Particle characterisation

Abstract

To provide a method of automatically selecting an optimal value or range for at least one measurement parameter for particle characterization by laser diffraction.SOLUTION: The method comprises receiving a plurality of measurements and associated measurement parameters, where the measurements are performed on at least one particulate sample 6 by using laser diffraction to determine at least one particle characteristic, and where the measurements are obtained using a plurality of values for the at least one measurement parameter. The method further comprises using a processor 14 to automatically select, based on the at least one particle characteristic and the associated measurement parameters, an optimal value or range of at least one measurement parameter for performing particle characterization by laser diffraction.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method and apparatus for particle characterisation. [Background technology]

[0002] Laser diffraction particle characterization is a technique for characterizing particles. In laser diffraction particle characterization, a light beam (which may originate from an LED or laser diode) is incident on a sample having particles suspended in a fluid diluent. The particles scatter the light beam, producing scattered light. The scattered light pattern (i.e., the distribution of scattered light intensities at different scattering angles relative to the incident light beam) characterizes certain properties of the particles. For example, particle size or size distribution can be inferred from the scattered light pattern by detecting the intensity of light scattered at multiple different scattering angles.

[0003] Determining particle size by laser diffraction is not always easy. There are many potential sources of error. To date, it has been difficult for unskilled users to develop an appropriate measurement method to minimize possible errors in the measurement. It is therefore an object of the present invention to alleviate this challenge by automating at least some aspects of the development of a method for particle characterization by laser diffraction. Summary of the Invention

[0004] According to one aspect of the present invention, there is provided a method for automatically selecting an optimal value or range for at least one measurement parameter for particle characterization by laser diffraction, the method including receiving a plurality of measurements and associated measurement parameters, the plurality of measurements being performed on at least one particulate sample by using laser diffraction to identify at least one particle characteristic, the plurality of measurements being obtained using a plurality of values ​​for the at least one measurement parameter, and the method further including using a processor to automatically select, based on the at least one particle characteristic and the associated measurement parameters, an optimal range of at least one sample preparation parameter and / or measurement parameter for performing particle characterization by laser diffraction.

[0005] The measurement parameters may include sample preparation parameters and / or settings for a laser diffraction instrument (configured to perform laser diffraction analysis on the sample).

[0006] The at least one particle characteristic may include particle size (eg, Dv10, Dv50, D90 or other measurement of particle size).

[0007] Each measurement may be performed on successive aliquots of the same sample, or on samples of the same type. For example, samples may be of the same type if they are obtained from a particular product or product type, or from a particular environment (such as, but not limited to, a manufacturing environment). A particulate sample may have multiple particles. A particulate sample may have a diluent liquid or may be dry.

[0008] The method may further include performing the measurement by illuminating the sample with a light beam from the light source, thereby generating scattered light from interaction of the light beam with particles of the sample, detecting a distribution of scattered light intensity over a range of different scattering angles, and using a processor to determine particle characteristics from the distribution of scattered light intensity over the range of different scattering angles.

[0009] Automatically selecting an optimal value or range of at least one measured parameter may include identifying a value or range of the measured parameter that results in stable particle characteristics.

[0010] The method may include binning the measurements into a plurality of measurement parameter bins corresponding to different values ​​of the measurement parameter, determining a slope or variance between at least one particle characteristic for each measurement parameter bin, and selecting an optimal value range for the measurement parameter based on the slope or variance for each measurement parameter bin.

[0011] Binning the measurements may include using a clustering algorithm.

[0012] Selecting an optimal value or range of values ​​for the measurement parameter may include identifying a measurement parameter bin with a minimum slope or minimum variance, and identifying a measurement parameter bin having a slope or variance below a threshold defined as a multiple of the minimum slope or minimum variance. may have

[0013] The at least one measurement parameter may include obscuration of the light beam by the sample, and the method may include selecting an optimal range of obscuration of the light beam by the sample, and the plurality of measurements may include measurements performed at different values ​​of obscuration of the light beam by the sample.

[0014] The method may include identifying a range of shading that results in low variance in at least one particle property.

[0015] At least one measurement parameter may be selected from light blocking, stirring amount, stirrer speed, sonication intensity, sonication time, feed rate, hopper gap, gas dispersant pressure, measurement time (e.g., measurement time between each of illumination with red light and illumination with blue light), and sample mass.

[0016] The at least one particle characteristic may include particle size, and the at least one measured parameter may include shading, and the method may include limiting an optimum value or range of shading based on the measured particle size.

[0017] Optionally: i) if the particle size is below a lower threshold, the optimum range of light blocking may be limited to be below an upper threshold; and / or ii) If the particle size is greater than an upper threshold, the optimum range of light blocking may be limited to be greater than a lower threshold.

[0018] The upper or lower threshold may be Dv50. The upper threshold for Dv50 may be 0.5 to 2 microns, for example 1 micron. The upper limit for light blocking may be 8% to 15%, for example 10% or 11%. The lower limit for light blocking may be 2% to 6%, for example 4%.

[0019] The method may include prompting a user to provide an estimated particle density. The at least one measured parameter may include an amount of agitation (e.g., stirrer speed, ultrasonic intensity, sonication time). The method may include limiting the value or range determined as the optimal amount of agitation to below the threshold in response to the estimated particle density being equal to or greater than a threshold.

[0020] The estimated particle density threshold is 2 g / cm 3 The threshold stirring intensity may be a stirrer speed of 3000 rpm.

[0021] The processor may be configured as follows: i) provide guidance to users on correct sampling procedures; ii) providing guidance to the user on sample preparation and dispersant selection, optionally in response to information input by the user; iii) automatically determining whether a stable dispersion of the sample exists by checking for light blocking and / or by automatically performing at least one laser diffraction measurement; iv) automatically identifying appropriate measurement conditions (e.g., stirring speed / stirrer, light blocking, sonication) by performing at least one laser diffraction measurement; v) Automatic method validation and robustness testing.

[0022] Steps i) to v) may be performed in sequence.

[0023] According to a second aspect, a non-volatile machine-readable medium is provided for configuring a processor to receive a plurality of measurements and associated measurement parameters, the measurements being performed on at least one particulate sample by using laser diffraction to determine at least one particle characteristic, the plurality of measurements being obtained using multiple values ​​of the at least one measurement parameter, and the processor to automatically select an optimal value or range of the at least one measurement parameter for performing particle characterization by laser diffraction based on the at least one particle characteristic and the associated measurement parameters.

[0024] The processor may be configured to carry out the method according to the first aspect, including any features thereof.

[0025] According to a third aspect, there is provided a laser diffraction apparatus comprising a sample cell, a light source configured to illuminate the sample cell with a light beam, thereby generating scattered light from interaction of the light beam with particles in the sample cell, a plurality of light detectors configured to detect a distribution of scattered light intensities over a range of different scattering angles, and a processor configured to determine particle characteristics from the distribution of scattered light intensities over the different scattering angles, wherein the processor is configured to receive a plurality of measurements and associated measurement parameters, the measurements being performed on at least one particulate sample to determine at least one particle characteristic using laser diffraction, the plurality of measurements being obtained using a plurality of values ​​of the measurement parameters, and the processor is configured to automatically select an optimal range of at least the measurement parameters for performing particle characterization by laser diffraction based on the at least one particle characteristic and the associated measurement parameters.

[0026] The processor may be configured to obtain at least some of the plurality of measurements by continuously varying at least one measurement parameter and performing a laser diffraction analysis for each value of the at least one measurement parameter.

[0027] The laser diffraction device may further include a display, and the processor is configured to use the display to guide a user through the set of measurements to provide at least a portion of the plurality of measurements and associated measurement parameters.

[0028] The laser diffraction device may have a particle dispersion unit, and the measurement parameters may include particle dispersion unit settings, and the processor is configured to automatically control the particle dispersion unit using the particle dispersion unit settings.

[0029] The processor may be configured to perform the method according to any of claims 1 to 10.

[0030] The present invention will now be described, by way of example only, with reference to the accompanying drawings, in which: [Brief explanation of the drawings]

[0031] [Figure 1] Schematic diagram of the apparatus for particle analysis by laser diffraction [Figure 2] Flow chart of dispersion stability evaluation method [Figure 3] Illustration of automatic classification of dispersion instability [Figure 3a] Detailed view of the inset graph in Figure 3 [Figure 3b] Detailed view of the inset graph in Figure 3 [Figure 3c] Detailed view of the inset graph in Figure 3 [Figure 3d] Detailed view of the inset graph in Figure 3 [Figure 3e] Detailed view of the inset graph in Figure 3 [Figure 3f] Detailed view of the inset graph in Figure 3 [Figure 3g] Detailed view of the inset graph in Figure 3 [Figure 3h] Detailed view of the inset graph in Figure 3 [Figure 3i] Detailed view of the inset graph in Figure 3 [Figure 3j] Detailed view of the inset graph in Figure 3 [Figure 3k] Detailed view of the inset graph in Figure 3 [Figure 3l] Detailed view of the inset graph in Figure 3 [Figure 4] Graph showing particle size from successive measurements with increasing stirrer speed [Figure 5] Exemplary Method of Stirrer Rate Titration [Figure 6] Graph showing particle properties measured using different laser obscurations [Figure 7] Graph showing particle characteristics for a subset of shadings [Figure 8] Graph showing the average particle characteristics for each cluster of measurements in Figure 7 [Figure 9] Three graphs showing the gradients of each cluster of particle characteristics shown in Figure 7. [Figure 10]The relationship of particle size to light shading shows that there is a stable region where the gradient of particle size to light shading is low (flat). [Figure 11] Exemplary Method for Dark Titration [Figure 12] Graph showing dispersive titration [Figure 13] Exemplary Methods for Ultrasonic Titration [Figure 14] Graph showing pressure titration [Figure 15] Flow diagram illustrating an exemplary method [Figure 16] Flow diagram illustrating an exemplary method [Figure 17] Flow diagram illustrating an exemplary method DETAILED DESCRIPTION OF THE INVENTION

[0032] Referring to Figure 1, a laser diffraction device 10 is shown having a light source 2 (e.g., a laser or LED) that generates a light beam 4. The light beam 4 is incident on a sample cell 8 that contains a sample 6. The sample cell 8 may include a flow cell that receives a diluent containing suspended sample from a dispersion unit 7. The sample 6 has particles suspended in a diluent fluid.

