line abso be A measu ed in he gi en angula posi ions a e
displayed in Fig. 5.2.3. Wi h he accoun o he abso be line con ibu ion he
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
83!
ob ained spec a cha ac e ise he ene gy dis ibu ions o gamma adia ion emi ed a
he gi en angles. The ene gy dis ibu ion in minimum (b) exhibi s only one line (Fig.
5.2.3b). Fo b e i y we will call i he main line. The FWHM o his line is 2 (10
neV). This alue was ound by sub ac ing he wid h o he single line abso be om
he measu ed FWHM. In he angula ange o he lowe peak o he ocking cu e e
(Fig. 5.2.2a posi ion a), he main line is accompanied by a sa elli e line (Fig. 5.2.3a)
wi h a smalle ene gy. The lines a e sepa a ed by ene gy in e al ( ) 9 . Wi h
inc ease o de ia ion om he B agg angle he in ensi y o he low ene gy sa elli e
inc eases.
In he angula ange o he highe peak o he ocking cu e a sa elli e wi h
he ene gy la ge han he ene gy o main line appea s, see, o example, he wo
ene gy dis ibu ions aken a he peak and a i s igh slope (Figs. 5.2.3c,d). The heo y
p edic s ha ene gy o his sa elli e should be e y close o he ene gy o he main
line. Due o he limi ed ene gy esolu ion in he expe imen he sa elli e was no
esol ed in he measu ed ene gy dis ibu ions o adia ion. Only he g owing
asymme y o he main line (Fig. 5.2.3c) om he side o bigge ene gies indica es he
p esence o he sa elli e. The di e ence in ene gy be ween he main line and he
sa elli e is 2.5 . The sa elli e heigh is ising as he de ia ion om he B agg
angle inc eases. In he angula posi ion 'd' Fig. 5.2.2 he sa elli e e en domina es he
main line. The ene gy dis ibu ion o he emi ed adia ion in his angula ange looks
jus like a single b oad line (Fig. 5.2.3d).
Fig. 5.2.4 shows he ime dis ibu ions o he emi ed adia ion in he same
angula posi ions o he c ys al ela i e o inciden beam. In he ange o smalle
angles, o example on he op o he lowe peak (posi ion 'a' Fig. 5.2.2), he ime
dis ibu ion (Fig. 5.2.4a) exhibi s quan um bea s (Rü e e al., 1990), whe e posi ions
o he maxima a e con olled by sepa a ion o he main line and he sa elli e in he
ene gy spec um (Fig. 5.2.3a). In his pa icula case he ime dis ibu ion ollows he
dependence (Smi no e al., 2011):
(5.2.1)
whe e, τ is an e ec i e decay ime, which is de e mined by he cohe en gamma- ay
emission. The pe iod o quan um bea s gi en by equa ion is 98 ns.
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
84!
The e o e, o he wo lines wi h = 9 he i s bump should be loca ed a 49
ns and second bump a 147 ns. This is app oxima ely wha he measu ed ime
dis ibu ion shows (Fig. 5.2.4a).
Figu e 5.3.4
Time dis ibu ions o nuclea B agg di ac ion o i on bo a e (333) e lec ion measu ed in di e en
angula posi ion o he c ys al ela i e o he inciden beam. Oli e ci cles shows expe imen al poin s,
ed line - heo e ical i , blue dash line shows i o da a using Eq. (5.2.2). Measu emen s pe o med a
ou angula posi ions (Fig. 5.2.2b): small peak (a), minimum (b), big peak (c) and he igh slope o
he big peak (d).
In his case, he quan um bea s a e dumped due o a sho ened li e ime - ,
which is ela ed o an enla ged wid h o he ene gy lines as . Tha 's why
bea s a enua e as e . In he minimum o he ocking cu e (posi ion 'b' Fig. 5.2.2) he
ime dis ibu ion does no show an exponen ial decay (Fig. 5.2.4b), which is expec ed
in case o a Lo en z-like line. This indica es ha single line is no a simple Lo en z-
like line. In he minimum, he ime dis ibu ion can be app oxima ed by he
dependence (Smi no e al., 2011):
(5.2.2)
In his case bump posi ion is de ined by alue o , i.e., . The i o da a by
Eq. 5.2.2 is plo ed by a dash line on Fig. 5.2.2b. Thus, he ime dis ibu ion in he
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
85!
minimum o he ocking cu e is an o e dumped bea s case, whe e quan um bea s a e
o e dumped by exponen ial decay.
In he ange o bigge angles, o example in big peak posi ion o igh slope
(posi ions 'c' and 'd' on Fig. 5.2.2b) he ime dis ibu ion shows an in e media e case
(Figs. 5.2.4c,d) be ween o e dumped bea s case and quan um bea s case. The e ime
dependences al eady do no ollow Eq. (5.2.2). In his case, posi ion o he bump is
s ill possibly mainly de e mined by .
One a iable pa ame e was used in he i ing o ene gy and ime
dis ibu ions o e lec ed adia ion, namely, he angle o incidence o he exci ing
synch o on X adia ion. The good i o all he ene gy and ime dis ibu ions was
ob ained, see Fig. 5.2.3 and Fig. 5.2.4. This indica es a good ag eemen be ween he
expe imen al esul s and he heo e ical p edic ion.
The ime dependence would ollow he Eq. 5.2.1 i he main line is c ea ed in
he p ocess o des uc i e in e e ence be ween wo emission lines, in case whe e
ene gy di e ence be ween wo lines, i.e, con e ge o 0. In such case esul ing
line doesn' ollow Lo en zian dis ibu ion. In i s app oxima ion, main line can be
desc ibed by squa ed Lo en zian dis ibu ion(Smi no e al., 2011):
(5.2.3)
whe e is he wid h o each o he in e e ing lines and is a alue o esonance
ene gy. Howe e , i he main line would exac ly ollow he squa ed Lo en zian
dis ibu ion, hen o main line wi h wid h , i.e., o 70 ns, peak ( ) in
ime spec a should be a 140 ns and no a 25 ns like we ha e a measu ed ime
spec um (Fig. 5.2.4b). Because, i u ns ou ha in eali y he main line is ollowing
mo e complica ed dependence. The shi o he peak o lowe imes indica es ha ails
o main line dependence d op slowe han o squa ed Lo en zian dis ibu ion.
5.2.5. Summa y
We ha e s udied p ope ies o gamma adia ion emi ed by nuclei in i on
bo a e c ys al se in he icini y o B agg angle o i pu e nuclea e lec ion. I is
ound ha he emission in ensi y measu ed a di e en angula posi ions o he c ys al
5.2!Angula , spec al and empo al p ope ies o nuclea adia ion 57Fe !
Synch o on Mössbaue sou ce.!
!
!
!
86!
ela i e o he exci ing synch o on adia ion beam exhibi s ex ao dina y beha io
nea he Néel empe a u e. App oaching Néel empe a u e, he angula dependence o
he emission in ensi y ans o ms d as ically in shape and wid h. The ocking cu e
s ongly b oadens, spli s and acqui es a double-peak shape. On he o he hand, he
ene gy and ime dis ibu ions o he emi ed adia ion s ongly depend on he c ys al
se ing in he emission ange. In pa icula , he ene gy dis ibu ion measu ed in he
minimum be ween he wo peaks consis o only one line wi h wid h o ∼ 10 neV.
Analysis o ime dis ibu ion showed ha in i s app oxima ion ha line obey squa ed
Lo en zian dis ibu ion. The ob ained esul s a e in e y good ag eemen wi h
heo e ical p edic ions. The esul s ob ained in his wo k a e highly impo an o o
de eloping o new synch o on based echniques wi h equi emen s o neV
bandwid h o incoming adia ion.
Acknowelegmen s
Au ho s a e g a e ul o J.-P. Celse o his help wi h p epa a ion o he
expe imen . Also we would like o hank Alexy Bosak and Ilya Se guee o use ul
scien i ic discus sion du ing pape p epa a ion.
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
87!
5.3 The 57Fe Synch o on Mössbaue sou ce a he ESRF
!
Vasily Po apkin,a,b* Aleksand I. Chumako ,a,c Gennadii V. Smi no ,c Jean-Philippe
Celse,a Rudol Rü e ,a Ca he ine McCammonb and Leonid Dub o inskyb
!
aEu opean Synch o on Radia ion Facili y, BP 220, F-38043 G enoble, F ance,
bBaye isches Geoins i u , Uni e si ä Bay eu h, D-95440 Bay eu h, Ge many, and
cNa ional Resea ch Cen e "Ku cha o Ins i u e", 123182 Moscow, Russia.
*-co esponding au ho
Published in jou nal o Synch o on Radia ion Vol 19 pa 4 p. 559-569
5.3.1 Abs ac
We desc ibe he design o a 57Fe Synch o on Mössbaue Sou ce (SMS) o
ene gy-domain Mössbaue spec oscopy using synch o on adia ion a he Nuclea
Resonance beamline (ID 18) a he Eu opean Synch o on Radia ion Facili y (ESRF).
The SMS is based on a nuclea esonan monoch oma o employing pu e nuclea
e lec ions o an i on bo a e (57FeBO3) c ys al. The sou ce p o ides 57Fe esonan
adia ion a 14.4 keV wi hin a bandwid h o 15 neV, which is unable in ene gy o e a
ange o abou ±0.6 µeV. In con as o adioac i e sou ces, he beam o gamma
adia ion emi ed by he SMS is nea ly ully esonan and ully pola ized, has high
b illiance and can be ocused o a 10x5 µm2 spo size. Applica ions include, among
o he s, he s udy o e y small samples unde ex eme condi ions, o example a
ul ahigh p essu e o combined high p essu e and high empe a u e, and hin ilms
unde ul ahigh acuum. The small c oss sec ion o he beam and i s high in ensi y
allow o apid collec ion o Mössbaue da a. Fo example, he measu emen ime o a
spec um o a sample in a diamond an il cell (DAC) a ~100 GPa is a ound 10 min;
whe eas such an expe imen wi h a adioac i e poin sou ce would ake mo e han one
week and he da a quali y would be conside ably lowe . The SMS is op imized o
highes in ensi y and bes ene gy esolu ion, which is achie ed by collima ion o he
inciden synch o on adia ion beam and hus illumina ion o he high-quali y i on
bo a e c ys al wi hin a na ow angula ange a ound an op imal posi ion o he ocking
cu e. The SMS is pe manen ly loca ed in an op ics hu ch and is ope a ional
immedia ely a e mo ing i in o he inciden beam. The SMS is an in-line
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
88!
monoch oma o , i.e., he beam emi ed by he SMS is di ec ed nea ly exac ly along
he inciden synch o on adia ion beam. Thus, he SMS can be easily u ilized wi h all
exis ing sample en i onmen s in he expe imen al hu ches o he beamline. Due o a
e y s ong supp ession o elec onic sca e ing o pu e nuclea e lec ions (~10-9)
SMS ope a ion does no equi ed any ga ing o he p omp elec onic sca e ing. Thus,
SMS can be u ilized in any mode o s o age ing ope a ion.
5.3.2 In oduc ion
Con en ional Mössbaue spec oscopy has been a aluable ool o decades in
s udying he magne ic and elec onic p ope ies o a ious ma e ials. Howe e , he
echnique is no well sui ed o he s udy o samples wi h diame e s less han ~100 µm.
Focusing o adia ion om adioac i e sou ces is ex emely di icul (Yoshida e al.,
2009), and esul s in a low coun a e and a educ ion o spec al quali y due o an
inc ease o he backg ound. This leads o longe measu ing imes in o de o ob ain
easonable quali y spec a, and e en p ecludes some s udies unde ex eme
condi ions, o example he in es iga ion o mine als ele an o he deep Ea h a
p essu es o mo e han 100 GPa. Thus, he absence o ocusing possibili ies e a ds
p og ess in esea ch ields whe e mic on-scale samples a e in ol ed, o example,
high-p essu e geophysics and geochemis y.
The ime-domain analog o adi ional Mössbaue spec oscopy is ealized ia
ime- esol ed nuclea o wa d sca e ing o synch o on adia ion. Synch o on
adia ion has e y high b illiance and allows o ex eme ocusing; hence NFS is an
excellen ool o s udy mic on-sized samples. Howe e , NFS is no well sui ed o
s udy highly complex phases, which con ain i on in di e en spin s a es, alence
s a es and c ys allog aphic si es. The ime spec um o NFS esul s om an
in e e ence o wa es cohe en ly sca e ed by all i on a oms in he sample; he e o e i
also con ains he c oss- e ms o he sca e ing ampli udes o i on a oms in di e en
s a es. Thus, in o de o de i e he hype ine pa ame e s o one i on s a e o si e, one
needs o i he hype ine pa ame e s o all s a es. In con as , abso p ion, by na u e is
an in-cohe en p ocess, esul s in an ene gy spec um, which is linea supe posi ion o
cons i uen componen s. In mos cases his enables he i ing o each i on si e o s a e
independen ly om all o he s; hence ene gy-domain Mössbaue spec oscopy is mo e
sui able o s udying highly complex phases. Thus o s udies o mic on-scale
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
89!
abso be s wi h complex Mössbaue spec a, he op imum app oach would be o
employ high- esolu ion ene gy-domain Mössbaue spec oscopy in combina ion wi h
high b illiance and ex eme ocusing o synch o on adia ion. In sho , wha is
needed is a synch o on-based sou ce o Mössbaue adia ion.
