Light Scattering

1 1 Kc2 = + 2 A2 c2 + .... R( ) P( ) M w K = optical constants c2 = polymer concentration P( ) = particle scattering factor, known for various particle geometries 1 2 A2 = 2 = 2 nd virial coefficient 2 V1 R( ) = excess Raleigh ratio = Rsolution Rsolvent
Light X-ray Neutron
~ excess scattering intensity We will derive this. Note the nice set of variablesthat we would like to be able to determine

Scattering arises from ()2 polarizability fluctuations ()2 electron density variations (b)2 neutron scattering length variation

Scattering arises from Density Fluctuations

A dilute gas in vacuum Consider small particles:
(situation:~ point scatterers)

dp << radiation
Particles dp Incident radiation

Scattered Intensity at scattering angle to a detector r away from sample: 1+ I 8 ( cos ) = polarizability of molecule I = I0 = incident beam intensity r For N particles in total volume V (assume dilute, so no
4 2 0 2 4 2

coherent scattering)

I' =

N I V

N = 1 + 4 V

= dielectric constant = n2, () = frequency dependent

Fundamental relationship: index of refraction polarizability

N n = 1+ 4 V
Can approximate

n gas

dn 1+ c dc

dn = refractive index increment dc

2

c = conc. of particles per unit volume

dn dn n 2 = 1 + c 1 + 2 c + . . . gas dc dc 1 (dn / dc) c Solving gives = 2 (N /V ) the polarizability

So by analogy for a polymer-solvent solution:

dn n n0 + c2 dc 2

2 n 2 n0 + 2 n0

dn c2 dc 2

Rayleigh and the Molecular Weight of Gases

2 2 1+ N 8 ( cos ) (dn / dc) c I' = V 4 r 2 2 (N /V ) 2

simplifying

1+ I 0 2 2 ( cos 2 ) dn 2 2 ' I = c 4 2 r (N /V ) dc

and since

N c = V M / N AV

This expression contains several parameters dependent on scattering geometry, so we define Rayleigh Ratio, R as
2

which equals

dn 2 2 M c dc R= 4 N AV
dn 2 2 dc K 4 N AV
2

I' R= I 0 ( cos 2 )/r 2 1+

Or just

R = KMc
Note, for polymer-solvent solution:

Where K is a lumped optical constant

2 dn 2 2 n 0 dc 2 K 4 N AV

Rayleigh measured the molecular weight of gas molecules using light scattering!

Scattering from Fluctuations II

A dilute gas of polymer chains in solution
polymer coil (solute) solvent Incident radiation

Debye: Re-identify fluctuations as chains in a solvent and extend Rayleighs idea to polymers in solution
pure solvent

coil/solvent

dp

Vsolution

Now for polymer coils: ~ dp Recognize 4 features in a binary component system: 1. Each cell has on average, the same number of solvent molecules but there are variations. Fluctuations in solvent density will give rise to some (weak) scattering (subtract off pure solvent scattering). 2. Fluctuations in the number of solute molecules (chains) will give rise to significant scattering 3. Fluctuations in the concentration of solute create osmotic forces 4. Polymer chains are large and cannot treat them as point scatterers P() 1

The Features of Excess Scattering

Define R = Rsolution Rpure solvent = Excess Rayleigh Ratio Feature 2. Remaining scattering arises from fluctuations in solute concentration 2 dn I 0 2 2 ( cos 2 ) dn 2 2 1+ 2 R depends on I' = c dc and (c 2 ) 2 4 r 2 (N /V ) dc
Feature 1.

Einstein-Smoluchowski mean squared concentration fluctuation per unit vol.

(c )
2

Feature 3. A local osmotic pressure will arise due to local concentration differences, this effect acts to suppress solute concentration fluctuations. Note in the gas-vacuum system, such an effect is not present. 1 ( ) 1 2 = RT + c 2 + L 2 V1 2 c2 M 1 = RT + 2A 2c 2 + L c 2 M

RTc2 VN AV ( /c2 )

Feature 3 contd
2 2 dn 2 n 0 dc c 2 2 2

R solution R solvent =

1 N AV + 2A 2c 2 + L M
4

Polymer in solution
1 R = Kc 2 1 + 2A c + L 2 2 M

Gas in vacuum
2 dn 2 M c dc R= 4 N AV 2

similar to the Rayleigh scattering for gases but with a new term depending A2

Scattering from Polymer Solutions

Feature 4. Polymer chains are large and can not be assumed as point scatterers for visible light ~ 6,328 He-Ne laser
1001000

Coil size ~ typically 100-1000

depending on molecular weight of polymer Therefore need to consider self-interference of monomers in polymer coil on scattered intensity

Therefore P() term is important

Scattering of Polymer Solutions

Introduce finite size chain scattering factor P()
R = Kc2 1 + 2A2 c2 +L M w 1 P( )

