Transcript Folie 1 - uni
“Light Scattering from Polymer Solutions and Nanoparticle Dispersions”
By: PD Dr. Wolfgang Schaertl
Institut für Physikalische Chemie, Universität Mainz, Welderweg 11, 55099 Mainz, Germany
schaertl@uni-mainz.de
Parts from the new book of the same title, published by Springer in July 2007 Slides are found at: http://www.uni-mainz.de/FB/Chemie/wschaertl/105.php
1. Light Scattering – Theoretical Background
1.1. Introduction
Light-wave interacts with the charges constituting a given molecule in remodelling the spatial charge distribution:
E
0 cos 2
x
2
c t
Molecule constitutes the emitter of an electromagnetic wave of the same wavelength as the incident one (“elastic scattering”)
m E E s
Note: usually vertical polarization of both incident and scattered light (vv-geometry)
Particles larger than 20 nm: - several oscillating dipoles created simultaneously within one given particle - interference leads to a non-isotropic angular dependence of the scattered light intensity - particle form factor, characteristic for size and shape of the scattering particle - scattered intensity I ~ N i M i 2 P i (q) (scattering vector q, see below!) Particles smaller than /20: - scattered intensity independent of scattering angle, I ~ N i M i 2
Particles in solution show Brownian motion (D = kT/(6 h R), and < D r(t) 2 >=6Dt) => Interference pattern and resulting scattered intensity fluctuate with time
1.2. Static Light Scattering
Scattered light wave emitted by one oscillating dipole
E s
2
m
t
2 1
r c D
2 4 2
r c D
2
E
0 exp
i
2
t
kr D
Detector (photomultiplier, photodiode): scattered intensity only!
I s
s
E s
2 sample
I 0 r D
I
detector Light source I 0 = laser: focussed, monochromatic, coherent Sample cell: cylindrical quartz cuvette, embedded in toluene bath (T, n D )
Standard Light Scattering Setup of the Schmidt Group, Phys.Chem., Mainz: sample, bath laser detector on goniometer arm
Standard Light Scattering Setup of the Schmidt Group, Phys.Chem., Mainz:
Standard Light Scattering Setup of the Schmidt Group, Phys.Chem., Mainz:
Scattering volume: defined by intersection of incident beam and optical aperture of the detection optics Important: scattered intensity has to be normalized
Scattering from dilute solutions of very small particles
(e.g.
nanoparticles or polymer chains smaller than /20)
(“point scatterers”)
Fluctuation theory:
I
: (
c c
)
T N
contrast factor Ideal solutions, van’t Hoff:
c
kT M
Real solutions, enthalpic interactions solvent-solute:
b
2 4
2
0 4
N L
c n D
,0 2 (
n D
c
) 2
K
in cm 2 g -2 Mol 1
kT
(
M
2
A c
2 ...
) Absolute scattered intensity of ideal solutions, Rayleigh ratio ([cm -1 ]):
R
Kc R
1
M I
0 4 4
2
0 4
N L n D
,0 2 (
n D
c
) 2 (
I solution
I solvent
)
r D
2
V
and
R
I solution
I solvent
I I std
Scattering standard I std : Toluene ( I abs = 1.4 e-5 cm -1 ) Reason of “Sky Blue”! (scattering from gas molecules of atmosphere)
Real solutions, enthalpic interactions solvent-solute expressed by 2nd Virial coeff.:
Kc R
1
M
2
A c
2 ...
Scattering from dilute solutions of larger particles
- scattered intensity dependent on scattering angle (interference) The scattering vector
q
(in [cm -1 ]) , length scale of the light scattering experiment:
k
0
q k q
4
n D
sin(
2 )
q
= inverse observational length scale of the light scattering experiment :
q
q
-scale
qR
<< 1
qR
< 1
qR
≈ 1
qR
> 1
qR
>> 1 resolution whole coil topology topology quantitative chain conformation chain segments information mass, radius of gyration cylinder, sphere, … size of cylinder, ...
helical, stretched, ...
chain segment density comment e.g. Zimm plot
Scattering from 2 scattering centers – interference of scattered waves
k
0
A
B C
k
0
r ij k k AB
BC
???
