Texture classification of particle size distribution data measured with laser diffraction method using different water types

The particle size analysis (PSA) by laser diffraction (LD) method can be used for monitor or control of particulate process, because it has the advantages of shorter measurement time and good repeatability, and a variety of commercial instruments is available. However particle size distribution (PSD) measured by LD method depends a great deal on not only optical detector configuration and calculation procedure but also on the system of sample loading into the measuring zone. From this fact, the validation of PSA by LD method should be done using reference particles (RP), whose size distribution is better to have a range over one decade of size, according to ISO 13320-1. For this purpose, the Association of Powder Process Industry and Engineering, Japan (APPIE) distribute the spherical barium titanate glass particles as RP of JIS Z 8900-1, whose size ranges are 1 - 10 µm (MBP 1 - 10), 3 - 30 µm (MBP 3 - 30) and 10 - 100 µm (MBP 10 - 100). This paper shows why LD method needs to check its performance by using RP, and then reports the results of the round robin test of two kinds of RP (MBP 1 - 10 and MBP 10 - 100) and silica RP candidate with 0.1 1.0 µm size measured by LD instruments, which was conducted by the Technical Group of Measurement and Control in APPIE. PSD results measured by LD instruments were almost same as each other for both RP samples. MBP 1 - 10 sample was well dispersed in water without detergent, but a few drops of detergent sometimes needed for dispersing MBP 10 - 100 sample. For MBP 1 - 10 sample, PSD by LD method was slightly smaller than that measured by scanning electro microscopy (SEM) or electro sensing zone (ESZ) methods. For MBP 10 - 100 sample, PSD by LD method agreed well with that by SEM or ESZ methods. Silica sample can be supplied to the users as the dry powder, which can be re-dispersed in water with small amount of surfactant. From the results of the round robin test using silica sample, PSD measured by LD method roughly agreed with EM method.

Application of the Arya and Paris (AP) model to estimate the soil water retention curve requires a detailed description of the particlesize distribution (PSD) but limited experimental PSD data are generally determined by the conventional sieve-hydrometer (SH) method. Detailed PSDs can be obtained by fitting a continuous model to SH data or performing measurements by the laser diffraction (LD) method. The AP model was applied to 40 Sicilian soils for which the PSD was measured by both the SH and LD methods. The scale factor was set equal to 1.38 (procedure AP1) or estimated by a logistical model with parameters gathered from literature (procedure AP2). For both SH and LD data, procedure AP2 allowed a more accurate prediction of the water retention than procedure AP1, confirming that it is not convenient to use a unique value of for soils that are very different in texture. Despite the differences in PSDs obtained by the SH and LD methods, the water retention predicted by a given procedure (AP1 or AP2) using SH or LD data was characterized by the same level of accuracy. Discrepancies in the estimated water retention from the two PSD measurement methods were attributed to underestimation of the finest diameter frequency obtained by the LD method. Analysis also showed that the soil water retention estimated using the SH method was affected by an estimation bias that could be corrected by an optimization procedure (OPT). Comparison of a-distributions and water retention shape indices obtained by the two methods (SH or LD) indicated that the shape-similarity hypothesis is better verified if the traditional sieve-hydrometer data are used to apply the AP model. The optimization procedure allowed more accurate predictions of the water retention curves than the traditional AP1 and AP2 procedures. Therefore, OPT can be considered a valid alternative to the more complex logistical model for estimating the water retention curve of Sicilian soils.

Application of the Arya and Paris (AP) model to estimate the soil water retention curve requires a detailed description of the particlesize distribution (PSD) but limited experimental PSD data are generally determined by the conventional sieve-hydrometer (SH) method. Detailed PSDs can be obtained by fitting a continuous model to SH data or performing measurements by the laser diffraction (LD) method. The AP model was applied to 40 Sicilian soils for which the PSD was measured by both the SH and LD methods. The scale factor was set equal to 1.38 (procedure AP1) or estimated by a logistical model with parameters gathered from literature (procedure AP2). For both SH and LD data, procedure AP2 allowed a more accurate prediction of the water retention than procedure AP1, confirming that it is not convenient to use a unique value of for soils that are very different in texture. Despite the differences in PSDs obtained by the SH and LD methods, the water retention predicted by a given procedure (AP1 or AP2) using SH or LD data was characterized by the same level of accuracy. Discrepancies in the estimated water retention from the two PSD measurement methods were attributed to underestimation of the finest diameter frequency obtained by the LD method. Analysis also showed that the soil water retention estimated using the SH method was affected by an estimation bias that could be corrected by an optimization procedure (OPT). Comparison of a-distributions and water retention shape indices obtained by the two methods (SH or LD) indicated that the shape-similarity hypothesis is better verified if the traditional sieve-hydrometer data are used to apply the AP model. The optimization procedure allowed more accurate predictions of the water retention curves than the traditional AP1 and AP2 procedures. Therefore, OPT can be considered a valid alternative to the more complex logistical model for estimating the water retention curve of Sicilian soils.

