001-2312PDG.

pdf

1 PARTICLE SIZE ANALYSIS BY LASER LIGHT DIFFRACTION

2 The method is based on the ISO standards 13320, 9276-1(1998) and 9276-1/AC1
3 (2004). Figure 2 of this chapter is based on Figure 1 of JIS Z 8825 (2013).

4 INTRODUCTION
5 The laser light diffraction technique used for the determination of particle-size
6 distribution is based on the analysis of the diffraction pattern produced when particles
7 are exposed to a beam of monochromatic light. Historically, the early laser diffraction
8 instruments only used scattering at small angles. However, the technique has since been
9 broadened to include laser light scattering in a wider angular range and application of
10 the Mie theory, in addition to the Fraunhofer approximation and anomalous diffraction.
11 The technique cannot distinguish between scattering by single particles and scattering
12 by clusters of primary particles, i.e. by agglomerates or aggregates. As most particulate
13 samples contain agglomerates or aggregates and as the focus of interest is generally on
14 the size distribution of primary particles, the clusters are usually dispersed into primary
15 particles before measurement.
16 For non-spherical particles, an equivalent sphere-size distribution is obtained because
17 the technique assumes spherical particles in its optical model. The resulting particle-size
18 distribution may differ from those obtained by procedures based on other physical
19 principles (e.g. sedimentation, sieving).
20 This chapter provides guidance for the measurement of size distributions of particles in
21 different dispersed systems, for example, powders, sprays, aerosols, suspensions,
22 emulsions, and gas bubbles in liquids, through analysis of their angular light-scattering
23 patterns. It does not address specific requirements of particle size measurement of
24 specific products.

25 PRINCIPLE
26 A representative sample, dispersed at an adequate concentration in a suitable liquid or
27 gas, is passed through a beam of monochromatic light, usually a laser. The light
28 scattered by the particles at various angles is measured by a multi-element detector.
29 Numerical values representing the scattering pattern are then recorded for subsequent
30 analysis. These scattering pattern values are then transformed, using an appropriate
31 optical model and mathematical procedure, to yield the proportion of total volume to a
32 discrete number of size classes, forming a volumetric particle-size distribution.

33 INSTRUMENT
34 The instrument is located in an environment where it is not affected by electrical noise,
35 mechanical vibrations, temperature fluctuations, humidity or direct bright light.
36 An example of a Fourier set-up of a laser light diffraction instrument is given in Figure
37 1. Other optical arrangements as shown in figure 2 may be used.

1
 001-2312PDG.pdf

38

39
1. Obscuration 5. Scattered light not 9. Working distance of lens (4)
detector collected by lens (4)
2. Scattered beam 6. Particle ensemble 10. Multi-element detector
3. Direct beam 7. Laser light source 11. Focal distance of lens (4)

40 4. Fourier lens(es) 8. Beam processing optics

41 Figure 1. - Example of a Fourier set-up of a laser light diffraction instrument

42
43

44
1. Laser light source 4. Fourier lens(es) 6. Transmission or
obscuration detector
2. Reverse Fourier 5. Low-angle detector(s), either 7. High-angle detector
lens(es) bespoke design or pixel array array
3. Measurement cell
or zone
45 Figure 2. - Example of optical arrangements in laser light diffraction instrument
46
47 The instrument comprises light source(s), beam processing optics, a sample
48 measurement zone (or cell), a Fourier/reverse Fourier lens, and a multi-element detector
49 for measuring the scattered light pattern. A data system is also required for
50 deconvolution of the scattering data into a volumetric size distribution and associated
51 data analysis and reporting.
52 The particles can enter the laser beam in 2 positions. In the Fourier set-up the particles

2
 001-2312PDG.pdf

53 enter the parallel beam before the collecting lens and within its working distance. In the
54 so-called reverse Fourier set-up, the particles enter after the collecting lens and thus, in a
55 converging beam. The advantage of the Fourier set-up is that a reasonable path length
56 for the sample is allowed within the working distance of the lens. The reverse Fourier
57 set-up requires small path lengths but enables the measurement of scattered light at
58 larger angles.
59 The interaction of the incident light beam and the ensemble of dispersed particles results
60 in a scattering pattern with different light intensities at various angles. The total angular
61 intensity distribution, consisting of both direct and scattered light, is then focused onto a
62 multi-element detector by a lens or a series of lenses. These lenses create a scattering
63 pattern that, within limits, does not depend on the location of the particles in the light
64 beam. Hence, the continuous angular intensity distribution is converted into a discrete
65 spatial intensity distribution on a set of detector elements.
66 It is assumed that the measured scattering pattern of the particle ensemble is identical to
67 the sum of the patterns from all individual scattering particles present in random relative
68 positions. Note that only a limited angular range of scattered light is collected by the
69 lens(es) and, therefore, by the detector.