[0033] The dispersion unit 7 may be configured for wet dispersion (with liquid diluent) and / or dry dispersion (with gas diluent). Although both wet and dry dispersion units are illustrated, either may be omitted depending on the type of dispersion being analyzed.

[0034] If the dispersion unit is configured for wet dispersion, it typically comprises a tank 71 and a stirrer 72 or sonicator 73 to improve dispersion of the sample particles in the diluent.

[0035] If the dispersion unit is configured for dry dispersion, the dispersion unit 7 may have a hopper 76 for receiving the sample and a sample dispenser 74 for dispensing the sample from the hopper 76 into a stream of dilution gas (e.g., into a venturi 75). The sample dispenser 74 may have a vibrating tray. The stream of dilution gas may be supplied at a gauge pressure defined by a control valve (not shown), with higher pressures resulting in higher flow rates.

[0036] The operating parameters of the dispersion unit 7 (hopper gap, sample dispensing rate, gas pressure, stirrer speed, sonication intensity, sonication time, diluent circulation rate, etc.) may be controlled by the processor 14 .

[0037] A dispersion liquid comprising diluent and sample particles may be circulated from the dispersion unit 7 through the sample cell 8 .

[0038] The light beam 4 is scattered by interacting with particles in the sample, producing scattered light. The intensity of the scattered light varies with the scattering angle, which is the angle between the illuminating light beam and the scattered light. Multiple detectors 12 are positioned at different scattering angles. The detectors 12 are used to sample the intensity of the scattered light at multiple different scattering angles. Output from the detectors is provided to a processor 14.

[0039] Processor 14 is configured to determine particle characteristics from the detector output. For example, the distribution of scattered light intensity at different angles characterizes particle size. As a result, particle size (e.g., Dv10, Dv50, Dv90) or particle size distribution is inferred by processor 14 from the detector output.

[0040] Particle characterization by optical diffraction can be performed on dry samples, with particles dispersed by a diluent gas, or on wet samples, with particles suspended in a liquid diluent.

[0041] Wet / liquid dispersion Wet or liquid dispersion is the most common method of sample dispersion for particle characterization by light diffraction. For fine particles (e.g., a few microns or less in size), wet dispersion may be particularly suitable due to the significant reduction in interparticle adhesion forces resulting from the process of surface wetting. Wet samples can disperse agglomerated particles with relatively little energy input. However, not all samples are suitable for analysis by wet dispersion.

[0042] To produce reproducible results, any particle characterization technique relies on at least three factors: i) Representative sampling ii) a stable dispersion, and iii) Appropriate measurement conditions

[0043] Parameters that affect these factors are collectively referred to herein as "measurement parameters," a term that includes any parameters related to sampling and sample preparation. The importance of each of these factors depends to some extent on the particle size of the sample. For fine particles, dispersants and dispersion energy are more important, while for coarse materials, sampling is more important.

[0044] To assist users in obtaining reproducible results from particle characterization by laser diffraction, a computer-implemented tool is provided that guides users through the steps for developing an appropriate test method.

[0045] The software tool is configured to guide the user through the following (preferably in the order listed): 1. Correct sampling procedures 2. Sample Preparation and Dispersant Selection 3. Identifying a stable dispersion state 4. Identifying appropriate conditions for the method (e.g., stirring speed / agitation, light blocking / sample concentration, sonication time (if required)) 5. Method validation and robustness testing

[0046] In the initial phase, a graphical user interface may be provided to obtain relevant information from the user to assist the software in the analysis and experimental design. In this section, the user is prompted to enter information such as the expected / target particle size, agglomeration characteristics or primary properties to be measured, or other information useful for experimental design (e.g., a given dispersant, how soluble the sample is in the dispersant, etc.). Another important factor that may be defined is the target reliability and variability required from the test. For example, a laser diffraction test method may be defined to conform to USP, ISO, or other standards.

[0047] A table showing example input information is provided below: Table 1 shows example user-defined parameters.

[0048] [Table 1]

[0049] One option for defining method repeatability is to use ISO standard 11320.2020, which requires users to perform five consecutive measurements on a dispersed single-shot test sample. The ISO standard defines a minimum method repeatability for laser diffraction for six consecutive readings from one single-shot test sample. Dv10:RSD<3% Dv50:RSD<2.5% Dv90:RSD<4%

[0050] Intermediate precision may be defined, requiring measurements of 3 to 5 separate test samples, with the mean cumulative volume at the required percentile being within a pre-calculated tolerance limit U. lim must not exceed .

[0051] USP <429> defines the criteria for method reproducibility for laser diffraction. For USP standards, results from laser diffraction are considered adequate if they meet the following criteria for six readings: Dv50:RSD≦10% Dv10 and Dv90: RSD≦15% For Dv50 less than 10 μm, these maximum values ​​are doubled.

[0052] In some examples, Earthmover's Distance (EMD) may be employed in defining the target method reproducibility. EMD is a measure of the distance between two probability distributions over a region and may be directly related to the similarity between the data. Other options include user-defined custom percent variation of sample properties (such as Dv10, Dv50, and Dv90) between defined continuous values.

[0053] Representative Sampling After initial input from the user, the software tool will advise the user on the appropriate procedure for taking an aliquot or sample from the amount of material to be characterized. For example, the bulk of the particulates from which the sample is taken may be contained in a container or bottle. It is important that the sample is representative of the particulates the user intends to analyze. Particulates are prone to segregation. To avoid an unrepresentative sample, the software tool may advise the user to mix the sample thoroughly. An example of an appropriate method for mixing the bulk of the particulates is to roll the container in different orientations for at least two minutes. Shaking the container should be avoided as this may cause the particles in the sample to separate by size.

[0054] Representative sampling is an important factor, especially when measuring coarse particles. The advice provided is displayed in the workflow GUI in an appropriate format depending on the user's input on the sample. For example, if the sample is identified by the user as being derived from bulk powder, the workflow will recommend that the bulk powder be thoroughly mixed before sampling, using a riffler or by rolling it in different directions with both hands for approximately 20 seconds.

[0055] The next consideration is how much sample is required for the measurement. ISO standard guidance dictates a minimum of 400 particles for laser diffraction errors to be less than 5%. Using the user-provided density of the material and the user-provided expected particle size, the software tool may determine and provide to the user (e.g., via a display) the minimum sample mass required to obtain reproducible results.

[0056] Sample preparation After providing guidance on sampling, the software tool may guide the user in preparing the sample for addition to the optical diffraction particle characterization tool.

[0057] The software tool may suggest a list of liquid dispersants organized in terms of polarity and provide information for conducting tests to properly confirm wetting and solubility.

[0058] Light diffraction measurements determine the following for a dispersant: i) Provides good wetting of the sample (allows dispersion) ii) Do not lyse the sample iii) No air bubbles iv) have appropriate viscosity v) transparent to the irradiating beam vi) have a different refractive index than the sample vii) be chemically compatible with the materials used in the device

[0059] The most commonly used dispersant is water. However, water is not suitable for all samples due to poor wetting or sample dissolution. As a general rule of thumb, sample solubility in the dispersant can be minimized by using a dispersant of opposite polarity to the sample, but this is balanced by the demands of particle wetting, which becomes more difficult without the aid of a surfactant for large polarity differences. A list of possible dispersants is provided in Table 2, in order of decreasing polarity. Table 2 lists exemplary dispersants in order of decreasing polarity.

[0060] [Table 2]

[0061] One test to determine if a dispersant liquid is suitable is to place a few drops of a particulate sample into a beaker containing the dispersant. A visual check can determine whether the sample disperses easily in the diluent. This visual check is quicker than conducting measurements on various dispersants. Good wetting between the particles and the dispersant will indicate a uniform suspension of the particles in the liquid, while poor wetting will show droplets on top of the powder and may indicate significant aggregation and settling.

[0062] Additionally, it is possible to visually determine whether the sample is soluble in the diluent.

[0063] If the user is unsure, the software can prompt the user to analyze the sample, and the change in shading is monitored during agitation (e.g., stirring or sonication). The stability of the sample in the diluent can additionally (or alternatively) be analyzed visually by leaving the sample in the diluent for a period of time.

[0064] When using water as a dispersant, surfactants may be suggested by the software tool to improve stability (e.g., to prevent clumping).

[0065] Surfactants are added in small amounts, typically one drop per liter of dispersant. Adding too much surfactant can cause foaming due to the action of stirring or pumping the sample, introducing air bubbles into the system. Air bubbles can cause scattering that is difficult to distinguish from scattering due to particles, which can bias the measurement. Antifoaming agents can be added to prevent foaming, but because this may contain particulates, it is important to add the antifoaming agent to the dispersant before measuring background.

[0066] The software tool may prompt the user to: 1. Add a small drop of surfactant to the sample placed in a dry beaker and mix to form a thick paste. 2. Add the dispersant and mix again.

[0067] This approach avoids sample clumping.

[0068] Examples of commonly used surfactants are shown in Table 3. Table 3 shows surfactant information.