The objective of this study was to develop pedotransfer functions ( PTF s) for converting soil particle‐size distribution ( PSD ) data from the laser diffraction method ( LDM ) to the classical sieve–pipette method ( SPM ) for use on a wide range of temperate soil types. Four hundred soil samples, representative of E uropean soil types and climate zones, were selected from the LUCAS ( L and U se/ L and C over A rea F rame S urvey) topsoil database and their PSD s were determined with LDM and SPM . The LDM measurements were made on samples with (i) their organic matter ( OM ) removed and (ii) their OM content present. The ranges of PSD obtained with the two pretreatment methods enabled clay–silt and silt–sand boundaries from LDM (6.6 and 60.3 µm for soil with OM , respectively, and 5.8 and 69.2 µm for soil without OM , respectively) to be optimized. Optimization of the boundaries of the fractions considerably improved the prediction performance of SPM PSD from LDM PSD . Specific PTF s with different input requirements were developed for continental scale applications in E urope to convert data from LDM to SPM . The predictions of SPM clay, silt and sand contents were the most accurate with PTF s that used PSD from LDM and soil chemical properties ( R 2 0.83, 0.81, 0.94; RMSE 6.14, 7.91 and 6.58%, respectively). For the most accurate results no pretreatment for OM removal was required, but data on chemical properties were necessary. If no soil chemical data are available, the most accurate PTF s need input data of LDM PSD that originate from samples on which the OM content was removed prior to the PSD analysis. Highlights PTF s are developed to harmonize PSD data between laser diffraction ( LDM ) and sieve–pipette ( SPM ) methods. PTF s are derived from a representative dataset from E urope for application at the continental scale. Clay–silt and silt–sand boundaries for LDM without removing OM are at 6.6 and 60.3 µm, respectively. Clay–silt and silt–sand boundaries for LDM with OM removed are at 5.8 and 69.2 µm, respectively.

Particle size analysis is most fundamental analysis to powder technology. However, the validation or calibration of particle size analysis, especially the laser diffraction (LD) method, was not often carried out due to the difficulty to prepare the well-known sample. For example, the validation of LD method need spherical reference particles, of which size distribution is better to have a range over one decade of size. As this request, the Association of Powder Process Industry and Engineering, Japan distributed three kinds of standard reference particles (SRP) of spherical barium titanate glass; their size ranges are 1 – 10 µm, 3 – 30 µm and 10 – 100 µm. Those particles are also fitted to the reference particles in JIS Z 8900-1 Standard.In this paper, the particle size distribution (PSD) based on volume of SRP which was converted from the number-based PSD of SRP measured by a scanning electron microscope (SEM) was compared with PSD measured by LD instruments, which was conducted by the Technical Group of Measurement and Control in APPIE. PSD results measured by LD instruments were almost same as each other. PSD by LD method was slightly different from PSD measured by SEM. This discrepancy was discussed by using by a flow type image analysis method which could measure the size of particles in the aqueous solution. It was found that sample particles could not be dispersed completely in the aqueous solution even if using dispersing instruments such as an ultrasonic bath, and that this influence was not serious to the PSD result when using the suitable operating condition of ultrasonic bath.

