MS2000 Customer Training Course

Calvin Wong
Application Specialist
E-Mail: [email protected]
Contents

› Basic Principles
› Obtaining and Understanding the Size
Distribution
› Data quality
› Analysis Models
› Reporting the Results
› Method Screening – Wet Measurement
› Method Screening – Dry Measurement
Mastersizer 3000

Basic Principles
Why do we measure particle size?

› Predicting product performance

• Dissolution rate

• Content Uniformity

• Mouth Feel Assessment

• Viscosity

• Colour
Basic concepts of particle size

› The MS2000 is an instrument designed

to measure particle size.

› But what do we mean by “particle size”?

Concepts

› A very general definition of a “particle”…

“Any condensed-phase tridimensional discontinuity in

a dispersed system may generally be considered a
particle…”
(NIST, US Department of Commerce, Special Publication 960-
3).

› This covers:
▪ Droplets in air or in a liquid medium
▪ Solids in air or a liquid medium
▪ Gas bubbles in a liquid medium
Concepts

› People normally associate particles with images

such as these…

…but what are their sizes?

Concepts

You are given a regular-shaped object and a ruler and

asked to give a one-number indication of its size.

What would your reply be ?

Concepts

You may reply: “360x140x120mm”

Which might be correct but it is not one number.

It is not possible to describe the

3-dimensional object with a single number
Concepts – Equivalent spheres

The cylinder on the

far left has the same
volume as a sphere
of 213µm diameter.
Concepts – Equivalent spheres
› All particle size analysis techniques measure some property of
a particle and reports results as the equivalent spherical
diameter based on this measured parameter
Laser diffraction

How does it work?

Laser Diffraction – The light scattering pattern
from a group of particles
Laser Diffraction – Light Scattering
The Mastersizer 2000
The Mastersizer 2000

› The scattering intensity observed for sub-micron

particles is increased by using 466nm blue light
source
 The Mastersizer 2000

› The angular scattering data is presented in real-time in

the measurement window of the Mastersizer software.
› Detector numbers correspond to increasing scattering
angle…

Detectors 1-33: Small angle

Detectors 34-42: Large angle
Detectors 43-44: Extreme angle (Red)
Detectors 45-46 : Extreme angle (Blue)
Detectors 47-48: Backscatter (Red)
Detectors 49-50: Backscatter (Blue)
Detector 51- Extinction-Red
Detector 52- Extinction-Blue

Low Angle High Angle

The Mastersizer 2000
Typical Data Set – Coarse Particles

Data Graph - Light Scattering

1000
900
800
Light Energy

700
600
500
400
300
200
100
0
1 3 5 7 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51
Detector Number
Glass Beads, 19 Oct 1999 11:12:39
The Mastersizer 2000
Typical Data Set – Sub-Micron Particles
Mastersizer 2000

Obtaining and Understanding the Size

Distribution
Laser diffraction – getting the result

› Laser Diffraction measures the intensity of

scattered light over a range of angles, from an
ensemble of particles of unknown size
› However….
› Scattering theory predicts the angular scatter
from an ensemble of particles of known size and
optical properties
› How do we use scattering theory to arrive at a
size distribution from the laser diffraction data?
▪ This involves the use of a mathematical inversion
process
Calculation of the size distribution
Laser diffraction – getting the result

› Two sets of Data; measured scattering data and

theoretical fit data
› Residual is calculated area between the two
Understanding the size distribution

› The Mastersizer 3000 system is designed such

that equal volumes of particles of different sizes
produce a similar measured scattering intensity.

› The size distribution is reported as a volume

distribution as this best reflects the sensitivity of
the system.

› What does this mean in practice?

Understanding the size distribution
› Volume distributions, as the name suggests, are based on the
volume occupied by their constituent particles.
› One million 1µm spherical particles will occupy the same volume
as one 100µm particle.

› This should be remembered if you ever compare volume

distribution results with a number-based distribution.
› Conversion from one to another is often error prone!
Understanding the size distribution

› Imagine we have 3 particles with diameters of 1,2,3

units.