Such a sou ce can be designed o 57Fe using he p ope ies o pu e nuclea
e lec ions o an i on bo a e c ys al. Fo e lec ions o his ype elec onic di ac ion
is o bidden, bu nuclea di ac ion is allowed because o a speci ic pola iza ion
ac o o nuclea esonan sca e ing in he p esence o hype ine in e ac ions
(Smi no e al., 1969). Due o magne ic hype ine spli ing o he 57Fe nuclea le els
in he g ound and he i s exci ed s a es, he ene gy spec um o he e lec ed
adia ion consis s o se e al lines. Unde hese condi ions he nuclea a ay in a
c ys al beha es as a mul iline adia o . This is, o cou se, an incon enien p ope y o
pe o ming spec oscopic measu emen s. Howe e , when he i on bo a e c ys al is
hea ed close o i s Néel empe a u e o 348.35 K, he ene gy spec um o he e lec ed
adia ion collapses o a single line (Smi no e al., 1986). Unde hese condi ions, a
pa icula case o hype ine in e ac ion is ealized whe e magne ic dipole and elec ic
quad upole hype ine in e ac ions a e s ongly mixed (Smi no e al., 2000). The
combined mul ipa h quan um in e e ence in space, ene gy, and spin domains esul s
in he o ma ion o a pseudo-single-line esonance s uc u e in i on bo a e (Smi no e
al., 2011), which p o ides he basis o he c ea ion o a single-line Synch o on
Mössbaue Sou ce (SMS). The spec um o he emi ed adia ion has an ene gy
bandwid h close o he na u al line wid h o he Mössbaue esonance. The possibili y
o de elop such a sou ce was demons a ed a he Eu opean Synch o on Radia ion
Facili y (ESRF) in 1997 (Smi no e al., 1997). La e , he same app oach was
success ully implemen ed a SP ing-8 (Mi sui e al., 2007a; Mi sui e al., 2007b;
Mi sui e al., 2007c; Mi sui e al., 2009).
Fo ene gy-domain Mössbaue spec oscopy, one needs o scan he ene gy by
changing he ela i e eloci y o he sou ce and he sample using he Dopple e ec .
In SMS expe imen s, se e al me hods exis o achie e he Dopple e ec . The
simples is o mo e he sample (Smi no e al., 1997; Mi sui e al., 2007a; Mi sui e
al., 2007b); howe e his app oach is no con enien o mos expe imen s whe e
gene ally a complica ed and/o a hea y sample en i onmen is used. Ano he
possibili y is o use an addi ional silicon c ys al be o e he sample and o achie e he
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
96!
o i s quali y. Thus, in o de o p ese e a good quali y o he ins umen al unc ion
and a high in ensi y, one needs o ha e a high quali y i on bo a e single c ys al, a
small di e gence o he inciden beam, and a su icien ly accu a e o a ional s age o
p ecise and s able angula posi ioning o he c ys al.
Figu e 5.3.4
(a) The ocking cu es o pu e nuclea (333) e lec ion (in absolu e e ical scale) and o he elec onic
Umweg e lec ion (a bi a y e ical scale) o he 57FeBO3 c ys al o empe a u es in he icini y o
he Néel poin . The a ows show he angula posi ions whe e he Mössbaue abso p ion spec a we e
measu ed. (b) and (c) The Mössbaue abso p ion spec a o he single line abso be measu ed a angula
posi ions 1 and 2, espec i ely, a a empe a u e o he i on bo a e c ys al o 348.90 K.
App oaching he Néel empe a u e, he angula wid h o he ocking cu e
inc eases. Acco dingly, he dip be ween he wo peaks o he ocking cu e whe e he
single line ins umen al unc ion can be ob ained, becomes mo e dis inc and b oad,
making i easie o main ain he c ys al a he op imal angula posi ion. Fu he mo e,
he b oadening o he ocking cu e is accompanied by a na owing o he ene gy
spec um o adia ion e lec ed by he c ys al a he exac B agg angle (Smi no e al.,
2011). Thus, inc easing empe a u e imp o es he ins umen al unc ion o he SMS.
Howe e , his imp o emen is accompanied by a s ong dec ease o he in ensi y o
he e lec ed adia ion; he e o e a easonable comp omise be ween he quali y o he
ins umen al unc ion and he coun a e has o be ound. Unde o he wise equi alen
condi ions, a good quali y o he c ys al and a p ope collima ion o he inciden beam
should allow one o achie e his comp omise a highe in ensi y.
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
97!
Fig. 5.3.4a shows he ocking cu es o he (333) e lec ion o wo di e en
empe a u es in he icini y o he Néel poin . The i s ocking cu e (g een ci cles) is
ob ained a he empe a u e 348.90 K whe e an op imal comp omise be ween he
ene gy wid h o he ins umen al unc ion and he in ensi y o he e lec ed beam is
achie ed. The second ocking cu e shown in Fig. 5.3.4a (blue iangles) is measu ed
a a empe a u e 0.05 K abo e he p e ious one. This small change o empe a u e
inc eases he wid h o he ocking cu e om 33 µ ad o 37 µ ad, which sligh ly
elaxes he equi emen s o collima ion o he beam, quali y o he c ys al, and
angula s abili y. Howe e , his lowe ing o he ole ance o angula dis o ions is
accompanied by ~40% loss o in ensi y. Thus, e en small imp o emen s o he beam
collima ion and c ys al quali y could p o ide a emendous gain in he e iciency o
he SMS.
The ins umen al unc ions o he SMS we e e alua ed using a single line
K2Mg57Fe(CN)6 abso be wi h an a ea densi y o he esonan 57Fe iso ope o 0.50
mg/cm2 and a line wid h o 2.1 Γ0 (Γ0 = 0.097 mm/s). Fig. 5.3. 4b shows he
abso p ion spec um measu ed when he i on bo a e was se o he angula posi ion a
he cen e o he dip on he ocking cu e (Fig. 5.3.4a). The ob ained line wid h o 4.3
Γ0 es i ies ha he line wid h o he ins umen al unc ion a he op imal angula
posi ion is abou 2.2 Γ0 (a e abso be wid h decon olu ion).
A he angula posi ions o bo h peaks o he ocking cu e, he ins umen al
unc ion acqui es undesi able sa elli e peaks. The same e ec happens when he i on
bo a e c ys al is no o su icien quali y: he double-peak s uc u e is washed ou and
he ins umen al unc ion measu ed a he exac B agg angle acqui es sa elli e peaks
due o he e ec i e angula a e aging. Unde hese condi ions, a single line
abso p ion spec um can be ob ained a he “backup” angula posi ion a he high-
angle slope o he ocking cu e (Fig. 5.3.4a). Howe e , he ene gy wid h o he
ins umen al unc ion a he backup posi ion is la ge : he measu ed line wid h o 6.7
Γ0 (Fig. 5.3.4c) shows ha he line wid h o he ins umen al unc ion is abou 4.6 Γ0
(a e abso be wid h decon olu ion).
5.3.3.6. Umweg e lec ions
E en i a pa icula e lec ion is o bidden o elec onic di ac ion, he X- ay
beam may ind a way o mul iple sca e ing in he c ys al la ice. The X- ays may
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
98!
expe ience di ac ion on wo o mo e se s o c ys alline planes so ha he beam is
e lec ed in he same di ec ion as i he e lec ion would be allowed. Such “bypass”
e lec ions a e called Umweg e lec ions (Renninge , 1937). These e lec ions a e
cha ac e ized by a b oad ene gy bandwid h ypical o elec onic di ac ion. The
Umweg e lec ions may add in ense non- esonan adia ion o he e lec ed beam, hus
c ea ing a s ong non- esonan backg ound in he Mössbaue spec um. In o de o
a oid his backg ound, one needs o escape B agg condi ions o one o mo e se s o
planes in ol ed in Umweg e lec ions while p ese ing he B agg condi ion o pu e
nuclea e lec ion. This can be achie ed by o a ing he i on bo a e c ys al a ound he
axis o he sca e ing ec o , i.e., by scanning he azimu hal angle o he c ys al. Fo
he c ys al o i on bo a e, he azimu hal dependence o he (NNN) Umweg e lec ions
has six- old symme y and 12 symme y poin s, which co espond o he and
c ys allog aphic di ec ions. Fig. 5.3.5 shows he azimu hal dependence o he
Umweg e lec ions in he icini y o he di ec ion o he (111) pu e nuclea
e lec ion. Fo his measu emen , he c ys al was hea ed well abo e i s Néel
empe a u e, whe e nuclea sca e ing was comple ely supp essed and he moni o ed
in ensi y o igina ed only om elec onic sca e ing o he Umweg e lec ions. Nea
he symme y poin , he ails o he neighbo ing Umweg e lec ions o una ely
cancel each o he ou (Fig. 5.3.5). Thus, a his pa icula azimu hal posi ion he non-
esonan backg ound in he abso p ion Mössbaue spec a aken wi h he SMS is
supp essed almos comple ely.
The in ensi y o p omp elec onic sca e ing in Umweg e lec ions is much
highe han he in ensi y o esonan sca e ing o pu e nuclea e lec ions and may be
used o ad an age. In pa icula , wi h Umweg e lec ions i is much easie o align he
sample inside he DAC o he beam and o adjus ocusing op ics.
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
99!
Figu e 5.3.5
The azimu hal dependence o Umweg e lec ions o he pu e nuclea (111) e lec ion o he i on bo a e
c ys al in he icini y o he azimu hal posi ion along he
€
[101]
c ys allog aphic di ec ion. The da a
we e aken a 353.15 K.
5.3.3.7. Ene gy modula ion.
The u nace wi h he i on bo a e c ys al is moun ed on a s anda d Mössbaue
ansduce , which mo es he c ys al in he plane pa allel o he c ys al su ace. The e
a e se e al app oaches o unde s and how he ene gy modula ion o he e lec ed
beam occu s: conside ing he SMS as a monoch oma o , one no es ha he mo ion o
he c ys al along i s su ace does no change he ene gy o he X ays. Indeed he
u ilized pu e nuclea (NNN) e lec ions a e symme ic. The e o e, he incoming and
ou going beams o m equal angles wi h he plane o mo ion and he Dopple shi in
equency o he incoming adia ion is compensa ed by exac ly he opposi e Dopple
shi o ou going adia ion. Thus he equency o he incoming adia ion is
ansmi ed o he same equency o he ou going adia ion. In he case o elec onic
sca e ing he a omic sca e ing ampli ude is cons an wi hin he ange o he Dopple
shi . Howe e , nuclei can only be exci ed by he equency equal o hei own
esonan equency. By mo ing nuclei wi h di e en eloci ies, one can e ec i ely
change he esonan equency o he nuclei. The bandwid h o he esonan ene gy is
much less han he ange o he Dopple shi s. On he o he hand, he bandwid h o
incoming adia ion is la ge han he Dopple shi ange. The e o e, nuclei being
mo ed wi h di e en eloci ies e ec i ely selec om he incoming beam he
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
100!
adia ion componen s wi h di e en equencies. In his way he ene gy modula ion
o he e lec ed adia ion is achie ed. Al e na i ely, one may conside he SMS as a
cohe en adia o o nuclea gamma ays. The 57Fe nuclei o he i on bo a e c ys al a e
exci ed by he whi e spec um o inciden adia ion o any eloci y o he mo ing
c ys al. A e exci a ion, he nuclei emi nuclea gamma adia ion. The ene gy o his
adia ion is equal o he ene gy o he nuclea esonan ansi ion. Simila o a
con en ional adioac i e sou ce, he ansla ional mo ion o he i on bo a e c ys al
case a Dopple shi and leads o he ene gy modula ion o he emi ed adia ion.
In con as o o he app oaches o ob ain he Dopple shi (Mi sui e al., 2007c),
mo ing he i on bo a e c ys al along he plane o he c ys al su ace has an impo an
ad an age: his mo ion does no cause a spa ial displacemen o he e lec ed beam.
The e o e, he ene gy modula ion does no comp omise ocusing and does no shi
he beam ela i e o he sample (Mi sui e al., 2007d).
F om he poin o iew o oscilla o y dynamics, ideally he u nace should be
loca ed a he cen e o ine ia o he mo ing ame o he ansduce . Howe e , his
equi emen is di icul o combine wi h he di ac ion geome y. To allow o
di ac ion, he u nace is a ached o he mo ing ame abo e he ansduce (Fig.
5.3.6). Unde such condi ions he ame is ob iously no balanced, i.e., i s cen e o
mass is shi ed upwa ds om he axis o he ansduce . This shi leads o he
appea ance o an ine ial o ce o que. Unde hese condi ions, he ansla ion
mo emen o he ame would be accompanied by ocking mo ions, which would
esul in a a ia ion o he angle be ween he inciden adia ion and he c ys al du ing
he c ys al mo ion. This would case (i) b oadening o he ins umen al unc ion and
(ii) oscilla ions o he in ensi y o he e lec ed beam. Ob iously, he quali y o
Mössbaue spec a wi h such a sou ce would be signi ican ly lowe ; i s ly, because o
diminished ene gy esolu ion and secondly, because he base line (i.e., he Mössbaue
spec um collec ed wi hou sample) would no longe be la .
In o de o a oid hese d awbacks, he ame is balanced by a coun e weigh
which is moun ed on he bo om side o he ame below he ansduce . The dis ances
om he axis o he ansduce o he coun e weigh and o he u nace a e equal.
Fo adjus men o he mass o he coun e weigh , he u nace is exchanged by a
silicon c ys al wi h a mass equal o he mass o he u nace. The angula s abili y o
he ame du ing i s ansla ional mo ion is e alua ed by moni o ing he in ensi y o
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
101!
he beam e lec ed by he silicon c ys al as a unc ion o he eloci y o he ame.
Wi h a p ope ly adjus ed mass o he coun e weigh , he measu ed base line is
s aigh . In o de o each he highes sensi i i y o he adjus men , he base line is
measu ed a he hal -maximum posi ion o he ocking cu e o he silicon c ys al.
A e adjus men o he coun e weigh , he es ima ed esidual a ia ion o he
angula posi ion o he ame is abou 1 µ ad. This is 10 imes smalle han he wid h
o he angula ange nea he exac B agg angle o he pu e nuclea (333) e lec ion o
he i on bo a e c ys al whe e he single-line spec um o he ins umen al unc ion can
be achie ed.
5.3.3.8. Mechanics o he SMS
In o de o ope a e he SMS, he ollowing o a ions and ansla ions o he i on
bo a e c ys al a e equi ed: i s ly, one needs a e y accu a e o a ion s age o adjus
he B agg angle o he c ys al. Secondly, in o de o a oid Umweg e lec ions, an
azimu hal o a ion is needed. Finally, he il o a ion o he c ys al is needed o keep
he e lec ed beam in he e ical plane. All hese o a ional s ages should be p ecise
and s able. In addi ion, wo ansla ions (a ho izon al and a e ical one) a e needed o
adjus he c ys al posi ion ela i e o he inciden synch o on adia ion beam.
