ASTM E3247-20

This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the
Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.
Designation: E3247 − 20
Standard Test Method for
Measuring the Size of Nanoparticles in Aqueous Media
Using Dynamic Light Scattering
This standard is issued under the fixed designation E3247; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision.Anumber in parentheses indicates the year of last reapproval.A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.
1. Scope 1.5 Units—The values stated in SI units are to be regarded
as standard. Where appropriate, c.g.s. units are given in
1.1 This test method addresses the determination of nano-
addition to SI.
particle size (equivalent sphere hydrodynamic diameter) using
1.6 This standard does not purport to address all of the
batch-mode (off-line) dynamic light scattering (DLS) in aque-
safety concerns, if any, associated with its use. It is the
ous suspensions and establishes general procedures that are
responsibility of the user of this standard to establish appro-
applicable to many commercial DLS instruments. This test
priate safety, health, and environmental practices and deter-
method specifies best practices, including sample preparation,
mine the applicability of regulatory limitations prior to use.
performance verification, data analysis and interpretation, and
1.7 This international standard was developed in accor-
reporting of results. The document includes additional general
dance with internationally recognized principles on standard-
information for the analyst, such as recommended settings for
ization established in the Decision on Principles for the
specific media, potential interferences, and method limitations.
Development of International Standards, Guides and Recom-
Issues specific to the use of DLS data for regulatory submis-
mendations issued by the World Trade Organization Technical
sions are addressed.
Barriers to Trade (TBT) Committee.
1.2 The procedures and practices described in this test
2. Referenced Documents
method, in principle, may be applied to any particles that
exhibit Brownian motion and are kinetically stable during the
2.1 ASTM Standards:
course of a typical experimental time frame. In practice, this
E1617Practice for Reporting Particle Size Characterization
includes particles up to about 1000 nm in diameter, subject to
Data
limitations as described in the test method.
E2490Guide for Measurement of Particle Size Distribution
of Nanomaterials in Suspension by Photon Correlation
1.3 This test method does not provide test specimen prepa-
Spectroscopy (PCS)
ration procedures for all possible materials and applications,
E2456Terminology Relating to Nanotechnology
nor does it address synthesis or processing prior to sampling.
E3144Guide for Reporting the Physical and Chemical
The test specimen (suspension) preparation procedures should
Characteristics of Nano-Objects
provide acceptable results for a wide range of materials and
E3206Guide for Reporting the Physical and Chemical
conditions. The analyst must validate the appropriateness for
Characteristics of a Collection of Nano-Objects
their particular application.
2.2 ISO Standards:
1.4 This test method is applicable to DLS instruments that
ISO 22412Particle size analysis—Dynamic light scattering
implement correlation spectroscopy. Analysts using instru- (DLS)
ments based on frequency analysis may still find useful
3. Terminology
information relevant to many aspects of the measurement
process, including limits of applicability and best practices.
3.1 Definitions:
On-line (flow-mode) DLS measurements are not treated here
3.1.1 aliquot, n—a representative portion of a whole, as-
specifically and may have additional limitations or issues
sumed to be taken with negligible sampling error. adapted
relative to batch-mode operation.
from ISO 11074
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
This test method is under the jurisdiction of ASTM Committee E56 on contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
Nanotechnology and is the direct responsibility of Subcommittee E56.02 on Standards volume information, refer to the standard’s Document Summary page on
Physical and Chemical Characterization. the ASTM website.
Current edition approved Sept. 1, 2020. Published October 2020. DOI: 10.1520/ Available fromAmerican National Standards Institute (ANSI), 25 W. 43rd St.,
E3247-20. 4th Floor, New York, NY 10036, http://www.ansi.org.
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
E3247 − 20
3.1.2 average hydrodynamic diameter, x¯ ,n—the en- 3.1.6.1 Discussion—Therawproductofthecorrelatoristhe
DLS
semble average diameter that reflects the central tendency of intensity correlation function. Alternatively, a correlator can
the underlying population of particles as determined in dy- compare the signals arriving at identical times from two
namic light scattering. adapted from ISO 22412 different sources, a method referred to as cross-correlation.
Software aided computation of the correlation function is also
3.1.2.1 Discussion—The average hydrodynamic diameter
feasible, where sufficient computational capacity exists.
can be obtained from the size distribution calculated by
different methods, including, for instance, the cumulants
3.1.7 correlogram, n—a graphical representation of the
method combined with the Stokes-Einstein equation or any
correlation function with delay time, τ,onthe x-axis.