Core Ideas LDM was evaluated by SPM and SEM for soil PSD determination. LDM significantly overestimated sand and silt but underestimated clay contents compared to SPM. LDM resulted in soil texture class shifts in nearly half of the soil samples in contrast to SPM. Linear conversions were significant for sand and silt but not for clay contents from LDM to SPM. LDM held statistically similar median particle sizes with SEM for PSDs within the clay fraction. The laser diffraction method (LDM) has been increasingly applied for quantifying soil particle size distribution (PSD), owing to its advantages of rapid analysis, high reproducibility, and continuous PSD measurement for a wide range of size fractions. However, some ambiguities exist regarding the comparability of results with those obtained using other classical methods. The objective of the current study was to evaluate LDM‐derived PSDs via comparisons with PSDs obtained with the standard sieve–pipette method (SPM) and from the absolute method of microscopy. A total of 277 soil samples were collected at different soil depths in a typical cropland in the northeast mountainous region of Beijing and analyzed with both SPM and LDM. Due to time and labor constraints, scanning electron microscopy (SEM) was performed on 100 samples randomly selected for the PSDs within the clay fraction withdrawn by SPM. The results manifested on the average 18.9% underestimation of clay content and 25.3% overestimation of silt content by LDM compared to SPM. These disagreements directly caused the shifts of soil texture class in 44.8% of the soil samples. Significant linear regression equations were generated to convert LDM–derived sand and silt contents to SPM–derived ones ( p < 0.01). The linear conversions for the clay content were only significant for the calibration samples, but possessed negative coefficients of determination for the validation set. According to SEM, silt‐sized particles were wrongly included in the clay fraction identified by SPM. Eliminating such particles, the clay contents corrected by SEM were significantly lower when assuming the shape of clay particles < 2 µm as plates or discs with constant thickness–diameter ratio of 1/10, and higher when considering the clay particles as spheres for volume calculation, in contrast to those measured by LDM ( P < 0.01). Detailed volume‐based PSDs within the clay fraction were further compared between SEM and LDM, revealing dissimilar PSD patterns but statistically similar median particle diameters. These findings suggest the effectiveness of LDM in soil PSD determination. Future work is needed to systematically quantify the impact of other possible factors such as clay mineralogy and refractive index on LDM‐derived PSDs.

Recent studies have shown that soil particle size analyses using laser diffraction method (LDM) can give compatible results compared with traditional sedimentation based methods, if the clay-silt particle size cutoff is transformed. Additionally, procedures including separation of the sand fraction by wet sieving and running a well dispersed sample of only fractions smaller than sand during laser diffraction measurement, have given promising results. The main purpose of the present study was to test a combination of these approaches for determining cutoff transformed LDM values on 44 soil samples from agricultural sites spread over Sweden, including its compatibility with the sieve and pipette method (SPM). Furthermore, these results were compared with results of transformed LDM values based on pedotransfer functions between measured LDM and SPM. Also LDM related aspects concerning scattering parameters, repeatability and organic matter calculations were studied. To find the optimum clay-silt cutoff, Lin’s concordance correlation coefficient (Lin’s CCC) was calculated. The highest value (0.977) was found with the 3.409–3.905 µm bin (a refractive index of 1.52 and an absorption coefficient of 0.1 was used). The pedotransfer-transformed LDM approach showed equally high Lin´s CCC as the cutoff-transformed approach for the different soil particle fraction size classes. With the cutoff-transformed LDM approach, 36 out of 44 samples were assigned to the same texture class as SPM, and with the pedotransfer-transformed LDM, the corresponding number was similar (34 out of 44 samples). The results here are promising for application in routine soil analyses, but more specific transformed clay-silt cutoffs and pedotransfer functions for LDM versus SPM should ideally be established for different types of soils. For this, microscopy and image analysis methods to help understand and quantify the influence of particle shapes on obtained particle size distributions are useful.

The validation of particle size analysis by laser diffraction (LD) method should be done using reference particles, whose size distribution is better to have a range over one decade of size, according to ISO 13320‐1. Two kinds of samples met this request, 1–10 μm and 10–100 μm samples, were distributed from the Association of Powder Process Industry and Engineering, Japan (APPIE).In this paper, the results of the round robin test of these two samples by LD method were reported together with the comparison of the size distributions measured by electrical sensing zone and scanning electron microscope methods. 1–10 μm sample was well dispersed in water without detergent, but a few drops of detergent sometimes needed for dispersing 10–100 μm sample. For 1–10 μm sample, the mean particle size by LD methods was slightly smaller than that by SEM method, and was agreed with each other results by seven different LD instruments. For 10–100 μm sample, the mean particle size by LD methods agree well with that by SEM method, but the discrepancy of one instrument in larger size range became larger than the results of other instruments.