2 For a Gaussian coil P ( ) 2 u 1+ eu u

(Debye, 1939)

where

4 n 0 2 u = sin Rg 2

q = scattering vector,

u = q 2 Rg

In the limit of very small ,

Useful approximation for small, nonzero

P( => 0) = 1

P ( )

1 u u 1 and 1+ for u1/ 2 << 1 3 P ( ) 3

Scattering in a Polymer-Solvent System

In general, P() 1, c2 0, A2 0 therefore rewrite excess Rayleigh ratio as
Kc2 1 u = + 2A 2c 2 + L 1 + R M 3
u = q 2 Rg
2

2 16 2 n 0 2 2 Kc 2 1 = + 2A2c 2 + L1+ sin Rg 2 2 R M 3

Equation is valid for dilute solutions and scattering angles such that

u1/ 2 << 1
Next: how to plot scattered intensity vs. c2 and versus q2 to extract M and A2

Zimm Plot Analysis of Light Scattering Data

Light Scattering Experiment

qo

Sample

~
Collect Scattered Intensity at typically 8 angles: = 30, 37.5, 40, 60, 75, 90, 105, 120 Measure R typically for 5 solute concentrations:

Detector

c2 , c2 , c2 , c2 , c2 Scattering Measurements: 1 2 3 4 5 1. Pure (no dust!) solvent Rsolvent 2. Polymer solutions [c2,i , I(i)] Rpolymer 3. Construct Zimm Plot
Note:

R = Rpolymer-Rsolvent

Kc2 R

is a function of 2 variables, c2 and .

Zimms cool idea was to plot y(p,q) vs (p+q) to separate out dependence on each variable: called Zimm plot

Construction of a Zimm Plot

The Master Equation:
2 16 2 n 0 sin2 Kc 2 1 2 R2 = + 2 A2 c 2 + ...1+ g R M w 3 2

do double extrapolation:

0o c2 0

A2 & M w <Rg2> & M w

fixed for a set of concentrations

conc.

Kc2 R(i , ci )

Mole/gram Units

dilute
an g small

large

angle

c fixed for a set of angles 2

le

sin2(/2) + k c2

dimensionless axis

Zimm Plot
Kc2 R ( i , ci )
30o 2 x 10-3 1.5 x 10-3 3.18 2.73 2.29 2.10 1.92 37.5o 3.26 2.76 2.33 2.14 1.95 45o 3.25 2.81 2.37 2.17 1.98 60o 3.45 2.94 2.53 2.32 2.16

Sample Data Set

75o 3.56 3.08 2.66 2.47 2.33 90o 3.72 3.27 2.85 2.64 2.51 105o 3.78 3.4 2.96 2.79 2.66 120o 4.01 3.57 3.12 2.93 2.79 135o 4.16 3.72 3.29 3.10 2.96 142.5o 4.21 3.75 3.38 3.21 3.11 150o 4.22 3.78 3.37 3.2 3.12

c2
g/cm
3

1 x 10-3 0.75 x 10-3 0.5 x 10-3

dn 2 2 2 2 n 0 dc K= 4 N AV n 0 = 1.5014 dn = 0.106 cm 3 g1 dc

For c2 = 0.002, = 30

Kc2 = 3.18 R( i , ci )

= 5.461x105 cm
N AV = 6.02 x1023 mole1

Plotting the Data

vertical data set =30 c2 varies (decreases)

Extrapolating to zero concentration

horizontal data sets variable c2=2x10-3

( .267, 3.18)

Kc2 R

Image removed due to copyright restrictions. Please see, for example, http://web.umr.edu/~WLF/MW/orangeline.gif, from http://web.umr.edu/~WLF/MW/Zimm.html

Plot typical data point :

= 30,c2 = 2 103 g / cm 3

These points are determined by extrapolating the equations of each regression line

y-axis: x-axis:

Kc2 = 3.18 R

30 3 sin 2 + k(210 ) 2 = 0.067 +100(210 3 ) = 0.267

Pick constant to spread the plot

x = 0.067

Plotting the data contd

Extrapolating to zero scattering angle
(.267, 3.18)

100c2
Image removed due to copyright restrictions. Please see, for example, http://web.umr.edu/~WLF/MW/yellowline.gif, from http://web.umr.edu/~WLF/MW/Zimm.html

Extrapolate the extrapolated data to obtain:

Mw , A2
Kc2 R

Rg

Image removed due to copyright restrictions. Please see, for example, http://web.umr.edu/~WLF/MW/first.gif, from http://web.umr.edu/~WLF/MW/Zimm.html

Rg

An example:

A2, Rg2 and

Kc2 R0 =( 1 + 2A2c2+ Mw
Kc

Mw
) ( 1+
0
2

Note: units on y-axis are mole/g. The x-axis is dimensionless. <s2>z )