number of deconvolution algorithms that produce an intensity-
3.1.7.1 Discussion—The y-axis can be presented in one of
weighted particle size distribution. Note that when obtained
several forms: (1) the normalized intensity autocorrelation,
from the cumulants method the average hydrodynamic diam-
(2)
g (τ), for which the baseline value at τ = ∞ approaches 1 and
eter is often referred to in the literature as the z-average
the y-intercept at τ = 0 can obtain a maximum value of 2; (2)
diameter (z-avg), a legacy term. Enclosure by angular brackets
(2)
g (τ) – 1, where the baseline value approaches 0 and the
^x & indicates the arithmetic mean derived from replicate (2) 0.5
DLS
y-intercept has a maximum of 1; (3) g (τ)–1 , sometimes
measurements. Note this measurand can be scattering angle
(1)
denoted as G (τ), which also varies from 0 to 1; and (4) the
dependent and can be influenced by other factors besides size. (1)
normalized electric field correlation function, g (τ), with the
3.1.3 baseline, n—the measured far point of a correlation
same limiting values as the previous two functions, and the
function, typically determined from one or more channels of only function that is instrument-independent. Because of the
the correlator positioned at very large delay times, τ.
exponential decay associated with correlation functions, natu-
ral logarithmic plots are occasionally used.
3.1.3.1 Discussion—Essentially the square of the time-
averaged scattered intensity; baseline can also be calculated
3.1.8 count rate, n—theaveragenumberofphotonsdetected
from the square of the total measured photon counts over the
per unit time typically expressed in kilo-counts per second
course of the experiment divided by the total number of time
(kcps).
intervals. In this test method, the baseline is the value of
~2! 0.5 3.1.9 cumulants, n—a method for approximating the first
@g τ 21# determined from the far point channels (see
~ !
order (electric field) autocorrelation function determined in a
correlation function, correlogram, and signal-to-noise ratio).
DLS experiment as a polynomial expansion in delay time, τ.
3.1.4 combined standard uncertainty, n—standard measure-
3.1.9.1 Discussion—As described in ISO 22412, cumulants
ment uncertainty that is obtained using the individual standard
produces an estimate of the mean scattered light intensity-
uncertainties associated with the input quantities in a measure-
weighted harmonic mean and width of the underlying particle
ment model. ISO/IEC Guide 99
size distribution.
3.1.5 correlation function (correlation coeffıcients), n—the
3.1.10 cumulative undersize distribution, n—showstherela-
primary output or product of a digital correlator.
tive amount at or below a specified particle size, where the
3.1.5.1 Discussion—Essentially,therawdatageneratedbya
value at 50 % represents the median size.
dynamic light scattering experiment that is subjected to analy-
3.1.10.1 Discussion—Obtained by integration of the differ-
sis in order to derive a characteristic particle size or size
ential (discrete) size distribution. The percentile sizes derived
distribution, or both. The correlator measures the intensity-
from the cumulative distribution are commonly used in indus-
intensity correlation, an exponentially decaying function of
try to specify product size characteristics. Typically presented
timewitheachindividualdatapointreferredtoasacorrelation
as diameters, such as d , d and d , the percentiles d
10 50 90 x
coefficient. Also called the autocorrelation function because it
represent the value at which x % of the underlying population
compares photon counts from the same source at time t with
lies at or below size d.
those at time t + τ, where τ is the delay or lag time; it is
3.1.11 decay rate (decay constant) Γ,n—the characteristic
typically baseline-normalized such that the value at very large
rate at which an exponentially decaying correlation function
τ approaches unity. By convention, upper case G is used to
decreases toward its baseline, expressed in units of inverse
represent the unnormalized function, while lower case g
(2)
time.
indicates normalized values. The superscript 2, as in g (τ),
indicates a second order (intensity) correlation function, from 3.1.11.1 Discussion—The rate of decay is related to the
translational diffusion coefficient of the particles, D, by the
which the first order (electric field) correlation function is
~1! ~2! 0.5
relationship Γ5Dq , where q is the modulus of the scattering
= =
calculated according to: g ~τ! β5@g ~τ!21# , where β is
vector.
the y-intercept of the function on the right-hand side extrapo-
(1)
lated to τ = 0. For non-interacting Brownian particles, g (τ)is
3.1.12 differential (discrete) size distribution, n—shows the
related to the translational diffusion coefficient, D, such that
relative amount at each size value or interval.
~1! 2
ln g ~τ!52Dq τ and q is the modulus of the scattering vector.
3.1.12.1 Discussion—In dynamic light scattering, the differ-
3.1.6 correlator, n—adigitalcorrelatorpartitionstimeintoa ential or discrete distribution typically shows the % intensity
series of clock intervals of small (typically sub-microsecond) on the y-axis and size on the x-axis, from which the mode and
duration and constructs the sums of the products of photon mean of each particle population can be identified. The
counts registered at different intervals separated by increasing distribution can be represented as a continuous function or as
delay or lag time, τ. a histogram. In each case, the y-value yields the relative
E3247 − 20
amountoftheweightedpopulationatthecorrespondingsizeor intendeduse.Withregardstothesizerangeandthepresenceof
in the corresponding size bin. size-related properties, this term is a subject of controversy in
the field. The use of 100 nm as a reference point does not
3.1.13 diffusion coeffıcient D, n—mean squared displace-
suggest that materials or products with dimensions above 100
ment of a particle per unit time. ISO 13099
nm cannot or do not exhibit dimension-dependent properties.