Fig. 5.3.6 shows a schema ic iew o he mechanics o he SMS. The bo om
pa is a la ge 2-ci cle segmen (Hube 2-ci cle segmen s 5203.80), and he i on bo a e
c ys al is loca ed a he cen e o o a ion o his segmen . One ci cle is u ilized o
B agg angle adjus men s and is equipped wi h a 20:1 gea box. The accu acy o
angula posi ioning o his o a ion s age is abou 0.5 µ ad. The o he ci cle is u ilized
o il angle adjus men s; he accu acy o angula posi ioning o his s age is abou 2
µ ad. The e ical ansla ion s age (Hube z-s age 5103.A20-90) is moun ed on he 2-
ci cle segmen ; he accu acy o linea posi ioning p o ided by his s age is be e han
5 µm. On he e ical ansla ion s age a wedge is moun ed which suppo s he s age
o azimu hal o a ion. The wedge angle is equal o he B agg angle o he chosen pu e
nuclea e lec ion o he i on bo a e c ys al. The 1-ci cle goniome e (Hube 1-ci cle
goniome e 409) is used o azimu hal o a ion and is equipped wi h a 10:1 gea box.
The accu acy o angula posi ioning p o ided by his o a ion s age is abou 5 µ ad.
The Mössbaue ansduce (Wissel MVT-1000) equipped wi h he ame and he
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
102!
u nace is moun ed on he azimu hal s age. The en i e mechanics a e moun ed on a
ho izon al ansla ion s age which is used o adjus he ho izon al posi ion o he
c ys al ela i e o he beam and o mo e he en i e se up in and ou o he beam. The
accu acy o linea posi ioning p o ided by his s age is abou 5 µm.
Figu e 5.3.6
The mechanics o he Synch o on Mössbaue Sou ce: (1) u nace; (2) ansduce ; (3) 1-ci cle
goniome e o azimu hal o a ion; (4) e ical ansla ion s age; (5) 2-ci cle segmen o B agg and il
o a ions; (6) coun e weigh o he ame; (7) wedge.
5.3.4 Op ical scheme
5.3.4.1. In-line se up
Apa om he i on bo a e c ys al, he op ical scheme o he SMS includes wo
addi ional op ical elemen s. The i s op ical elemen is an addi ional monoch oma o .
The beam a e he high-hea -load monoch oma o is so in ense ha i may p oduce a
empe a u e g adien o e he su ace o he i on bo a e c ys al. In o de o a oid he
empe a u e g adien , an addi ional monoch oma o is needed whose pu pose is o
dec ease he ene gy bandwid h o he beam inciden o he i on bo a e c ys al and hus
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
103!
o dec ease he hea load. Fo his pu pose he bandwid h can be chosen o be only
mode a ely (no ex emely) small. Unde hese condi ions, he addi ional
monoch oma o can be ela i ely easily op imized o an essen ially high h oughpu .
The e o e, he inclusion o he addi ional monoch oma o essen ially does no
dec ease he in ensi y o he beam p o ided by he sou ce. The second addi ional
op ical elemen is a de lec o . The de lec o is needed in o de o make he SMS an in-
line monoch oma o . The de lec o o a es he beam inciden o he i on bo a e c ys al
in he e ical plane so ha he beam emi ed by he c ys al is di ec ed nea ly
ho izon ally, almos exac ly along he synch o on adia ion beam om he high-hea -
load monoch oma o .
The choice o he addi ional monoch oma o and de lec o depends on he
u ilized pu e nuclea e lec ion. In his pape , we desc ibe he in-line op ical schemes
o he (111) and (333) pu e nuclea e lec ions o he i on bo a e c ys al.
Figu e 5.3.7
The op ical scheme o a high-p essu e expe imen wi h DAC using he Synch o on Mössbaue Sou ce
based on he (333) pu e nuclea e lec ion. U – undula o ; HHLM – high-hea -load monoch oma o ;
CRL – compound e ac i e lens; SMS – he Synch o on Mössbaue Sou ce: HRM – high esolu ion
monoch oma o , De – Si (311) de lec o , IB – he i on bo a e c ys al inside he u nace wi h he ou
magne s and moun ed on he Mössbaue ansduce ; KBM - Ki kpa ick-Baez mi o s; DAC – diamond
an il cell; D – a alanche pho o diode de ec o .
(333) e lec ion: The pu e nuclea (333) e lec ion o i on bo a e is con enien
o he de lec o e lec ion (311) o silicon. Fo 14.4 keV adia ion, he B agg angle o
he silicon (311) e lec ion is 15.23 deg ee and nea ly ma ches ha o he i on bo a e
(333) e lec ion (15.49 deg ee). Thus, he angle be ween he beam om he SMS and
he ho izon al plane is only 0.52 deg ees. This enables he beam o each all
downs eam expe imen al hu ches and o use all sample en i onmen s a ailable a he
beamline. This de lec o can be easily combined wi h any addi ional in-line
monoch oma o . This is a signi ican ad an age o he SMS based on he (333) pu e
nuclea e lec ion o i on bo a e. In his wo k we use an in-line high- esolu ion
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
104!
monoch oma o (HRM) wi h a bandwid h o ~15meV. I is a double c ys al
monoch oma o wi h wo asymme ic (975) e lec ions o silicon c ys als.
Fig. 5.3.7 shows a ypical op ical scheme o a high-p essu e expe imen wi h a
DAC using SMS based on he (333) pu e nuclea e lec ion. The undula o (U) is he
sou ce o he synch o on adia ion beam. The high-hea -load monoch oma o
(HHLM) dec eases he ene gy bandwid h o he beam o ~2 eV, and he compound
e ac i e lens (CRL) is used o dec ease he di e gence o he beam inciden o he
SMS. The high- esolu ion monoch oma o (HRM) dec eases he ene gy bandwid h o
he beam u he o ~15 meV. The de lec o (De ) di ec s he beam o he i on bo a e
c ys al (IB) whe e he inal monoch oma iza ion wi hin ~10 neV occu s. The beam
om he SMS is ocused by he Ki kpa ick-Baez mi o s (KBM) on he sample
loca ed in he DAC. The gamma adia ion ansmi ed h ough he sample is
moni o ed by he a alanche pho o diode de ec o (D).
Figu e 5.3.8
The op ical scheme o a high-p essu e expe imen wi h a DAC using Synch o on Mössbaue Sou ce
(SMS) based on he (111) pu e nuclea e lec ion. U - undula o ; HHLM – high-hea -load
monoch oma o ; CRL – compound e ac i e lens; SMS – he Synch o on Mössbaue Sou ce: MRM
– de lec o and medium esolu ion monoch oma o wi h ~100meV bandwid h, IB – he i on bo a e
c ys al inside he u nace equipped wi h ou magne s and moun ed on he Mössbaue ansduce ; KBM
– Ki kpa ick-Baez mi o s; DAC – diamond an il cell; D – a alanche pho o diode (APD) de ec o .
(111) e lec ion: Ano he easonable op ion o he SMS is he (111) pu e
nuclea e lec ion. The angula wid h o his e lec ion is h ee imes la ge han ha
o he (333) one. The e o e, he equi emen s o he angula s abili y, c ys al quali y,
and collima ion o he inciden beam a e mo e elaxed. As desc ibed abo e, unde
o he wise equi alen condi ions his helps o achie e a na owe ins umen al unc ion
and highe in ensi y. Howe e , he B agg angle o he (111) e lec ion is h ee imes
smalle han ha o he (333) one. The e o e, he a ea illumina ed by he inciden
beam on he su ace o he i on bo a e c ys al is h ee imes la ge . Thus, highe
quali y i on bo a e c ys als a e equi ed. Fu he mo e, o he (111) e lec ion o i on
bo a e he e is no con enien e lec ion o silicon o he de lec o wi h a su icien ly
5.3!The 57Fe Synch o on Mössbaue sou ce a he ESRF!
!
!
!
105!
close B agg angle. The de lec o o he (111) e lec ion can be ob ained using wo
e lec ions: Si(422) and Si(531). The di e ence be ween he B agg angles o hese
e lec ions is 5.111 deg ees while he B agg angle o he i on bo a e (111) e lec ion is
5.108 deg ees. Thus, he beam e lec ed by he {Si(422) + Si(531)} de lec o and he
i on bo a e (111) e lec ion de ia es om he ho izon al plane by an angle less han
100 µ ad. This makes he SMS based on he (111) pu e nuclea e lec ion an in-line
monoch oma o . This de lec o also dec eases he ene gy bandwid h o ~100 meV.
Thus, i ac s bo h as a de lec o and an addi ional monoch oma o . The ad an age o
his se up is ha one may combine he SMS expe imen wi h any o he echnique such
as inelas ic sca e ing o di ac ion: he beam a e he SMS based on he (111) pu e
nuclea e lec ion ma ches he pa h o he inciden synch o on adia ion beam wi hin
an accu acy o 100 µ ad. Thus, one can mo e he SMS ou o he beam and exchange
i wi h any o he in-line monoch oma o wi hou eadjus ing he ocusing op ics o
sample posi ion.
Fig. 5.3.8 shows he ypical op ical scheme o a high-p essu e expe imen wi h
a DAC using he SMS based on he (111) pu e nuclea e lec ion. The medium
esolu ion monoch oma o (MRM) wi h silicon e lec ions Si(422) and Si(531) se es
as bo h an addi ional monoch oma o and a de lec o . All elemen s be o e and a e
he SMS a e he same as o he se up based on he (333) pu e nuclea e lec ion.
Al hough mos o he p oblems encoun e ed du ing sou ce de elopmen ha e
been sol ed, he e is one unexpec ed e ec ha occu s, namely, bending o he
ini ially la i on bo a e c ys al. The bending appea s a e abou one week o
ope a ion and only when he c ys al is hea ed up o i s Néel empe a u e. Al hough
he ini ial c ys al s a e eco e s a e a pe iod o elaxa ion, he bending e ec
comp omises sys em s abili y and equi es u he in es iga ion.
Figu e 5.3.9
The p o iles and sizes o he ocal spo in (a) ho izon al and (b) e ical di ec ions. The a ows show
he ull wid hs a he hal maximum o he p o iles.
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
112!
5.4 No spin ansi ion o e ic i on in he lowe man le
V.Po apkin1,2, C.McCammon2*, L.Dub o insky2, K.Glazy in2, A. Kan o 2, I.
Kupenko2, C. P esche 2, R. Sinmyo2, G. Smi no 3, S. Popo 3, A. Chumako 1,3 and
R.Rü e 1
1Eu opean Synch o on Radia ion Facili y, BP 220, F-38043 G enoble, F ance
2Baye isches Geoins i u , Uni e si ä Bay eu h, D-95440 Bay eu h, Ge many
3Na ional Resea ch Cen e "Ku cha o Ins i u e", 123182 Moscow, Russia
*-co esponding au ho
published in Na u e Communica ion (2013) 4 1427
5.4.1. Abs ac
I on has he abili y o adop di e en elec onic con igu a ions (spin s a es) in he
dominan lowe man le phase, magnesium silica e pe o ski e, which can signi ican ly
in luence man le p ope ies and dynamics. P e ious s udies ha e sugges ed ha e ic
i on (which cons i u es a leas hal o he i on in lowe man le silica e pe o ski e
(McCammon, 1997; F os e al., 2004;McCammon, 2005)) unde goes a high-spin o
low-spin ansi ion (Zhang e al., 2006; S ackhouse e al., 2007; Ca alli e al., 2010;
Hsu e al. 2011; Ca alli e al., 2011) which has been sugges ed o be he cause o
some obse ed seismic eloci y anomalies (Ca alli e al., 2011;Hsu e al. 2011). He e
we apply a new synch o on-based me hod o Mössbaue spec oscopy o ou
di e en composi ions o lowe man le silica e pe o ski e a high p essu e and high
empe a u e and demons a e unambiguously ha Fe3+ emains in he high-spin s a e
a condi ions h oughou he lowe man le. Ou esul s ule ou he possibili y ha
la e al he e ogenei ies o geophysical pa ame e s in he mid-lowe man le (bulk sound
speed, elec ical conduc i i y) can be explained by i on spin ansi ions in silica e
pe o ski e, and ad oca e ins ead a composi ional o he mal a ia ion.
5.4.2. In oduc ion
The s uc u e o i on-con aining magnesium aluminum silica e pe o ski e
(he ea e e e ed o as FeAlP ) con ains wo si es, a la ge 8-12 coo dina ed si e
(“A”) which is p ima ily occupied by Mg2+ and Fe2+, and a smalle oc ahed al si e
(“B”) which is dominan ly occupied by Si4+ and o a lesse ex en by Al3+. Al hough
he e is disag eemen in Fe3+ spin ansi ion p essu es epo ed by p e ious s udies,
b oadly a high-spin (HS) ( i e unpai ed d elec ons) o low-spin (LS) (one unpai ed d
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
113!
elec on) ansi ion is p edic ed when Fe3+ occupies he B-si e, while Fe3+A is
p edic ed o emain in he high-spin s a e a all p essu es h oughou he lowe man le
(Zhang e al., 2006; S ackhouse e al., 2007; Ca alli e al., 2010;Hsu e al. 2011;
Ca alli e al., 2011). E en hough Fe3+ is p edic ed o occupy exclusi ely he A-si e in
lowe man le composi ions o FeAlP (Vanpe eghem e al., 2006), a HS-LS ansi ion
o Fe3+ was epo ed o occu in FeAlP a high p essu e, po en ially due o exchange
o Fe3+ om he A- o he B-si e (Ca alli e al., 2011, Fujino e al., 2012). Up o now,
howe e , only me hods which do no p o ide an unambiguous in e p e a ion o he
da a ha e been used, so we ha e de eloped a new me hod which enables indi idual
spin and alence s a es o be iden i ied, which p o ides a clea answe o he ques ion
o whe he Fe3+ unde goes a HS-LS ansi ion in FeAlP .