Droplet size and morphology analyses of dry liquid

With regard to the differences between soil particle size distribution (PSD) obtained by sieve-sedimentation methods (SSMs), e.g., the sieve-pipette method (SPM) and the laser diffraction method (LDM), usually, the clay fraction content in LDM measurements is lower than in SSM. Two groups of reasons can be identified for this. Firstly, differences resulting from the features of the methods themselves. Secondly, differences resulting from the soil sample preparation (after sampling) and pretreatment (disaggregation). These differences not only cause difficulties in the PSDs comparability but also make it difficult, and sometimes even impossible, to apply the LDM results to soil texture classification. The solution to this difficulty may be to use pedotransfer functions and standardization of measurement procedures. The aim of this work was to validate the pedotransfer function proposed by Makó et al. (2017) on a Polish soil physics database for recalculating the results from LDM to compare them with SPM results and assess how the soil preparation and pretreatment influenced the obtained PSD and resulting soil texture classification. Using the pedotransfer function, 74 % accuracy was achieved, which is comparable with the results of other pedotransfer functions reported in the literature. Comparing the PSD results for the methods of sample preparation found in the literature after sampling (analysis of fresh, air dry and oven-dried – in 105 °C – soil) and pretreatment before the measurement (breaking soil aggregates chemically and physically) allows recommending the conducting of LDM PSD measurements using air dry soils and soil ultrasound disaggregation.

In order to determine the mechanical properties of small fibres, such as the tensile strength and modulus, or to determine the interface properties between fibre and matrix in composites, an accurate and reliable fibre diameter measurement method is required. Due to the limitation of the wavelength of light, fibre diameters measured optically with a microscope are often not sufficiently accurate. Scanning electron microscopy (SEM) measurement can yield exact results, but is very time-consuming. Therefore, a practical alternative is the use of the laser diffraction (LD) method [1]. However, the question is whether the results obtained from this fast method are in accordance with the exact SEM results. Li and Tietz [2] found a systematic difference between fibre diameters measured by SEM and diameters measured using the LD method. They obtained a 5% larger value from the LD method and proposed an error correction. This result could not be confirmed by our investigation. The principles of our LD equipment are illustrated in Figs 1 and 2. The beam diameter of the low-power (0.8 mW) HeNe laser used was reduced by two lenses. The first (convex) lens with a 60 mm focal length focused the laser beam and the second (concave) lens with a 40 mm focal length reparallelized the beam. With this two-lens layout the diameter of the beam could be reduced up to a quarter of the initial incident diameter. Furthermore, a polarization filter and a blind were used to suppress reflections from the lenses. The fibre sample was adjusted perpendicular to the laser beam. The interference pattern was recorded by a photodiode on a sledge simply driven by an analogue y t writer. The direct coupling of the moving diode to the writer avoided transmission faults of the interference pattern. The fibre diameter, d, could be calculated from the laser diffraction equation

Particle size distribution (PSD) is a major soil characteristic, which is essential and commonly used for the development of pedotransfer functions (PTFs) to estimate the water retention of soils. The laser diffraction method (LDM) became a popular alternative to the standard sieve‐hydrometer method (SHM) of PSD measurement. Unfortunately, PSDs determined with LDM and SHM methods differ sometimes substantially. Moreover, it is claimed that the laser diffraction method underestimates finer fractions in favor of coarser fractions. Several authors have tried to elaborate on methods to recalculate LDM PSD into its SHM counterparts, but no universal methodology has been developed to this date. In this paper, we use PSD determined by LDM directly for PTF development and compare it with the classical PTF approach based on PSD measured by SHM. Four different PTF models based on LDM particle size distribution data were developed, with different PSD characteristics taken as the models' input variables. The possibility of using alternative PSD characteristics, such as deciles, area moment mean and volume moment mean, for PTF development was examined. The accuracy of PTF models constructed on the basis of LDM‐measured PSD was comparable with that of the developed models using texture data obtained from SHM, giving approximately the same RMSE and R2 values. Our study shows that LDM‐measured particle size distribution may be directly used for PTF developments without any recalculations to their sieve‐hydrometer counterparts.