3.1.13.1 Discussion—Characterizes the random transla-
3.1.22 number-weighted size, n—a distribution or mean
tional or “Brownian” motion of particles in a liquid medium.
This quantity is used in the Stokes-Einstein equation to where each particle is given equal weighting irrespective of its
size.
calculatetheequivalentspherehydrodynamicsize.InDLS,the
diffusion of non-interacting individual particles is observed 3.1.22.1 Discussion—Counting techniques, such as image
underconditionswheretheconcentrationissufficientlylow.As analysis or resistive pulse measurement will yield a number-
the concentration increases, interparticle interactions increase, weighted distribution or mean.
yielding a concentration-dependent or ensemble diffusivity.
3.1.23 polydispersity index PI, n—dimensionless measure
Non-spherical particles can also exhibit measurable rotational
of the broadness of the size distribution. ISO 22412
diffusion. These effects can lead to bias in the measurements.
3.1.23.1 Discussion—PI refers in this standard to the value
3.1.14 dynamic light scattering (DLS), n—method in which derived from a cumulants analysis of DLS data as defined in
particles undergoing Brownian motion in a liquid suspension
ISO 22412.
are illuminated by a laser and the change in intensity of the
3.1.24 qualification, n—proof with reference material that
scatteredlightisusedtodetermineparticlesize. adapted from
an instrument is operating in agreement with the manufactur-
ISO/TS 80004-6
er’sspecifications.Alsoreferredtoasperformanceverification.
3.1.15 expanded uncertainty, n—product of a combined
3.1.25 relative refractive index, n—ratio of the absolute
standard uncertainty and a coverage factor larger than unity.
refractive index (RI) of the particles to the real part of the
3.1.15.1 Discussion—Acoverage factor of two represents a
suspending medium.
confidence interval of approximately 95 %, assuming the
3.1.26 sample, n—a material or suspension from which test
degrees of freedom are sufficiently high (>10).
specimens or aliquots are obtained.
3.1.16 high effıciency particulate air, HEPA, adj—noting or
3.1.26.1 Discussion—In a DLS experiment, higher particle
using an air filter designed to remove at least 99.97 % of
relative RI translates to greater light scattering intensity,
airborne particles greater-than-or-equal-to 0.3 µm in diameter.
subject to other properties such as particle size, shape, concen-
3.1.17 hydrodynamic diameter d ,n—the calculated diam- tration and light absorption (imaginary part of particle RI).
H
eter of a theoretical hard sphere that diffuses in solution at the
3.1.27 signal-to-noise ratio (S/N), n—in this standard, de-
same rate as the analyte particle.
fined for DLS experiments in the following manner: [((inter-
3.1.17.1 Discussion—Hydrodynamic diameter is a generic
cept + 1)) ⁄ ((baseline + 1))] – 1, where intercept is the value
term that is not specific to an analysis method or measurement
~2! 0.5 ~2!
of @g ~τ!21# extrapolated to τ=0,and baseline is @g ~τ!
technique. 0.5
21 determined from the far point channels of the correlator
#
3.1.18 intensity-weighted size, n—a distribution or mean and should be close to 0 in magnitude.
whereeachparticleisweightedbyitslightscatteringintensity.
3.1.28 standard uncertainty u, n—measurement uncertainty
3.1.18.1 Discussion—DLS yields an intensity-weighted dis-
expressed as a standard deviation. ISO/IEC Guide 99
tributionormean.Intensityisahigherorderweightingrelative
3.1.29 test specimen, n—an aliquot used for measurement
to number or volume. For instance, using the Rayleigh
purposes.
approximation, the relative contribution for particles much
3.1.30 volume-weighted size, n—a distribution or mean
smallerthanthewavelengthoflightwillbeproportionaltosize
where each particle is weighted by its volume.
raised to the 6th power.
3.1.30.1 Discussion—Equivalent to mass-weighting if the
3.1.19 method validation, n—the process used to confirm
particle density is uniform. The relative contribution of each
that an analytical procedure employed for a specific test is
particle will be proportional to size-cubed.
suitable for its intended purpose.
3.1.31 y-intercept (intercept), n—the extrapolated τ =0
3.1.20 modulus of the scattering vector q, n—the absolute
value for a measured correlation function.
value of the momentum transfer or scattering vector,
3.1.31.1 Discussion—The y-interceptvaluemaybereported
πn θ
q5 4 sin ,expressedinunitsofinverselength,where θ
S D S D
differently, depending on the specific form of the correlation
λ 2
is the scattering angle, n is the refractive index of the function used in its determination. For instance, the y-intercept
~2! 0.5
suspending liquid at the laser wavelength and λ is the can be determined from a plot of the function @g ~τ!21# or
wavelength in vacuo. g (τ)–1,bothofwhichyieldalimiting y-interceptvalueof1.