5.4.3. Expe imen al Me hods
5.4.3.I. In oduc ion
De ec ing spin ansi ions o Fe3+ in FeAlP p esen s a signi ican challenge. X-
ay emission spec oscopy p o ides in o ma ion on he bulk spin numbe , bu canno
sepa a e indi idual con ibu ions. In con as ene gy domain 57Fe Mössbaue
spec oscopy gene ally enables an unambiguous esolu ion o all hype ine pa ame e s
which can be used o in e spin s a es; howe e high p essu e measu emen s using
con en ional adioac i e poin sou ces equi e ex emely long coun ing imes
(gene ally mo e han one week pe spec um). Thi d gene a ion synch o on sou ces
o e a solu ion in he o m o ime-domain Mössbaue spec oscopy (i.e., nuclea
o wa d sca e ing); howe e his me hod is no well sui ed o ma e ials wi h a la ge
numbe o componen s (such as FeAlP ) due o he non-uniqueness o i ing models.
To sol e his p oblem, we ha e de eloped an ene gy-domain synch o on Mössbaue
sou ce (SMS), which o e s a numbe o ad an ages: high lux, a beam diame e o a
ew mic ons and ze o backg ound. SMS allows o apid measu emen o ene gy-
domain Mössbaue spec a unde ex eme condi ions wi h a quali y gene ally
su icien o unambiguously decon olu e e en highly complex spec a ( u he de ails
o he me hod a e gi en in he Supplemen a y In o ma ion).
In o de o in es iga e he spin s a e o i on in lowe man le silica e pe o ski e,
we s udied ou di e en silica e pe o ski e samples: Mg0.63Fe0.37Si0.63Al0.37O3+δ
sample #1 (~80% Fe3+/ΣFe); Mg0.63Fe0.37Si0.63Al0.37O3+δ sample #2 (~70% Fe3+/ΣFe);
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
114!
Mg0.78Fe0.2Al0.05Si0.97O3+δ (~50% Fe3+/ΣFe); Mg0.94Fe0.06SiO3+δ (~20% Fe3+/ΣFe). We
collec ed SMS spec a a oom empe a u e and p essu es up o 122 GPa using
diamond an il cells, wi h o wi hou lase annealing o he samples. De ails o sample
syn hesis and expe imen me hodology a e gi en in he Supplemen a y In o ma ion.
5.4.3.II. Me hods
57Fe Synch o on Mössbaue sou ce (SMS) spec a we e collec ed on beamline
ID18 a he Eu opean Synch o on Radia ion Facili y (ESRF) du ing ope a ion in
uni o m mode (7/8 illing) wi h he beam ocused o oughly 10×10 µm2 using a
Ki kpa ick–Baez mi o . Fu he de ails o he SMS me hod a e gi en in he
Supplemen a y In o ma ion. The eloci y scale was calib a ed ela i e o α-Fe oil,
and spec a we e collec ed o e 10-60 min each. Spec a we e i ed using he
p og am MossA (P esche e al., 2012). The dimensionless e ec i e Mössbaue
hicknesses we e app oxima ely 40 and 20 o he Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e
samples #1 and #2, and 8 and 3 o he Mg0.78Fe0.2Al0.05Si0.97O3 and Mg0.94Fe0.06SiO3
pe o ski e samples, espec i ely. Fu he de ails o sample syn hesis and expe imen al
me hodology a e gi en in he Supplemen a y In o ma ion.
5.4.4. Resul s and Discussion
SMS spec a o Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e, which con ains i on
dominan ly as Fe3+, a e ex emely well esol ed (Fig. 5.4.1). We i he da a o h ee
quad upole double s, one assigned o Fe3+ and wo assigned o Fe2+ (high QS and low
QS) based on hei cen e shi s. Visually he e appea s o be no change o he Fe3+
double o e he en i e p essu e ange (Fig. 5.4.1 and Supplemen a y Fig. 5.4.S4).
The hype ine pa ame e s (cen e shi [CS] and quad upole spli ing [QS]) in all
samples s udied a e he same wi hin expe imen al e o , and he e is no change in
Fe3+/ΣFe o indi idual samples o e he en i e p essu e ange o he expe imen (Fig.
5.4.2). The hype ine pa ame e s o he Fe3+ double a e consis en wi h he high-spin
s a e (Gü lich e al., 2011), and hei smoo h a ia ion wi h p essu e indica es ha
Fe3+ does no unde go any spin ansi ions wi hin he en i e p essu e ange. No ably
he QS alue epo ed o low-spin Fe3+ om bo h expe imen al (Ca alli e al., 2010;
Ca alli e al., 2011) and heo e ical (Hsu e al., 2011) s udies ( ed dashed line, Fig.
5.4.2a) is mo e han wice ou obse ed alues.
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
115!
The hype ine pa ame e s o he low QS Fe2+ double (blue double in Fig. 5.4.1)
co esponds o he high-spin s a e (McCammon e al., 2008), while he double wi h
high quad upole spli ing (black double in Fig. 1) co esponds ei he o in e media e-
spin (IS) Fe2+ (McCammon e al., 2008) o a dis o ion o he si e occupied by high-
spin Fe2+ (Hsu e al., 2011). I espec i e o he in e p e a ion o he Fe2+ spin s a e,
ou conclusions ega ding he absence o a spin ansi ion in Fe3+ emain he same.
Figu e. 5.4.1:
SMS spec a o Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #1 a oom
empe a u e showing hei e olu ion wi h p essu e (a) 2.5 GPa; (b) 37.9 GPa; (c)
67 GPa; (d) 93 GPa. The da a we e i o one Fe3+ double ( ed) and wo Fe2+ double s
(blue and black), and he i esidual is shown below each spec um. A ea asymme y
is due o p e e ed o ien a ion o he sample, and he eloci y scale is gi en ela i e o
α-i on.
Single-c ys al X- ay e inemen s (Glazy in e al., 2011) and c ys al chemical
calcula ions (see Supplemen a y In o ma ion) ha e shown ha Fe3+ occupies only he
A-si e o he pe o ski e s uc u e in ou samples; hence ou da a demons a e ha
Fe3+A does no unde go a high-spin o low-spin ansi ion up o a leas 122 GPa. This
conclusion is in ag eemen wi h esul s om expe imen al (Ca alli e al., 2010; Ca alli
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
116!
e al., 2011) and heo e ical (Zhang e al., 2006; S ackhouse e al., 2007;Hsu e al.
2011) s udies. Howe e se e al s udies (Ca alli e al., 2010; Ca alli e al., 2011;
Fujino e al., 2012) ha e sugges ed ha he educed olume o LS Fe3+B could lead o
a edis ibu ion o Fe3+ om he A- o he B-si e in he pe o ski e s uc u e wi hin he
lowe man le. To es his hypo hesis, we lase annealed ou samples a a numbe o
p essu es and collec ed SMS spec a bo h du ing and a e hea ing. Visually SMS
spec a collec ed a e lase annealing showed no change o hose aken be o e hea ing
(Supplemen a y Fig. 5.4.S5) and hype ine pa ame e s emain unchanged (s a
symbols; Fig. 5.4.2). We he e o e conclude ha he e is no exchange o Fe3+ be ween
he A- and he B-si e a lowe man le p essu es and empe a u es, in ag eemen wi h
he esul s o a complemen a y s udy using high-p essu e single-c ys al X- ay
di ac ion wi h lase hea ing unde aken on he same composi ion as wo o ou
samples (Glazy in e al., 2011).
The coupled subs i u ion Mg2+A + Si4+B ↔ Fe3+A + Al3+B is conside ed o be he
dominan mechanism o inco po a ion o Fe3+ in o FeAlP in he lowe man le
(Vanpe eghem e al., 2006; B odhol e al., 2000; Saikia e al., 2009); hence Fe3+ is
expec ed o only occupy he A-si e in he pe o ski e s uc u e o lowe man le
composi ions. Elec ical conduc i i y measu emen s o single-phase Mg0.9Fe0.1SiO3
pe o ski e and a py oli e man le assemblage all show a dec ease in conduc i i y
abo e 50 GPa (Oh a e al., 2008; Oh a e al., 2010a; Oh a e al., 2010b), compa able
o he elec ical conduc i i y dec ease caused by HS-LS spin c osso e o Fe2+ in
(Mg,Fe)O (Yoshino e al.,2011). The d op in silica e pe o ski e conduc i i y has been
a ibu ed o a HS-LS ansi ion o Fe3+ (Oh a e al., 2008; Oh a e al., 2010a; Oh a e
al., 2010b); howe e he small p opo ion o Fe3+ in he Mg0.9Fe0.1SiO3 pe o ski e
sample (~ 10%), he esul s om his s udy and a companion s udy (Glazy in e al.,
2012) ha he e is no shi o Fe3+ om he A-si e o he B-si e a lowe man le
condi ions, and he o e whelming e idence ha no Fe3+ spin ansi ion occu s when
Fe3+ occupies he A-si e in FeAlP implies ha he d op in elec ical conduc i i y
canno be a ibu ed o a spin ansi ion o Fe3+. A mo e likely explana ion is he
p essu e-induced HS-IS ansi ion o Fe2+ in FeAlP obse ed using X- ay emission
spec oscopy (Bad o e al., 2004) and nuclea esonance me hods (McCammon e al.,
2004; Lin e al., 2004).
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
117!
Figu e 5.4.2:
P essu e a ia ion o FeAlP hype ine pa ame e s (cen e shi [CS] and
quad upole spli ing [QS]) (a) high-spin Fe3+; (b) high QS Fe2+. Sample da a a e
indica ed as ollows: Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #1 ( ed and blue
ci cles); Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #2 (o ange and blue squa es);
Mg0.78Fe0.2Al0.05Si0.97O3 pe o ski e (b own and da k blue ci cles); Mg0.94Fe0.06SiO3
pe o ski e (pu ple and g een ci cles). The expec ed quad upole spli ing o low-spin
Fe3+ is shown as a ho izon al ed dashed line. (c) P essu e a ia ion o Fe3+/ΣFe as
de e mined om he ela i e a eas: pe o ski e sample #1 ( ed ci cles);
Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #2 (o ange squa es);
Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e (b own ci cles); Mg0.94Fe0.06SiO3 pe o ski e
(pu ple ci cles). Ho izon al lines a e guides o he eye. Values ob ained a e lase
hea ing o Mg0.63Fe0.37Si0.63Al0.37O3 pe o ski e sample #2 and Mg0.78Fe0.2Al0.05Si0.97O3
pe o ski e a e indica ed by g ey s a s. In all cases, pa ame e s emained unchanged
om hose be o e hea ing wi hin expe imen al e o , demons a ing ha he e is no
high-spin o low-spin ansi ion in Fe3+ a he p essu e- empe a u e condi ions o he
lowe man le.
Elec omagne ic induc ion da a o e an impo an complemen o seismic da a
wi h espec o lowe man le modeling, since he o me a e mo e sensi i e o
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
118!
empe a u e and i on con en , while he la e be e cons ain he mine alogy
(Ve hoe en e al, 2009). Labo a o y elec ical conduc i i y da a o he ele an
mine als a e an impo an componen o he app oach, and i is c ucial ha hey a e
ep esen a i e o beha io in he dep h ange o in e es . Ea lie labo a o y-based
elec ical conduc i i y models inco po a ing FeAlP conduc i i y alues a he op o
he lowe man le epo ed ha elec ical conduc i i y in he lowe man le is ela i ely
homogeneous (Xu e al, 2000), while a mo e ecen model inco po a ing he la es
expe imen al conduc i i y alues ound a dec ease in elec ical conduc i i y be ween
1300 and 1800 km dep h (Oh a e al., 2010). Indeed a ecen one-dimensional
in e sion o 32 yea s o geomagne ic da a shows a dec ease in conduc i i y below
1400 km, and a h ee-dimensional in e sion o he same da ase shows a ia ions o
mo e han one o de o magni ude in he mid pa o he lowe man le, some o which
a e co ela ed wi h as and slow egions o seismic omog aphy models (Ta i s e al.,
2010). Based on he esul s o his s udy we can ule ou he possibili y ha seismic
anomalies in he mid-lowe man le egion a e caused by a Fe3+ spin ansi ion in
FeAlP , and comp essibili y s udies ha e al eady shown ha he e is no p essu ed-
induced change o elas ici y in his dep h ange o Fe2+- ich silica e pe o ski e
(McCammon e al., 2008; Saikia e al., 2009; Lundin e al., 2008). Such a ia ions in
elec ical conduc i i y a e mo e likely caused by composi ional o empe a u e
he e ogenei y, bu may also be due o mino phases such as wa e o ca bona es
(Ta i s e al., 2010). A inal esolu ion o he con o e sy su ounding i on spin
ansi ions in FeAlP h ough Mössbaue spec oscopy ( his wo k) and single c ys al
X- ay di ac ion (Glazy in e al., 2012) p o ides an impo an s ep in de eloping
quan i a i e models o join in e sion o elec omagne ic and seismic da a ha place
imp o ed cons ain s on lowe man le mine alogy, composi ion and he mal s a e.
Acknowledgemen s
We acknowledge he Eu opean Synch o on Radia ion Facili y o p o ision o
synch o on adia ion acili ies (ID18) and we would like o hank Jean-Philippe Celse
o addi ional echnical assis ance and Ilya Se guee o help wi h da a analysis
so wa e. The p ojec was pa ly suppo ed by unds om he Ge man Science
Founda ion (DFG) P io i y P og amme SPP1236, he PROCOPE exchange
p og amme, and he Ge man Fede al Minis y o Educa ion and Resea ch (BMBF).
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
119!
5.4.5. Supplemen a y In o ma ion
5.4.5.I. Synch o on Mössbaue sou ce
Ene gy-domain Mössbaue spec oscopy p o ides di ec access o hype ine
pa ame e s om which i on alence and spin s a e can be de e mined. I has a la ge
ad an age compa ed o nuclea o wa d sca e ing (NFS) and X- ay emission
spec oscopy (XES) in ha each con ibu ion occupies a nea ly unique ene gy ange
which gene ally enables an unambiguous esolu ion o all componen s. In con as ,
NFS and XES p o ide bulk in o ma ion which combines he e ec s o all
con ibu ions. This ad an age o Mössbaue spec oscopy makes i an ideal ool o
s udy sys ems whe e i on exis s in di e en spin and alence s a es, and di e en
c ys allog aphic posi ions.