Laser diffraction method (LDM) provides a rapid solution for obtaining soil particle size distribution (PSD) especially in the analysis of a larger number of soil samples. However, there are disadvantages associated with the use of LDM for PSD measurements over the traditional sieve-pipette method (SPM). The present study is aimed to assess the suitability of LDM as a routine method for determining soil PSD, evaluate the precision or its repeatability of PSD parameters, and establish a simplified protocol for transforming the LDM results into SPM ones. The soil samples (n = 43) from 13 Chinese provinces were analyzed, and the results indicated that the relative errors for clay, silt, and sand fractions by the LDM were 35.34, 27.38, and 19.41%, respectively, compared with those by the SPM in the condition of limited sample numbers and large between-sample variation; samples could best be run at least twice during the LDM analysis to reduce the error caused by a limited sample volume; and a soil particle refractive index of 1.50 and a soil particle absorption index of 0.01 were found to be optimal for the Mie theory model. With relatively limited sample numbers and apparent textural difference between the samples, the distinct incompatibilities were observed in the present work between the PSD obtained by the LDM and SPM. However, depending on the specific research purpose, the deviations between the LDM and SPM may be considerably reduced with an increase in the sample capacity or a decrease in the spatial scale.

It is expected in the future that soil particle size distribution (PSD) measurements by laser diffraction method (LDM) may replace sieve-pipette sedimentation methods (SPM) as they are faster, require less sample, and are accurate and reproducible. LDM measurement result is a continuous function of PSD, which can facilitate the conversion between the various limits (by countries, by scientific field) of the calculated particle size fractions (PSF – e.g. clay, silt, sand). Currently, there is no standard method for LDM PSD measurement. Many different types of instruments and preparation devices are currently used in laboratories, with various sample preparation, pre-treatment and measurement methods (duration, chemical and/or mechanical dispersion, settings, etc.). In soil LDM PSD measurements, researchers put relatively little emphasis on the choice of the type of aqueous media used. Thus, it is still questionable to what extent the results of the LDM measurement depend on the selection of the dispersion method and the aqueous media. For our research, eight soil samples with various physical and chemical properties were collected in Hungary. The particle size fractions (clay, silt, sand) determined with LDM (Malvern Mastersizer 3000) measured in three types of aqueous media (distilled, deionized and tap water), in different combinations of two dispersion methods (no treatment, ultrasonic or chemical dispersion with Calgon and their combination) were compared. For the comparison, PSF results of the conventional sieve pipette method (SPM) were used as a reference. Our results showed that LDM measurement can achieve various degrees of dispersion with different preparations, in many cases only partial dispersion, disaggregation, sometimes re-aggregation, and flocculation of soil particles were observed as compared to full preparation (in SPM). The “disaggregation pattern” of the soil samples also depended on the quality of the aqueous media and the properties of the soil investigated, because several types and degrees of interactions could occur in the various soil-liquid-dispersant/disaggregation effect systems.

The existing classifications of soil texture are based on the sedimentation data. The aim of this article is to consider the ways to develop soil textural classification on the basis of particle-size distribution (PSD) data obtained by the laser diffraction method. A detailed comparison of PSD data obtained by the classical pipette method and the method of laser diffractometry has been performed. We have shown the reproducibility of the laser diffraction method and the effect of the oxidation stage on the soil texture class. This study is based on eight genetic soil types (overall, 32 full-profile soil pits) forming a zonal soil sequence from Podzols (Subpolar Urals) to ferrallitic soil (southwest Oceania) and differing in their mineralogical compositions, textures, and elementary pedogenetic processes. The direct use of the Kachinskii and USDA classifications with the data of the laser diffraction method leads to mistakes in determining the soil texture class in 43 and 65% of cases, respectively. The increasing complexity of recalculation, introduction of new variables, and accounting for interlaboratory errors allow us to determine correctly the texture class according to the Kachinskii and USDA classifications in no more than 70 and 72% of soil samples, respectively. The most simple and effective approach to solve the classification problem for the laser diffraction method is to calibrate existing classifications directly on the basis of data on soil samples, for which the texture class was determined by the field method.