› If these were measured by microscopy – a technique

that produces a number distribution - then we would
produce the following distribution…
Understanding the size distribution

Number Distribution
35

30
Frequency (number %)

25

20

15

10

0
1 2 3

Diameter
Understanding the size distribution

Volume Distribution
80

70
Frequency (volume %)

60

50

40

30

20

10

0
1 2 3

Diameter
Understanding the size distribution

› It is normal practice to present volume

distributions with a logarithmic X-axis (abscissa)


Size distribution statistics

› The two most important weighted means are:

› D[3,2] or Sauter mean diameter

▪ This is the surface area weighted mean

› D[4,3] diameter
▪ This is the volume weighted mean
Size distribution statistics

› D[4,3] is sensitive to the presence of large

particles

D[4,3] 11.2

D[4,3] 16.9
Size distribution statistics

› D[3,2] is sensitive to the presence of small

particles

D[3,2] 35.571 µm
D[3,2] 27.457 µm
Size distribution statistics

› The median is the exact midpoint of the distribution

› The mode is the most commonly occurring size
class
Size distribution statistics

› The median of a particle size distribution is the

size above [ and below ] which we can find 50%
of all particles
› It is commonly referred to as:
▪ The 50th percentile or the
▪ Dv50 or the
▪ D[v,0.5]
› The “v” emphasises that the median was
obtained from a volume distribution
› Other percentiles such as the Dv10 and Dv90
can also be specified
Percentiles

› The Dv10, Dv50 and Dv90 are the default

percentiles shown in the analysis report.

› These percentiles are commonly used to

describe a particle size distribution but are by no
means the only ones that can be used.

› Percentiles allows us to detect changes in our

distributions that might have occurred due to
changes in sampling, dispersion or measurement
conditions.
Other size distribution parameters
› The span of a distribution provides a measure of
the distribution width:

Span = (Dv90-Dv10)/Dv50

› The span has a value of 1 for a symmetrical

distribution

› The uniformity is another parameter describing

distribution spread.
▪ For narrow distributions this will be a small number
Mastersizer 2000

Data Quality
Data quality - introduction
› What do we mean by data?

▪ Data is the fundamental light scattering caused by your

sample
• It is not the particle size result
• It is independent of the optical model.
▪ A stable result requires stable data
Data quality - introduction
› What Constitutes Quality Data?
› A good background measurement, taking into account
▪ cleanliness of cell windows and dispersant
▪ alignment of the system
▪ stability over time
› Sample data
▪ gaining sufficient signal to noise
▪ avoiding negative data
▪ limiting the effect of multiple scattering
▪ avoiding beam steering

In general this applies to all laser diffraction instruments

Data components -Types of data

Background data

Sample data
Background data and system cleanliness
› A good measurement requires a clean, stable background
› This should show progressive decrease across the detector
range

Less than 100 units

on detector 1

Less than 20 scattering

units by detector 20
A clean background – wet system

› A good clean background on a wet system should look very similar

to this...

▪ Less than 100 units on the 1st detector

▪ Limited fluctuations over time
A clean background - dry system

The flow of air through a dry cell will cause a slightly higher
degree of fluctuation over time than seen in a wet system
▪ The same decrease in the signal is observed across the detector
range.
▪ The same maximum acceptable value of 100 applies
Poor background - material stuck to the
windows
› A ‘hump’ in the data is often an indication of
material stuck to the cell windows

Dirty Windows?
Poor background – contaminated dispersant
› Particulate contamination causes fluctuations in
the background over time
› Bubbles in the dispersant can also result in a
similar background
Poor background – misaligned system.

› Castellation indicates a mis-aligned system.

› This may be caused by:
▪ Dirt on the cell windows. Address these and then re-align.
▪ Changes in the temperature of the dispersant

Castellation
Poor backgrounds –
thermal gradients in organic dispersants
› The sample dispersion unit may be at 30°C+, while the solvent
is generally much cooler
› This gives rise to thermal gradients resulting in fluctuations in
the laser beam, alignment problems and poor backgrounds

Thermal gradient in dispersant

How will I recognise thermal gradients?

- By wild instability of the scattering on the first few detectors….