70 DEVELOPMENT OF THE PROCEDURE

71 The measurement of particle size by laser diffraction can give reproducible data, even in
72 the sub-micron region, provided the instrument used and the sample tested are carefully
73 controlled to limit variability of the test conditions (e.g. dispersion medium, method of
74 preparation of the sample dispersion).
75 It is applicable to particle sizes in the range of approximately 0.1 µm to 3 mm. With
76 special instrumentation and conditions, the applicable size range can be extended above
77 3 mm and below 0.1 µm.
78
79 Sampling. The sampling technique must be adequate to obtain a representative sample
80 of a suitable volume for the particle-size measurement. Sample splitting techniques such
81 as rotating riffler or the cone and quartering method may be applied.
82 Evaluation of the dispersion procedure. A preliminary examination of the sample is
83 carried out visually or with the aid of a microscope, to estimate its size range and
84 particle shape. The dispersion procedure must be adjusted to the purpose of the
85 measurement. The purpose may be such that it is preferable to deagglomerate clusters
86 into primary particles as far as possible, or it may be desirable to retain clusters as intact
87 as possible. In this sense, the particles of interest may be either primary particles or
88 clusters.
89 For the development of a procedure it is highly advisable to check that comminution of
90 the particles does not occur, and conversely, that dispersion of particles or clusters is
91 satisfactory. This can usually be done by changing the dispersing energy and monitoring
92 the change of the particle-size distribution. The measured size distribution must not
93 change significantly when the sample is well dispersed and the particles are neither
94 fragile nor soluble. Moreover, if the manufacturing process (e.g. crystallisation, milling)
95 of the material has changed, the applicability of the procedure must be verified (e.g. by
96 microscopic comparison).
97 Sprays, aerosols and gas bubbles in a liquid are measured directly, provided that their

3
 001-2312PDG.pdf

98 concentration is adequate, because sampling or dilution generally alters the particle-size

99 distribution.
100 In other cases (such as emulsions, pastes and powders), representative samples may be
101 dispersed in suitable liquids. Dispersing aids (wetting agents, stabilisers) and/or
102 mechanical forces (e.g. agitation, sonication) are often applied for deagglomeration or
103 deaggregation of clusters and stabilisation of the dispersion. For these liquid dispersions,
104 a recirculating system is most commonly used, consisting of an optical measurement
105 cell, a dispersion bath usually equipped with a stirrer and ultrasonic elements, a pump,
106 and tubing. Non-recirculating, stirred cells are useful when only small amounts of a
107 sample are available or when special dispersion liquids are used.
108 Dry powders can also be converted into aerosols through the use of suitable dry powder
109 dispersers, which apply mechanical force for deagglomeration or deaggregation.
110 Generally, the dispersers use the energy of compressed gas or the differential pressure of
111 a vacuum to disperse the particles to an aerosol, which is blown through the
112 measurement zone, usually into the inlet of a vacuum unit that collects the particles.
113 However, for free flowing, coarser particles or granules the effect of gravity may be
114 sufficient to disperse the particles adequately.
115 If the maximum particle size of the sample exceeds the measuring range of the
116 instrument, the material that is too coarse can be removed by sieving and the mass and
117 percentage of removed material are reported. However, after pre-sieving, note that the
118 sample is no longer representative, unless otherwise proven.
119 Optimisation of the liquid dispersion. Liquids, surfactants, and dispersing aids used to
120 disperse powders must:
121 – be transparent at the laser wavelength and practically free from air bubbles or
122 particles;
123 – have a refractive index that differs from that of the test material;
124 – be non-solvent for the test material (pure liquid or pre-filtered, saturated solution);
125 – not alter the size of the test material (e.g. by solubility, solubility enhancement, or
126 recrystallisation effects);
127 – favour easy formation and stability of the dispersion;
128 – be compatible with the materials used in the instrument (such as O-rings, gaskets,
129 tubing, etc.);
130 – possess a suitable viscosity to facilitate recirculation, stirring and filtration.
131 Surfactants and/or dispersing aids are often used to wet the particles and to stabilise the
132 dispersion. For weak acids and weak bases, buffering of the dispersing medium at low
133 or high pH respectively can assist in identifying a suitable dispersant.
134 A preliminary check of the dispersion quality can be performed by visual or microscopic
135 inspection. It is also possible to take fractional samples out of a well-mixed stock
136 dispersion. Such stock dispersions are formed by adding a liquid to the sample while
137 mixing it with, for example, a glass rod, a spatula or a vortex mixer. Care must be taken
138 to ensure the transfer of a representative sample and that settling of larger particles does
139 not occur. Therefore, a sample paste is prepared or sampling is carried out quickly from
140 a suspension maintained under agitation.
141 Optimisation of the gas dispersion. For sprays and dry powder dispersions, a