[0069] [Table 3]

[0070] Compatibilizing agents may be used instead of (or in addition to) surfactants. Compatibilizing agents work by charging the particle surfaces, causing them to repel one another (i.e., giving them a zeta potential that aids dispersion). Compatibilizing agents are typically added in large amounts, usually 1 gram per liter. A list of commonly used compatibilizing agents is given below: -Sodium hexametaphosphate (e.g., Calgon) -Sodium pyrophosphate -Trisodium phosphate -ammonia -Sodium oxalate -Calcium chloride Many of these are solid substances that dissolve in the dispersant, and after preparation the solution must be filtered to remove impurities.

[0071] The software tool may prompt the user to prepare a slurry to be added to the optical diffractometer by mixing a small amount of concentrated sample, a dispersant, and any additives.

[0072] Once the particles are dispersed, the sample may then be added to the dispersion unit of the optical diffractometer without the addition of a surfactant. To prevent the sample from settling in the beaker, the software may instruct the user to:

[0073] 1. Using a pipette, pipette the sample up and down continuously while stirring. 2. Use a pipette to add the sample to the reservoir of the dispersion unit.

[0074] Stability assessment The next part of the workflow for the software tool is to evaluate the sample's stability in the optical diffractometer. The software tool may prompt the user to add the sample to the laser diffraction instrument (e.g., to the dispersion unit) to generate an appropriate amount of light blocking. For expected particle sizes less than 1 micron, an appropriate light blocking range may be 3-5% light blocking. For expected particle sizes between 1 and 10 microns, an appropriate light blocking range may be 5-10% light blocking. For particle sizes greater than 10 microns, a 10-20% light blocking range may be appropriate.

[0075] The software tool may automatically determine the appropriate light blocking range from preliminary analysis results obtained when the sample is added to the diluent in the dispersion unit and the resulting dispersion is circulated through the sample cell.

[0076] If the user provides an expected particle size, the software may suggest a light-blocking range based on the defined range above. The software tool may then prompt the user to add a sample to this light-blocking range and take a series of short-duration measurements (e.g., 2-10 s, e.g., 5 s) over a period of several minutes (e.g., 2-5 min, e.g., 3 min). Once all results are collected, particle size (e.g., Dv10, Dv50, Dv90) may be plotted against time and analyzed to confirm the stability of particle properties.

[0077] A common problem in optical diffraction analysis is when material is dissolving. This is identifiable by a slight decrease in light blocking due to the reduction and eventual disappearance of fine particles. In this situation, the coarse fraction becomes more dominant. When light blocking versus Dv10 is plotted on a live trend graph, if dissolution is present, laser light blocking will decrease and Dv10 will increase.

[0078] The software automatically identifies whether the sample is dissolving and its dissolution rate. The software then suggests an appropriate time frame for the sample to stabilize and be suitable for laser diffraction measurement, using, for example, ISO standards for variability (or user-targeted input criteria). Additionally, the software tool may notify the user that the sample may be dissolving, but with some caveats (e.g., maximum measurement duration). If the sample is dissolving too quickly, the software tool may prompt the user to consider using surfactants or additives, or changing the dispersant.

[0079] The second most common problem is clumping of material. If the sample is clumped, a tail of large particles will appear in the distribution. Another common sign of clumping is a gradual decrease in light interception. This will be evident from the trend of light interception vs. Dv90. As the sample clumps, light interception will decrease and Dv90 will increase. If the software identifies that the sample is clumped, it may prompt the user to first increase the agitation speed to disperse the coarse particles, and if that does not work, it may instruct the user to try using ultrasound.

[0080] The software tool may guide the user to add the sample to the dispersion unit. The software tool may guide the user to add the sample to the dispersion unit. A predetermined number of measurements (e.g., five) may then be taken automatically under software control. The software may analyze the results and, by trending the data (as described above), identify whether the sample is unstable or shows signs of the common problems described above (dissolution or aggregation). If this occurs, the software may guide and suggest ways to improve the dispersion; for example, using surfactants or sonication as needed. Once the sample is stable (within ISO limits or other user-defined limits), the software may guide the user to the next step and perform the necessary titration.

[0081] The next part of the workflow may involve automatically running a series of tests to identify appropriate settings for a particular sample type for analysis using the optical diffractometer. For wet measurements (using liquid dispersants), the main settings to consider are the stirring speed, the appropriate light blocking range, and whether ultrasound is required.

[0082] To identify each parameter, a titration may be performed in which the parameter is varied and the particle characteristics are automatically determined. The advantage of having a software tool perform these parameter titrations automatically is that the software provides a simple procedure for automatically identifying the appropriate settings. Typically, a minimum of five measurements may be sampled from each parameter in each titration.

[0083] 2 is a flow diagram illustrating a method for evaluating the stability of a dispersion for laser diffraction analysis. In step 501, a software tool may automatically control a laser diffraction instrument to take a series (e.g., at least five, e.g., six) of light diffraction measurements of the dispersion. In step 502, the software may analyze whether the results are stable, for example, against a predefined standard, such as an ISO standard (e.g., as described herein) or other similar standard. If the results are determined to be stable, the software tool may report this to the user and proceed to the next step (e.g., stirrer speed titration).

[0084] If the results are not stable against predefined criteria (e.g., ISO criteria), the software may perform an automatic analysis of the results to identify the cause in step 504. The causes of a non-stable dispersion may include specific classifications such as deagglomeration, dissolution, flocculation, and sedimentation. Depending on the result of a successful classification, the software may report appropriate remedial actions to improve the stability of the dispersion in step 505. If the software cannot classify the reason why the dispersion is unstable, a suggestion (e.g., displayed to the user) may be made in step 506 to use a dry dispersion, or the user may be requested to reconsider the dispersion in other ways.

[0085] FIG. 3 shows a set of criteria (and associated exemplary inset graphs) that may be used to automatically identify each class of unstable dispersion. Other approaches may be used; these examples are illustrative. Particle size (Dv10, Dv50, Dv90, etc.), light obstruction, and total scattering across all detectors are three measures that may be used to classify unstable dispersions. Disaggregation may be identified if there is a negative trend in particle size 601, a positive trend in laser light obstruction 602, and a negative trend in total scattering 603. Dissolution may be identified if there is a positive trend in particle size 604, a negative trend in light obstruction 605, and a negative trend in total scattering 606. Aggregation may be identified if there is a positive trend in particle size 607, a negative trend in light obstruction 608, and a positive trend in total scattering 609. Precipitation may be identified if there is a negative trend in particle size 610, light obstruction 611, and total scattering 612.

[0086] Figure 3a shows the inset graph 601 of Figure 3 in more detail. Particle size (Dv10, Dv50, Dv90) is plotted on the y-axis over 20 consecutive measurements on the x-axis. A decreasing trend in particle size, especially Dv90 particle size, may indicate that larger particles (with aggregates) are breaking down. A threshold slope, or threshold % change in particle size over a series of measurements, may be defined as a first indicator for deagglomeration.

[0087] Figure 3b shows the inset graph 602 of Figure 3 in more detail. The % light blocking is plotted against the measurement number over 20 measurements. There is a trend of increasing light blocking, consistent with the large particles breaking down into smaller particles. The threshold slope, or threshold % change in light blocking over a series of measurements, may be defined as a second indicator for disaggregation.

[0088] Figure 3c shows the inset graph 603 of Figure 3 in more detail. The light energy at each of several detectors is shown over a set of five measurements (higher detector numbers correspond to higher scattering angles). Arrow 603a points from the earlier curve to the later curve. Scattering at lower angles decreases in the later measurements, indicating that larger particles are breaking down (i.e., disaggregating). The net effect of disaggregating larger particles is to reduce the total amount of scattered light energy detected by the detectors (i.e., the trend of the area under the curve for successive measurements is negative). A threshold slope in the total scattering detected by all detectors, or a threshold % change in the total scattering detected by all detectors over a series of measurements, may be defined as a third indicator of disaggregation.

[0089] If all three indicators for deagglomeration are met, the software may report to the user that the dispersant is deagglomerated and that they should consider using a surfactant and / or ultrasonic titration.

[0090] Figure 3d shows the inset graph 604 of Figure 3 in more detail. The particle size distribution is shown (top), with volume percent on the y-axis and size class on the x-axis. A set of five curves corresponding to five consecutive measurements is shown. The trend in these measurements is a decrease in the volume percent comprising smaller particles. Dv10 particle size is plotted against the number of measurements (bottom), with Dv10 tending to increase over consecutive measurements. The pattern of increasing particle size, or more specifically, the pattern of increasing Dv10, may be defined as a first indicator for dissolution. For example, a threshold slope, or a threshold percent change in particle size (e.g., Dv10) over a series of measurements, may be defined as a first indicator for dissolution.

[0091] Figure 3e shows the inset graph 605 of Figure 3 in more detail. Laser light blockage is plotted against sample number over five consecutive measurements. Light blockage decreases with successive measurements. The threshold slope, or threshold % change in light blockage over a series of measurements, may be defined as a second indicator for lysis.

[0092] Figure 3f shows the inset graph 606 of Figure 3 in more detail. The light energy at each of several detectors is shown over a set of five measurements (higher detector numbers correspond to higher scattering angles). Arrow 606a points from the earlier curve to the later curve. Scattering at higher angles decreases in the later measurements, indicating that smaller particles are being removed (i.e., lysed). The net effect of removing smaller particles is to reduce the total amount of scattered light energy detected by the detectors (i.e., the trend of the area under the curve for successive measurements is negative). The threshold slope in total scattering detected by all detectors, or the threshold % change in total scattering detected by all detectors over the series of measurements, may be defined as a third indicator of lysis.

[0093] If all three indicators of dissolution are met, the software may report to the user that the dispersion is dissolved and that a change in dispersant and / or the use of a surfactant should be considered.