4.0 = 0 line mole g

-1

The intercept occurs at a y-value ( R2) of 1.36 x 10-6 mole/g. Hence, Mw = 0 7.35 x 105 g/mole. Now, consider the c2= 0 line. Where c2 = 0, Kc2 R0 = 1 ( 1+ Mw 1 ( Mw
0

3.0

0.1375 1.25

KC2/DR x 10

< s2>z )

2.0

c2= 0 line
0.55 0.4315 100c2 + Sin2 2 0.5

slope =

<s2>z

Measure the slope of the c2 = 0 line.

y

From intercept,

Mw

1.0

1.0

slope =

= x

5.5 x 10-7mole/g 0.4375

M w = 735,000 g/mol

= 1.26 x 10-6 mole/g

<s2>z =

3( 1.26 x 10-6 mole/g ) (7.35 x 105g/mole)(5.461 x 10-5)2 16 2(1.5014)2

= 2.33 x 10-11cm2
Figure by MIT OCW.

getting A2, Rg2 and M w

(i)
r
2

< r6 s> =
2

= 6 < Rg >
Pretty big molecules!

= 6 (2.33 x 10-11 cm2) = 1.40 x 10-10 cm2 = 1.182x10-5

cm = 1,182 A

2 1/ 2

(ii) Now consider the = 0 line. At = 0 (neglecting higher order c2 terms)

Kc 2 1 = + 2 A2 c2 R M w

slope = 2 A2 The density of polystyrene (2) is 1.05g/cm3. The molar volume of benzene V1 = MW/1. The molecular weight of benzene is 78.11 g/mole and its density is 0.8787g/cm3 (from CRC Handbook of Chemistry and Physics).
from the Zimm plot
3 1.25 x106 mole /g 4 mole cm = 9.09x10 = 2 A2 slope = g2 0.1375 x102 g /cm 3

A 2 = 4.55x10

mole cm 3 g2

A2, Rg and M w
2
(ii) The relationship between A2 and is given by:
1 = A 2 V1 2 2 2

Substituting the values,

3 1 4 mole cm = 4.55x10 g2 2

78.11g /mole 2 1.05 g /cm 3 ) ( 0.8787 g /cm 3

= 0.455
(So, of course, the polymer solution used for light scattering will be a single phase since < 1/2 for miscibility of solvent and polymer).

Noncrystalline Materials
The structure of noncrystalline materials (i.e. polymer glasses, amorphous polymer melts) is characterized by short range order (SRO) SRO develops due to excluded volume and locally dense packing (glasses ~ are only 10% less dense than crystals) Pair distribution function g(r) is a dimensionless function used to quantify SRO. In polymers SRO is primarily due to covalent intra-molecular bonds and neighboring intermolecular packing. Properties of noncrystalline polymers are heavily influenced by * , the characteristic relaxation time relative to an experimental observation time, t. liquid (melt) t * rubbery t * *(T) ? t * glassy

Structural Features of Noncrystalline Polymers

SRO in Polymers

C
fixed
3

Due to covalent intra-molecular bonds

Due to chain-chain inter-chain distances ~ 5

C1 C2 C1 C3 C1 C4

= constant = constant = varies

gives rise to peaks at large q

Inter-chain
C1 C1 ~ 5

c1

C1

Typical X-ray signature of noncrystalline materials broad overlapping peaks from multiple distances:

I(q)

interintraq

Pair Distribution Function g(r)

1001000
Figure by MIT OCW.

The Pair Distribution Function g(r) addresses the distances between the centers of mass of pairs of units. Since glasses and liquids are isotropic, the magnitude of the inter-unit distance is of interest. The scalar distance rij between molecule i and molecule j is: rij = ri - rj

g(r) contd
rij = ri - rj Characterize the set of distances {rij} from an average unit i to every other unit j =1 .. N. g(r) counts the number of units dn in a small spherical shell sampling volume element of size dv at each distance r from a reference unit, dv = 4r2dr The statistical average of these numbers for many units chosen as the reference is divided by the Figure showing the pair-distribution functions for gas, liquid or glass, and average unit density monatomic crystal removed due to copyright restrictions.
See Figure 2.5 in Allen, S. M., and E.L. Thomas. The Structure of Materials. New York, NY: J. Wiley & Sons, 1999.

1 dn (r , r + dr ) g (r ) = dv (r , r + dr )

Features of g(r) in Glasses & Liquids

g(r)

At several unit diameters, the average number of units/vol. becomes constant

1.0

2Ro

Figure by MIT OpenCourseWare.

Due to excluded volume, g(r) = 0 for distances less than 2Ro Liquids and glasses are strongly correlated at the shortest distance between 2 units, the maximum occurs at slightly > 2Ro, this largest peak is the average distance to the first shell to the nearest-neighbor unit. C1 C2 C1 C2 g(r)intra C3 g(r)inter Superposed correlations