4
 001-2312PDG.pdf

142 compressed gas free from oil, water and particles may be used. To remove such
143 materials from the compressed gas, a dryer with a filter can be used. Any vacuum unit
144 should be located away from the measurement zone, so that its output does not disturb
145 the measurement.
146 Determination of the concentration range. In order to produce an acceptable signal-
147 to-noise ratio in the detector, the particle concentration in the dispersion must exceed a
148 minimum level. Likewise, it must be below a maximum level in order to avoid multiple
149 scattering. The concentration range is influenced by the width of the laser beam, the path
150 length of the measurement zone, the optical properties of the particles, and the
151 sensitivity of the detector elements.
152 In view of the above, measurements must be performed at different particle
153 concentrations to determine the appropriate concentration range for any typical sample
154 of material. (Note: in different instruments, particle concentrations are usually
155 represented by differently scaled quantities, e.g. obscuration, optical concentration,
156 proportional number of total mass).
157 Determination of the measurement time. The measurement time, the reading time of
158 the detector and the acquisition frequency, is determined experimentally in accordance
159 with the required performance of the analytical procedure. Generally, the measurement
160 time is chosen to allow a large number of detector scans or sweeps at short time
161 intervals.
162 Selection of an appropriate optical model. Most instruments use either the Fraunhofer
163 or the Mie theory, though other approximation theories are sometimes applied for
164 calculation of the scattering matrix. The choice of the theoretical model depends on the
165 intended application and the different assumptions (size, absorbance, refractive index,
166 roughness, crystal orientation, mixture, etc.) made for the test material. If the refractive
167 index values (real and imaginary parts for the used wavelength) are not exactly known,
168 then the Fraunhofer approximation or the Mie theory with a realistic estimate of the
169 refractive index can be used. The former has the advantages that it is simple and it does
170 not need refractive index values; the latter usually provides less-biased particle-size
171 distributions for small particles. For instance, if the Fraunhofer model is used for
172 samples containing an appreciable amount of small, transparent particles, a significantly
173 larger amount of small particles may be calculated. In order to obtain traceable results, it
174 is essential to document the refractive index values used, since small differences in the
175 values assumed for the real and imaginary part of the complex refractive index may
176 cause significant differences in the resulting particle-size distributions. Small values of
177 the imaginary part of the refractive index (e.g. 0.01 i - 0.03 i) are often applied to allow
178 the correction of the absorbance for the surface roughness of the particles. It is noted, in
179 general, that the optical properties of the substance to be tested, as well as the structure
180 (e.g. shape, surface roughness and porosity), bear upon the final result.
181 VALIDATION
182 Typically, the validity of a procedure may be assessed by the evaluation of its
183 specificity, linearity, range, accuracy, precision and robustness. In particle-size analysis
184 by laser light diffraction, specificity as defined by ICH is not applicable as it is not
185 possible to discriminate between different components in a sample, nor is it possible to
186 discriminate agglomerates from dispersed particles unless properly complemented by
187 microscopic techniques. Exploring a linear relationship between concentration and