3.1.21 nanoparticle, n—for purposes of this standard, im- 3.1.32 z-average, n—commonly used for the intensity-
plies that at least two external physical dimensions are smaller weighted average as applied to the diffusion coefficient or
-7
than about 100 nm (<10 m). adapted from E2456 particle size isolated in a DLS experiment using the cumulants
method of analysis.
3.1.21.1 Discussion—The length scale may be a hydrody-
namic diameter or a geometric length appropriate to the 3.1.32.1 Discussion—Abbreviated as z-avg this is a legacy
E3247 − 20
is not required for the implementation of this test method.
term commonly used to identify the intensity-weighted har-
monic mean diameter derived from a cumulants analysis of
6. Interferences
DLS data as described in ISO 22412.This standard recognizes
the common use of this term, but recommends that x¯ be
6.1 Interference can result from sample absorption at the
DLS
used for reporting DLS-based population mean size results, so
wavelength of measurement. Absorption reduces scattered
longastheprocedureusedinitscalculationisclearlyidentified
intensity and thereby reduces sensitivity. Absorption can also
(for example, cumulants, NNLS, or CONTIN).
cause temperature increase, which alters the local viscosity
affecting diffusivity. Refer to X2.1 for additional details and
4. Summary of Test Method
recommendations.
4.1 An appropriate volume and concentration of sample
6.2 Fluorescence or autofluorescence by sample compo-
suspension is placed in a clean, optically transparent cuvette,
nents is a potential interference if the excitation wavelength is
well-plate or other instrument-specific sample holder appropri-
at or near the wavelength of measurement. This interference
ate for DLS measurements.
might be difficult to identify without prior knowledge of the
component excitation behavior.
4.2 The intensity autocorrelation function is computed from
the measured time-dependent fluctuation of scattered light
6.3 Interference can result from number fluctuations (time-
using conditions set by the analyst or recommended in this test
dependent variation of particle concentration in the scattering
method. Calculations are based in part on the cumulants
volume), particle-particle interactions (for example, at rela-
analysis. The mean intensity-weighted equivalent sphere hy-
tively high concentrations or when strong electrostatic forces
drodynamic diameter is reported along with the polydispersity
are present) and multiple scattering (at relatively high concen-
index as defined in the standard. Results obtained by applica-
trations – this effect increases with particle size). In this
tion of a deconvolution analysis algorithm, such as a con-
context,‘low’and‘high’concentrationaresituational(material
strained inverse Laplace transform, are reported in a manner
and instrument dependent). Refer to X2.3 for further details
that enables all critical fitting parameters to be known and
and recommendations related to concentration effects and their
includes the original intensity-weighted distribution result. A
mitigationpriortoanalysis(forexample,conductingadilution
representativenormalizedautocorrelationfunctioningraphical
series analysis to identify an appropriate concentration range).
format is reported for each aliquot. Appendix X1 provides a
6.4 Salts, added deliberately or unintentionally (for
brief general overview of the DLS technique as applied in this
example, via contamination of glassware), can significantly
test method. Appendix X2 provides general good practice
impact the measured size and stability of the analyte if the
guidance supplemental to Guide E2490. Appendix X3 lists
particles are electrostatically stabilized. Refer to X2.4 for
currently available reference materials and quality control
further details and recommendations related to electrolyte
materials that might be applicable to DLS. Appendix X4
concentration.
defines the application of a random effects ANOVA model to
6.5 DLS cannot differentiate between the analyte of interest
estimate the grand average measurand and the combined
and contaminant particles that also scatter light. The presence
measurementuncertainty.AppendixX5isareportingchecklist
of adventitious or undesired particles (for example, dust) can
to aid the user in verifying compliance with this test method,
lead to measurement artifacts and bias. Proper handling of test
and Appendix X6 is an example data report that meets the
specimenspriortoanalysisiskeytoavoidingcontaminationby
reporting requirements.
adventitiousparticlesandobtainingresultsthatcorrectlyreflect
the properties of the analyte. Refer to X2.5 for further details
5. Significance and Use
and specific recommendations regarding particle contamina-
5.1 Particle size is a key property of manufactured or
tion.
engineered nanoparticles used in a wide range of applications.