[0094] Figure 3g shows the inset graph 607 of Figure 3 in more detail. The particle size distribution is shown (top), with volume % on the y-axis and size class on the x-axis. Five curves corresponding to five consecutive measurements are shown, with arrows 607a pointing from earlier to later measurements. The trend in the measurements is an increase in volume % associated with larger particles. Agglomerations appear in the consecutive measurements, in the form of larger particles. A plot of Dv90 particle size (primary y-axis) and light shading (secondary y-axis) versus measurement number (bottom) is also shown, showing an increasing trend for Dv90 particle size and a decreasing trend for light shading. The pattern of increasing particle size, and more specifically, the pattern of increasing Dv90, may be defined as a first indicator of agglomeration. For example, a threshold slope, or a threshold % change in particle size (e.g., Dv90) over a series of measurements, may be defined as a first indicator for agglomeration.

[0095] Figure 3h shows the inset graph 608 of Figure 3 in more detail. Dv90 particle size (primary y-axis) and light blocking (secondary y-axis) are plotted (bottom) against the number of measurements, showing an increasing trend for Dv90 particle size and a decreasing trend for light blocking. The pattern of decreasing light blocking may be defined as a second indicator for agglutination. For example, the threshold slope, or threshold % change in light blocking over a series of measurements, may be defined as a second indicator for agglutination.

[0096] Figure 3i shows the inset graph 609 of Figure 3 in more detail. The light energy at each of several detectors is shown over a set of five measurements (higher detector numbers correspond to higher scattering angles). Arrow 609a points from the earlier curve to the later curve. Scattering at lower angles increases in the later measurements, indicating that larger particles are being created (i.e., by agglomeration). The net effect of creating larger particles is to increase the total amount of scattered light energy detected by the detectors (i.e., the trend of the area under the curve for successive measurements is positive). The threshold slope in the total scattering detected by all detectors, or the threshold % change in the total scattering detected by all detectors over the series of measurements, may be defined as a third indicator of agglomeration.

[0097] If all three indicators of flocculation are met, the software may automatically report this to the user. Optionally, the software may automatically increase the stirrer speed (or increase agitation) and repeat the measurement. The software may, for example, suggest a stirrer speed titration. The software may automatically suggest to the user to perform an ultrasonic titration. The software may further suggest at least one of changing the dispersant, adding a surfactant, or an additive.

[0098] Figure 3j shows the inset graph 610 of Figure 3 in more detail. It shows the particle size distribution, with volume % on the y-axis and size class on the x-axis. Five curves corresponding to five consecutive measurements are shown, with arrows 610a pointing from earlier to later measurements. The trend in these measurements is a slight increase in the volume % associated with smaller particles. Larger particles settle slightly faster than smaller particles. The pattern of decrease in particle size (e.g., Dv10, Dv50, Dv90) may be defined as a first indicator for settling. For example, a threshold slope, or a threshold % change in particle size over a series of measurements, may be defined as a first indicator of settling.

[0099] Figure 3k shows the inset graph 611 of Figure 3 in more detail. Laser light blocking is plotted against sample number over five consecutive measurements. Light blocking decreases with each successive measurement. The threshold slope of light blocking, or threshold % change over the series of measurements, may be defined as a second indicator of sedimentation.

[0100] Figure 3l shows inset graph 612 of Figure 3 in more detail. The light energy at each of several detectors is shown over a set of five measurements (higher detector numbers correspond to higher scattering angles). Arrow 612a points from the earlier curve to the later curve. Sedimentation results in a decrease in scattering at all angles. The threshold slope in the total scattering detected by all detectors, or the threshold % change in the total scattering detected by all detectors over the series of measurements, may be defined as a third indicator of sedimentation.

[0101] If all three indicators of settling are met, the user may be informed that the dispersion is subject to settling, and the measurement may be automatically repeated (or suggested to the user) with higher agitation (e.g., stirrer speed) and / or the user may be prompted to consider changing the dispersant.

[0102] If the software is unable to classify the instability as being due to one of the above categories, the software may report this to the user and advise them to consider a dry measurement or to reconsider whether a more stable dispersion can be obtained in other ways.

[0103] Model Selection The software tool may prompt the user to select a model. Model options may include a general model, a narrow model, or a validation model. The model may include at least one of an inversion approach used to obtain particle size or particle size distribution from the scattering data (measured by the detector), a smoothing factor, and a selection of the detector whose scattering data is used to determine the particle size (or size distribution).

[0104] (1) General Purpose Model - This model is suitable for most samples, including natural and ground materials. The software may guide the user to select this model unless the sample is known to meet the more specific requirements of a more specialized analysis mode. The general purpose model may use non-negative least squares for the inversion.

[0105] (2) Narrow Model - This model is suitable for samples with one or more narrow modes, each with a magnitude much smaller than 10 years. It is not intended for use with broad distributions that exhibit multiple peaks. The model may be similar to the generalized model. This model may use non-negative least squares for inversion.

[0106] (3) Validation Latex Model - This model is designed to allow the analysis of one or more very narrow latex size standards, such as those used during instrument validation. This model may employ the ratio of forward to backscatter to determine particle size. A smoothing factor may be employed to keep the particle distribution unimodal.

[0107] The software may use rules (or may request input from the user) to suggest the correct analytical model. The analytical model may be automatically selected by the software when the first measurement is made. The model to use may be prompted each time a measurement is made. A generic model may be used as the default (suitable for many materials), or a narrow model may be automatically selected if the span is less than 1.00. A validation model may be automatically selected if the span is less than 0.25.

[0108] Selecting the Lighting Type The light source in some laser diffraction instruments has two or more wavelengths (e.g., red laser light and blue light, which may be provided by an LED). In some instruments (e.g., the Mastersizer 3000), blue light illumination can improve measurement resolution for particles smaller than 1 μm in size.

[0109] In certain embodiments, the software automatically (or by default) performs measurements using a shorter wavelength light source (e.g., blue) if the particle size (e.g., determined in an initial analysis or entered by a user) is less than 1 micron. The software can also compare the observed blue light blocking to the observed red light blocking. The higher the blue light blocking, the smaller the particle size (e.g., less than 10 nm).

[0110] Stirrer velocity titration The agitator in the wet dispersion unit (from which the sample is circulated through the sample cell for optical diffraction analysis) is required to ensure that the dispersion is uniform and that the sample passing through the measurement cell is representative. For larger or denser materials, an agitation rate titration is performed to ensure that all particles in the sample are suspended. For emulsion samples (where the particles have immiscible droplets), an agitation rate titration can determine at what speed the droplets begin to break up under the action of the agitator.

[0111] During a stirrer speed titration, the software automatically performs a series of measurements at different stirring speeds. The optimal stirring speed range or value is saved in memory and / or presented to the user at the end of the measurement. The user is first prompted to add the sample to the dispersant unit at an appropriate light blocking ratio (e.g., 5% for very fine particles, 10% for coarse particles). Depending on the sample information entered by the user, the first measurement is performed at a relatively low stirring speed (e.g., 750 RPM for emulsions, 1500 RPM for other materials; however, these examples are merely illustrative and will vary depending on the stirrer design and the optical diffractometer). The software gradually increases the stirring speed (e.g., in 500 RPM increments if measuring primary particles by default), waits a few minutes for the system to stabilize (the default delay is 5 minutes, but a different delay time can be used), and then performs the measurement.

[0112] The software may be configured to identify changes in particle size and recommend a series of measurements (default 250 RPM increments). The software controls the optical diffractometer and performs the series of measurements autonomously. The series of measurements determined by the software may respond to parameters such as the addition of surfactants or the use of diluents other than water. For example, faster stirring speeds can generate bubbles, which can lead to inaccurate and misleading results.

[0113] By the end of the series of measurements, the software analyzes the change in particle properties (e.g., particle size) and the residual % between each setting. An optimal agitation speed may be identified, and an appropriate range of agitation speeds may be suggested that provides stable results and within a user-defined range of variation (e.g., based on ISO standards).

[0114] An algorithm may be used to determine the appropriate stirrer speed (or intensity of mechanical stirring, if other stirring methods are used instead of a stirrer).

[0115] Very high intensity agitation may be dedicated to high density particles that require more force to adequately distribute the sample throughout the sample cell. Particle density may be a parameter that the user can provide to the processor (e.g., via an input device such as a keyboard in the initial stage described above). A threshold particle density may be defined, below which the agitation intensity may be limited to a predetermined threshold, e.g., 2 g / cm. 3 may be defined as a particle density threshold below which particle densities may limit the stirrer speed recommended by the processor (according to an algorithm) to 3000 RPM or less, thereby preventing high stirrer speeds from being recommended when not necessary.

[0116] The optimal stirrer speed (or range of stirrer speeds, or range of stirring intensities) may be selected by first finding a region where particle size stabilizes. A stirrer speed may be considered to result in stable particle properties if the gradient of the particle property between successive measurements at that stirrer speed is less than a predetermined threshold. Multiple particle characterizations (e.g., to determine particle size) may be performed at different stirring intensities (e.g., stirrer speeds). The stirring intensities may be binned. For example, in the example stirrer speed, a bin width of 500 rpm may be used (although other bin widths, such as 250 rpm or 1000 rpm, may also be used). The gradient of the particle property versus stirring intensity across each bin (e.g., Dv10, Dv50, etc.) may be calculated. Bins with a gradient less than a minimum threshold (e.g., 0.01) may be identified as constituting the optimal range of stirring intensities.

[0117] Figure 4 shows particle size (Dv10, Dv50, Dv90 shown on the y-axis) as the stirrer speed (shown on the x-axis) is gradually increased for successive measurements on the same sample. The particle size stabilizes at stirrer speeds above ~2400 rpm. For this sample, the algorithm may recommend stirrer speeds above 2400 rpm or even above 2500 rpm.