› After a few minutes the background should stabilise as the temperature
differences equilibrate.
› If it does not stabilise, fit the cover of the dispersion unit to reduce
surface evaporation.
Poor backgrounds – a final check...
› After background subtraction, the inspection window shows;
▪ Random scattering data
▪ Positive and negative data
› Any ‘blocks’ of scattering will indicate dispersant contamination
Sample data – adding the right amount of
sample
Terms used in this section:

Obscuration: The amount of laser light blocked and /or

scattered by the sample: a guide to the sample concentration

Signal-to-noise ratio: The magnitude of the sample data in

relation to the background data

Multiple scattering: The phenomenon caused by light

scattered by a particle being scattered by other particles when
too much sample is added
Sample addition

How much sample should be added to the dispersion unit?

▪ Too little:
• Signal to noise ratio may be poor, or:
• Not enough sample may have been added to be representative of
the bulk - particularly if the sample is very polydisperse

▪ Too much:
• Multiple scattering may affect the reported particle size distribution -
particularly if the material is small (typically < 10 microns)

What is the ‘correct obscuration?’

Sample addition

› How much sample should be added to the dispersion unit?

▪ How many particles are present?

› 60μm Glass Beads

▪ Polydisperse
▪ Few particles/ml
▪ More sample
improves sampling

› 2μm Latex
▪ 1000’s particles per
drop
▪ Multiple scattering
may occur at low
obscurations
Sample addition –
Guide to target obscuration values
› Wet measurement › Dry measurement

▪ If the sample is fine, a ▪ Fine 0.5%~3%

low obscuration is best, ▪ Coarser 0.5%~6%
~5-10% will improve the ▪ Keep the obscuration
signal to noise ratio within the required values
▪ If the sample is coarser, by enabling the
obscuration can be obscuration filter in the
increased to 5%~12% software.
▪ If the sample is very ▪ Too high an obscuration
polydisperse, 15-20% reduces dispersion
may be needed to efficiency and leads to
improve sampling false peaks
▪ Too high an obscuration
will give multiple
scattering
Signal to Noise ratios – large particles
(42.58 micron glass beads measured at an obscuration of only 7%)

Maximum signal ~250units

High
High
sample
sample
signal
to signal
background
to
background
ratio ratio

Note that the signal to noise ratio is usually high for large particles
because these scatter light more strongly.

Consequently, signal to noise ratio is less of an issue for large particles.

Signal to noise ratios - small particles
(1 micron Latex measured at an obscuration of 5% )

Maximum signal ~8units

The signal to noise ratio is the amount of sample data relative to the
background data.
Because small particles scatter light weakly, it is important that the
background does not swamp the data signal.
However, in this graph, the data is good since it falls where there is little or
no overlap between the sample data and the background data.
Multiple Scattering

Detector
Wet cell
High detector
number
High detector number
Scattering indicates
small particles

Low detector
number

Increasing the obscuration (concentration) makes multiple scattering more likely.

This leads to higher angle scattering which corresponds to smaller particles.
Wet analysis - multiple scattering

….leading to exaggerated fines being interpreted

If in doubt, carry out an obscuration titration to determine the

effect of measuring at increasing obscurations upon the
resultant particle size distribution.
Beam Steering
› The peak on the right of this graph suggests the presence of large particles.
› However this could be caused by Beam steering, resulting from:
• Thermal instability following the use of ultrasound in solvents
• Partial dissolution of the sample in the dispersant - changing its
refractive index
› Or real large particles?

Are these
real?
Negative data - introduction

› If the background is measured before the dispersant

stabilises then
▪ The background is over-subtracted
▪ The inspection stage shows significant negative data
▪ If this occurs, skip back and re-measure the
background
Dry measurement data quality

› For dry measurements, noise on detectors 1 to 10 can be

problematic, particularly when fine powders are measured
▪ The problem is caused by fluctuations in the
background data owing to small temperature variations
in the air flow
▪ The jagged display see here is symptomatic of these
fluctuations
▪ Measuring a longer background can improve this
Noisy background data
How to improve dry measurement data quality?
› Check:
▪ Are the cell windows clean?
▪ Is the air pressure correct?
▪ Does the air filter need changing?
• Is there oil droplet contamination or moisture in the air supply?
▪ Is the feeder earthed against static electricity ?
▪ Is the vacuum bag full?

› Is the sample flow even?