6.6 In general, DLS is highly sensitive to the presence of
For purposes relevant to evaluations of safety, effectiveness,
performance, quality, public health impact, or regulatory status large particles, due to a strong size dependence of scattered
intensity. However, this sensitivity depends on the scattering
of products, the correct measurement and uniform reporting of
size and related parameters under use conditions, or during the propertiesandconcentrationsofallparticlespresentandonthe
instrument configuration used. For this reason, small numbers
manufacturing process, are critical to suppliers, analysts, regu-
lators and other stakeholders. of very large particles (for example, adventitious particles,
aggregates or agglomerates) within a population of very small
5.2 This test method is intended principally for the analysis
particles, can effectively dominate the scattering signal. Com-
of nanoparticles in aqueous suspension with dimensions be-
mercialinstrumentstypicallyusesomeformof“dustrejection”
tween about 1 nm and 100 nm, but may be applied to diffusive
to mitigate this problem. On the other hand, if these large
colloidal particles even if their dimensions fall outside the
species are of practical concern for the intended use, then the
nanoscale range (up to 1000 nm).
analyst must be aware of any process their measurement
5.3 For more detailed guidance on DLS measurements,
system uses to reduce the effects of large particles. For
includingoperationalaspects,refertoAppendixX2ofthistest
instance,someinstrumentsautomaticallyrejectapercentageof
method.
the measurement runs with the highest scattered intensity, the
assumption being that runs affected by the presence of very
NOTE 1—The user is also referred to Guide E2490, which provides
broadguidancefortheapplicationofDLStonanomaterials.GuideE2490 large particles will be removed from analysis. The erratic
E3247 − 20
appearance of large size peaks in a calculated DLS size toX2.6forfurtherdetailsandspecificrecommendationsonthe
distribution, can indicate the presence of large particle choice and use of measurement cuvettes.
contaminants, while the analyte peak should be relatively
stable with respect to size and amplitude. Using a backscatter 7. Apparatus
configuration also minimizes the effects of large particle
7.1 Commercial instruments are available from numerous
scattering, which is strongly oriented toward forward (low)
manufacturers for the measurement of hydrodynamic size by
angles.
photon counting and autocorrelation coupled with cumulants
analysis and deconvolution algorithms for the analysis of
6.7 For particles much smaller than the wavelength of light
polydisperse materials. Instruments typically operate at a fixed
(of order 60 nm or less for red lasers), scattering increases as
angle (for example, 90° or 173°), at multiple fixed angles
the 6th power of size (Rayleigh approximation). This strong
(between about 13° and 173°) or use a goniometer to select
dependenceonsizecontinuesabovethisthreshold,butismore
anglesoverabroadrange(fromabout10°to160°). Most have
complex as particle size approaches the wavelength of light
default settings or automated optimization of measurement
(Mie scattering). Therefore, polydisperse test specimens can
parametersthatrequireminimalinputfromtheanalyst.Manual
yield results that do not accurately reflect the true distribution
operation is also an option provided on most instruments. The
of sizes present.
basicDLSinstrumentincludesacoherentlasersource(linearly
6.8 Proteins,polymersandothermacromolecularspecies,if
polarized, visible wavelength), optics to focus and direct the
presentinsignificantquantities,canscattersufficientlytoaffect
laserbeamandtodetectscatteredphotonsoververyshorttime
themeasuredcorrelationfunctionandtherebybiasresults.Just
intervals, a sample chamber capable of maintaining constant
as with adventitious particles, DLS does not differentiate
temperature 60.2°C,aphotodetector(forexample,avalanche
between types of scatterers. These species are generally small
photodiode), a correlator (either stand-alone or integrated into
and are typically weak scatterers, so their impact will depend
the control PC) and a PC with control software to perform
strongly on their concentration. At relatively low
measurements and to analyze data.
concentrations, their contribution can be insignificant, while at
7.2 DLS instruments that use substantially different mea-
higher concentrations they can produce a bias in the results.
surement principles or optical configurations are available,
The potential impact is also dependent on the size,
including frequency domain (power spectrum) analysis, cross-
concentration, and composition of the analyte itself.
correlation spectroscopy and heterodyne detection; these alter-
6.8.1 Macromolecular species can also interact with the
nativeapproachesarebeyondthescopeofthistestmethod,but
analytetoaltertheanalytesizeorstructure,therebyinfluencing
they should produce comparable results for optically dilute
the measured hydrodynamic size.
(transparent) nanoparticle suspensions. The cross-correlation
6.8.2 Dissolved macromolecular species, if present in suf-
configurationmitigatesmultiplescatteringeffects,whilepower
ficient quantity, can increase the apparent viscosity of the
spectrumanalysisusingheterodyne(referencebeam)detection
aqueous medium. If this effective change in viscosity is
might improve sensitivity to very weak scatterers.
significant and not accounted for, it can yield a size that is
7.3 A HEPA-filtered hood or cabinet is useful for sample
larger than reality due to the inverse dependence of diffusivity
preparation and storage but is not required. Similarly, a
on viscosity.
centrifugemaybeusefulforseparatingthesuspensionmedium
(as supernatant) from analyte particles or to remove large
6.9 Air bubbles present a potential interference issue for
adventitious particle contaminants; use is situational and must
DLS measurements. Degassing of aqueous suspensions occurs
be fit for the intended purpose and clearly justified.
spontaneously when subjected to a temperature increase,
thereby generating bubbles. If present, scattering from such
7.4 Appropriate filters for removal of unwanted or adventi-
large inhomogeneities can interfere with the intended
tious particles that are large compared with the analyte or for
measurement, leading to drastic and usually obvious impacts
preparationofdiluents;useissituationalandmustbefitforthe
on the results (for example, highly atypical correlograms,
intended purpose and clearly justified.
extremely high baselines, very poor data quality metrics).