[0118] From this set of stirrer speeds on the stable region, the standard deviation within each set of measurements at each stirrer speed bin may be calculated, with the stirrer speed bin with the lowest standard deviation having the most measurements and therefore being the one that should be recommended to the user.

[0119] An example of an algorithm for determining the optimum range of stirrer speeds and the best stirrer speed is given below.

[0120] Measurements are taken at several different stirring speeds If particle density < threshold: Remove stirring speeds < 3000 RPM Bin the data into stirring speed bins For each agitation speed bin: Calculate the mean and standard deviation of d10, d50, and d90 Calculate the gradients of d10, d50, and d90 For all gradients < 0.01: Score function = calculate the sum of all standard deviations Adding agitation speed bins to the optimal agitation speed range Best stirring speed = minimum value of the score function Returns the best stirring speed and the optimum stirring speed range

[0121] The list of best stirrer speeds determined by the algorithm is compared to the list determined by an experienced user in Table 4 below. Table 4 shows a comparison of the stirrer speed values ​​obtained from the experienced users and the stirrer speed values ​​calculated by the algorithm.

[0122] [Table 4]

[0123] In most cases, the algorithm can predict a stirrer speed that is close to or equal to the speed recommended by an experienced user.

[0124] An exemplary method for stirrer speed titration is shown in Figure 5. In step 801, the user is asked whether the sample is an emulsion. If the answer is "no," the software checks (in step 802) whether the expected particle size (Dv50) is greater than 1 micron. If the expected particle size is greater than 1 micron, the user is prompted (in step 803) to add the sample to 7-10% light. If the expected particle size is less than 1 micron, the user is prompted to add the sample to 3-5% light. In step 806, a series of laser diffraction measurements is automatically performed at successively increasing stirrer speeds (e.g., 1500, 2000, 2500, 3000, 3400 rpm). In step 807, the software checks whether there is a wide range of stirrer speeds (e.g., 1500 rpm) over which the results stabilize. If the answer is "yes," the software suggests (in step 808) an optimal range of stirrer speeds that results in the least change in reported particle size. If the answer in step 807 is "No," the user may be prompted to run a more closely spaced set of stirrer speeds, for example, using a stirrer speed intermediate those defined above. The user may be presented with this option even if the answer in step 807 was "Yes."

[0125] If the sample is an emulsion (i.e., the answer in step 801 is "yes"), in step 810 the software checks whether the expected particle size is <1 micron. If the answer in step 810 is "yes," the user is prompted to add the sample to 3-5% shading, and a series of measurements is automatically performed at successively increasing stirrer speeds in step 811. If the answer in step 810 is "no," the user is prompted to add the sample to 7-10% shading, and a series of measurements is automatically performed at successively increasing stirrer speeds in step 811. The series of increasing stirrer speeds in step 811 may be, for example, 1200, 1300, 1400, 1500, 1600, and 1700 rpm. In step 812, the software may check whether there is a decrease in particle size at higher stirrer speeds. If the answer in step 812 is "no," the software may perform further measurements at successively higher stirrer speeds (e.g., up to 2000 rpm, or until the particles begin to break down). If the particles are breaking down and the answer in step 812 is "yes," the maximum stirrer speed may be limited in step 814 to a speed at which the particles do not break down.

[0126] A final range of stirrer speeds may be selected by software, the final range being the range of stirrer speeds where the particle size stabilizes, eg, the particle size change is below a threshold value.

[0127] Dark titration Light loss is a good indicator of how much sample is circulating in the dispersion unit. Light loss titration determines how much sample should be added to the diluent in the dispersion unit. If too much sample is added, a phenomenon called multiple scattering can occur, resulting in a decrease in the particle size determined by the instrument.

[0128] Adding too little sample can result in large particles in the sample not being represented, and noise in the scattering data can lead to unstable Dv90. Depending on the initial information provided on the expected particle size, the software may suggest an appropriate initial light blocking range. If agitation rate titration has previously been performed, the particle size has already been determined, and the software may make initial light blocking recommendations based on the previously determined particle size. In some embodiments, the software may prompt the user to add sample to the dispersion unit, and the software may control the laser diffraction device to perform an initial analysis of particle size (e.g., Dv50) and, optionally, an initial analysis of particle size distribution (e.g., a polydispersity measurement such as Dv90-Dv10).

[0129] Material type Wet measurement shading Very fine particles (e.g., Dv50<1μm) <5% Fine particles (e.g., Dv50 1-100μm) 5-10% Coarse particles (e.g., Dv50>100μm) 10-20% Polydisperse particles (e.g., Dv50 1-500 μm) 10-20%

[0130] Based on measurements taken from the sample, an algorithm may be used to determine an appropriate light blocking range for the sample.

[0131] An example of a light shading titration for a beverage is shown in Figure 6. Figure 6 shows a plot of particle size (101 Dv10, 102 Dv50, 103 Dv90) on the y-axis against light shading on the x-axis.

[0132] Drinks contain small particles, 1 μm in diameter or smaller. For particles in this size range, the software may be configured to ignore shading values ​​greater than 10% due to the risk of multiple scattering. Similarly, for particles 10 μm or larger, shading values ​​less than 5% are ignored due to the low signal-to-noise ratio. After this filtering, the data points shown in Figure 7 remain.

[0133] Figure 7 contains measurements from three light-blocking bins. Automatic binning of the data does not guarantee that the data will be consistently included in all three bins. Because light blocking within each group varies significantly between measurements, if the bin width is fixed, light blocking measurements within one group may straddle the bins. If the bin width is set wide enough to prevent this, the binning process may inadvertently combine groups of light-blocking measurements.

[0134] To prevent this, a k-means clustering algorithm (or other robust clustering method) may be used to sort the data into k clusters. To group the data into the correct number of clusters, the user may be prompted to select the number of shading values ​​at which the particles were measured (e.g., corresponding to a particular loading of sample into a volume of diluent in the dispersion unit).

[0135] The average size of each cluster is shown in Figure 8. Statistics for determining optimal shading may be derived from the particle sizes in Figure 8. Stability is determined by the particle size plateau, and an algorithm may be implemented that looks for the minimum gradient between particle properties (e.g., Dv50) among bins of shading, returning the optimal shading value. The gradients for each of the three sizes are shown in Figure 9.

[0136] The particle properties (Dv10, Dv50, Dv90) all show different trends. For light blocking, the maximum acceptable slope may be set as a multiple of the minimum slope. The multiple may be a number between 3 and 5, e.g., 4. Light blocking below this threshold may be considered acceptable. The range of acceptable light blocking values ​​may be defined as the range of light blocking values ​​found to be within this factor of the lowest variation slope.

[0137] Not all samples will return an acceptable light-blocking range using this approach. A range of light-blocking values ​​is required because it may be difficult for a user to precisely target a single light-blocking value by adding a specific amount of sample to the dispersion unit. Targeting a range of light-blocking values ​​may be much easier for the device user. To ensure a range of light-blocking is returned, the lowest slope multiple defining the acceptable slope range may be increased incrementally (e.g., in steps of 1) if fewer than two light-blocking values ​​are returned. The incremental increase continues until the algorithm returns multiple light-blocking values. To determine the range of acceptable values, the lower light-blocking value may be rounded down to the nearest integer percentage value, and the highest may be rounded up. Following this approach, the resulting light-blocking values ​​returned by the algorithm for different sample types are shown below. Table 5 shows a comparison of light-blocking values ​​obtained from experienced users with those calculated by the exemplary algorithm.

[0138] [Table 5]

[0139] The results show that the algorithm accurately estimated the optimal laser blocking range in most cases.

[0140] The features of the techniques described above for identifying optimal light blocking values ​​or ranges may also be applied to determine other measurement parameters.

[0141] Figure 10 shows the stable region of light blocking. The particle size (Dv50) in this example is relatively small, ~1 micron, so low values ​​of light blocking are required to stabilize the particle properties. On the graph, the stable region is defined as the region between 3% and 5% light blocking.

[0142] FIG. 11 shows a flow diagram of an exemplary method for performing a light blocking titration. In step 601, the user is prompted to: i) add sample until a 2-3% light blocking range is reached; ii) then add sample until a 4-5% light blocking range is reached; and iii) then add sample until a 7-9% light blocking range is reached. After each light blocking range is reached, the software (by controlling the laser diffraction device) automatically performs multiple (e.g., six) laser diffraction analyses. After the measurement in the 7-9% light blocking range, the software analyzes the data in step 602, looking for evidence of multiple scattering. If multiple scattering is detected, no further measurements are taken, and the method proceeds to step 604. If multiple scattering is not detected in step 602, the user is prompted to add sample to 10-12%, 15-17%, and 19-21% light blocking, and measurements are taken again at each level. In step 604, the measurements at each light blocking range are analyzed and stability is assessed (as described above). In optional step 604, the software determines whether the sample is stable over more than two-thirds of the light-blocking range tested. If the answer is "yes," the software suggests the range of light-blocking that results in the least change in particle size, for example, based on the lowest relative standard deviation (or as described above). If the answer is "no," the software may suggest (or automatically perform) a more detailed light-blocking titration with smaller increments of light-blocking in step 606. In step 607, the user may be provided with the option of a more detailed light-blocking titration 606. In step 608, the optimal range of light-blocking (based on either the coarse or the more detailed titration) is reported to the user.

[0143] ultrasonic titration Depending on the type of measurement the user indicates they desire, ultrasonic titration may be appropriate. If the user indicates they wish to characterize aggregates, ultrasonication is not recommended because ultrasonication is likely to disrupt aggregates. If the user indicates they are interested in primary particle characteristics, ultrasonic testing may be recommended. A dispersion titration, including ultrasonic titration, may allow the user to determine the optimal range of power and duration required to disperse the sample. If the dispersion titration is flat and no particle size reduction is observed during sonication, the sample may be fully dispersed without the use of ultrasonication. Conversely, if particle size continues to decrease during sonication with ultrasonication (e.g., a stable particle size is not reached), primary particles may be being broken down by the ultrasonication.