AB BC AB
ij
cos
ij
cos
BC
k
0
ij
0
ij r ij k
2
2
2
cos
2
AB
leads to phase difference: 2 interfering waves with phase difference D :
E s
exp(
ik r
)
E s
exp D 2
E s
exp(
ik r
2
BC
2 2
D
ij I s
I
2 2 1 1 exp
iqr ij
Scattered intensity due to Z pair-wise intraparticular interferences, N particles:
Nb
2
i Z
1
j Z
1 exp
i
r j
Nb
2
i Z
1
j Z
1 exp
iqr ij
orientational average and normalization lead to: 1 2 1
Z
2
i Z
1
j Z
1 exp
iqr ij
1
Z
2
i Z
1
j Z
1 sin
qr ij
1
Z
2
i Z
1
j Z
1 1 1 6 2
q r ij
2 ...
replacing Cartesian coordinates r i by center-of-mass coordinates s i we get:
i Z
1
j Z
1
r ij
2
i Z
1
j Z
1
s j
s i
2
i Z
1
j Z
1
s j
2 2
j
s i
2 2 2 2
Z s
1 3 2
s q
2 ...
s 2 , R g 2 = squared radius of gyration . finally yields the well-known Zimm-Equation (series expansion of P(q), valid for small R):
Kc R
1 2
A c
2 ...
Kc R
1
M
( 1 1 3 2
s q
2 ) 2
A c
2
The Zimm-Plot, leading to M, s (= R g ) and A 2 :
6,0 5,5 5,0 4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,0 5,0
c = 0
Kc R
q = 0
10,0 15,0
(q
2
+kc) / 10
10
cm
-2
1
M
( 1 1 3 2
s q
2 ) 2
A c
2 example: 5 c, 25 q
20,0
Zimm analysis of polydisperse samples yields the following averages:
The weight average molar mass
M w
k K
1
k K
1
N M M k k k N M k k
The z-average squared radius of gyration:
s
2
z R g
2
z
k K
1
k K
1 2
N M s k k k
2
N M k k
2 Reason: for given species k, I k ~ N k M k 2
Fractal Dimensions
:
R d f
if
q
R g
1 log :
M
2 :
q
2
d f
log log 2
d f
log
q
d f
log
q
: topology cylinder, rod thin disk homogeneous sphere ideal Gaussian coil Gaussian coil with excluded volume branched Gaussian chain d f 1 2 3 2 5/3 16/7
Particle form factor for “large” particles
1 2
NZ b
2 1
Z
2
i Z
1
j Z
1 exp
iqr ij
1
Z
2
i Z
1
j Z
1 sin
qr ij
for homogeneous spherical particles of radius R: 9 6 sin
qR
cos 2 10 0 10 -1 Zimm!
10 -2 10 -3 10 -4 10 -5 0 2 4 first minimum at
qR
= 4.49
6 qR 8 10 12
1.3. Dynamic Light Scattering
Brownian motion of the solute particles leads to fluctuations of the scattered intensity change of particle position with time is expressed by van Hove selfcorrelation function, DLS-signal is the corresponding Fourier transform (dynamic structure factor)
s
n t n r t
) isotropic diffusive particle motion
s s
s
) [ 2
3 D
R
2 ] 3 2 exp ( 2
D
R
2 2 ) mean-squared displacement of the scattering particle: D
R
2 6
D s
D s
kT f
kT
6
h
R H
Stokes-Einstein, Fluctuation - Dissipation
The Dynamic Light Scattering Experiment - photon correlation spectroscopy
1
D q s
2
exp
, 2 Siegert relation: t
s
exp(
D q s
2
)
s s
2
1 note : usually the “coherence factor” f c smaller than 1, i.e.: is
2
c
s
2 f c increases with decreasing pinhole diameter, but photon count rate decreases!
DLS from polydisperse (bimodal) samples
s
0 exp 2
q D s
dD s
A
1 exp 2
q D s
1
A
2 exp 2
q D s
2
log
Data analysis for polydisperse (monomodal) samples
”Cumulant method“, series expansion, only valid for small size polydispersities < 50 % ln
s
1 1 2!
2 2 1 3!