▪ if the sample obscuration is high, try lowering the feed rate or
adjusting the hopper height
▪ try changing the height of the hopper, different basket, ball bearing
▪ try a different feed tray: often one tray design will deliver a more even
sample flow for a particular material

› Use Fine Powder Mode when material less than 10 microns is present
Summary - data quality
› Background data
▪ Make sure that:
• Material is not stuck to the cell windows
• There is no dispersant contamination
• There are no thermal gradients
• That the system has been properly aligned

› Sample data
▪ Check that
• There are reasonable signal to noise levels
• There is no multiple scattering
• There is no negative data
• There is no noisy data
• The inner detector data is free from castellation
• There is no beam steering
Mastersizer 3000

Optical Properties
Scattering models - Fraunhofer v Mie

› Laser diffraction requires a model that accurately

defines the light scattering behaviour of all
particles
› There are currently two popular choices
available, Mie Theory and the Fraunhofer
Approximation
› Mie theory has been found to be more accurate
over a wider range of sizes – particularly for
particles less than 50 microns in size.
Scattering models - Fraunhofer v Mie

› Older instruments used the less accurate

Fraunhofer Approximation. This was due to
limited computer processing power.

› Since 1986 the preferred model has been Mie

Theory which correctly predicts the scattering at
all wavelengths of light at all angles.
Scattering models - Mie theory
› Unlike Fraunhofer, Mie theory:
▪ Accounts for the interaction of light with matter
• Assumes that spherical particles are present

▪ Is valid for all wavelengths, scattering angles and sizes

of particle

▪ Correctly predicts scattering intensities

• (Correctly predicts secondary scattering)
▪
▪ “For particles smaller than about 50µm the Mie theory
offers the best general solution.”
-ISO 13320-1
 Predicted light scattering- Mie theory

Scattered light

Incident light

Absorption

Scattered light
Predicted light scattering - Mie Theory

› The Mie scattering model takes into account light

passing through the particle.

Scatterin
gIntensit Angle of
y scatter

Particle size

Angle of scatter
Scattering models– Fraunhofer Approximation

› Accounts only for classical diffraction around a

disc
Scatterin
gIntensit Angle of
y scatter

Particle size

Angle of scatter
Scattering models- Fraunhofer Approximation

› Disadvantages
▪ Will produce incorrect answers when…
▪ Particles are <50µm
▪ The angle of scatter becomes large and secondary
scattering occurs.
▪ The relative refractive index is small (<1.3) - this
equates to a particle refractive index of 1.73 in water.

› Claimed Advantage
▪ “No need to know the optical properties of your
material.” Which implies that optical properties are
required for Mie theory.
Using the Fraunhofer approximation

› Located under particle type

Optical properties – What are they?

› For the application of Mie theory, 3 properties

need to be known.
These are:
▪ The Refractive index of the dispersant.
▪ The Refractive index of the sample material
▪ The Imaginary Refractive index of the sample material
(often referred to as the absorption)
Optical properties – What are they?

› The Imaginary refractive index (absorption)

▪ This can be determined by looking at the dispersed

sample under a microscope by looking at its shape,
transparency and internal structure.

▪ Absorption is generally only required to a factor of 10

e.g. 0.1 or 0.01 not 0.023
Estimating absorption from particle appearance

Appearance Imaginary RI Example

0 Latices

0.001 Emulsions

0.01 Crystalline milled powders

0.1 Slightly colored powders

1.0+ Highly colored

(complementary) and metal
powders
Sources of Refractive Index values

› There are three routes to determining the

refractive index (RI) of a product

▪ Look it up
▪ Use a Refractometer – see appendix 1
▪ Use a Microscope – see appendix 2
Sources of refractive index information