7.5 Appropriate syringes for filtration and transfer of
Visual inspection of the sample cuvette is warranted in this
samples and test specimens as needed.
case. Allowing sufficient time for test specimens or sample to
7.6 Appropriate glassware and plasticware (for example,
equilibrate at or near the measurement temperature will also
microtubes, transfer tubes, bottles) for transfer and storage of
alleviate this problem.
diluent, samples, and test specimens.
6.10 The quality of the optical cell or cuvette can impact
7.7 Sample cells or cuvettes, and other peripheral equip-
DLS results; surface scratches, imperfections or other sources
ment as recommended by the manufacturer for the instrument
of undesired scattering can produce measurement artifacts that
used and fit for the intended purpose.
significantly impact results and are difficult to isolate.
Generally, quartz or optical grade glass cells yield the best
8. Reagents and Materials
results and are less prone to surface imperfections relative, for
instance, to plastic disposable cuvettes or 96 well plates. This 8.1 Filtereddeionized(demineralized)waterforpreparation
must be weighed against the convenience of disposable cu- of suspensions and solutions and for cleaning or rinsing
vettes and well plates for high-throughput applications. Refer glassware and measurement cuvettes.
E3247 − 20
8.2 Appropriate filtered aqueous solutions for dilution of 11. Performance Verification and Method Validation
concentratedsamplesifappropriateandfitfortheintendeduse
11.1 DLS is a first-principles technique, and as such does
(for example, buffers, electrolytes, suspending medium).
not require calibration. However, analysts shall periodically
measure appropriate reference materials or quality control
9. Sampling and Test Specimens
materials in order to provide qualification (that is, verification)
9.1 It is important that the analyzed aliquot (test specimen)
of correct instrument operation within manufacturer specifica-
isrepresentativeofthelargersampleorprocessingstreamfrom
tions. It is recommended, at a minimum, to verify instrument
which it is taken. The source material must be homogeneous
performance at the beginning of a new or substantially modi-
before any sampling takes place; this can be achieved using
fied protocol as part of the method validation process, follow-
procedures such as sample inversion, mild shaking or stirring,
ing lengthy periods of disuse, or after repairs or relocation of
or vortexing. The procedure used to ensure homogeneity must
the device. Verification may be repeated at the end of a series
befitforpurposeandappropriateforthetargetanalyte,without
of test measurements for additional assurance. In addition,
altering the state of dispersion or introducing bubbles to the
annual performance certification by the instrument manufac-
sample; generally, the least energetic approach shall be chosen
turer is strongly recommended. Additionally, validation of the
that achieves the desired uniformity. Application of ultrasonic
measurement method (including sample preparation) must be
treatment or other high energy dispersive procedures (for
demonstrated as fit for purpose for all aspects of its scope. If a
example, intense shaking, high speed stirring, and excessive
dilution series must be conducted to identify an appropriate
vortexing) is not appropriate for homogenization, where the
concentrationrangetomeetinstrumentrequirementsandavoid
goal is to achieve a uniform distribution of analyte, not to
concentration-dependent effects, this test shall be included in
disrupt agglomerates or otherwise alter the native dispersion
the validation protocol. Validation is achieved using test
state. The procedure chosen should be included in the method
materials equivalent (or similar) to the product or sample to be
validation process.
tested, which are processed and prepared in a similar fashion.
9.2 In general, stable suspensions of nanoparticles are Method validation must be performed only once unless some
“self-homogenized” due to Brownian motion, although some
part of the procedure or product (test material) is changed, or
stratificationcanoccuroverlongtimeperiods(daystomonths) major instrument operating conditions have been altered.
if left standing. Analysis and comparison of several indepen-
11.1.1 Correct calibration of the temperature sensor is
dentlysampledaliquotswillprovidevalidationthatsamplingis necessary to convert the measured diffusion coefficient to size.
representative. If large (non-colloidal) particles or aggregates Thiscalibrationisgenerallytheresponsibilityoftheinstrument
are present, they will sediment rapidly and are not appropriate manufacturer.Temperaturecalibrationissuesshouldbeconsid-
for DLS analysis (which requires that the particles are pre- ered if the basic instrument verification process fails.
