Synchrotron and neutron experiments Angus P. Wilkinson School of Chemistry and Biochemistry Georgia Institute of Technology Atlanta, GA 30332-0400 Thanks are due to Alan Hewat and Ian Swainson for many of the slides
Jan 30, 2016
Synchrotron and neutron experiments
Angus P. WilkinsonSchool of Chemistry and Biochemistry
Georgia Institute of Technology
Atlanta, GA 30332-0400
Thanks are due to Alan Hewat and Ian Swainson for many of the slides
Outline Comparison of X-ray and neutron scattering Applications of neutron diffraction
– “Light” elements– Magnetism– High Q data– Penetration
What is a synchrotron and why use one? Resonant scattering and the determination of complex
cation distributions Where X-rays meet neutrons – in the high energy
regimen Summary
A comparison of X-rays and neutrons
X-rays Neutrons
Atomic scattering power varies smoothly with atomic number
Atomic scattering power varies erratically with atomic number
Atomic scattering power decreases as the scattering angle increases
Atomic scattering power is constant as the scattering angle changes
Largely insensitive to magnetic moments
Scattered by magnetic moments
Readily available as intense beams Low intensity beams
Typically, strongly absorbed by all but low Z elements
Weakly absorbed by most materials
Relative Scattering Powers of the Elements
Locating “light elements”
Structure of the 90K high Tc superconductor
– Left -by X-rays(Bell labs & others)
– Right -by Neutrons(many neutron labs)
The neutron picture gave a very different idea of the structure -important in the search for similar materials.YBa2Cu3O7 drawing from Capponi et
al. Europhys Lett 3 1301 (1987)
Hydrogen in metals
Hydrogen storage in metals– Location of H among heavy
atoms– No single crystals
Laves phases eg LnMg2H7 (La,Ce)– Binary alloys with large/small
atoms– Various arrangements of
tetrahedral sites can be occupied by H-atoms
– Up to 7 Hydrogens per unit
Gingl, Yvon et al. (1997) J. Alloys Compounds 253, 313.Kohlmann, Gingl, Hansen, Yvon (1999) Angew. Chemie 38, 2029. etc..
Hydrogen – a blessing and a curse Neutrons see hydrogen well – perhaps too well. Neutron incoherent scattering is an isotropic “random” scattering
of neutrons. This is the basis of some techniques (quasi-elastic neutron scattering) but is a killer for neutron, at least powder, diffraction.– Deuterate to avoid problems. This can be difficult and may change what you
want to examine. For example, cement hydration in H2O is different from that in D2O
Unit of b is fm. Unit of cross-section is 4b2 in barns (100 fm2). si c
% bc bi c i s a
H 99.985 -3.741 25.27 1.758 80.27 82.03 0.3326D 0.015 6.671 4.04 5.592 2.051 7.643 0.000519
Form factor fall off X-ray scattering amplitude is strongly dependent on sin
making it very difficult to get good quality x-ray data at high sin– This can give problems with determining “thermal parameters”
Neutrons give good signal at high sin
High Q data Time-of-flight neutron
diffraction facilitates the collection of data to very high Q (small d-spacing)– No form factor fall off– Highest flux at short
wavelength Similar experiments
can also be done with very high energy synchrotron radiation
Cu K Mo KNi metal, synchrotron radiation, GE detector
From Peter Chupas
The magnetic structure of MnO MnO, NiO and FeO order antiferromagnetically After taking into account the arrangement of unpaired spins the
unit cell is twice as big as the atomic arrangement would suggest– So you get extra peaks in the neutron diffraction pattern
Powder neutron diffraction data for MnO
Extra peaks are only present in the neutron diffraction pattern at temperatures where the unpaired spins are ordered (below Neel temperature).
Neutrons are penetrating
Neutrons can pass through a reasonable thickness of metal. This makes it easier to build sample environments– No Be windows or other special approaches
needed– V and some alloys such as TiZr have essentially
zero coherent scattering cross section and do not give any Bragg peaks
Radiant Furnace• Al vacuum body
• Water-cooled base
• W or Ta radiant elements
• Mo-foil heat shields
• 6 kW of power
• Turbo vac. 10-7 Torr base pressure, 5e-6 at 2000K
• Gas inserts, static or purge
Courtesy of I. Swainson
Cryomagnet
• 1.5K to RT • 200mK-1.5K He3
• Up to 9T vertical field
Courtesy of I. Swainson
Pressure with neutrons Pressure is problematic for
neutrons, due to low flux Usually need large sample
volume and P = F/A acts against you
But improvements in neutron optics, new sources (SNS etc) combined with advances in preparing large high strength single crystals (diamonds and Moissanite) for large volume gem anvil cells and the availability of devices such as the Paris-Edinburgh cell are expanding the accessible area of PT space
Gas pressure cell made from aluminum. Max P ~ 0.5 GPa
Element Mean a
Ce 0.63
Pr 11.5
Nd 50.5
Pm 168.4
Sm 5922
Eu 4530
Gd 49700
Tb 23.4
Dy 994
Ho 64.7
Er 159
Tm 100
Yb 34.8
Lu 74
Hf 104.1
Absorption – an isotopic problem
• Other (non-REE) absorbers include Cd and B• 11B, 7Li however are relatively cheap to buy.
Isotope % a
152Gd 0.2 735 154Gd 2.1 85 155Gd 14.8 61100 156Gd 20.6 1.5 157Gd 15.7 259000 158Gd 24.8 2.2 160Gd 21.8 0.77
Neutron are not without absorption problems!