› Look them up
▪ Your Malvern manual has a RI appendix, so does ISO
13320
▪ For many dispersants, you need look no further than
the supplier’s label
▪ CRC Handbook – refractive indices can be found for
many materials.
▪ Internet:
• Google Search including Google Scholar, the new
scientific search engine from Google.
• http://www:luxpop.com
• http://www.webelements.com/webelements/elements/te
xt/periodic-table/phys.html
Choosing optical properties
› You can estimate the optical properties based on
typical values of similar materials. (A Refractive
Index is generally only required to 2 decimal
places e.g. 1.42 not 1.4234)
› Some families of similar materials are:
▪ Plastics and elastomers = 1.38 - 1.57
▪ Organic compounds = 1.4 - 1.7
▪ Inorganic salts = 1.52 - 1.8
▪ Metal Oxides = 1.6 - 2.5
› Use the estimated refractive index as a starting
point and examine the fit to confirm the suitability
of the value chosen.
Assessing the fit
› Once you have chosen an optical model, assess
the result and the data “fit” to determine its
suitability
› The data fit should only be used as a guide to
the correctness of the chosen RI.
▪ The best data fit does not always give the right result
› Is the result sensible for what you know about
the manufacturing/production route?
▪ Milled powders - should show tail of fines
▪ Emulsions - typically log-normal distribution
▪ Classified materials - display a cut–off at either end of
the distribution
Using the data fit

› Here we see an overlay of the measured

scattering data and the data predicted based on
the optical properties.
› The precision with which these two curves
overlay is known as the “data fit”
› The ‘goodness’ of this fit is known as the
‘Residual’
Inspecting fit data: refractive index

› A poor fit in the low detectors indicates an

incorrect choice of refractive index

Poor fit indicates

incorrect choice of
refractive index
.
Inspecting fit data : absorption index

› Misfits to the extinction detectors 51 and 63

indicate incorrect absorption index.
Poor data fit here
indicates poor choice
of absorption.
Weighted and un-weighted data fits
Example - weighted

Example - unweighted
Residuals - good “rule of thumb”
› For a normal size distribution, the residual should typically
be less than 1%
› For a good fit, the weighted and unweighted residuals
should be of similar orders of magnitude
› A low residual is good, but the distribution should be
believable

Poor
agreement

Good
agreement
Residuals for very narrow distributions
› Residuals less than 1.0 are not always achievable.
› Narrow distributions can give higher residuals than
expected

Residuals greater than 1.0

However, the distribution

is correct
Assessing the fit

› The user is seeing an “unexpected” mode of

small material.
› The optical properties used were:
▪ RI:1.4, Absorption: 0.01
Assessing the fit using 1.4/0.01

Weighted fit

Poor fit
Un-weighted fit – incorrect RI
Assessing the fit using 1.59/0.01

Weighted fit

Improved fit
Un-Weighted fit
Example 1 – Looking at the results

› Sample is calcium carbonate

▪ Reference RI is between 1.53 and 1.63
Mastersizer 2000

Analysis models
Choices of model
General purpose model

› The general purpose model is designed to fit

wide size distributions
› This model is the most commonly used, and is
appropriate for most milled or precipitated
products, emulsions etc.
› For an unknown sample, it should ALWAYS be
the first analysis model used.
Single narrow mode model

› This model fits material with a particle size

distribution of one decade of size or less
▪ e.g. 0.1-1µm, 1-10µm, 10-100µm, 100-1000µm.
› Such materials include latex standards, or
products which have been classified (giving them
narrow distributions).
Multiple narrow modes model

› This model is designed for known mixtures of

material, each less than one decade of size.
› It will be more sensitive than the General
Purpose model, but should be used with care as
it is less robust.
 Multiple narrow modes example

• This PVC sample contains a small proportion of fines, which

the general purpose model failed to resolve.
• As both modes are <1 decade of size, multiple narrow modes
is an appropriate model.

Multiple narrow modes

model detects very small
amount of fines General purpose used
Particle shape - irregular or spherical?
› Irregularly shaped small particles (<1µm) can
cause light to depolarise more strongly in one
direction than another
› Selecting “Irregular” enables the software to
correctly interpret the high angle scattering that
this phenomenon causes.
› Not all laser diffraction instruments have this
capability.
› “Irregular” is the best starting point as most
milled materials are irregular in shape.
› Spherical samples are rarer but will include glass
beads, latex spheres and emulsions.
Calculation sensitivity

› 2 choices are available, normal or enhanced

▪ Normal sensitivity - for routine use.
▪ Enhanced sensitivity - a more sensitive analysis, but
may be less robust.
Fine powder mode

› This mode is only available when using a dry

powder sample dispersion unit
› This removes the noise on the inner detectors
caused by temperature gradients in dry powder
feeder measurements
(See data quality section)
› It is only used in measurements of powders
which are finer than 600 microns
Mastersizer 3000

Method screening – Wet measurements

Wet Dispersion - Advantages

› Wet dispersion allows the user to maximise

measurement stability by changing dispersants,
ultrasonication and surfactants/additives.
Good wet dispersion – what to look for

› Start by selecting what you feel will be the most

appropriate dispersant and performing a set of
repeat measurements (same aliquot) before,
during and after ultrasonication.