dominately diffusive (colloidal) and do not settle significantly
11.2 For the purpose of instrument performance
during the time-frame of the measurement). A series of
verification, polystyrene latex and silica reference materials
replicate measurements on the same aliquot will provide
withnominaldiametersbetween(20and100)nmareavailable
confirmation that the sample is stable over the relevant time
from commercial suppliers and government organizations. For
frame, if results vary randomly and a clear trend is not
qualification below 20 nm, and in the absence of available
apparent. If following mild sample or test specimen agitation,
reference materials, proteins such as cytochrome c and bovine
sediment or stratification becomes visually apparent over the
serum albumin are recommended. These materials are com-
timespanofseveralminutes,thenthesampleisnotappropriate
mercially available, are relatively inexpensive, and their hy-
for DLS measurement and an alternative technique should be
drodynamic size has been thoroughly characterized and re-
considered.
ported in the peer-reviewed literature. Polystyrene latex
reference materials are also available in larger sizes, and these
10. Preparation of Apparatus
may be useful for applications involving polydisperse samples
10.1 Follow instrument manufacturer instructions for pow-
orsamplescontainingparticleslargerthan100nm.Itshouldbe
ering up device, and for time necessary to achieve laser and
emphasized that instrument performance verification does not
thermal stability necessary for analysis.
necessarily require use of a reference material of the same or
10.2 Clean glass or quartz cuvettes with filtered demineral- similar size to the analyte. However, use of two or more
ized water and store dry. Periodic use of commercial cleaning
independent reference materials varying in size or
agentsformulatedspecificallyforopticalcellsandcomponents
composition, or both, is recommended as good practice, and
is recommended to remove difficult residues, but care must be
usingtwothatbrackettheanalytesize,ifpossible,providesthe
taken to remove all traces of the cleaning detergent as this can
highest level of confidence. Refer to Appendix X3 for sug-
impact analyte properties. Keep dry cleaned cuvette sealed/
gested reference materials and quality control materials.
capped until needed. If available, store cuvette under HEPA
11.3 The procedure for instrument performance verification
filtered air (for example, in a clean bench).
using a reference material is adapted from ISO 22412, as
10.3 Remove dust and other debris from plastic disposable follows:
cuvette or well-plate surfaces using a mild stream of clean 11.3.1 Measure the reference material 5 times under repeat-
(preferably filtered) compressed air or inert gas prior to use. ability conditions and calculate the average of the cumulants-
Alternatively, disposable cuvettes can be rinsed with filtered based intensity-weighted average hydrodynamic diameters
demineralizedwaterpriortouseatthediscretionoftheanalyst. ^x¯ & and their standard deviation u (in nm).
DLS DLS
E3247 − 20
11.3.2 Divide the expanded uncertainty stated for the refer- 11.5 For instrument performance verification, as part of
ence material mean size (in nm) by the coverage factor k to method validation, preference should be given to reference
materials that are similar to the target analyte or test material
obtain the standard uncertainty u ; the value of k is typically
RM
2, representing a 95 % confidence interval. If the stated when available.
uncertainty is expressed as a standard deviation (in nm), then
11.6 For regulatory purposes, analysts shall follow estab-
conversion is not necessary, and the stated uncertainty then
lished guidance for method validation as appropriate for DLS
equals u .
RM analysis (see for example ICH (1), FDA 1996 (2), and FDA
11.3.3 Calculate the tolerance u (in nm) for the mea-
2015 (3)).Validationprocedureswrittenforanalyticalmethods
meas
sureduncertaintybydividing u obtainedabovebythestated shall be adapted for DLS analysis, though some aspects and
RM
mean diameter (in nm) for the reference material and then
criteria might not be commutable (for example, limit of
multiplying this value by ^x¯ &. detection, limit of quantification, response linearity). Aspects
DLS
relating to repeatability, reproducibility and robustness should,
11.3.4 Combine the results of the previous two steps qua-
however, be fully commutable, and as such must be addressed
dratically and multiply by a coverage factor k = 2 to obtain the
during method development and validation.
2 2
=
expanded uncertainty U, where U52 u 1u .
RM meas
11.6.1 The evaluation of robustness should be addressed
11.3.5 Calculate the absolute difference between ^x¯ & and
during the method development phase. It should demonstrate
DLS
the reference stated mean value (in nm). If this difference is ≤
the reliability of an analysis with respect to deliberate varia-
U, then the performance verification bias test passes.
tions in method parameters. If measurements are sensitive to
variations in analytical conditions, the analytical conditions
11.3.6 If the relative standard deviation of the measure-
u should be suitably controlled, or a precautionary statement
DLS
ments × 100 % is < 2% then the performance verifica-
^x¯ &
should be included in the method.
DLS
tion repeatability test passes.
12. Procedure
11.3.7 Failure of verification tests indicate that a problem
12.1 Always wear appropriate personal protective gear (for
may exist with the instrument, the measurement cell, the
example, gloves, lab coat, goggles) and take appropriate
reference material, or the preparation of the test specimen.
precautions when handling nanomaterials.
