Structure prediction and structure determination of solids via investigation of their energy landsca

1
Structure prediction and structure determination
of solidsvia investigation of their energy
landscape

• J. C. Schön
• Max-Planck-Institut für Festkörperforschung
• D-70569 Stuttgart

Funding DFG, BMBF Together with H. Putz, M.
Wevers, . Cancarevic, M. Jansen
2
Motivation
Structure determination from limited information
Prediction of possible (meta)stable compounds
for chemical systems
Structure of chemical systems on an atomic
level
Dynamics of structural transformations
Calculation of phase diagrams
3
History
4
History
5
History
6
History
7
History
8
Physics
9
(No Transcript)
10
(No Transcript)
11
Energy landscape concepts
12
(No Transcript)
13
Energy landscape concepts
14
(Schön, Putz, Jansen, J. Phys. Cond. Mat., 1996)
15
Energy landscape and structures of MgF2
16
(Wevers, Schön, Jansen, J. Phys. A, 2001)
17
Structure prediction general procedure
18
Structure prediction through global exploration
of energy landscapes of chemical systems
(Schön, Jansen, Angew. Chem., 1996)
(Schön, Jansen, Z. Krist., 2001)
E.g.
E.g.
E.g.
19
Structure prediction through global exploration
of energy landscapes of chemical systems
20
Local optimization - Heuristic algorithm
Volume Optimization
Cell Optimization
Position Optimization
(Cancarevic, Schön, Jansen, Mat. Sci. Forum,
2004)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
A5B5, a Predicted New Type of Structure
Na5Cl5
(Schön, Jansen, Comp. Mat. Sci., 1995)
26
Energy landscape and structures of Na3N
(Schön, Wevers, Jansen, J. Mat. Chem., 2001)
(Jansen, Schön, ZAAC, 1998)
27
E(V) curves for promising structure candidates in
the system Na3N
28
Powder diffraction pattern for Na3N (Rietveld
refinement) and structure candidate (anti-ReO3
type)
D. Fischer and M. Jansen, Angew. Chem., 2002
29
(No Transcript)
30
(Fisher, Cancarevic, Schön, Jansen, ZAAC, 2004)
31
Temperature-dependent X-ray powder diffraction of
K/N
K3N anti-TiI3 structure type hex., P 63/mcm (No.
193) a780, c759 pm, ?120 o Z2, V120
cm3/mol, K (6g), N (2b) d(K-N)278 pm
32
Energy landscape and structures of Ca3SiBr2
(Putz, Schön, Jansen, ZAAC, 1999)
33
Variation of the composition
Ca5Si2Br2
34
(Cancarevic, Schön, Jansen, Mat. Sci. Forum,
2004)
(Schön, Wevers, Jansen, J. Mat. Chem. ,2001)
35
M2S (MLi, Na, K, Rb, Cs)
36
Incorporation of external information - Cost
functions and constraints
37
Incorporation of external information - Cost
functions and constraints
38

Primary building units
(Schön, Jansen, Acta Cryst. Suppl. , 1999)
(Schön, Jansen, Z. Krist., 2001)
39

N2
40

SnO
41

KNO2
42

MgCN2
43
E(V) curves for candidates in the system MgCN2
(Schön, Jansen, Z. Krist., 2001)
(Mellot-Draznieks, Girard, Ferey, Schön,
Cancarevic, Jansen, Chem. Eur. J., 2002)
44
Fixed Angles
Fixed Distances
45
E(V) curves for candidates in the system Li4CO4
Type-I
Type-IV
Type-II
Type-V
Type-VI
Type-III
(Mellot-Draznieks, Girard, Ferey, Schön,
Cancarevic, Jansen, Chem. Eur. J., 2002)
46
(No Transcript)
47
(No Transcript)
48
(No Transcript)
49
(No Transcript)
50
Structure determination
(Putz, Schön, Jansen, J. Appl. Cryst., 1999)
51
Pareto-optimization of CaCO3 (30 atoms)
52
Examples
53
K2(CN2)
54
(No Transcript)
55
Fields of application
56
Technical Details
57
Heuristic algorithm - Programming details
Scanner.V-1.0 to V-1.9
G42 History
Load.V-1.0 to V-3.5
Load.P-1.0 to P-1.1
KPLOT CMPZ
Filter.V-1.0 to V-2.2
Hartree.B-1.0 to B-2.1
BILLY
Hartree.L-1.0 to L-2.1
Hartree.F-1.0 to F-2.1
StuCRY.V-1.0 to V-1.1
Coordy.V-1.0 to L-1.0a
CRYtoKPL.V-1.0 to V-1.7
58
?
? 10
? 10