3 3 ...
first Cumulant
1
D q s
² second Cumulant
2
D s
2 yields inverse average hydrodynamic radius
D s
2
q
4 yields polydispersity
D
D s
2
D s D s R H
1 2
1 2 2 for samples with average particle size larger than 10 nm:
D app
i
i
i
2
i
2
i
i
P q
D i
note:
i
i
i
2
i
D app
D s z
1
K R g
2
z q
2
Cumulant analysis – graphic explanation:
monodisperse sample polydisperse sample
D y/ D x=-D s q 2 large, slow particles D y/ D x=-D s q 2 linear slope yields diffusion coefficient small, fast particles slope at =0 yields apparent diffusion coefficient, which is an average weighted with n i M i 2 P i (q)
D app vs. q 2 :
D s z
2,0x10 -14 1,5x10 -14 1,0x10 -14 5,0x10 -15 0,0 0 1x10 10 2x10 10 q 2 /cm -2 3x10 10 4x10 10
Explanation for D app (q):
D app
i
i
i
2
i
2
i
P q i
D i
for larger particle fraction i, P(q) drops first, leading to an increase of the average D app (q)
q
1
q
2
10
0
10
-1
10
-2
10
-3
10
-4
10
-5
0,00
R = 60 nm R = 80 nm R = 100 nm
0,01 0,02 q [nm
-1
] 0,03 0,04
ln(g1( 50 ))=P1+P2* +P3/2 * ^2 PI = SQRT(P3/P2^2) Ni 40 20 Ni 15 30 20 10 10 5 0,0 0 0 lng1 5 10 R i /nm 15 Data: Data2_lng1 Model: cumulant Chi^2 = 3.7224E-8 P1 P2 P3 0.00882
±0.00003
-10790.57918 ±0.23957
896471.16145 ±926.09523
D app (90°)=2.04e-11 m 2 /s, entspr. R = 10.5 nm PI = 0.09, D R/R=10% (Normalvert.) -0,2 0,00000 /s 20 0 0 5 10 R i /nm 15 20 0,00002 0,00 -0,02 -0,04 -0,06 -0,08 -0,10 -0,12 -0,14 -0,16 -0,18 -0,20 0,000000 lng1 D app (90°)=1.59e-11 m 2 /s, entspr. R = 13.5 nm PI = 0.20, D R/R=30% (Normalvert.) 0,000004 0,000008 Data: Data2_lng1 Model: cumulant Chi^2 = 3.3258E-10 P1 P2 P3 0.0079 ±5.5823E-6 -8423.55623
±0.25513
2723184.05649 ±4894.69843
/s 0,000012 0,000016 27 0,000020
100 80 60 Ni 40 20 0 0 5 10 15 20 R i /nm 25 30 35 40 1,20E-011 1,00E-011 8,00E-012 6,00E-012 4,00E-012 0,0001 0,0002 0,0003 0,0004 0,0005 q 2 /nm -2 0,0006 0,0007 2,80E-008 2,78E-008 2,76E-008 2,74E-008 2,72E-008 2,70E-008 2,68E-008 2,66E-008 2,64E-008 2,62E-008 0,0008 2,60E-008 0,0001 0,0002 0,0003 0,0004 0,0005 q 2 /nm -2 0,0006 0,0007 28 0,0008
100 80 60 Ni 40 20 0 0 2,00E-012 1,50E-012 50 100 150 R i /nm 200 250 300 2,00E-007 1,80E-007 1,60E-007 1,40E-007 1,00E-012 1,20E-007 5,00E-013 0,0001 0,0002 0,0003 0,0004 0,0005 q 2 /nm -2 0,0006 0,0007 0,0008 1,00E-007 0,0001 0,0002 0,0003 0,0004 0,0005 q 2 /nm -2 0,0006 0,0007 29 0,0008
10 5 20 15 Ni 0 0 4,40E-013 4,20E-013 4,00E-013 3,80E-013 3,60E-013 200 0,0001 0,0002 0,0003 400 R i /nm 600 0,0004 0,0005 q 2 /nm -2 800 1000 0,0006 0,0007 5,70E-007 5,65E-007 5,60E-007 5,55E-007 5,50E-007 5,45E-007 5,40E-007 5,35E-007 5,30E-007 5,25E-007 0,0008 5,20E-007 0,0001 0,0002 0,0003 0,0004 0,0005 q 2 /nm -2 0,0006 0,0007 30 0,0008
Combining static and dynamic light scattering, the
r
-ratio:
r
R g R H
topology homogeneous sphere hollow sphere ellipsoid random polymer coil cylinder of length l, diameter D r -ratio 0.775
1 0.775 - 4 1.505
1 3 ln
l D
0.5
for polydisperse samples:
r
R g
2
Z
Z
Strategy for particle characterization by light scattering - A
Sample topology (sphere, coil, etc…) is known yes Dynamic light scattering sufficient (“particle sizing“) no Static light scattering necessary Time intensity correlation function decays single-exponentially yes no Only one scattering angle needed, determine particle size (R H ) from Stokes-Einstein-Eq.