5
 001-2312PDG.pdf

188 response, or a mathematical model for interpolation, is not applicable to this procedure.
189 Rather than evaluating linearity, this procedure requires the definition of a concentration
190 range within which the result of the measurements does not vary significantly.
191 Concentrations below that range produce an error due to a poor signal-to-noise ratio,
192 while concentrations above that range produce an error due to multiple scattering. The
193 range depends mostly on the instrument hardware. Accuracy is confirmed through an
194 appropriate instrument qualification and comparison with microscopy, while precision
195 may be assessed by means of a repeatability determination.
196 The attainable repeatability of the procedure mainly depends on the characteristics of the
197 material (homogeneity – representative sampling, milled/not milled, robust/fragile,
198 width of its size distribution, etc.), whereas the required repeatability depends on the
199 purpose of the measurement. Mandatory limits cannot be specified in this chapter, as the
200 repeatability (independent sample preparations) may vary appreciably from one
201 substance to another. However, it is good practice to aim at acceptance criteria for
202 repeatability such as srel ≤ 10 per cent [n = 6] for any central value of the distribution
203 (e.g. for x50). Values at the sides of the distribution (e.g. x10 and x90) are oriented
204 towards less stringent acceptance criteria such as srel ≤ 15 per cent [n = 6].
205 For any characteristic parameter (e.g. x10, x50, x90) that is below 10 µm the acceptance
206 criteria may be doubled.
207 Robustness may be tested during the selection and optimisation of the dispersion media
208 and forces. The change of the dispersing energy may be monitored by the change in the
209 particle-size distribution.

210 MEASUREMENT
211 Precautions. The main sources of variation are the characteristics of the test material,
212 sample preparation conditions, and the individual analyst’s technique. Efforts are made
213 to reduce contributions from these sources.
214 The instructions given in the instrument manual are followed:
215 – never look into the direct path of the laser beam or its reflections;
216 – earth all instrument components to prevent ignition of solvents or dust explosions;
217 – check the instrument set-up (e.g. warm-up, required measuring range and lens,
218 appropriate working distance, position of the detector, no direct bright light);
219 – in the case of wet dispersions, avoid air bubbles, evaporation of liquid, schlieren or
220 other inhomogeneity in the dispersion ; similarly, avoid improper mass-flow from the
221 disperser or turbulent air-flow in the case of dry dispersions ; such effects can cause
222 erroneous particle-size distributions.
223 Measurement of the light scattering of dispersed sample(s). After proper alignment
224 of the optical parts of the instrument, a blank measurement of the particle-free
225 dispersion medium must be performed using the same procedure as that used for the
226 measurement of the sample. The background signal must be below an appropriate
227 threshold. The detector data are saved in order to subtract them later from the data
228 obtained with the sample. The dispersed sample is then measured according to the
229 procedure developed.
230 For each detector element, an average signal is calculated, sometimes together with its
231 standard deviation. The magnitude of the signal from each detector element depends

6
 001-2312PDG.pdf

232 upon the detection area, the light intensity and the quantum efficiency. The co-ordinates
233 (size and position) of the detector elements together with the focal distance of the lens
234 determine the range of scattering angles for each element. Most instruments also
235 measure the intensity of the central (unscattered) laser beam. In this case the ratio of the
236 intensity from the sample to that from a blank is inverse correlated to the proportion of
237 scattered light and hence the particle concentration.
238 Conversion of scattering pattern into particle-size distribution. This deconvolution
239 step is the inverse of the calculation of a scattering pattern for a given particle-size
240 distribution. The assumption of spherical particle shape is particularly important as most
241 algorithms use the mathematical solution for scattering from spherical particles.
242 Furthermore, the measured data always contain some random and systematic errors,
243 which may vitiate the size distributions. Several mathematical procedures have been
244 developed for use in the available instruments. They contain some weighting of
245 deviations between measured and calculated scattering patterns (e.g. least squares),
246 some constraints (e.g. non-negativity for amounts of particles), and/or some smoothing
247 of the size distribution curve.
248 The algorithms used are specific to each make and model of equipment, and are
249 proprietary. The differences in the algorithms between different instruments may give
250 rise to differences in the calculated particle-size distributions.
251 Replicates. The number of replicate measurements (with individual sample
252 preparations) to be performed depends on the required measurement precision. It is
253 recommended to set this number in a substance-specific procedure.