Con en ional ene gy-domain Mössbaue spec oscopy uses adioac i e sou ces
whose b illiance is e y low; hence high-p essu e s udies using diamond an il cells
(DACs) equi e long measu ing imes since beam ocusing is no possible in a
labo a o y se ing. This educes he quali y o he esul s due o p essu e g adien s,
inc eased backg ound, and es ic s he maximum p essu e a which measu emen s a e
possible. These p oblems can be sol ed by combining he ad an ages o high-
b illiance hi d gene a ion synch o ons (high lux, ex eme ocusing o he beam)
wi h he esolu ion o ene gy-domain Mössbaue spec oscopy o p oduce a
synch o on sou ce o Mössbaue adia ion.
A synch o on Mössbaue sou ce (SMS) p o ides a high-b illiance beam o
synch o on adia ion wi h an ene gy bandwid h o ~15 neV. In addi ion, SMS has
se e al u he p ope ies ha a adioac i e sou ce does no possess. The SMS beam is
pola ized up o 99%, and i consis s 100% o ecoilless adia ion wi h ze o
backg ound. These p ope ies enable apid and p ecise measu emen s o Mössbaue
spec a o samples unde ex eme condi ions.
The possibili y o de elop such a sou ce was i s demons a ed a he nuclea
esonance beamline ID18 (Rü e & Chumako 1996) a he Eu opean Synch o on
Radia ion Facili y (ESRF) in 1997 (Smi no e al., 1997).
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
120!
Figu e 5.4.S1
Op ical scheme o SMS expe imen o high-p essu e s udies. U - undula o ; HHLM
– high-hea load monoch oma o ; HRM – high esolu ion monoch oma o ; SMS: Si –
Si (311) c ys al; IB – enclosed u nace (moun ed on a eloci y ansduce ) wi h he
i on bo a e c ys al inside; KBM - Ki kpa ick-Baez mi o s; DAC – diamond an il
cell; D – a alanche pho o diode de ec o .
Fig. 5.4.S1 shows he expe imen al se up. The synch o on beam is emi ed he
undula o (U), passes h ough high-hea load monoch oma o (HHLM) wi h ene gy
bandwid h o ~2 eV, and a high- esolu ion monoch oma o (HRM) wi h ene gy
bandwid h o ~15 meV. This cascade o monoch oma o s is equi ed o dec ease he
hea load on he i on bo a e c ys al in o de o p e en a empe a u e g adien . The
SMS i sel is a monoch oma o composed o wo c ys als. The key elemen o he
sou ce is an i on bo a e c ys al (FeBO3) en iched in he 57Fe iso ope. The c ys al is
used in (333) pu e nuclea e lec ion. Fo such e lec ion, elec onic di ac ion is
o bidden, while nuclea di ac ion is allowed. The c ys al is placed in an ex e nal
magne ic ield o ~110 Oe and hea ed close o i s Néel empe a u e o ~75.65°C.
Unde hese condi ions he hype ine magne ic s uc u e collapses o a single line. The
c ys al he e o e e lec s synch o on adia ion wi hin an ene gy bandwid h o ~20
neV. Fu he de ails o he physics o he p ocess a e gi en in (Smi no e al., 2011).
The u nace is enclosed and special a angemen s o he magne s a e used in o de o
ensu e homogenous hea ing and magne ic ield o s able ope a ion. The u nace wi h
he c ys al inside is moun ed on a eloci y ansduce in o de o modula e he ene gy.
The pu pose o he Si (311) c ys al is o di ec he beam exi ing he SMS in a di ec ion
pa allel o he incoming synch o on beam, which allows o a mo e con enien
ins alla ion o di e en ypes o sample en i onmen such as DAC, c yos a , u nace,
e c. The SMS is ollowed by a ocusing mi o (KBM), which allows ocusing o he
5.4!No spin ansi ion in e ic i on in he lowe man le!
!
!
!
121!
synch o on beam o a oughly 10 x 10 µm spo size on he sample loca ed in he
DAC. The ansmi ed γ-quan a a e moni o ed by an a alanche pho o diode de ec o .
Du ing he SMS expe imen he linewid h o he sou ce is con olled be o e and
a e each sample measu emen using K2Mg57Fe(CN)6, whose Mössbaue spec um
consis s o a single line (Fig. 5.4.S2). The eloci y scale is calib a ed using 25 µm
hick na u al α-i on oil (Fig. 5.4.S3).
Figu e 5.4.S2
SMS spec um o K2Mg57Fe(CN)6 used o moni o he sou ce linewid h.
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
128!
5.5. Po able Double-Sided Lase -Hea ing Sys em o
Ene gy-Domain Mössbaue Spec oscopy a Synch o on
and Single C ys al Di ac ion Expe imen s wi h
Diamond An il Cells
I. Kupenko1,*, L. Dub o insky1, N. Dub o inskaia2, C. McCammon1, K. Glazy in1, E.
Byko a1,2, T. Bo a Balla an1, R. Sinmyo1, A.I. Chumako 3, V. Po apkin1,3, A.
Kan o 1,3, R. Rü e 3, M. Han land3, W. C ich on3, M. Me lini4
1Baye isches Geoins i u , Uni e si ä Bay eu h, D-95440 Bay eu h, Ge many
2Ma e ial Physics and Technology a Ex eme Condi ions, Labo a o y o
C ys allog aphy, Uni e si y o Bay eu h, D-95440 Bay eu h, Ge many
3Eu opean Synch o on Radia ion Facili y, BP 220, F-38043 G enoble, F ance
4Dipa imen o di Scienze della Te a, Uni e si à degli S udi di Milano, Via Bo icelli
23, 20133 Milano (I aly)
*-co esponding au ho
published in Re iew Scien i ic Ins umen s (2012) 83 124501
5.5.1. Abs ac
Diamond an il cell (DAC) echnique coupled wi h lase hea ing is a majo me hod o
s udy ma e ials a s a ic mul imegaba p essu es and a high empe a u es. Recen
p og ess in expe imen al echniques, especially in high-p essu e single c ys al X- ay
di ac ion, equi es po able lase hea ing sys ems, which may hea and mo e he
DAC du ing da a collec ion. We ha e de eloped a double-sided lase hea ing sys em
o DACs which can be moun ed on a ~0.1 m2 a ea and has a weigh o ~12 kg. The
sys em is easily ans e able be ween di e en in-house o synch o on acili ies and
can be assembled and se up du ing a ew hou s. The sys em was success ully es ed
a he ID09a and ID18 beamlines o he Eu opean Synch o on Radia ion Facili y
(ESRF). We demons a e applica ion o he sys em on example o a single c ys al X-
ay di ac ion in es iga ion o (Mg0.87,Fe3+0.09,Fe2+0.04)(Si0.89,Al0.11)O3 silica e
pe o ski e and ene gy-domain Mössbaue spec oscopy s udy o (Mg0.8Fe0.2)O
e ope iclase using ecen ly de eloped Synch o on Mössbaue Sou ce a high
p essu es and empe a u es.
5.5.2. In oduc ion
S udies o ma e ials a ex eme p essu e and empe a u e condi ions a e e y
impo an o na u al sciences such as physics (Lin e al., 2004; G ego yanz e al.,
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
129!
2005; Dewaele e al., 2008), chemis y and ma e ial syn hesis (Ze e al., 2009;
Sch öde e al., 2011; Palyano e al., 2011), and especially o geosciences, because
hey gi e an oppo uni y o simula e condi ions o he deep Ea h’s in e io
(Mikhaylushkin e al., 2007; Kolesniko e al., 2009; Wu e al., 2009).
The diamond an il cell echnique is a e y use ul and easy- o-use ool o
in es iga ions unde high p essu e. I was ini ia ed in he la e 1950s, and by now has
become he mos success ul me hod o p essu e gene a ion in a mul i megaba
p essu e ange (E eme s, 1996; Dub o inskaia e al., 2010). The lase hea ing in
DACs was i s in oduced by Ming and Basse (Ming & Basse , 1974) and is well
desc ibed in he li e a u e (Boehle , 2000). The e a e nume ous acili ies, including
specialized beamlines a he hi d-gene a ion synch o ons (Shen e al., 2001; Schul z
e al., 2005; Hi ose, 2006; P akapenka e al., 2008), whe e he DAC lase -hea ing
echnique is coupled wi h di e en analy ical me hods o in si u in es iga ions such
as he Raman spec oscopy o X- ay di ac ion (Boehle e al., 2008; San o o e al.,
2005). Howe e , un il ecen imes all hese lase -hea ing sys ems we e linked o
ce ain equipmen o o a beamline.
Mode n scien i ic challenges demand highe lexibili y in capabili ies o
esea ch echniques including lase hea ing in DACs. As a esul , since 2009 po able
DAC lase -hea ing sys ems ha can be easily mo ed be ween di e en analy ical
acili ies, including ans e om in-house o a synch o on o be ween synch o on
beamlines, began o eme ge (Boehle e al., 2009; Dub o insky e al., 2009). S ill
he e was no a po able double-sided lase -hea ing sys em ha would mee all
scien i ic needs. The sys em desc ibed by Boehle e al., (2009) allows double-sided
lase hea ing bu equi es a a he la ge space (0.4 m2) o he suppo . I has ne e
been es ed o applica ions in single c ys al s udies. Mo eo e , he design o his
sys em (Boehle e al., 2009) has a d awback – simul aneous empe a u e and X- ay
measu emen s a e no possible because op ical componen s o collec ing he mal
adia ion block and/o shadow he X- ay p ima y beam o he sca e ed signal. The
sys em desc ibed by Dub o insky e al., (2009) was success ully used in ou ine in
house expe imen s (Gu e al., 2011; Pa akhonskiy e al., 2011), he in si u synch o on
X- ay abso p ion and di ac ion in es iga ions (Na ygina e al., 2011), and (wi h
some modi ica ions) in synch o on single-c ys al X- ay di ac ion expe imen s
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
130!
(Dub o insky e al., 2010). I s main disad an age was one-sided lase -hea ing and,
hus, limi ed applica ions o only op ically hin samples.
He e we epo he de elopmen o a double-sided lase hea ing sys em o
single-c ys al di ac ion s udies and o he expe imen s, which equi e mo emen o
he DAC du ing a da a collec ion. The new sys em p o ides wi h a mo e
homogeneous empe a u e dis ibu ion wi hin he sample (e.g. i allows highe
empe a u es wi hou possible des uc ions due o he mal s esses), and pe mi s he
use o hicke samples ( hus p o iding an oppo uni y o use he sys em in wo k wi h
ma e ials which a e di icul o hea ).
5.5.3. Design o he lase -hea ing sys em
The sys em consis s o wo majo componen s – he sou ces o lase ligh and
wo uni e sal lase -hea ing heads (UniHeads) (Fig. 5.5.1). The UniHeads a e based on
he ini e cu ing lase head (P eci ec KG) and hei unc ions in he po able lase
sys em a e o ocus incoming lase beams on he sample wi hin he DAC, o p o ide
wi h a high magni ica ion imaging o he sample in he DAC wi h coaxial
illumina ion, and o gi e access o he mul iwa eleng h spec o adiome y o
empe a u e measu emen s (Shen e al., 2010).
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
131!
Fig 5.5.1
Schema ic diag am o he op ical componen s o UniHeads.
5.5.3.1. Lase
As a lase sou ce we ha e used wo SPI100 modula ed ibe lase s wi h
maximum ou pu powe o 100 W. The weigh o he bo h lase s is 40 kg and he
exci a ion wa eleng h is 1071 nm. An addi ionally in eg a ed ed lase sou ce se es
as an aid o ini ial beam a ge ing. The ou pu o he SPI lase s has a Gaussian powe
dis ibu ion wi h a diame e o ~3 mm a 1/e2. The ou pu powe le el can be
moni o ed and con olled by an ex e nal analog in e ace. The ibe lase s can be
ope a ed in a con inuous o pulse mode wi h he equency up o 100 kHz and
minimum ime o he pulses o less han 10 µs. The lase s equi e no wa e -cooling
and use only a s anda d elec ical plug.
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
132!
5.5.3.2. Focusing op ics and beam a ge ing
The ou pu 1071 nm lase ligh is collec ed by a π-shape (MolTech GmbH),
90° bended using a specially coa ed beam spli e (Beam Spli e 1 in Fig. 5.5.1: he
e lec i i y abo e 1050 nm is >99.99%, i is anspa en in he 400-900 nm
wa eleng h ange), and ocused by he UniHead se o lenses wi h an 80-mm wo king
dis ance. The o iginal 3 mm in diame e lase beam wi h he Gaussian in ensi y
dis ibu ion is con e ed in o a la op beam wi h up o ~50 µm FWHM (P akapenka
e al., 2008).
In o de o di ec he lase beam o he sample and make i coaxial wi h he
inciden X- ay beam, we employ ca bon mi o s moun ed a app oxima ely 45° angle
o he UniHeads’ axis (Fig. 5.5.1, 2). The ca bon mi o s a e made ou o polished
glassy ca bon pla es o 1 o 2 mm hickness. They a e coa ed a he op ical acili y
Lab o he Uni e si y o Bay eu h by a 100 nm hick sil e ilm and, on op o he
sil e laye , by a 5 nm hick silica ilm. In case o expe imen s using he X- ay
di ac ion, he ca bon mi o o he downs eam UniHead is mounded unde an angle
which is sligh ly below 45° (usually abou 41-42 o
) so ha and he mi o appea s
ou side o he pa h o he di ec X- ay beam. Thus, we a oid he undesi able
di ac ion o X- ay on he ca bon mi o . The X- ay sca e ing om glassy ca bon o
he ups eam mi o is blocked by he DAC’s body.
