PDF analysis of CdS Nanoparticles

1
PDF analysis of CdS Nanoparticles

• Reinhard B. Neder
• Institut für Mineralogie
• Universität Würzburg, Germany

2
Topics

• Nanoparticles
• Experimental determination of PDF
• Interpretation of PDF
• Simulation of PDF
• Results for CdS-Glutathion

3
Nanoparticles

• Extended clusters 1 nm to 300 nm
• Properties intermediate between molecule and
  crystal
• Optical properties depend on size

4
Open topics for Nanoparticles

• determination of size

? TEM preparation core/shell statistics ?
XRD Debye-Scherrer invalid size disorder ?
Exciton depends on bulk theory ? small
angle available edges
5
Invalidity of Scherrer equation
6
Open topics for Nanoparticles

• determination of size

? TEM preparation core/shell ?
XRD Debye-Scherrer invalid size disorder ?
Exciton depends on bulk theory ? small
angle available edges
7
Open topics for Nanoparticles

• structure

small crystal ltgt glass ltgt unique structure
homogeneous structure ? heterogeneous
structure
8
Open topics for Nanoparticles

• structure

? TEM lattice planes gt well ordered ?
XRD too few, broad peaks ? EXAFS local order
only
9
Open topics for Nanoparticles

• structure

? TEM lattice planes gt well ordered ?
XRD too few, broad peaks ? EXAFS local order
only
Wu et al. J. Phys. Chem B (2000), 106, 4569
10
Open topics for Nanoparticles

• structure

? TEM lattice planes gt well ordered ?
XRD too few, broad peaks ? EXAFS local order
only
CeO2
Wu et al. J. Phys. Chem B (2000), 106, 4569
11
Open topics for Nanoparticles

• structure

? TEM lattice planes gt well ordered ?
XRD too few, broad peaks ? EXAFS local order
only
12
CdS-Glutathione
contradictory size information 15 to 30 Å
13
Synthesis
Precipitation form aqueous solution Glutathione
CdCl
monodisperse powder
14
Data Collection
BW5, HASYLAB
?0.088 Å E140 keV T15 K sealed capillary Qmax
30 Å-1
15
Experimental Data
Inset corresonds to experiment with Cu-K?
Only three broad maxima
16
Normalized Structure Factor
17
Experimental PDF
narrow first maximum at 2.525 Å
broad, asymmetric second maximum at 4.11 Å
18
Experimental PDF
19
Experimental PDF
weak maxima at 1.5 Å
20
Analysis of first Maximum
R 2.525 Å ? 0.063 Å N 3.4
ONE Cd-S distance
21
Analysis of second Maximum
R 4.13 Å ? 0.15 Å N 5.8 R 3.85 Å ?
0.12 Å N 2.7
TWO Cd-Cd distances gt two Cd-S-Cd angles
Cd-Sorganic -Cd 100 Cd-Sinorganic -Cd
109
22
Summary of direct Interpretation
RAMAN spectroscopy No H-S modes gt
Glutathione bound to Cd

• Chem. analysis Cd1 S0.5 Glutathione0.5

PDF 1. peak Cd-S 2.525 Å ? 0.063 Å
2. peak Cd-Sinorganic-S 109(5) Cd-Sorga
nic -S 100(4)
PDF longer distances highly disordered
PDF longest distances Diameter 18 Å
23
Aspects of PDF Simulation
Finite particle size
Structure more cystal like? more glass like
?
Proper density Proper coordination
distribution Proper distance distributions
Best strategy trial and error RMC evoluti
onary algorithm
24
PDF of Nano versus Bulk
Bulk Number of interatomic pairs increases with
r2
DISCUS uses r !
4 ? ?0 r
25
PDF of Nano versus Bulk
Nanoparticle A longest vector exists
DISCUS uses r !
4 ? ?0 r
4 ? ?0 r tanh(shape(r-diameter))
Diameter
26
Crystal of nanoparticles
Periodic boundary conditions lead to PDF
maxima well beyond the particle diameter.
27
Crystal of nanoparticles
Periodic boundary conditions with random
orientation destroy PDF maxima beyond the
particle diameter.
28
Individual nanoparticle
PDF with modified background function
29
Simulation of PDF
Experimental PDF
Distorted zincblende structure
30
Simulation of PDF
Growth algorithm
S- or Cd- terminated all lt111gt
N(Cd) 50 S on surface fault probability
starting group
31
Simulation of PDF
Relaxation of the structure by Monte Carlo
constrains
lt Cd-Sorganic-C 100(4)
Cd-S 2.523(63)Å
lt Cd-Sinorganic-C 109(5)
32
Simulation of PDF
Decoration by Glutathione
33
Calculated PDF
70 Cd atoms
16 Å diameter
60 fault propability
average of many particles
34
Further Issues
Fit to data by evolutionary algorithm easy to
adapt to multiple steps needed growth,
relaxation, glutathione, PDF calculation and
averaging different structures
Fit to data by RMC requires large supercell to
avoid inter-particle vectors gt adjust ?0,
avoid shift of particle, limit movement to
within particle starting model ? different
particles ?
to be learned...
35
Acknowledgements
V.I. Korsounski K. Hradil J.Neuefeind P. Jovari
Ch. Barglik-Chory G. Müller Th. Proffen
Financial support by German Science Foundation
SFB 410