(in case there are no particle interactions (polyelectrolytes!) Applicability of commercial particle sizers!
Sample is polydisperse or shows non-diffusive relaxation processes!
to determine “true” average particle size, extrapolation q -> 0 - to analyze polydispersity, various methods
Strategy for particle characterization by light scattering - B
Sample topology is unknown, static light scattering necessary
Kc R q
2 yes no Particle radius between 10 and 50 nm: analyze data following Zimm-eq. to get:
M W R g M z w A
2 Particle radius larger than 50 nm and/or very polydisperse sample: use more sophisticated methods to evaluate particle form factor Dynamic light scattering to determine
R H
R H
1
z
1 Estimate (!) particle topology from
r
R g R H
2. Static Light Scattering – Selected Examples
1. Galinsky, G.;Burchard, W. Macromolecules 1997, 30, 4445-4453 Samples:
Several starch fractions prepared by controlled acid degradation of potatoe starch ,dissolved in 0.5M NaOH Sample characteristics obtained for very dilute solutions by Zimm analysis: sample LD11 LD16 LD12 LD19 LD18 LD17 LD13 10 -6
M w
(g/mol) 0.92
1.87
5.20
14.5
43 64 97
R
(nm) 36 48
g
70 113 180 190 233 10 4
A 2
[(mol cm 3 )/g 2 ] 1.00
0.60
0.28
0.13
0.082
0.060
0.025
Normalized particle form factors universal up to values of
qR g
= 2
Details at higher q (smaller length scales) – Kratky Plot:
C
form factor fits: 1 2 1
1
C
6 2 2
C
related to branching probability, increases with molar mass
Are the starch samples, although not self-similar, fractal objects?
q
d f
log
d f
log
q
- minimum slope reached at qRg ≈ 10 (maximum q-range covered by SLS experiment !) - at higher q values (simulations or X-ray scattering) slope approaches -2.0 - characteristic for a linear polymer chain (C = 1). - at very small length scale only linear chain sections visible (non-branched outer chains)
2. Pencer, J.;Hallett, F. R. Langmuir 2003, 19, 7488-7497 Samples:
uni-lamellar vesicles of lipid molecules 1,2-Dioleoyl-
sn
-glycero-3-phosphocholine (DOPC) and 1-stearoyl-2-oleoyl-
sn
-glycero-3-phosphocholine (SOPC) by extrusion
Data Analysis:
monodisperse vesicles
R o R i
R o
3 3
R i
3 2 2
R o
3 1
o
R i
3
qR o
1
i qR i
2 1
sin
x x
2 cos
x x
thin-shell approximation sin
qR
2 small values of
qR
, Guinier approximation exp 2
q R g
2 3
R g
2 3
R o
2 5 1 1
R R i o R R i o
5
3
typical
q
-range of light scattering experiments: 0.002 nm-1 to 0.03 nm -1 vesicles prepared by extrusion: radii 20 to 100 nm => first minimum of the particle form factor not visible in static light scattering
particle form factor of thin shell ellipsoidal vesicles, two symmetry axes (a,b,b)
0 1 sin
qu
2
dx u
2
a x
2
b
2 1
x
2
x
cos
k k
0 prolate vesicles, surface area 4 (60 nm) 2 oblate vesicles, surface area 4 (60 nm) 2
anisotropy vs. polydispersity:
monodisperse ellipsoidal vesicles
b a
sin
qR
2
dR
1
a
2
b
2
R R
2
b
2 polydisperse spherical vesicles 0 0
R
2
,
R
2
sin
qR
2 static light scattering from monodisperse ellipsoidal vesicles can formally be expressed in terms of scattering from polydisperse spherical vesicles !