› What should you look for?

Good wet dispersion -
Trend graph

Step 1: Measurements are Step 2: Measurements Step 3: Measurements

taken pre ultrasound are taken during are taken post
ultrasound ultrasound

No initial change in the There is still no The particle size

particle size is change in the particle has not changed
observed size
Good wet dispersion - Records view
› Look at the records view and ensure that obscuration is
stable and not increasing or decreasing with time

No obvious trend
in obscuration

› Note: you can also plot obscuration on the trend graph

Good wet dispersion - Data overlay
› Overlay the data plots to verify that the data is stable and
that poor optical property choices are not masking subtle
changes.

No change
in data

You can use the zoom

function to look at
data in a specific
region
Wet dispersion – the process
› Dispersion starts with the break-up of
agglomerates by:
▪ Wetting the sample
▪ Stirrer action
▪ Ultrasonication
› What are the signs of dispersion?
▪ Particle size reduction
▪ An increase in obscuration
▪ Increased light scatter on the higher
detectors on the data plot
▪ Coarse fractions in the distribution reduce
and may disappear
Evaluating a dispersion process

› What happens when the sample requires

ultrasonic dispersion?
Ultrasound
started
Rapid decrease in
size until…
Post Ultrasound
Pre Ultrasound

Particle size is
decreasing gradually. No change in size
Stability so dispersion is
This indicates
reached here stable
dispersion
Evaluating a dispersion process–
inspecting the obscuration
› The Records view shows that the obscuration increases
when ultrasonication is introduced
Ultrasound started. Note
the increase in
obscuration as more and
smaller particles become
present due to the
dispersion of
agglomerates
Evaluating a dispersion process–
Inspecting light scattering data
› Ultrasonic titration - measurements were taken
during ultrasound
Loss of signal on
inner rings indicates
loss of larger
particles

Trend across repeat measurements

Evaluating a dispersion process–
Inspecting the results
› Overlay the results of the ultrasonic titration to
see where changes have occurred – do the
results make sense?

Agglomerates
dispersing
Dissolution

› Symptoms:
▪ Obscuration diminishes.
▪ The fines become finer and disappear
▪ The coarse fraction reduces in size and
become more dominant as the fine
material disappears.
Dissolution – Using trend graphs
› The Trend graph shows size increasing with time and
obscuration decreasing – we are losing small particles

Obscuration will decrease with

time due to loss of material

Dv10 Increases over time

Dissolution – Inspecting the records view

› The Records view helps you to see what is

happening Obscuration is
dropping
quickly with
time.

D(0.1) is increasing.
Fines disappear first.
Dissolution – light scattering intensity
decreases

› The level of scattering signal recorded is

decreasing - this indicates that obscuration is
decreasing

Loss of signal
here indicates loss
of small particles
Dissolution – Possible remedies

› Try other dispersants if your sample does

appear to be dissolving.
› Select the dispersant in which there is the least
dissolution – determined by measurement
repeatability
› Slight dissolution can sometimes be tolerated
but measurements should be performed quickly!
› Saturated solutions can be used but only as a
last resort – these require special care.
Dissolution –
Alternative dispersants

› Some common dispersants in order of their polarity are

shown below:

Dispersant Polarity Index

Water 9.0

Ethanol 5.2

Propan-2-ol 3.9

Cyclohexane 0.2
Heptane 0.0
Agglomeration

› Symptoms:

▪ A tail of large particles appears on the

distribution
▪ The obscuration may decrease slightly
▪ Use of ultrasonication may get rid of the
tail and increases the obscuration

▪ Caution: Use of ultrasonication may cause

agglomeration!
Agglomeration – Size increases
› Here the sample can be seen to be increasing in size
over time – agglomeration is taking place