[0144] Particle size should remain stable after sonication. If particle size begins to increase due to particle reagglomeration, additives may be required to stabilize the dispersion. The software tool may guide the user to add the sample to appropriate shading and set the dispersion unit to an appropriate agitation speed (which may be automatically determined as described above). If user input indicates that the sample particles are fragile, the ultrasound intensity may be set to a low level. The software tool may control the optical diffraction tool to take a series of measurements before ultrasound is applied, and then apply ultrasound for an appropriate initial time (e.g., 30 seconds, 50% intensity). If a consistent decrease in particle size is observed, the sonication intensity may be gradually reduced, and the sample may be recharacterized to identify an appropriate intensity for sonication. The particle size and shading for each measurement may be automatically determined by the software using the optical diffractometer.

[0145] After determining the appropriate ultrasonic treatment intensity, the duration of ultrasonic treatment may be increased stepwise (e.g., between 10 and 30 seconds, e.g., 20 seconds) until the appropriate duration of ultrasonic treatment is identified. The appropriate duration of ultrasonic treatment may be defined as when the particle size stabilizes, i.e., when the primary particles are not crushed but the agglomerates are broken down. For example, Dv10, Dv50, and Dv90 may change less than a predetermined threshold when the particle size stabilizes.

[0146] The software may prompt the user to confirm the preliminary conclusions about stability automatically determined by the software. To facilitate this, the user may be automatically presented with the software's preliminary conclusions and plots showing trends in Dv10, Dv50, and / or Dv90 with respect to sonication time, and particle size distributions may be provided for different durations (which may reveal what is happening to the sample during sonication, such as breaking down aggregates). The advantage of this approach is that users are trained to recognize and question trends in the measurement data, thereby rapidly improving their expertise.

[0147] FIG. 12 shows a dispersion titration with ultrasonic titration. Sample sizes Dv10 201, Dv50 205, and Dv90 209 are plotted against record number. At each successive record number, an additional amount of agitation is applied (i.e., each successive record number has an additional increment of agitation or sonication time). In a first stage 221, the sample is agitated to disperse. The Dv90 sample size 209 does not stabilize during agitation, indicating that the sample has aggregates that break down as a result of agitation. In a second stage, ultrasonication is applied to break down larger particles, and the Dv90 particle size stabilizes (at point 210). The stability of a particle characteristic (e.g., Dv90) may be based on a convergence criterion (e.g., a % change between successive measurements below a predetermined value) or a threshold slope. At point 210, the ultrasonic duration is found to be sufficient for the sample to stabilize. The sonication time for this measurement may be reported (on a display) and / or stored by the software tool as optimal for the sample. This process may be repeated for multiple samples to select the optimal sonication time range.

[0148] Alternatively, the particle size elbow may be defined as defining the optimal duration of sonication. In this method, a line may be defined on a graph plotting a sample property (e.g., Dv90) against sonication time between the sample property corresponding to the minimum sonication time and the sample property corresponding to the maximum sonication time. The optimal sonication time may be defined as the measurement of the greatest distance from the line.

[0149] An example algorithm that defines this approach is given below.

[0150] #The algorithm starts after several measurements are performed at several ultrasound durations Bin the data into sets of ultrasound durations Calculate the average of the d50 for each period For each ultrasound duration: d50_line = Linear interpolation of d50 using minimum and maximum duration values distance=d50_measured-d50_line The optimal sonication time is at the longest distance Return the optimal ultrasonic exposure time

[0151] FIG. 13 shows a flow diagram of an exemplary method for ultrasonic titration. In step 701, three initial measurements are taken without ultrasound. In step 703, ultrasound is applied at the current ultrasound level for a predetermined period (e.g., a starting level of 50% power, each period 30 seconds), and the software looks for a stable measurement of low-angle scattering (e.g., a change in scattered light intensity less than a threshold change in angle less than 10 degrees between the previous and current ultrasound doses). This sonication and measurement is repeated in step 703 until low-angle scattering stabilizes (or until the total sonication time reaches a threshold). In step 704, the results of the series of sonication times are analyzed to determine at least one particle size (e.g., Dv10, Dv50, Dv90). Step 705 involves verifying that these particle size results are stable. If the answer in step 705 is "yes," an appropriate ultrasound duration and power are selected from the titration (e.g., as described above), and the method proceeds to step 706. In step 706, repeat measurements (e.g., three times) are taken on additional sample aliquots to confirm the selected settings are appropriate. In step 707, the user is asked whether they want to try a higher ultrasonic power (which may result in a faster measurement time). If the answer is "yes," the method gradually increases the ultrasonic power in step 708 and returns to step 701 to repeat the sonication step at a higher ultrasonic power (e.g., 100%). If the answer is "no" in step 705, the method proceeds to step 709, where the software checks whether particle size is increasing. If particle size is increasing, the software concludes (in step 710) that the dispersant is inappropriate and recommends an additive or alternative dispersant. If particle size is not increasing, the software checks for a decrease in particle size in step 711. If a continuously decreasing particle size is found, the software concludes that the sample is ultrasonically milled and decreases the ultrasonic power in step 712, before returning to repeat the measurement of step 701 at a lower ultrasonic power.If step 711 does not find a continuous decrease in particle size, the software increases the ultrasonic power in step 708 and repeats the measurements from step 701. If the sonication power reaches 100% without achieving stable results, the software may recommend external sonication.

[0152] Dry / gas diluent dispersion For some samples, such as reactive or easily soluble samples, a dry (vapor phase diluent) may be appropriate.

[0153] To assist users in obtaining reproducible results from particle characterization by optical diffraction, a computer-implemented tool (i.e., a software tool) is provided that guides users through the process of developing a suitable test method for dry dispersions. The tool may be configured to guide the user as follows (preferably in the order shown): 1. Correct sampling procedures 2. Identify the appropriate feed rate 3. Identify the correct air pressure 4. Compare the results with a well-dispersed wet sample (optional) 5. Method validation and robustness testing

[0154] Representative Sampling Sample preparation before adding to the system is important. More than half of the problems encountered when measuring samples by laser diffraction are due to poor sample preparation. Software tools advise users on the correct procedure for removing the sample from the container. Advice may include: If the sample is a powder, larger particles tend to move to the top of the container and smaller particles to the bottom. If the sample is stored in a container, it should be thoroughly mixed before measurement. Shaking the container should not be done, as this often leads to further particle separation. Instead, hold the container in both hands and gently roll it over for 20 seconds, repeatedly turning it over. This works best if the container is half full. Another consideration is whether the material is free-flowing in its dry state. Good pouring characteristics indicate a non-cohesive powder, which will usually disperse without issue in a dry powder feeder, while cohesive materials tend to stick together or cake, biasing measurements. Dry the sample in an oven to remove moisture and eliminate clumping, but be careful not to damage it. If it is clear that using an oven will damage the sample, use a desiccator. Fresh samples that have not had time to absorb atmospheric moisture are always preferred and usually yield better results. If hygroscopic samples are required to be transported through a system over any distance, they should be sealed in silica gel bags as soon as practical in the pipe.

[0155] Software tools can display good sampling procedures as videos or animations. Another option is a series of images outlining the actions to be followed for sampling (e.g., mixing samples stored in a container).

[0156] Feed rate / hopper gap titration In dry dispersions, wetting is not a criterion, so the choice of dispersant is less important. In dry dispersions, the gas phase dispersant may comprise air, dry nitrogen, or other inert gases.

[0157] Gas-phase dispersion of particulate samples for particle characterization by laser diffraction measurements is typically performed by introducing the material from a hopper into a venturi. The sample is delivered from the hopper to the venturi by a vibrating sample tray. This technique is not required, and any sample dispenser can be used to deliver the sample into the dispersion gas stream (which then passes through a sample cell, where the particles scatter light and the light is detected at various scattering angles). Alternative sample dispensers include screw-feed dispensers.

[0158] The amount of material passing through the measurement cell is recorded by laser obscuration measurement. An appropriate feed rate may be automatically selected corresponding to an appropriate range of obscuration for the majority of the measurement period. For dry dispersions, the obscuration amount should be less than 10% of the laser obscuration. For fine, cohesive samples, lower obscuration is required to maximize disperser efficiency, while coarser materials can be measured at higher obscuration. If poor dispersion is observed for fine or cohesive materials, lowering the upper obscuration limit may be required to improve disperser efficiency. The following target obscuration may be appropriate for materials with different sizes: 0.1-3% for fine materials (<10 μm), and 0.5-6% for coarse materials (>100 μm).

[0159] The appropriate light blocking range varies depending on the size and nature of the sample. Ideally, measurements should be performed when the powder flow on the dispenser (e.g., vibrating tray) is uniform and the laser blocking is always within the target range. However, this may be difficult with cohesive powders, which may not be uniformly dispersed. In such situations, using a microtray and reducing the gap between the hopper and the dispenser tray can be effective. The stability of the sample flow may be optimized by adjusting the configuration of the sample tray and hopper. For flowable materials, reducing the hopper height may prevent the material from flowing out of the tray too quickly. For highly cohesive materials, increasing the hopper height may improve flowability. Basket and ball-bearing hoppers can be used to improve the flow of cohesive or lumpy materials.

[0160] If the sample is not flowable, for example, if the observed light blockage varies by more than a predetermined amount over a predetermined time, the software tool may be configured to recommend a basket and ball bearing hopper.