5.5.3.3. Illumina ion, obse a ion and he spec o adiome y module
The illumina ion o he sample is ealized due o he buil -in LED in he
UniHead (Fig. 5.5.1). The LED is con olled by an ex e nal powe supply. Fo
obse a ion o he sample in he DAC he high- esolu ion µEYETM CCD came as a e
used. To p e en o e sa u a ion o he came as by he e lec ed lase ligh , KG3 IR
il e s (SCHOTT GmbH) a e exploi ed. The so wa e o he came a allows enla ging a
pa o he obse ed a ea and acing a ce ain posi ion in he image. The UniHead
equipped wi h a modi ied op ical ou pu module o empe a u e measu emen s
consis s o a µEYETM CCD came a, a ocusing lens p ojec ing a cen al pa (o abou
10 µm in diame e ) o he image o he hea ed spo on o he end o he op ical ibe ,
and a beam-spli e cube (50/50, Edmund Scien i ic Inc) (Pippinge e al., 2011). The
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
133!
beam-spli e cube is mo able and can be ixed in wo posi ions. I i is ixed ou side
o he op ical pa h, hen all he ligh adia ed by he hea ed spo ansmi s in o he
op ical ibe . The cube posi ion wi hin he op ical pa h allows 50% o he ligh o be
ans e ed o he CCD came a o isual obse a ions. In ou expe imen s we used an
Ocean Op ics QE65000 spec ome e o Ac on SP2300 spec ome e (P ince on
Ins umen s) wi h he PIXIS400 CCD de ec o . To p e en he incidence o he lase
ligh on o he de ec o , 1064 nm no ch il e s (Edmund Op ics Inc.) a e used. The
empe a u e measu emen s a e pe o med by he spec o adiome y me hod (Shen e
al., 2010) i.e. by i ing he mal adia ion signals in a gi en wa eleng h (usually 600
o 850 µm) ange o he Planck adia ion unc ion. The so wa e de eloped by A.
Kan o and I. Kan o allows on-line moni o ing o he empe a u e. The sys em
esponse spec a a e calib a ed by mel ing o pla inum (Pippinge e al., 2011).
5.5.3.4. Sys em alignmen
All pa s, he UniHeads, a 3D s age o a DAC holde , and he holde wi h he
DAC a e moun ed on a common aluminum pla e (0.35x0.30 m2; he hickness o 0.15
m). In he expe imen s desc ibed below he ups eam UniHead was se up ho izon ally
and he downs eam UniHead was se up e ically (Fig. 5.5.2). The downs eam
UniHead is moun ed igidly on he pla e and has no deg ees o eedom. The sample
in he DAC should be aligned wi h espec o his UniHead using he h ee-
dimensional s age. The ed alignmen lase o he SPI100 can be used o simpli y he
p ocess. The ups eam UniHead is moun ed on i s own h ee dimensional s age and
should be aligned o he sample wi hin he DAC a e wa ds. Focusing o he lase
beams on he sample is p oceeded due o adjus able lenses o he π-shape s
(P akapenka e al., 2008). The las pa o he p ocedu e is he alignmen o he
spec ome e s. Fo his pu pose he end o he op ical ibe coming o he spec ome e
is connec ed o he isible (in ou expe imen s 532 nm) diode lase whose spo is
ocused on he sample. By means o he adjus able sc ews (Fig. 5.5.2c) o he
spec oscopic module he spo is placed exac ly a he posi ion o he cen e o he
hea ed a ea. Acco ding o ou expe ience, he moun ing o he sys em and i s
alignmen equi es wo o h ee hou s.
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
134!
a)
b)
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
135!
c)
Fig. 5.5.2
The double-side po able lase -hea ing sys em moun ed o expe imen s a ID09a (a,b)
and ID18 beam lines a ESRF (c). 1, holde wi h DAC; 2, ca bon mi o s; 3, ocusing
op ics; 4, π-shape ; 5, beam-spli e cube; 6, CCD came a; 7, LED; 8, h ee
dimensional s ages; 9, adjus able sc ews o spec ome e ocusing; 10, MAR555
de ec o .
5.5.4. Examples o applica ion o he po able lase -hea ing sys em
5.5.4.1. Mössbaue spec oscopy in es iga ion o (Mg0.8Fe0.2)O
Hype ine pa ame e s de e mined om Mössbaue spec a s ongly depend on
p essu e and empe a u e. Thus he ene gy domain 57Fe Mössbaue spec oscopy is
no only one o he bes me hods o s udy i on’s alence and spin s a e a high
p essu e, bu he p obe o choice o es empe a u e homogenei y in lase -hea ed
DACs. Howe e , con en ional ene gy- esol ed Mössbaue spec oscopy u ilizes
adioac i e sou ces wi h low b illiance. The e o e high-p essu e Mössbaue s udies
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
136!
equi e oo long collec ion ime o be e ec i ely applied o in es iga ions in a lase -
hea ed DAC. Recen ly de eloping o a Synch o on Mössbaue Sou ce (SMS) a he
Nuclea Resonance beamline (ID 18) (Chap e 5.2) sol ed his p oblem. The SMS
allows one o pe o m ene gy-domain Mössbaue measu emen s using synch o on as
a sou ce o gamma adia ion. In con as o adioac i e sou ces, he beam emi ed by
he SMS is nea ly ully esonan , has high b illiance and can be ocused o a 10x5 µm2
spo size. Tha opens a possibili y o as and obus measu emen s in DACs a high
p essu es and, i coupled wi h lase hea ing, a high empe a u es. In o de o es he
double-sided lase hea ing sys em wi h he SMS we selec ed (Mg0.8Fe0.2)O
( e ope iclase, Fp20), since i s Mössbaue spec a a high p essu e and empe a u e
a e al eady known om ou p e ious wo k using an ex e nally hea ed diamond an il
cell and a con en ional in-house Mössbaue sou ce (Kan o e al., 2009). Pelle s o
Fp20 o 30 o 40 µm in diame e wi h he hickness o abou 15 µm we e loaded along
wi h a Ne medium (Ku noso e al., 2008) in o a p essu e chambe ( he ini ial
diame e o 100 µm and he heigh o abou 30 µm) made in a Re gaske . The sample
was comp essed using a symme ic pis on-cylinde ype DAC p oduced a BGI. All
measu emen s we e ca ied ou a ID 18 using he SMS, and each spec um ook
oughly 10 minu es o collec . The Mössbaue spec um o he high-spin phase (P <
50 GPa) consis s o a single quad upole double a ising om oc ahed ally coo dina ed
Fe2+ (Fig. 5.5.3a) (Kan o e al., 2006). Upon lase hea ing om only one side (o
du ing double-sided hea ing when he X- ay beam and he lase spo a e no well
aligned), he spec um becomes asymme ic (Fig. 5.5.3b). Since we know om
p e ious expe imen s on (Mg0.8Fe0.2)O ha empe a u e causes a dec ease in
quad upole spli ing, we in e p e he spec um in Fig. 5.5.3b as indica ing a
empe a u e g adien in he sample be ween he ela i ely cold ma e ial (g ey double )
and a ho e egion (black single ), co esponding o empe a u es o 450 K and 1000
K, espec i ely. Du ing double-sided hea ing wi h he co ec alignmen o he X- ay
beam and lase spo , he Mössbaue spec um becomes mo e symme ic (Fig. 5.5.3c)
a he empe a u e o 1300 K. We ound ha empe a u e g adien s also a ise om he
double-sided lase hea ing o hick samples (abo e abou 20 µm), whe e he in e io
o he sample emains colde han he su ace.
5.5!Po able Double-Sided Lase -Hea ing Sys em o SMS and Single!
C ys al Di ac ion Expe imen s wi h DACs!
!
!
137!
Fig. 5.5.3
Synch o on Mössbaue sou ce spec a o Fe0.2Mg0.8O a 29 GPa: (a) oom
empe a u e; (b) lase hea ing wi h misaligned beam; (c) double-sided lase hea ing
wi h good alignmen . The empe a u es de e mined om he cen e shi a e 450 K
(g ey double in (b)) and 1000 K (black single in (b)) and 1300 K (single in (c)). The
i esiduals a e shown below each spec um.
5.5.4.2. X- ay single c ys al di ac ion o Fe,Al ich silica e pe o ski e in
double-side lase hea ed DAC
I on and aluminum bea ing magnesium silica e pe o ski e (Mg,Fe,Si,Al)O3 is
likely o be he main componen o he Ea h lowe man le. I on and aluminum could
signi ican ly a ec p ope ies o silica e pe o ski e, especially due o elec onic
ansi ions in Fe2+ and Fe3+. Recen single c ys al di ac ion s udies(Dub o insky e
al., 2010; Bo a-Balla an e al., 2012) e eal ha e ous and e ic silica e
pe o ski es do no demons a e any sign o i egula changes in he beha io o he
mola olume, la ice pa ame e s, he mean bond dis ances o he (Si,Al)O6 oc ahed a
and (Mg,Fe)O8-polyhed a which could be ela ed o he high-spin – low-spin (HS-LS)
c osso e . Howe e , he e is one hypo hesis which has ne e been es ed be o e based
on single c ys al X- ay di ac ion da a: acco ding o Ca alli e al.,(2010) high
6.!Bibliog aphy!
!
!
144!
E eme s, M.I. (1996) High P essu e Expe imen al Me hods. 408 (Ox o d Science:
Ox o d, 1996)
Fa ugia, L.J. (1999) WinGX sui e o small-molecule single-c ys al c ys allog aphy.
Jou nal o Applied C ys allog aphy, 32, 837-838.
Fei, Y., A. Ricolleau, M. F ank, K. Mibe, G. Shen, and V. P akapenka (2007), High-
P essu e Geoscience Special Fea u e: Towa d an in e nally consis en p essu e
scale. P oceedings o he Na ional Academy o Sciences, 104, 9182-9186.
Fo man, R. A., G. J. Pie ma ini, J. D. Ba ne , and S. Block (1972), P essu e
Measu emen Made by he U iliza ion o Ruby Sha p-Line Luminescence.
Science, 176, 284-285
F os D.J., Liebske C., Langenho s F., McCammon C.A., T ønnes G.R, and Rubie
D.C. (2004) Expe imen al e idence o he exis ence o i on- ich me al in he
Ea h's lowe man le. Na u e 428, 409-412
F os , D.J. (2008) The Uppe Man le and T ansi ion Zone. Elem. Mag. 4, 171-176
Fujino, K. D. Nishio-Hamane, Y. Se o, N. Sa a, T. Nagai, T. Shinmei, T. I i une, H.
Ishii, N. Hi aoka, F. Q. Cai, K.-D. (2012) Tsuei. Spin ansi ion o e ic i on in
Al-bea ing Mg–pe o ski e up o 200 GPa and i s implica ion o he lowe man le.
Ea h Plane . Sci. Le ., 317-318, 407-412 .
Fy e, W.S. (1960) The possibili y o d-elec on coupling in oli ine a high p essu es.
Geochim. Cosmochim. Ac a, 19, 141–143
Ga g, A. B., V. Vijayakuma , and B. K. Godwal (2004) Elec ical esis ance
measu emen s in a diamond an il cell o 40 GPa on y e bium. Re . Sci. Ins um.,
75, 2475-2578.
Ga ney E.S. and Ande son D.L. (1973) E ec o low-spin Fe2+ on he composi ion
o he lowe man le. J. Geophys. Res. 78, 7005-7014
Ge dau E., Rü e R., Winkle H., Tolksdo W., Klages C. P., and Hannon J. P.
(1985) Nuclea B agg di ac ion o synch o on adia ion in y ium i on ga ne .
Phys. Re . Le ., 54 , 835-838.
Glazy in, K. (2011) E ec o i on oxida ion s a e on he bulk sound eloci y o he
lowe man le. Disse a ion. h p://opus.ub.uni-bay eu h.de/ oll ex e/2011/944/.
Glazy in, K. e al. (2012) I on in oxides, silica es and alloys unde ex eme p essu e-
empe a u e condi ions. Na u e Communica ion (submi ed).
Gold eich P. and Wa d W.R. (1973) The o ma ion o plane esimals. As ophys. J.,
183, 1051-1062
Goncha o A.F., S uzhkin V.V., Jacobsen S.D. (2006) Reduced adia i e
conduc i i y o low spin (Mg,Fe)O in he lowe man le. Science, 312, 1205–1208
6.!Bibliog aphy!
!
!
145!
Goodman J. and Pindo B. (1999) Secula ins abili y and plane esimal o ma ion in
he dus laye . Ica us, 148, 537-549
G eenwood N.N. and Gibb T.C. (1971) Mossbaue Spec oscopy, pp. 314–315, and
he e e ences he ein, Chapman and Hall, London
G ego yanz, E., Deg ya e a, O., Somayazulu, M., Hemley, R. & Mao, H.-kwang
(2005) Mel ing o Dense Sodium. Phys. Re . Le ., 94, 185502.
Gu, T., Wu, X., Qin, S. & Dub o insky, L. (2011) In si u high-p essu e s udy o FeP:
Implica ions o plane a y co es. Physics o he Ea h and Plane a y In e io s,
184, 154-159.
Gü lich P., Bill E., T au wein A. X., (2011) Mössbaue Spec oscopy and T ansi ion
Me al Chemis , Sp inge .
Haas M., (2007) Pola i on e ec s a p opaga ion o synch o on adia ion in
Mössbaue media. Phys. Le . A, 361 , 391-395.
Haase, J., Goh S. K., Meissne T., Ali eza P. L., and Rybicki D. (2009), High
sensi i i y nuclea magne ic esonance p obe o an il cell p essu e expe imen s.
Re . Sci. Ins um., 80, 073905.
Ha , S., and Zindle A. (1986) In sea ch o a bulk-Ea h composi ion. Chemical
Geology, 57, 247-267.
Hannon J.P. and T ammell G.T. (1999) Cohe en γ- ay op ics. Hype ine. In e ac .,
123/124 , 127-274.
Hedlin M.A.H., Shea e P.M., and Ea le P.S. (1997) Seismic e idence o small-scale
he e ogenei y h oughou he Ea h's man le. Na u e 387, 145-150
Higne e, O., Pe en, J.-C., Al a o, V., Chinchio, E., & F eund, A.K., (2001) SPIE
Con e ence P oceedings 45, 4501-4510.