Courtesy of I. Swainson
Synchrotron radiation
High intensity Plane polarized Intrinsically collimated Wide energy range Has well defined time
structure
Advantages of using a synchrotron
The high level of intrinsic collimation and high intensity of the source facilitates the construction of powder diffractometers with unrivaled resolution– More information in the powder pattern
Can achieve good time resolution, although not combined with ultrahigh resolution
Can do experiments at short wavelengths– Facilitates collection of high Q (small d-spacing) data, and reduces
or eliminates problems due to absorption Can do resonant scattering
– Chose a wavelength close to an absorption edge and tune the scattering power of the elements in you samples
Diffractometer Geometry
Crystal analyzer gives very good resolution, low count rate and is insensitive to sample displacement, useable with flat plate or capillary
Soller slits give modest resolution, good count rate and insensitivity to sample displacement
Simple receiving slits give good count rate, easily adjustable resolution, can be used with flat plate or capillary
11BM high resolution diffractometer12 channel analyzer system
Complex materials Many real materials do not have just one species on a given
crystallographic site– YBa2Cu3O7-x
» Can have both oxygen and oxygen vacancies on a given site
– Zeolites, Mx[Si1-xAlxO2]» Extraframework cations M occupy sites that may also have vacancies and water
present» Al may not be randomly distributed over all available sites
– NiFe2O4
» What is the distribution of nickel and ion over the tetrahedral and octahedral sites in the spinel?
It can be difficult to pin down the distribution of species over the available sites
Information from diffraction data Bragg scattering provides a measure of the scattering density at a
particular crystallographic site
With one diffraction data set it can be very difficult /impossible to estimate, xi ni and Ui for multiple species on nominally the same site– typically we assume that the xi and Ui are the same for all species at
nominally the same site» This may be a gross approximation!
– to estimate individual ni the species must differ in scattering power, even then more than two species can not be handled
» Determining Mn/Fe distribution in MnFe2O4 using neutrons is easy
)](2exp[)]/(sin8exp[ 222iiii
iiihkl lzkyhxiUfnF
Scattering contrast In some cases lab x-ray data does not generate enough
contrast to solve a problem– Ni/Fe distribution and other “neighboring element problems”
Neutrons may generate the needed contrast– But not for Ni/Fe!
More than one data set with different scattering contrast levels may be needed – Differing scattering contrast data set per species on the site?
» constraints on composition and site occupancy reduce this requirement
– Can get these additional data sets by isotopic substitution and neutron scattering or by resonant x-ray scattering
Resonant x-ray scattering and isotopic substitution
Isotopic substitution is very expensive Different isotopically substituted samples may not
be the same! Resonant x-ray scattering makes use of the same
sample for all measurements Reliable resonant scattering factors can be
awkward to get Absorption and restricted d-spacing range can be a
problem with resonant scattering measurements
The X-ray scattering factor
The elastic scattering is given by,
For a spherical atom,
)(")(')(),( EifEfQfQEf o
f” “mirrors” the absorption coefficient
f’ is intimately related to the absorption coefficient
Absorption and anomalous scattering
aEhe
mcEf
202
)("
dEEE
EEfEf
022
0 )(
)("2)('
Examples – Cs8Cd4Sn42
Cd location in the type I clathrate Cs8Cd4Sn42 – Is the Cd randomly distributed over all the available framework
sites?– Distribution of Cd effects Seebeck coefficient and thermoelectric
performance– Cd absorbs neutrons
Cd and Sn have similar atomic number – essentially indistinguishable by X-ray scattering unless X-rays
have energy close to absorption edge– collect data at 80 keV, Cd K-edge and Sn K-edge
» more good data improves reliability of the results » Scattering factors estimated from absorption measurements
Chem. Mater. 14, 1300-1305 (2002).
Sn scattering factors in Cs8Cd4Sn42
Anomalous scattering terms calculated from Kramers-Kronig transformation of absorption data
1.08
1.1
1.12
1.14
1.16
1.18
1.2
29.18 29.19 29.2 29.21 29.22 29.23 29.24
Sn K-edge scan for Cs8Cd
4Sn
42
Energy / keV
Data collected at 29.194 keV
0.5
1
1.5
2
2.5
3
3.5
4
4.5
-10
-9.5
-9
-8.5
-8
-7.5
-7
-6.5
-6
29180 29190 29200 29210 29220 29230 29240
f" f'
Energy / ev
Resonant scattering and Cs8Cd4Sn42
Selecting an X-ray energy close to an absorption edge distinguishes Cd from Sn
0
10
20
30
40
50
0 0.2 0.4 0.6 0.8 1
Real part of scattering factors
Cd 80 keVSn 80 keVCd at Cd edgeSn at Cd edgeCd at Sn edgeSn at Sn edge
sin(theta)/lambda
Diffraction data recorded at up sin/ ~0.7Å-1
Location of Cd in Cs8Cd4Sn42
Cd is located only on 6c sites – From analysis of data collected at 80 keV and both the Cd and Sn K-edges
Type I framework. 6c site (red), 16i site (grey) and 24k site (green)
Powder XRD at high energy High energy X-rays offer many of the advantages
associated with neutrons – along with a lot more flux!– Can use massive sample environment due to penetrating
nature of X-rays– Can map out phase and stress distributions inside parts due
to penetrating power– Systematic errors due to absorption and extinction are
eliminated– Can make measurements to very high Q
» provides a lot of structural detail
Summary Synchrotron based instruments offer very high
resolution, excellent peak to background ratio, high data rates, low absorption and the ability to tune an elements scattering power
Synchrotron instruments are expensive and the data is often harder to analyze than that obtained using neutrons
Neutrons excellent for low Z element problems Neutrons usually the tools of choice for magnetism