[0161] Regardless of the type of sample dispenser, the software may be configured to control the sample dispenser to gradually increase the feed rate until the sample is within the appropriate laser blocking range (to identify an initial feed rate). The software may then be configured to take a series of measurements (at three different feed rates) within a ±5% range around the initially identified feed rate. This may be automated, with the user simply following instructions for when to add sample to the hopper (if required). Once the software has acquired a series of data, an algorithm may be used to determine which settings are most suitable by looking at the overall stability of the data. At this point, the identified hopper gap and feed rate are repeated, and the user is prompted to take three sample aliquots to confirm they are correct.

[0162] An example of a suitable algorithm for identification may include determining the minimum standard deviation between subsequent measurements and selecting the corresponding feed rate as the optimal feed rate (or selecting the combination of feed rate and hopper gap as optimal). Exemplary results employing this approach are shown below. Table 6 shows a comparison of feed rate values ​​obtained from an experienced user with those calculated by the exemplary algorithm.

[0163] [Table 6]

[0164] These results show that the algorithm can correctly identify the optimal feedrate to use or select an acceptable feedrate that is adjacent to the value selected by an experienced user.

[0165] One thing to note is that some of these measurements were taken not only for multiple hopper gap values, but also for multiple feed rate values. This titration allows the algorithm to extract these parameters as a pair. The algorithm is able to correctly identify the correct hopper gap when there is a titration on this value, as shown in Table 7. Table 7 shows a comparison of the hopper gap obtained from an expert user and the hopper gap calculated by the exemplary algorithm.

[0166] [Table 7]

[0167] Pseudocode for an example algorithm is shown below.

[0168] #After several measurements have been performed at several feed rates (and possibly several hopper gaps), the algorithm begins Bin the data into pairs of feed rate and hopper gap For each pair: Calculate the mean and standard deviation of d10, d50, and d90 Calculate the gradients of d10, d50, and d90 For all gradients < 0.01: Score function = calculate the sum of all standard deviations Optimal feed rate and hopper gap = minimum value of the score function Returns the optimum feed rate and hopper gap

[0169] In an alternative embodiment, the feed rate may be automatically set to achieve a target (or target range) of light occlusion (a suitable range of light occlusion is defined above). For example, closed-loop control employing at least one of proportional, integral, and derivative error feedback may be employed. After each measurement of light occlusion, the feed rate may be increased if the light occlusion is below the target level and decreased if the light occlusion is at or above the target level (this example corresponds to proportional error feedback).

[0170] In this embodiment, the user may select the desired hopper gap and shading, and the feed rate may be automatically selected to achieve that shading level.

[0171] If the material is fine (Dv50 less than 10 μm), the target laser obscuration may be set to 1%. If the material is coarse (Dv50 greater than 10 μm), the target laser obscuration may be set to 3%. If the user does not know the particle size, the active feed rate may be set to 1% of the target laser obscuration range. The workflow first guides and advises the user where it is likely to be required to use a basket and ball-bearing hopper. A default hopper gap may be set (e.g., 2 mm) to initially determine the feed rate. The active feed rate control increases the feed rate until the measured value approaches the target laser obscuration. As the measured value falls within the obscuration range (e.g., within a preset tolerance from the target value), the software measures until the hopper is empty. After the measurement, the software analyzes the requested average feed rate and suggests whether the hopper gap needs to be increased or decreased. The software performs measurements at different hopper gaps, and the algorithm analyzes all the results and suggests the optimal hopper gap. In this case, the optimal hopper gap is selected based on the minimum standard deviation between measurements and the minimum absolute gradient between different setpoints. At this point, repeat the setup with three different aliquots to ensure the identified settings are reproducible.

[0172] Air pressure titration Dry samples are typically dispersed (in an optical diffractometer) using a compressed air-driven venturi. The dispersion of a dry sample may be assessed by performing a pressure titration and comparing it to a well-dispersed wet measurement. Pressure titrations generally show a decrease in particle size with increasing pressure as agglomerates disperse. However, particle size can also decrease due to particle attrition, and these two processes can overlap.

[0173] A standard venturi utilizes two mechanisms available for dispersing dry powders: (a) velocity gradients due to shear stress and (b) particle-particle collisions. For most materials, these two mechanisms are sufficient to disperse the sample. However, highly agglomerated and stubborn samples may require the use of a high-energy venturi, which utilizes the more energetic dispersion mechanism of particle-wall collisions.

[0174] The software tool may be configured to recommend to the user when a higher-energy venturi is deemed necessary. The measurements required to perform an air pressure titration are at least partially automated. The user adds the sample to the hopper, and the software controls the optical diffractometer and sample dispenser to perform the desired measurements. For example, the software may control the instrument to perform a series of measurements at 4, 3, 2, 1, 0.5, and 0.1 bar (gauge pressure). The employed algorithm may take into account the number of peaks observed in the particle size distribution measured by the instrument from the dispersed sample to identify whether the sample is not fully dispersed. For example, peaks or shoulders corresponding to large particles (e.g., at low air pressure) may be identified as undispersed aggregates. Such peaks or shoulders should decrease or disappear at higher air pressures. If not, they represent genuine particles.

[0175] The software may require increasing the feed rate at lower air pressures to keep measurements within the target laser beam interruption. The algorithm developed may use the acquired data to identify where the data stabilizes and results have the least deviation, and look for trends between air pressures.

[0176] To determine the optimal air pressure to use for each sample, multiple measurements can be taken and binned into different air pressure bins. The air pressure bin with the smallest slope, and therefore on the size plateau, may be selected as the optimal air pressure for each sample. The results are included in Table 7 below. Table 8 shows a comparison of air pressures obtained from experienced users with those calculated by the exemplary algorithm.

[0177] [Table 8]

[0178] Pseudocode for an example algorithm is shown below.

[0179] #After several measurements at several air pressures, the algorithm starts #The workflow should include user input to check for expected monomodal samples. Bin the data into sets of air pressures For each pressure: Calculate the mean and standard deviation of d10, d50, and d90 If the sample is monomodal: Calculate the average psd Calculate the number of maxima in a psd If the number of maximum values ​​> 1: Ignore this air pressure value Calculate the gradients of d10, d50, and d90 For all gradients < 0.01: Score function = calculate the sum of all standard deviations Optimal air pressure = minimum value of the score function Returning optimal air pressure

[0180] Figure 14 shows an example of air pressure titration data, showing values ​​of Dv10 211, Dv50 215 and Dv90 219 with increasing air pressure. In this case, higher air pressures give more stable results.

[0181] Although reference is made to air pressure, any suitable gas may be used as the diluent in dry dispersion.

[0182] Comparison of dry dispersion results with wet dispersion results This is optional, as not all users have access to wet measurement data. To ensure the settings selected or recommended by the software tool are valid, a final test method for analyzing dry dispersions should be a comparison with the results of a properly dispersed wet sample (of the same material). While this is not always possible, it is highly recommended, as it ensures that particles are properly dispersed and not milled. Users can input measurements of a well-dispersed wet sample into the user interface, and the software can automatically compare this with the results obtained from barometric titration. The software then directly compares the results and suggests the optimal air pressure, which can then be used to recommend a final operating procedure for characterizing a specific sample type. Because measurements from dry dispersions are expected to differ slightly from wet measurements, a similarity threshold, expressed as a percentage, may be required. If the difference exceeds the threshold, the workflow can suggest possible root causes and ways for the user to improve the results.

[0183] Dry dispersions performed at different air pressures can be compared to measurements of wet dispersions. The air pressure that most closely matches the results of the wet dispersion measurements may be selected as the optimal air pressure. If all measurements from the dry dispersion indicate a smaller particle size (e.g., Dv50) than the wet dispersion, the software may indicate to the user that the sample is too brittle for dry dispersion. If all measurements from the dry dispersion indicate a larger particle size (e.g., Dv50) than the wet dispersion, the software may indicate to the user that the sample is not fully dispersed and recommend the use of a higher energy venturi or wet dispersion.

[0184] Method validation and robustness testing The settings determined or recommended by the software tool as described above may be fine-tuned and / or further verified by performing repeated measurements. The amount of sample used in a measurement can be another sampling step and thus affect the variability of the results. For example, vibration of the sample tray can cause sample segregation. Larger, more free-flowing particles within the sample may arrive at the measurement cell before finer, more cohesive particles. Therefore, it is important (at least for dry measurements) to ensure that the entire subsample on the tray is measured. During the method development process, it is useful to perform repeated shorter measurements to optimize parameters. However, once the measurement conditions are determined, measurements should be set long enough to measure all the material on the tray. The sample mass required to obtain reproducible results depends on the sample size and polydispersity. Samples containing larger particles or with a highly polydisperse distribution require measuring larger masses. Therefore, to ensure method reproducibility, multiple measurements using different sample masses may be performed. In this case, the software prompts the user to add different amounts of sample to the hopper. The software performs the maximum number of measurements possible before all the sample is exhausted. The software then analyzes the results and identifies the sample mass that will give the most reproducible and consistent data.

[0185] The optimal measurement parameters (for wet or dry dispersions) obtained from a series of titrations may be validated by performing multiple replicates and ensuring that the results are similar. A series of replicates is performed using different aliquots from the same sample to assess whether the parameters are valid. For method validation, the minimum number of measurements recommended by the software may be three aliquots, but the user may be able to select a larger value. Measurements may be automated (simply add the sample to the hopper). After replicate measurements are performed, the software evaluates whether all measurements are similar and within the ISO standard for reproducibility (or other defined standard for measurements provided by the user).

[0186] The software may automatically determine the sensitivity of measurement repeatability for a particular measurement parameter. For example, a measurement parameter may be varied from an optimal value or range until the measurement is no longer sufficiently repeatable and falls within the measurement specifications. The amount of deviation required for the measurement to vary may be recorded or saved, indicating how sensitive the measurement is to that parameter. In another approach, the measurement parameter may be varied by a fixed amount (e.g., 10%), and the parameter with the greatest change in measurement may be identified as the most sensitive parameter.