=> impossible to de-convolute contributions from vesicle shape and size polydispersity using SLS data alone !
combination of SLS and DLS:
DLS: intensity-weighted size distribution => number-weighted size distribution (fit a,b) => SLS: particle form factor
,
result:
polydisperse ( D R = 10%) oblate vesicles, a : b < 1 : 2.5 input for a,b – fits to SLS data
3. Fuetterer, T.;Nordskog, A.;Hellweg, T.;Findenegg, G. H.;Foerster, S.; Dewhurst, C. D. Physical Review E 2004, 70, 1-11 Samples:
worm-like micelles in aqueous solution, by association of the amphiphilic diblock copolymer poly-butadiene(125)-b-poly(ethylenoxide)(155)
Analysis of SLS-results:
monodisperse stiff rods asymmetric Schulz-Zimm distribution
polydisperse stiff rods
2
qL
0
qL
0 sin
ql dql
k
sin
qL
1
qL L w
2
2
k
1
L k
exp 1
k
1
M w
1
L L w M n
1
0
Koyama, flexible wormlike chains 1
l K
2 0
l K
l K
x
exp 1 3
q
' 2 sin
dx
Holtzer-plot of SLS-data :
q
R Kc
vs
.
q
plateau value = mass per length of a rod-like scattering particle
fit results:
(i) polydisperse stiff rods: (ii) polydisperse wormlike chains:
L w
389
L w
380
w w M n
1.2
M n
2.0,
l K
410
nm
Analysis of DLS-results:
1
n
3 0
S n
exp
D q T
2 2
n
2
n
1
D R
amplitudes depend on the length scale of the DLS experiment: - long diffusion distances (qL < 4): only pure translational diffusion S 0 - intermediate length scales (4 < qL < 15): all modes (
n =
0, 1, 2) present according to Kirkwood and Riseman:
D T
kT
3
h
L
ln
L
,
D R
9
D T L
2 polydispersity leads to an average amplitude correlation function!
DLS relaxation rates : linear fit over the whole q-range:
significant deviation from zero intercept, additional relaxation processes or “higher modes” at higher
q
results:
D z
2 2
nm s
1
R H
nm R g R H
2 R g from Zimm-analysis and calculations!
4. Wang, X. H.;Wu, C. Macromolecules 1999, 32, 4299-4301 samples:
high molar mass PNIPAM chains in (deuterated) water
reversibility of the coil-globule transition:
molten globule ? surface of the sphere has a lower density than its center
Selected Examples – Static Light Scattering: sample problem solution
branched polymeric nanoparticles vesicles (nanocapsules) worm-like micelles PNIPAM chains in water at different T self-similarity (fractals) ?; distinguish size polydispersity and shape anisotropy in P(q) ?
characterization: length, R (R H coil : no rotation-translation coupling if qL < 4) – globule - transition g /R H details at qR > 2 by Kratky plot (P(q) q 2 vs. q), fitting parameters for branched polymers, simulation of P(q) at qR > 10 (SLS: qR < 10) => not fractal !
combine DLS (only size polydispersity !) and SLS to simulate expt. P(q) details at higher q by Holtzer plot (I(q) q vs. q), fit P(q), R g from Zimm-analysis at small q values R g from Zimm-analysis, R H DLS, decrease in R and R g by / R H
3. Dynamic Light Scattering – Selected Examples
1. Vanhoudt, J.;Clauwaert, J. Langmuir 1999, 15, 44-57
sample: spherical latex particles in dilute dispersions sample nominal diameter/nm diameter ratio s2 19 s3 54 s4 91 s5 19, 91 4.8
s6 19, 54 2.8
Data analysis of polydisperse samples:
1. Cumulant method (
CUM
), polynomial series expansion: ln
f
0.5
g
1 ln polydispersity index
PI
2
2 2 2 2
2 0 0
B
D app
q
2 2
B
particle diameter is a so-called harmonic z-average:
d
i
i n d i i
6
n d i i
5 (only for homogeneous spheres)
M i
2
d i
6 s7 54, 91 1.7
2. non-negatively least squares method (
NNLS
):
2
j N
1
g
1
j
i M
1
b i
exp
i
j
2
M
exponentials considered for the exponential series, yielding a set of coefficients
b i
defining the particle size distribution for decay rates equally distributed on a log scale.