Hi ose, K. (2006) Pos pe o ski e phase ansi ion and i s geophysical implica ions.
Re iews o Geophysics, 44, RG3001.
Ho iuchi H., I o E., and Weidne D.J. (1987) Pe o ski e- ype MgSiO3: single-c ys al
X- ay di ac ion s udy. Am. Mine al. 72, 357-360
Hsu, H., Umemo o, K., Blaha, P. & Wen zco i ch, R. M. (2010) Spin s a es and
hype ine in e ac ions o i on in (Mg,Fe)SiO3 pe o ski e unde p essu e. Ea h
Plane . Sci. Le . 294, 19-26.
Hsu, H., Blaha, P., Cococcioni, M. & Wen zco i ch, R. M. (2011) Spin-s a e
c osso e and hype ine in e ac ions o e ic i on in MgSiO3 pe o ski e. Phys Re
Le , 106, 118501.
Huggins F. (1975) The 3d le els o e ous i on in silica e ga ne s. Am. Mine al., 60,
316-319.
6.!Bibliog aphy!
!
!
146!
Ingalls R. (1964) Elec ic- iled g adien enso in e ous compounds. Phys. Re .
133A, 787- 795.
I i une T. and Ringwood A.E. (1987) Phase ans o ma ions in p imi i e MORB and
py oli e composi ions o 25 GPa and some geophysical implica ions, pp. 221-230.
In High-p essu e esea ch in mine al physics (eds. Manghnani M. and Syono Y.)
Te a/Ame ican Geophysical Union, Tokyo/Washing on.
Jackson, I., and S. M. Ridgen (2000), Composi ion and Tempe a u e o he Ea h’s
Man le: SeismologicalModels In e p e ed h ough Expe imen al S udies o Ea h
Ma e ials in “The Ea h’s man le : composi ion, s uc u e, and e olu ion,” 1s
ed., edi ed by I. Jackson, Camb idge Uni e si y P ess, Camb idge, UK.
Jackson J.M., S u hahn W., Shen G., Zhao J., Hu M.Y., E andonea D., Bass J.D. and
Fei Y. (2005) A synch o on Mössbaue spec oscopy s udy o (Mg,Fe)SiO3
pe o ski e up o 120 GPa. Am. Mine al., 90, 199-205.
Jagou z, E., H. Palme, H. Baddenhausen, K. Blum, M. Cendales, G. D eibus, B.
Spe el, V. Lo enz, and H. Wanke (1979), The abundances o majo , mino and
aceelemen s in he ea h’s man le as de i ed om p imi i e ul ama ic nodules
in “P oceedings o he Ten h Luna and Plane a y Science Con e ence, Hous on,
Texas, Ma ch 19-23, 1979,” Pe gamon P ess, New Yo k.
Jones, J. H., and M. J. D ake (1986). Geochemical cons ain s on co e o ma ion in
he Ea h. Na u e, 322, 221-228
Kagan Y., A. M. A anas'e , and V. G. Kohn. (1979) On exci a ion o isome ic
nuclea s a es in a c ys al by synch o on adia ion. J. Phys. C Solid S a e Phys.,
12 , 615-631.
Kagan Y.. (1999) Theo y o cohe en phenomena and undamen als in nuclea
esonan sca e ing. Hype ine. In e ac ., 123/124 , 83-126.
Kaneshima S. and Hel ich G. (1999) Dipping low- eloci y laye in he mid-lowe
man le: E idence o geochemical he e ogenei y. Science, 283, 1888-1891.
Kan o , I., Dub o insky, L. & McCammon, C. (2006) Spin c osso e in (Mg,Fe)O: A
Mössbaue e ec s udy wi h an al e na i e in e p e a ion o x- ay emission
spec oscopy da a. Phys. Re . B, 73. 100101-100104
Kepple , H. & F os , D. J. (2005) in Mine al Beha iou a Ex eme Condi ions Vol. 7
EMU No es in Mine alogy (ed R. Mile ich) 1-30 (Eö ös Uni e si y P ess, 2005).
Kepple H., Kan o I., and Dub o insky L.S. (2007) Op ical abso p ion spec a o
e ope iclase o 84 GPa. Am. Mine al. 92, 433-436
6.!Bibliog aphy!
!
!
147!
Kepple H., Dub o insky L.S., Na ygina O.V., Kan o I. (2008) Op ical abso p ion
and adia i e he mal conduc i i y o silica e pe o ski e o 125 Gigapascals.
Science 322, 1529-1532.
Kimu a K., Lewis R.S., and Ande s E. (1974) Dis ibu ion o gold and henium
be ween nickeli on and silica e mel s: implica ions o he abundances o
side ophile elemen s on he Ea h and Moon. Geochim. Cosmochim. Ac a, 38,
683–701.
Klo z, S., J. C. Che in, P. Munsch, and G. Le Ma chand (2009), Hyd os a ic limi s o
11 p essu e ansmi ing media, Jou nal o Physics D: Applied Physics, 42,
075413.
Kolesniko , A., Ku che o , V.G. & Goncha o , A.F. (2009) Me hane-de i ed
hyd oca bons p oduced unde uppe -man le condi ions. Na u e Geoscience, 2,
566-570.
Ko bo á, M., Kadečko á, S., No ák, J., B ádle , J., Smi no , G.V., & Sh ydko,
Yu.V. (1985). J. o C ys al G ow h, 71, 607-614.
K o A.N., Keil K., Good ich C.A., Sco E.R.D., and Weisbe g M.K. (2003)
Classi ica ion o me eo i es, pp. 83-128. In Me eo i es, Come s and Plane s (ed.
A.M. Da is) Vol. 1 T ea ise on Geochemis y (eds. Holland H.D. and Tu ekian
K.K.), Else ie -Pe gamon, Ox o d.
Ku noso , A., I. Kan o , T. Bo a-Bola an, S. Lindha d , L. Dub o insky, A.
Kuzne so , and B. H. Zehnde (2008) A no el gas-loading sys em o
mechanically closing o a ious ypes o diamond an il cells. Re . Sci. Ins um.,
79, 045110.
Lewis, J. (2004) Physics and chemis y o he sola sys em, 2nd ed., Else ie
Academic P ess, Ams e dam, Bos on.
Li J., S uzhkin V.V., Mao H.-K., Shu J., Hemley R.J., Fei Y., Mysen B., Dea P.,
P akapenka V.B., and Shen G. (2004) Elec onic spin s a e o i on in lowe man le
pe o ski e. P oced. Na . Acad. Sci. 101, 14027-14030.
Li J., S u hahn W., Jackson J.M. S uzhkin V.V., Lin J.-F., Zhao J., Mao H..-K. and
Shen G. (2006) P essu e e ec on he elec onic s uc u e o i on in
(Mg,Fe)(Si,Al)O3 pe o ski e: a combined synch o on Mössbaue and X- ay
emission spec oscopy s udy up o 100 GPa. Phys. Chem. Mine als., 33, 575-585
Lin J.-F., Deg ya e a, O., P ewi , C. T., De a, P., Sa a, N., G ego yanz, E., Mao, H.
K. & Hemley, R. J. (2004) C ys al s uc u e o a high-p essu e/high- empe a u e
phase o alumina by in si u X- ay di ac ion. Na u e ma e ials, 3, 389-393.
Lin J.-F., S uzhkin V.V., Jacobsen S.D., Hu M.Y., Chow P., Kung J., Liu H., Mao
H.K., and Hemley R.J. (2005) Spin ansi ion o i on in magnesiowüs i e in he
Ea h's lowe man le. Na u e, 436, 377-380.
6.!Bibliog aphy!
!
!
148!
Lin J.-F., Ga iliuk A.G., S uzhkin V.V., Jacobsen S.D., S u hahn W, Hu M.Y.,
Chow P., and Yoo C.S. (2006) P essu e-induced elec onic spin ansi ion o i on
in magnesiowus i e- (Mg,Fe)O. Phys. Re . B 73, 113107.
Lin J.-F., Vankó G., Jacobsen S.D., Io a V., S uzhkin V.V., P akapenka V.B.,
Kuzne co A., and Yoo C.S. (2007) Spin ansi ion zone in Ea h’s lowe man le.
Science 317, 1740-1743.
Lin J.-F., Wa son H., Vankó G., Alp E.E., P akapenka V.B, De a P., S uzhkin V.V.,
Kubo A., Zhao J., McCammon C.A., and E ans W.J. (2008) In e midia e-spin
e ous i on in lowe mos man le pos -pe o ski e and pe o ski e. Na u e
Geoscience, 1, 688-691.
Lin J.-F. , E. E. Alp, Z. Mao, T. Inoue, C. McCammon, Y. Xiao, P. Chow, and J.
Zhao (2012) Elec onic spin and alence s a es o i on in he lowe -man le silica e
pe o ski e by synch o on Mössbaue spec oscopy, Am. Mine ., 97, 592-597,
Lundin, S. e al. (2008) E ec o Fe on he equa ion o s a e o man le silica e
pe o ski e o e 1 Mba . Phys Ea h Plane In e ioi , 168, 97-102.
Mao, H. K., J. Xu, and P. M. Bell (1986) Calib a ion o he Ruby P essu e Gauge o
800 kba Unde Quasi-Hyd os a ic Condi ions, J. Geophys. Res., 91, 4673-4676
McCammon C.A., Rubie D.C., Ross II C.R., Sei e F. and O’Neill H.S .C. (1992)
Mössbaue spec a o 57Fe0.05Mg0.95SiO3 pe o ski e a 80 and 298 K. Am.
Mine al., 77, 894-897.
McCammon, C. A. (1997) Pe o ski e as a possible sink o e ic i on in he lowe
man le. Na u e, 387, 694-696.
McCammon C.A. (2005a) The pa adox o man le edox. Science, 308, 807-808
McCammon C.A. (2005b) Man le oxida ion s a e and oxygen ugaci y: Cons ain s on
man le chemis y, s uc u e and dynamics, pp. 221-242. In Ea h's Deep Man le:
S uc u e, Composi ion, and E olu ion (eds. Hils R.D., Bass J., Ma as J., and
T ampe J.), Ame ican Geophysical Union, Washing on D.C.
McCammon C.A., Kan o I.Yu., Na ygina O., Rouque e J., Ponk a z U., Se guee I.,
Mezoua , M., P akapenka V., Dub o insky L.S. (2008) S able in e media e-spin
e ous i on in lowe man le pe o ski e. Na u e Geoscience, 1, 684-687
McDonough, W., and S.-S. Sun (1995) The composi ion o he Ea h, Chemical
Geology, 120, 223-253.
Me ill, L., and W. A. Basse (1974) Minia u e diamond an il p essu e cell o single
c ys al x- ay di ac ion s udies, Re . Sci. Ins um., 45, 290-294.
Mikhaylushkin, A. , S.I. Simak, L. Dub o insky, N. Dub o inskaia, B. Johansson,
I.A. Ab ikoso (2007) Pu e I on Comp essed and Hea ed o Ex eme Condi ions.
Phys. Re . Le . 99, 165505.
6.!Bibliog aphy!
!
!
149!
Ming, L.-chung & Basse , W.A. (1974) Lase hea ing in he diamond an il p ess up
o 2000°C sus ained and 3000°C pulsed a p essu es up o 260 kiloba s. Re . Sci.
Ins um., 45, 1115
Mi zababae R.M., G. V. Smi no , V. V. Sklya e skii , A. N. A em'e , A.
N. Iz ailenko , and A. V. Babko (1971) Pu ely nuclea di ac ion o esonan
gamma- ays due o hype ine quad upole in e ac ions. Phys. Le ., 37A , 441-
442.
Mi o, M., M. Hi aka, T. Kawae, K. Takeda, T. Ki ai, and N. Toyoshima (2001)
De elopmen o Minia u e Diamond An il Cell o he Supe conduc ing Quan um
In e e ence De ice Magne ome e . Jpn. J. Appl. Phys., 40, 6641-6644.
Mi sui, T., Se o, M., Kiku a, S., Hi ao, S., Ohishi, Y., Takei, H., Kobayashi, Y.,
Ki ao, S., Higashi aniguchi, S., & Masuda, R. (2007a) Gene a ion and Applica ion
o Ul ahigh Monoch oma ic X- ay Using High-Quali y 57FeBO3 Single C ys al.
Jpn. J. Appl. Phys. 46, 821-825.
Mi sui, T., Se o, M., Hi ao, M., Ohishi, Y., Kobayashi, Y., Higashi aniguchi, Y., &
Masuda, R. (2007b) Ul ahigh-P essu e Measu emen in he Mul imegaba Range
by Ene gy-Domain Synch o on Radia ion 57Fe-Mössbaue Spec oscopy Using
Focused X- ays. Jpn. J. Appl. Phys., 46, L382-L384.
Mi sui, T., Se o, M., Masuda, R., Ki iyama, K., & Kobayashi, K. (2007c) Synch o on
Radia ion Mössbaue Spec oscopy Using Dopple -shi ed 14.4 keV Single-line
57Fe-Mössbaue Radia ion. Jpn. J. Appl. Phys., 46, L703-L705.
Mi sui, T., Se o, M., & Masuda, R. (2007d) Va iable-F equency Nuclea
Monoch oma o Using Single-Line Pu e Nuclea B agg Re lec ion o Oscilla ing
57FeBO3 Single C ys al. Jpn. J. Appl. Phys., 46, L930-L932.
Mi sui, T., Hi ao, N., Ohishi, Y., Masuda, R., Nakamu a, Y., Enoki, H., Sakaki K., &
Se o, M. (2009) De elopmen o an ene gy-domain 57Fe-Mössbaue spec ome e
using synch o on adia ion and i s applica ion o ul ahigh-p essu e s udies wi h a
diamond an il cell. J. Synch o on Rad., 16, 723-729.
Mo gan, J. W. (1986) Ul ama ic Xenoli hs: Clues o Ea h’s La e Acc e iona y
His o y, J. Geophys. Res., 91, 12375-12387.