[0187] After a method validation is performed and each measurement parameter is evaluated for suitability, the settings may be transferred to a format that can be saved by a user as a standard operating procedure. Such a standard operating procedure may enable at least partially automated repeat measurements using the identified measurement parameters. For example, the method parameters may be saved in a file on a non-volatile, machine-readable medium. The file may include instructions for configuring the laser diffraction instrument to perform the optimal method for analyzing a sample and may provide guidance to the user on the steps that they are required to perform.

[0188] An exemplary workflow for determining optimal measurement parameters for optical diffraction analysis is shown in FIG. 15. In step 301, multiple measurement results are received by a processor. The measurements are obtained by laser diffraction analysis of a particulate sample, with measurements obtained using different measurement parameters (e.g., sample preparation may vary between measurements, sample concentration may vary, the amount of agitation applied to the dispersion with the particulate sample and diluent may vary, or any of the measurement parameters described above may vary between measurements). In step 302, the processor analyzes the measurements and determines an optimal range of at least one measurement parameter for laser diffraction to perform particle characterization on the sample material. Step 302 may employ any of the methods and algorithms described herein.

[0189] A further exemplary workflow for determining optimal measurement parameters for laser diffraction analysis performed on wet dispersions is shown in FIG.

[0190] In step 311, a user inputs information about the sample to be analyzed by laser diffraction into a computer's graphical user interface. The information is provided to a software tool that automatically determines at least one optimal measurement parameter (or optimal range for the measurement parameter) for performing the laser diffraction analysis. The information provided by the user may include at least some of the information shown in Table 1 and described above.

[0191] In step 312, the software tool provides the user with appropriate sample preparation advice, which may be determined depending on the information provided by the user in step 311. The sample preparation advice may include the information and guidance described above in the sections entitled "Representative Sampling" and "Sample Preparation."

[0192] In step 313, the software tool at least partially automatically evaluates the stability of the dispersion containing the sample. The software tool may prompt the user to add the sample to a dispersion unit of the laser diffraction instrument, and the software tool may automatically evaluate the stability of the dispersion, for example, as described in the section entitled "Evaluating Stability."

[0193] In step 314, the software tool at least partially automatically determines an optimal value of at least one measurement parameter (or an optimal range of values ​​for at least one measurement parameter) for the laser diffraction analysis. Examples of measurement parameters include stirring parameters (including stirrer speed and ultrasonic intensity and time) and light blocking (i.e., sample concentration). Techniques for determining the values ​​of the measurement parameters may include at least some of the features described above in the sections entitled "Stirrer Speed ​​Titration," "Light Blocking Titration," and "Ultrasonic Titration."

[0194] In step 315, the software tool at least partially automatically evaluates the stability and robustness of the laser diffraction measurements performed using the selected (optimal) measurement parameters. The approach for evaluating the method stability and robustness may include at least some of the features described above in the section entitled "Method Validation and Robustness Testing."

[0195] In step 316, the software tool may transfer the method into a format that the user can save as a standard operating procedure, as described in the section entitled "Method Validation and Robustness Testing."

[0196] A further exemplary workflow for determining optimal measurement parameters for laser diffraction analysis performed on dry dispersions is shown in FIG.

[0197] In step 321, a user inputs information about a sample to be analyzed by laser diffraction into a computer's graphical user interface. The information is provided to a software tool that automatically determines at least one optimal measurement parameter (or optimal range for a measurement parameter) for performing the laser diffraction analysis. The information provided by the user may include at least some of the information shown in Table 1 and described above.

[0198] In step 322, the software tool provides the user with appropriate sample preparation advice, which may be determined depending on the information provided by the user in step 311. The sample preparation advice may include the information and guidance described above in the sections entitled "Representative Sampling" and "Sample Preparation."

[0199] In step 323, the user, guided by the software tool, establishes a steady flow of sample into the gas-phase dispersant. The software tool may at least partially automate the appropriate determination of measurement parameters (such as feed rate and / or hopper gap). The software tool may employ at least some of the features described in the section entitled "Feed Rate / Hopper Gap Titration."

[0200] In step 324, the software tool at least partially automatically determines at least one optimal measurement parameter (or optimal range for at least one measurement parameter) for the laser diffraction analysis. An example of a measurement parameter that may be determined in this step is air pressure. Techniques for determining the measurement parameter may include at least some of the features described in the section entitled "Air Pressure Titration" above.

[0201] In (optional) step 325, the software tool at least partially automatically compares the measurement results obtained using the optimal measurement parameters with measurement results obtained from the same sample in a wet dispersion. This step may include at least some of the features described in the section above entitled "Comparing Dry and Wet Dispersion Results."

[0202] In step 326, the software tool at least partially automatically evaluates the stability and robustness of the laser diffraction measurements performed using the selected measurement parameters. Approaches for evaluation of method stability and robustness may include at least some of the features described above in the section entitled "Method Validation and Robustness Testing."

[0203] In step 327, the software tool may transfer the method into a format that the user can save as a standard operating procedure, as described in the section entitled "Method Validation and Robustness Testing."

[0204] While the foregoing describes exemplary embodiments, they are not intended to limit the scope of the present invention, which should be determined with reference to the appended claims. Features of algorithms applied to determine optimal measurement parameters of a particular type may be similarly applied to determine optimal measurement parameters of a different type.

Claims

1. 1. A method for automatically selecting an optimal value or range for at least one measurement parameter for particle characterization by laser diffraction, comprising: receiving a plurality of measurements and associated measurement parameters, the plurality of measurements being performed on at least one particulate sample by using laser diffraction to identify at least one particle characteristic, the plurality of measurements being obtained using a plurality of values ​​for the at least one measurement parameter; and automatically selecting, using a processor, an optimal value or range of at least one measurement parameter for performing particle characterization by laser diffraction based on at least one particle characteristic and associated measurement parameters.

2. illuminating the sample with light from a light source; generating scattered light from interactions of the light beam with particles of the sample; detecting a distribution of the intensity of the scattered light over different scattering angles; 10. The method of claim 1, further comprising performing the measurement by: using a processor to determine particle characteristics from the distribution of the scattered light intensities over the different scattering angles.

3. The method of claim 1 or 2, wherein automatically selecting an optimal range for at least one measurement parameter comprises identifying values ​​or ranges of the measurement parameter that result in stable particle characteristics.

4. binning the measurements into a plurality of bins corresponding to different values ​​of the measurement parameter; determining a slope or variance between at least one particle characteristic of each measurement parameter bin; and selecting an optimal value range for the measurement parameter based on the slope or the variance of the bins for each measurement parameter.

5. The method of claim 4 , wherein binning the measurements comprises using a clustering algorithm.

6. Selecting the optimum value range for the measurement parameter is Identifying the bin with the smallest gradient or smallest variance; and identifying bins with slopes or variances below a threshold defined as a multiple of the minimum slope or minimum variance.

7. 7. The method of claim 1, wherein the at least one measurement parameter is selected from light blocking, amount of stirring, stirrer speed, sonication intensity, sonication time, feed rate, hopper gap, gas dispersant pressure, measurement time, and sample mass.

8. the at least one particle characteristic includes particle size, and the at least one measurement parameter includes shading; 8. The method of claim 1, wherein the method comprises limiting an optimum value or range of light blocking based on measured particle size.

9. prompting a user to provide an estimated particle density; the at least one measured parameter includes an amount of stirring; 9. The method of claim 1, wherein the method comprises limiting the value or range determined as the optimal amount of stirring to be below a threshold value in response to the estimated particle density being greater than or equal to the threshold value.

10. receiving a plurality of measurements and associated measurement parameters, the plurality of measurements being performed on at least one particulate sample by using laser diffraction to identify at least one particle characteristic, the plurality of measurements being obtained using a plurality of values ​​for the at least one measurement parameter; A non-volatile machine-readable medium that configures a processor to automatically select, based on at least one particle characteristic and an associated measurement parameter, an optimal value or range of at least one measurement parameter for performing particle characterization by laser diffraction.

11. A sample cell; a light source configured to illuminate the sample cell with a light beam and generate scattered light from interactions of the light beam with particles within the sample cell; a plurality of photodetectors configured to detect a distribution of the intensity of the scattered light over different scattering angles; a processor configured to determine particle characteristics from the distribution of the scattered light intensities over different scattering angles; The processor: receiving a plurality of measurements and associated measurement parameters, the plurality of measurements being performed on at least one particulate sample by using laser diffraction to identify at least one particle characteristic, the plurality of measurements being obtained using a plurality of values ​​for the at least one measurement parameter; A laser diffraction device further configured to automatically select, based on at least one particle characteristic and an associated measurement parameter, an optimal value or range of at least one measurement parameter for performing particle characterization by laser diffraction.

12. 12. The laser diffraction apparatus of claim 11 , wherein the processor is configured to obtain at least a portion of the plurality of measurements by continuously varying at least one measurement parameter and performing a laser diffraction analysis for each value of the at least one measurement parameter.

13. It also has a display, 13. The laser diffraction apparatus of claim 11 or 12, wherein the processor is configured to use a display to guide a user through a set of measurements to provide at least a portion of the plurality of measurements and associated measurement parameters.

14. A particle dispersion unit is provided. Measurement parameters include particle dispersion unit settings, 14. The laser diffraction apparatus of claim 11, wherein the processor is configured to automatically control the particle dispersion unit using the particle dispersion unit settings.

15. 15. The laser diffraction apparatus of any one of claims 11 to 14, wherein the processor is configured to perform a method according to any one of claims 1 to 10.

Images