Agglomeration
indicated by
increase in Dv90
 Agglomeration – Inspecting the records view

› Looking at the record view indicates that the

obscuration has dropped slightly

Slight decrease Increase in Dv90

in obscuration
Agglomeration – possible remedies
› The dispersion needs to be stabilised by the addition of a
surfactant or additive.
› Surfactants and additives stabilise suspensions in the
following ways:
▪ By decreasing interfacial tensions and wetting the
sample
▪ By altering the charge on particle surfaces to aid
repulsion. (Electrostatic stabilisation)
▪ By adding long chain molecules which can adsorb
onto the particle surfaces and provide steric
stabilisation
▪ By adding a charged long chain material that provides
electrosteric stabiisation
› Note:
▪ Too much additive can be as ineffective as too little!
▪ Beware of foaming or bubble creation with surfactants
Agglomeration –surfactants
› Surfactants help to increase particle wetting and aid
particle dispersion in the following ways

▪ Non-ionic species improve wetting and can provide a

steric barrier to particle aggregation
• Example: Igepal CA-630; Tween 20 / 80; Span 20 / 80
▪ Ionic Surfactants absorb to particle surfaces providing
electrosteric stabilisation
• Examples:
• Anionics: SDS (sodium dodecylsulfate); AOT (sodium bis-2-
ethylhexylsulfosuccinate)
• Cationics: CTAB (cetyltrimethylammonium bromide)

› See ISO 14887:2000 “Sample Preparation – dispersing

powders in liquids” for a good general guide
Agglomeration – admixtures

› Admixtures can increase particle charge

▪ Specific adsorption of ions to particle
surface
▪ Common admixtures include:
Sodium Hexametaphosphate; Sodium
Pyrophosphate; Ammonium Citrate

› Remember pH adjustment can also be important

if your sample has an iso-electric point
Ultrasonic milling of particles
› Cause: Excessive ultrasonication – the length and
power that can be used depends on the fragility of
the material being measured
› Symptoms:
▪ A continuous, slow decrease in particle size
as ultrasonication continues.
▪ The main mode decreases in relative volume
and/or shifts to smaller sizes
› Does the size distribution show a separate tail of
coarse agglomerates? Is this reducing?
› Use a microscope where possible – are there
really agglomerates present?
› Does the shape of the particles change after
ultrasound? (Needles fracture)
Separation – Sedimentation and “creaming”

› It is essential that an homogenous dispersion

should be presented to the measurement cell
› Large, dense materials may settle out if agitation
and/or pump speed are insufficient
› Low density material may float or “cream”
› Indicators:
▪ The Coarse fraction disappears but obscuration is not
greatly reduced (settling)
▪ It is known that the material in question is dense
(settling).
▪ Or, that we have a low density emulsion (creaming)
Separation –
Pump or stirrer titration
How do we check?
By performing a pump/stirrer titration:

› With this procedure, we aim to prove that all the

particles, even the larger denser ones, are being
circulated by the sample dispersion unit
› Vary the pump and/or stirrer speeds and make
repeat measurements at each setting – there
should be a region of stable results in the series
of measurements obtained.
Separation – Pump or stirrer titration

› In this example the size is stable from 2250-rpm

onwards
180
Dv10
160 Dv50
Dv90
140

120
size / m

100

80

60

40

20

0
500 1000 1500 2000 2500 3000 3500 4000

stir speed / rpm

 General rules for good measurements
› Measurement & background times
▪ Backgrounds should be at least 10 seconds
or twice the measurement time
▪ Make short repeat measurements (2 to 5
seconds duration)
▪ If a tail of coarse particles is present or the
sample is polydisperse then you may need to
increase measurement times to ensure
homogenous sampling of the dispersion
 General rules for good measurements

› Ultrasonication of samples generates heat within

the sample – for liquids with a high vapour
pressure this can lead to density changes and
optical instabilities
▪ In such cases, ultrasonicate for a brief period (30-secs)
then allow to thermally equilibrate and re-measure –
repeat until stable
› If you do not have a dispersion unit with built-in
ultrasonication then this must be performed
externally with a bath or ultrasonic probe.
▪ If possible use the same aliquot at each stage of the
titration