Check that the reference material has not exceeded the stated
expiration date. If the reference material is within its stated
12.2 The suspending medium (also known as, aqueous
shelf-life, address other possible sources of error and contact
solution, diluent) shall be filtered prior to sample preparation
the manufacturer if non-instrument issues prove inconsequen-
using a 0.1 µm or smaller pore size membrane, and should be
tial.
testedforscatteringcontributionstothemeasuredsignal(count
rate)intheabsenceoftheanalyte.Asageneralrule,oneshould
11.3.8 The above procedure presumes that the stated refer-
filter the medium to at least the nominal size of the analyte to
ence value is the hydrodynamic mean diameter measured by
be measured. This may not be practical for particles smaller
DLS or another appropriate technique.
than 20 nm; in this case, the count rate for the suspending
11.3.9 Intheabsenceofappropriatereferencematerials,one
medium should be measured separately and compared to the
or more quality control materials with an established size (see
testspecimencountratetoensuretheabsenceof,ortoaccount
for example, Appendix X3) may be used in order to verify
for, interfering particles or other scattering entities present at
correct performance over a specific size range. In this case, the
significant concentrations. As a rule of thumb, the count rate
comparison between measurement and known value must
for the aqueous medium should be less than about 10 % of the
consider the material source and the method by which the
count rate for the test specimen (optionally, follow the manu-
known size was established.
facturer’s recommendation).
11.4 For comparative purposes and instrument verification,
12.3 If preparing a sample from dry material or a
particlesizeresultsobtainedfromDLSmeasurementsoftendo
concentrate, the concentration shall be adjusted to accommo-
not coincide with those obtained by other techniques (for
date the scattering properties of the sample and the optical
example, electron microscopy). This is due to different
requirementsoftheinstrument(thatis,accordingtoinstrument
method-defined measurands associated with different sizing
manufacturer specifications for acceptable count rate range).
techniques.
The presence of multiple scattering or particle interactions
NOTE 2—For instance, differences arise from different weighted aver-
mustbeconsideredwhenpreparingsamplesandtestspecimens
ages (for example, intensity versus number), as well as differences in the
for analysis (for recommendations, see 6.2 and X2.3). The
physicalpropertythatisactuallymeasured(forexample,particlediffusion
presence of agglomeration should also be considered during
versus projected particle area). For monodisperse hard core spheres larger
than about 60 nm in diameter, these differences tend to decrease with sample preparation.
increasing size; the presence of surface coatings (for example, ligands,
12.4 If a prepared sample (native product or material) is to
polymers) can significantly alter this trend.Additionally, DLS is sensitive
be analyzed, the count rate must fall within the instrument
to the presence of small quantities of large particles or clusters of smaller
particles, whereas electron microscopy-based results, for example, typi- manufacturer’s specifications (generally, the linear range for
cally reflect the size of primary particles and, in combination with
statistical counting limitations, may not represent a statistically relevant
samplingofclustersorlargerparticlespresent(forexample,agglomerates
Theboldfacenumbersinparenthesesrefertothelistofreferencesattheendof
and aggregates, but also oversize primary particles). this standard.
E3247 − 20
the detector). If it is too low, and the sample cannot be non-filtration alternatives, in X2.5. If using a filter, pre-rinse
concentrated (for example, by centrifugation), then the sample the filter with at least 0.5 mL of the suspending medium (if
is not appropriate for DLS analysis. If the count rate is too available) before loading sample and reattaching filter car-
high, one option is to dilute the test specimen with an tridge; if suspending medium is not available, use deionized
equivalent medium if available and if it does not alter the water. If a procedure is used to remove adventitious particles,
analyte state. The count rate can also be reduced to optimal it must be validated for the test material prior to measurement
levelsbyuseofanattenuatorlocatedbetweenthelaserandthe to ensure that the analyte is not being removed or otherwise
sample cell interface. Some instruments automatically deter- modified by the process, and the procedure must be reported.
mine and apply the required beam attenuation. It is also
12.8 Include all relevant controls in both the validation and
possible to reduce the laser power output by reducing the
measurement processes and apply the same rules and recom-
appliedvoltage,thoughthisrequiresrestabilizationofthelaser
mendations to controls as to analyte.
and is not common practice. In the absence of beam
attenuation, and if dilution is not possible or desired, if the 12.9 Loading Test Specimen:
count rate exceeds the stated specifications, then the test
12.9.1 If using a syringe to load test specimen, allow the
specimen is not appropriate for DLS analysis on that system.
first 4 drops to go to waste. If possible, use the next 4 drops to
NOTE 3—Instrument-specific acceptability of a test specimen’s scat-
pre-rinse the cuvette, and discard. The remainder can be used
tered intensity (count rate) does not exclude the possibility of multiple
for measurement.
scattering or particle interactions, both of which can bias the size
12.9.2 Ifusingapipettetoloadtestspecimen,fillthepipette
measur
...