254 REPORTING OF RESULTS

255 The particle-size distribution data are usually reported as cumulative undersize
256 distribution and/or as density distribution by volume. The symbol x is used to denote the
257 particle size, which in turn is defined as the diameter of a volume-equivalent sphere.
258 Q3(x) denotes the volume fraction undersize at the particle size x. In a graphical
259 representation, x is plotted on the abscissa and the dependent variable Q3 on the
260 ordinate. Most common characteristic values are calculated from the particle-size
261 distribution by interpolation. The particle sizes at the undersize values of 10 per cent,
262 50 per cent, and 90 per cent (denoted as x10, x50, and x90 respectively) are frequently
263 used. x50 is also known as the median particle size. It is recognised that the symbols d or
264 D are widely used to designate the particle size, thus the symbol x may be replaced by d
265 or D.
266 Moreover, sufficient information must be documented about the sample, the sample
267 preparation, the dispersion conditions, and the cell type. As the results depend on the
268 particular instrument, data analysis program, and optical model used, these details must
269 also be documented.

270 INSTRUMENT QUALIFICATION

271 Use the instrument according to the manufacturer's instructions and carry out the
272 prescribed qualifications at an appropriate frequency, according to the use of the
273 instrument and substances to be tested.
274 Operation qualification. Laser diffraction instruments, although assuming idealised
275 properties of the particles, are based on first principles of laser light scattering. Thus,

7
 001-2312PDG.pdf

276 calibration is not required. However, it is still necessary to confirm that the instrument is
277 operating correctly. This can be undertaken using any certified reference material
278 (CRM) that is acceptable in industrial practice, which is characterized by a
279 metrologically valid procedure for one or more specified properties, accompanied by a
280 certificate that provides the value of the specified property, its associated uncertainty
281 (ucrm) and a statement of metrological traceability.
282 The entire measurement procedure is examined, including sample collection, sample
283 dispersion, sample transport through the measurement zone, the measurement and
284 deconvolution procedure. It is essential that the total operational procedure is fully
285 described.
286
287 Certified reference materials (CRMs) consist of spherical or non-spherical particles
288 having homogeneous particle density and optical properties. They shall have a known
289 distribution over an appropriate size rage such that for example the x90 / x10 ratio is at
290 least 1.5. The real and imaginary part of the complex refractive index are specified for
291 the material if the Mie theory is to be applied in data analysis.
292 Preferably, these particles have certified values from laser light diffraction analysis. The
293 use of reference values from methods other than laser light diffraction may cause a
294 significant bias. The reason for this bias is that the different principles inherent in the
295 various methods may lead to different sphere-equivalent diameters for the same non-
296 spherical particle.
297 Although the use of CRMs is preferred, other well-defined reference materials may also
298 be employed. They consist of substances of typical composition and particle-size
299 distribution for a specified class of substances. Their particle-size distribution has
300 proven to be stable over time. The results must comply with previously determined data,
301 with the same precision and bias as for the CRM.
302
303 The response of a laser diffraction instrument is considered to meet the requirements if
304 the average value of x50 from at least 3 independent measurements does not deviate by
305 more than 3 per cent from the certified range of values of the reference material. I.e.
306 when a certified size of a reference material is stated as a range rather than a single
307 value, that range is expanded by the relevant percentage to determine the criterion for
308 measurement. The average values for x10 and x90 must not deviate by more than 5 per
309 cent from the certified range of values. For any characteristic parameter (e.g. x10, x50,
310 x90) that is below 10 µm the acceptance criteria may be doubled.
311
312 Performance Qualification. In addition to the operation qualification, the performance
313 of the system must be qualified at regular time intervals or as frequently as appropriate.
314 This can be undertaken using any suitable reference material as mentioned in the
315 previous paragraph.
316 The qualification of the system is based on the concept that the equipment, electronics,
317 software and analytical operations constitute an integral system, which can be evaluated
318 as an entity. Thus the entire measurement procedure is examined, including sample
319 collection, sample dispersion, sample transport through the measurement zone, and the
320 measurement and deconvolution procedure. It is essential that the total operational

8
 001-2312PDG.pdf

321 procedure is fully described.

322
323 In general, unless otherwise specified in the individual monograph, the response of the
324 system is considered to meet the requirements if the x50 value does not deviate by more
325 than 10 per cent from the range of values of the reference material. If optionally the
326 values at the sides of the distribution are evaluated (e.g. x10 and x90), then these values
327 must not deviate by more than 15 per cent from the certified range of values. . For any
328 characteristic value (e.g. x10, x50, x90) that is below 10 µm the acceptance criteria may be
329 doubled.
330 NOTE: Stricter requirements for instrument operation qualification are prescribed in
331 the corresponding section.