3. Exponential sampling (
ES
): See 2., decay rates chosen according to:
n
1 1 exp
n
max 4. Provencher’s CONTIN algorithm:
i
1
i
2
g
1
i
B
e
d
2
LB
2 Numerical procedure to calculate a continuous decay rate distribution B( ), also called Inverse Laplace Transformation, enclosed in most commercial DLS setups. 5. double-exponential method (
DE
):
g
1 1
b e
2
Results:
sample nominal diameter diameter ratio
- CUM (1.)
PI
– CUM (1.)
d1,d2
– NNLS (2.)
d1,d2
– ES (3.)
d1,d2
– DE (5.)
s2 s3 s4 19 ± 1.5 54 ± 2.7 91 ± 3 20.3
0.029
55.0
0.009
87.0
0.008
Bimodal samples s5, s6, s7: I 1 (q=0) = I 2 (q=0) s5 19, 91 4.8
36.9
0.248
18, 81 s6 19, 54 2.8
29.5
0.191
16, 50 19, 54 18, 54 s7 54, 91 1.7
69.0
0.069
Note: bimodal samples with d2/d1 < 2 (s7) beyond resolution of DLS !
2. van der Zande, B. M. I.;Dhont, J. K. G.;Bohmer, M. R.;Philipse, A. P.
Langmuir
2000
, 16, 459-464
sample (TEM-results): colloidal gold nanoparticles stabilized with poly(vinylpyrrolidone) (M = 40000 g/Mol) system length [nm]
L
D L [nm] diameter [nm]
d
D d [nm] aspect ratio L/d Sphere18 Sphere15 Rod2.6a
Rod2.6b
Rod8.9
Rod12.6
Rod14 Rod17.2
Rod17.4
Rod39 Rod49 18 15 47 39 146 189 283 259 279 660 729 5 3 17 10 37 24 22 60 68 20 18 15 17 15 20 15 16 17 15 3 3 3 3 3 3 3 3 3 2.6
2.6
8.9
12.6
14.0
17.2
17.4
39.0
49.0
DLS setup and data analysis:
Kr ion laser (647.1 nm far from the absorption peak of the gold particles (500 nm)) Measurements in vv-mode and vh-mode (depolarized dynamic light scattering DDLS) (v = vertical, h = horizontal polarization) intensity autocorrelation functions were fitted to single exponential decays, including a second Cumulant to account for particle size polydispersity
g
2
exp
2 vv-mode (only translation is detected): 2
D q T
2 depolarized dynamic light scattering (vh-mode) (translation and rotation are detected, no coupling in case qL < 5) 2
D q T
2 12
D R
translational diffusion coefficient D T determined from the slope, rotational diffusion coefficient D R from the intercept of the data measured in vh –geometry.
Results:
qL
< 5 q
2 / 10 14 m -2 q max L
> 5 ( ≈ 9) !
q
2 / 10 14 m -2
diffusion coefficients according to Tirado and de la Torre, using as input parameters length and diameter from TEM
D T
kT
3
h
L
ln
d L
0.100
d
2
D R
3
kT
h
L
3 ln
d L
0.050
2 system Rod8.9
Rod12.6
Rod14 Rod17.2
Rod17.4
Rod39 Rod49 10 -12 D T , exp [m 2 s -1 ] 6.0
4.9
2.9
4.0
3.5
1.2
0.7
10 -12 D T , calc [m 2 s -1 ] 8.4
7.4
5.2 6.0
5.6
2.9
2.8
D R , exp [s -1 ] 306 281 66 177 175 14 D R , calc [s -1 ] 2238 1258 396 563 452 46 30 values determined by DDLS systematically too small, because PVP-layer (thickness 10 – 15 nm) not visible in TEM !
Selected Examples – Dynamic Light Scattering: sample
bimodal spheres stiff gold nanorods
problem solution
size resolution length and diameter in solution =?; deviation TEM – DLS ?
- double exponential fits - size distribution fits - CONTIN ; only if R 1 /R 2 > 2 depolarized DLS (vh) => D rot standard DLS (vv) => D trans ; deviation TEM-DLS due to PVP stabilization layer