• Self consistency check and accuracy of the
  corrected data (HEURISTIC ALGORITHAM)
• Automatic CRYSTAL output analysis
• on-line check (remote control)
• publication ready outputs (automatic fitting,
  plotting, labeling, legend, word and LaTeX
  tables, post script and graphic files, kplot
  files , find primitive cell , .. etc.)
• generation of the report in HTML format
• bookkeeping
• e-mailing on the end of RUN
• Automatic CRYSTAL properties analysis
• INSUALTOR-CONDUCTOR
• band structure
• on-line distance analysis
• etc
• Tangent (Transition pressures calculation)

?
?
?
StuCRY (Stuttgart to CRYSTAL PPBS conversion)
Techniques Comparison HF, DFT (LDA-LYP,
LDA-VBH, LDA-PWGGA, BECKE-LYP, BECKE-PWGGA, B3LYP)
BASIS SETS Optimization 10
Coordy
CRYtoKPL
59
E(V) curves for promising structure candidates
for alkali metal sulfides
Vtrans 134.46 Å3
Vtrans 230,59 Å3
Vtrans 94,58 Å3
Vtrans 329.80 Å3
ICSD - V. H. Sommer, R. Hoppe, Z. anorg. alg.
Chem. 429 , 118-130 (1977)
A. Vegas, A. Grzechnik, M. Hanfland,C Muhle,M.
Jansen , Sol. Sta. Sci., 4 (8) 1077-1081 (2002)
Grzechnik A, Vegas A, Syassen K, Loa I, Hanfland
M, Jansen M, J. of Sol. Sta. Chem, 154 (2)
603-611 (2000)
Vegas A, Grzechnik A, Syassen K, Loa I, Hanfland
M, Jansen M, A. Cryst. B, 57 151-156 (2001)
ICSD - May K., ZEKGA, 94 , 412-413 (1936)
60
Additional Info
61
History of structure prediction and determination
in extended solids

• Global optimization methods, e.g. GenAlg (Holland
  1975) or SimAnn (Kirkpatrick et al. 1983, Cerny
  1985)

62
Energy landscape concepts

• Stability of locally ergodic regions

63
Structure prediction general procedure
64
(No Transcript)
65
Including BASIS SETS Optimization
66
K/N Codeposition
substrate sapphire ((0001), epitaxial
polished) temperature 77 K K effusion cell
(373 K) N2 MW-plasma source (1.6 sccm, 80 mA,
purity 5.0) process pressure 6 x10-5
mbar time 5 h sample metallic
color XRD temperature-dependent, potassium
nitride?
67
X-ray powder diffraction pattern 1. K/N phase
68
M2S (MLi, Na, K, Rb, Cs)
69
M2S (MLi, Na, K, Rb, Cs)
70

MgCN2
71
Movements of building units Shift Rotation
Exchange
Fixed Angles
Fixed Distances
Assignment of Fixed Charge
72
Structure determination
73
Structure determination
74
Structure determination
75
Fields of application

• Molecular crystals
• Packing of rigid/floppy molecules primary
  building units tiny energy differences between
  modifications (Problems energy function,
  temperature/pressure dependence) limited
  data-mining (preferred space groups, number of
  molecules in unit cell)
• "Atom"-based solids
• Type of bonding (empirical potential) ab initio
  programs needed for local optimization influence
  of kinetics regarding outcome no data-mining
  (except testing ICSD) primary building units
  possible usually few problems with
  temperature/pressure

76
Fields of application

• "Building-unit"-based solids
• Coordination polyhedra (secondary building units)
  on many scales specially optimized potentials
  available as alternative to ab initio
  calculations temperature/pressure influence
  excluded by construction
• Amorphous solids
• No unique structure (many possible amorphous
  "structures", each with structural bandwidth and
  controlled by kinetics) No periodicity (gigantic
  unit cells necessary - ab initio very expensive)
  need to simulate structure generation process