Mu akami M., Ohishi Y., Hi ao N., and Hi ose K. (2012) A pe o ski ic lowe man le
in e ed om high-p essu e, high- empe a u e sound eloci y da a. Na u e, 485,
90–94.
Mu akami, M., K. Hi ose, K. Kawamu a, N. Sa a, and Y. Ohishi (2004) Pos -
Pe o ski e Phase T ansi ion in MgSiO3. Science, 304, 855-858.
Mu hy, V. R. (1991) Ea ly Di e en ia ion o he Ea h and he P oblem o Man le
Side ophile Elemen s: A New App oach. Science, 253, 303-306.
Na ygina, O., L. S. Dub o insky, H. Samuel, C. A. McCammon, I. Y. Kan o , K.
Glazy in, S. Pasca elli, G. Aquilan i, and V. B. P akapenka (2011) Chemically
6.!Bibliog aphy!
!
!
150!
homogeneous spin ansi ion zone in Ea h’s lowe man le. Physics o he Ea h
and Plane a y In e io s, 185,107-111.
Na ygina, O. (2010) In es iga ion o he p ope ies o i on-bea ing alloys and silica es
and hei implica ions o he Ea h’s lowe man le and co e Disse a ion.
h p://opus.ub.uni-bay eu h.de/opus4-ubbay eu h/ on doo /index/index/docId/558
O’Neill, H. S. C., and H. Palme (2000) Composi ion o he Silica e Ea h:
Implica ions o Acc e ion and Co eFo ma ion in “The Ea h’s man le :
composi ion, s uc u e, and e olu ion” 1s ed., edi ed by I. Jackson, Camb idge
Uni e si y P ess, Camb idge [u.a.].
Oh a K., Onoda S., Hi ose K, Sinmyo R., Shimizu K., Sa a N, Ohishi Y, Yasuha a A.
(2008) The elec ical conduc i i y o pos -pe o ski e in Ea h's D'' laye . Science
320, 89-91.
Oh a, K., Hi ose, K., Shimizu, K., Sa a, N. & Ohishi, Y. (2010) The elec ical
esis ance measu emen s o (Mg,Fe)SiO3 pe o ski e a high p essu es and
implica ions o elec onic spin ansi ion o i on. Phys Ea h Plane In.i, 180,
154-158.
Oh a, K. e al. (2010) Elec ical conduc i i ies o py oli ic man le and MORB
ma e ials up o he lowe mos man le condi ions. Ea h Plane . Sci. Le ., 289,
497-502 .
Palme, H., and K. Nickel (1985) Ca/Al a io and composi ion o he Ea h’s uppe
man le. Geochimica e Cosmochimica Ac a, 49, 2123-2132.
Palme H. and Jones A. (2003) Sola sys em abundances o he elemen s, pp. 41-61. In
Me eo i es, Come s and Plane s (ed. Da is A.M.) Vol. 1 T ea ise on
Geochemis y (eds. Holland H.D.and Tu ekian K.K.), Else ie -Pe gamon, Ox o d.
Palme J. and O’Neill H.S .C. (2003) Cosmochemical es ima es o man le
composi ion, pp. 1-38. In The Man le and Co e (ed. Ca lson R.W.) Vol. 2 T ea ise
on Geochemis y (eds. Holland H.D. and Tu ekian K.K.), Else ie -Pe gamon,
Ox o d.
Palyano , Y.N., Kup iyano , I.N., Sokol, A.G., Khokh yako , A.F. & Bo zdo , Y.M.
(2011) Diamond G ow h om a Phospho us–Ca bon Sys em a High P essu e
High Tempe a u e Condi ions. C ys al G ow h & Design, 11, 2599-2605.
Pankhu s Q.A., N.S. Cohen, L.F. Ba quin, M.R. Gibbs, and G.V. Smi no
(2001) Momen can ing and s uc u al aniso opy in amo phous alloys:
expe imen s using synch o on Mossbaue adia ion. J. Non-C ys . Solids, 287,
81-87.
Pa akhonskiy, G., Dub o inskaia, N., Byko a, E., Wi h, R. & Dub o insky, L.
(2011) Expe imen al p essu e- empe a u e phase diag am o bo on: esol ing he
long-s anding enigma. Scien i ic Repo s, 1, 96.
6.!Bibliog aphy!
!
!
151!
Pippinge , T., Dub o insky, L., Glazy in, K., Mile ich, R. & Dub o inskaia, N.
(2011) De ec ion o mel ing by in-si u obse a ion o sphe ical-d op o ma ion in
lase -hea ed diamond-an il cells. Física de la Tie a, 23, 29-41.
Po apkin, V., McCammon, C., Glazy in, K., Kan o , A., Kupenko, I., P esche , C.,
Sinmyo, R., Smi no , G.V., Popo , S., Chumako , A.I., and Rü e , R.,
Dub o insky, L. (2013) E ec o i on oxida ion s a e on he elec ical
conduc i i y o he Ea h’s lowe man le. Na u e Communica ion, 4, 1427.
P akapenka, V.B., A. Kuboa, A. Kuzne so b, A. Laskinc, O. Shku ikhind, P. De aa,
M. L. Ri e sa & S. R. Su ona (2008) Ad anced la op lase hea ing sys em o
high p essu e esea ch a GSECARS: applica ion o he mel ing beha io o
ge manium. High P essu e Resea ch, 28, 225-235.
P esche , C., McCammon, C. & Dub o insky, L. (2012) MossA - a p og am o
analyzing ene gy-domain Mossbaue spec a om con en ional and synch o on
sou ces. Jou nal o Applied C ys allog aphy, 45 329-331.
Ringwood, A. E. (1979) O igin o he ea h and moon. Sp inge , New Yo k.
Ringwood A.E. (1989) Cons i u ion and e olu ion o he man le. Spec. Publ. Geol.
Soc. Aus allia, 14, 457-485
Rü e R., & Chumako A.I. (1996) Nuclea esonance beamline a ESRF. Hype fine
In e ac ., 97/98, 589-604.
Rü e H.D., R.Rü e , E. Ge dau, R. Holla z, A. I. Chumako , M. V. Zelepukhin, G.
V. Smi no , and U. an Bü ck (1990) Quan um bea dila ion due o magne isa ion
b eakdown. Hype ine In e ac ., 58, 2473- 2476.
Renninge , M. (1937) Ron genome ische. Bei age zu Kenn nis de La-
dungs e eilung im Diaman gi e . Z.K is ., 97, 107-121.
Rubie, D. C., F. Nimmo, and H. J. Melosh (2007) T ea ise on Geophysics. Else ie .
Sa ano V.S. (1969) E olu ion o he p o oplane a y cloud and o ma ion o he
Ea h and plane s (English ansla ion NASA TTF-667, 1972).
Saikia, A., Bo a Balla an, T. & F os , D. J. (2009) The e ec o Fe and Al
subs i u ion on he comp essibili y o MgSiO3 - pe o ski e de e mined h ough
single-c ys al X- ay di ac ion. Phys Ea h Plane In., 173, 153-161.
Sch öde , T. , M.N. Schneide , T. Rosen hal, A. Eisele, C. Gold, E.-W. Scheid , W.
Sche e , R. Be hold and O. Oeckle (2011) Nanos uc u es in me as able
GeBi2Te4 ob ained by high-p essu e syn hesis and apid quenching and hei
in luence on physical p ope ies. Phys. Re . B, 84, 184104.
Schul z, E. Mezoua , M., C ich on, W. A., Bauchau, S., Bla mann, G., And aul , D.,
Fique , G., Boehle , R., Rambe , N., Si aud, B. & Loubey e, P. (2005) Double-
6.!Bibliog aphy!
!
!
152!
sided lase hea ing sys em o in si u high p essu e–high empe a u e
monoch oma ic x- ay di ac ion a he ESRF. High P essu e Resea ch, 25, 71-83.
Se o, M., Masuda, R., Higashi aniguchi, S., Ki ao, S., Kobayashi, Y., Inaba, C.,
Mi sui, T., & Yoda, Y. (2009). J.Phys.: Con . Se ., 217, 012002.
Shen, G., Ri e s, M.L., Wang, Y. & Su on, S.R. (2001) Lase hea ed diamond cell
sys em a he Ad anced Pho on Sou ce o in si u x- ay measu emen s a high
p essu e and empe a u e. Re . Sci. Ins um., 72, 1273-1282.
Shen, G., Wang, L., Fe y, R., Mao, H.-kwang & Hemley, R.J. (2010) A po able
lase hea ing mic oscope o high p essu e esea ch. Jou nal o Physics:
Con e ence Se ies 215, 012191.
She man D.M. (1988) S uc u al and magne ic phase ansi ions in mine als, pp. 113-
128, In: Ad ances in physical geochemis y (eds. Ghose S, Coey J.M.D. and Salje
E.) Sp inge , New Yo k 7.
Sheld ick, G.M. (2008) A sho his o y o SHELX. Ac a c ys allog aphica. Sec ion A,
Founda ions o c ys allog aphy 64, 112-22.
Shull C.G. , W. A. S ause , and E. O. Wollan (1951) Neu on Di ac ion by
Pa amagne ic and An i e omagne ic Subs ances. Phys. Re ., 83 , 333-345.
Smi no , G. V., Sklya e skii, V. V., Voscanian, R. A. & A em'e , A. N. (1969)
Nuclea di ac ion o esonan γ adia ion by an an i e omagne ic c ys al. JETP
Le . 9, 70-73.
Smi no G.V., V.V.Mos o oi, Yu.V. Sh yd’ko, V.N. Selezne , and V.V.Rudenko
(1980) Sup ession o nuclea eac ion in 57FeBO3 c ys al. JETP, 51 603-606.
Smi no G.V. (1986) Cohe en e ec s in esonan di ac ion: Expe imen . Hype ine.
In e ac . 27 , 203-218.
Smi no , G.V., Zelepukhin, M.V., & an Bü ck, U. (1986) Pu ely nicela di ac ion
o Mössbaue adia ion in he c i ical egion nea Néel poin . JETP Le ., 43, 352-
355.
Smi no , G.V., an Bü ck, U., Chumako , A.I., Ba on, A.Q.R., & Rü e , R. (1997)
Synch o on Mössbaue sou ce. Phys. Re . B 55, 5811-5815.
Smi no , G. V. (2000) Synch o on Mössbaue sou ce o 57Fe adia ion. Hype fine
In e ac ., 125, 91-112.
Smi no G.V., U. an Bü ck, W. Po zel , P. Schindelmann , S.L. Popo , E.
Ge dau , Y.V. Sh yd’ko , H.D. Rü e , and O. Leupold (2005) P opaga ion o
nuclea pola i ons h ough a wo- a ge sys em: E ec o in e sion o a ge s.
Phys.Re .A 71 023804 .
Smi no G.V. , U. an Bü ck, J. A hu , G.S. B own , A.I. Chumako ,
A.Q.R. Ba on , W. Pe y, and S.L. Ruby (2007) Cu en s and ields e eal he
6.!Bibliog aphy!
!
!
153!
p opaga ion o nuclea pola i ons h ough a esonan a ge . Phys. Re . A, 76 ,
043811 .
Smi no G.V. (2009) Rocking cu es, ime and ene gy dis ibu ions o synch o on
Mőssbaue adia io o 57Fe. P ep in IAE-6614/9
Smi no , G.V., Chumako , A.I., Po apkin, V., Rü e , R., & Popo , S.L. (2011)
Mul ispace quan um in e e ence in a 57Fe synch o on Mössbaue sou ce. Phys.
Re . A, 84, 053851.
Speziale S., Milne A., Lee V.E., Cla k S.M., Pas e nak M.P., and Jeanloz R. (2005)
I on spin ansi ion in Ea h’s man le. P oced. Na . Acad. Sci., 102, 17918-17922
S anko , S., Rü e , R., Sladecek, M., Rennho e , M., Sepiol, B., Vogl, G., Spi idis,
N., Slezak, T., & Ko ecki, J. (2008) An ul a-high acuum sys em o in-si u
s udies. Re . Sci. Ins um., 79, 045108.
S ackhouse S., B odhol J.P., and P ice G.D. (2007) Elec onic spin ansi ions in
i on-bea ing MgSiO3 pe o ski e. Ea h Plane . Sci. Le ., 253, 282–290.
S ix ude L. and Li hgow-Be elloni C. (2005a) Mine alogy and elas ici y o he
oceanic uppe man le: o igin o he low- eloci y zone. J. Geophys. Res., 110,
B03204.
S ix ude, L., and C. Li hgow-Be elloni (2005b) The modynamics o man le mine als
- I. Physical p ope ies. Geophysical Jou nal In e na ional, 162, 610-632.
S u hahn W., Jackson J.M., and Lin J. (2005) The spin s a e o i on in mine als o he
Ea h’s lowe man le. Geophys. Res. Le ., 32, L12307.
Sch yd'ko Y.V. and G. V. Smi no (1989) Expe imen al s udy o ime and
equency p ope ies o collec i e nuclea exci a ions in a single c ys al (gamma-
ay esonance). J. Phys. : Condens. Ma e , 1, 10563-10584.
Tonks W.B. and Melosh H.J. (1992) Co e o ma ion by gian impac s. Ica us, 100,
326-346.
Ta i s, P. & Mandéa, M. (2010) The he e ogeneous elec ical conduc i i y s uc u e o
he lowe man le. Phys Ea h Plane In., 183, 115-125.
T ammell G.T. (1961) in: P oc. In e n. A omic Ene gy Agency Symp. on Chemical
E ec s o Nuclea T ans o ma ions. 1 , 75.
an Bü ck U., Smi no G. V., Mössbaue R. L., Mau us H. J., and Semioschkina N.
A. (1980) Enhanced nuclea esonance sca e ing in dynamical di ac ion o
gamma ays. J. Phys. C : Solid S . Phys., 13, 4511-4529.
an Bü ck U., Mössbaue R. L., Ge dau E., Rü e R., Holla z R., Smi no G. V.,
Hannon J.P. (1987) Nuclea B agg sca e ing o synch o on adia ion wi h s ong
speedup o cohe en decay, measu ed on an i e omagne ic 57FeBO3. Phys.
Re .Le ., 59, 355-358.
